diff options
Diffstat (limited to 'fs')
342 files changed, 15836 insertions, 8361 deletions
diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 620d934..8aa56bb 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -320,31 +320,21 @@ fail_option_alloc: struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, const char *dev_name, char *data) { - int retval = -EINVAL; struct p9_fid *fid; - int rc; + int rc = -ENOMEM; v9ses->uname = kstrdup(V9FS_DEFUSER, GFP_KERNEL); if (!v9ses->uname) - return ERR_PTR(-ENOMEM); + goto err_names; v9ses->aname = kstrdup(V9FS_DEFANAME, GFP_KERNEL); - if (!v9ses->aname) { - kfree(v9ses->uname); - return ERR_PTR(-ENOMEM); - } + if (!v9ses->aname) + goto err_names; init_rwsem(&v9ses->rename_sem); rc = bdi_setup_and_register(&v9ses->bdi, "9p"); - if (rc) { - kfree(v9ses->aname); - kfree(v9ses->uname); - return ERR_PTR(rc); - } - - spin_lock(&v9fs_sessionlist_lock); - list_add(&v9ses->slist, &v9fs_sessionlist); - spin_unlock(&v9fs_sessionlist_lock); + if (rc) + goto err_names; v9ses->uid = INVALID_UID; v9ses->dfltuid = V9FS_DEFUID; @@ -352,10 +342,9 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->clnt = p9_client_create(dev_name, data); if (IS_ERR(v9ses->clnt)) { - retval = PTR_ERR(v9ses->clnt); - v9ses->clnt = NULL; + rc = PTR_ERR(v9ses->clnt); p9_debug(P9_DEBUG_ERROR, "problem initializing 9p client\n"); - goto error; + goto err_bdi; } v9ses->flags = V9FS_ACCESS_USER; @@ -368,10 +357,8 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, } rc = v9fs_parse_options(v9ses, data); - if (rc < 0) { - retval = rc; - goto error; - } + if (rc < 0) + goto err_clnt; v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ; @@ -405,10 +392,9 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, fid = p9_client_attach(v9ses->clnt, NULL, v9ses->uname, INVALID_UID, v9ses->aname); if (IS_ERR(fid)) { - retval = PTR_ERR(fid); - fid = NULL; + rc = PTR_ERR(fid); p9_debug(P9_DEBUG_ERROR, "cannot attach\n"); - goto error; + goto err_clnt; } if ((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_SINGLE) @@ -420,12 +406,20 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, /* register the session for caching */ v9fs_cache_session_get_cookie(v9ses); #endif + spin_lock(&v9fs_sessionlist_lock); + list_add(&v9ses->slist, &v9fs_sessionlist); + spin_unlock(&v9fs_sessionlist_lock); return fid; -error: +err_clnt: + p9_client_destroy(v9ses->clnt); +err_bdi: bdi_destroy(&v9ses->bdi); - return ERR_PTR(retval); +err_names: + kfree(v9ses->uname); + kfree(v9ses->aname); + return ERR_PTR(rc); } /** diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index e99a338..bf495ce 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -130,11 +130,7 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, fid = v9fs_session_init(v9ses, dev_name, data); if (IS_ERR(fid)) { retval = PTR_ERR(fid); - /* - * we need to call session_close to tear down some - * of the data structure setup by session_init - */ - goto close_session; + goto free_session; } sb = sget(fs_type, NULL, v9fs_set_super, flags, v9ses); @@ -195,8 +191,8 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, clunk_fid: p9_client_clunk(fid); -close_session: v9fs_session_close(v9ses); +free_session: kfree(v9ses); return ERR_PTR(retval); diff --git a/fs/Makefile b/fs/Makefile index cb92fd4..cb20e4b 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -115,7 +115,6 @@ obj-$(CONFIG_AFS_FS) += afs/ obj-$(CONFIG_NILFS2_FS) += nilfs2/ obj-$(CONFIG_BEFS_FS) += befs/ obj-$(CONFIG_HOSTFS) += hostfs/ -obj-$(CONFIG_HPPFS) += hppfs/ obj-$(CONFIG_CACHEFILES) += cachefiles/ obj-$(CONFIG_DEBUG_FS) += debugfs/ obj-$(CONFIG_TRACING) += tracefs/ diff --git a/fs/adfs/super.c b/fs/adfs/super.c index a19c31d..4d4a0df 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -242,7 +242,7 @@ static struct kmem_cache *adfs_inode_cachep; static struct inode *adfs_alloc_inode(struct super_block *sb) { struct adfs_inode_info *ei; - ei = (struct adfs_inode_info *)kmem_cache_alloc(adfs_inode_cachep, GFP_KERNEL); + ei = kmem_cache_alloc(adfs_inode_cachep, GFP_KERNEL); if (!ei) return NULL; return &ei->vfs_inode; diff --git a/fs/affs/amigaffs.c b/fs/affs/amigaffs.c index a8f463c..5fa92bc 100644 --- a/fs/affs/amigaffs.c +++ b/fs/affs/amigaffs.c @@ -140,7 +140,7 @@ affs_remove_link(struct dentry *dentry) { struct inode *dir, *inode = d_inode(dentry); struct super_block *sb = inode->i_sb; - struct buffer_head *bh = NULL, *link_bh = NULL; + struct buffer_head *bh, *link_bh = NULL; u32 link_ino, ino; int retval; diff --git a/fs/affs/inode.c b/fs/affs/inode.c index a022f4a..1734950 100644 --- a/fs/affs/inode.c +++ b/fs/affs/inode.c @@ -346,7 +346,7 @@ affs_add_entry(struct inode *dir, struct inode *inode, struct dentry *dentry, s3 { struct super_block *sb = dir->i_sb; struct buffer_head *inode_bh = NULL; - struct buffer_head *bh = NULL; + struct buffer_head *bh; u32 block = 0; int retval; diff --git a/fs/affs/symlink.c b/fs/affs/symlink.c index f39b71c..ea5b69a 100644 --- a/fs/affs/symlink.c +++ b/fs/affs/symlink.c @@ -16,14 +16,12 @@ static int affs_symlink_readpage(struct file *file, struct page *page) struct inode *inode = page->mapping->host; char *link = kmap(page); struct slink_front *lf; - int err; int i, j; char c; char lc; pr_debug("follow_link(ino=%lu)\n", inode->i_ino); - err = -EIO; bh = affs_bread(inode->i_sb, inode->i_ino); if (!bh) goto fail; @@ -66,7 +64,7 @@ fail: SetPageError(page); kunmap(page); unlock_page(page); - return err; + return -EIO; } const struct address_space_operations affs_symlink_aops = { diff --git a/fs/afs/rxrpc.c b/fs/afs/rxrpc.c index 3a57a1b..b506428 100644 --- a/fs/afs/rxrpc.c +++ b/fs/afs/rxrpc.c @@ -85,7 +85,7 @@ int afs_open_socket(void) return -ENOMEM; } - ret = sock_create_kern(AF_RXRPC, SOCK_DGRAM, PF_INET, &socket); + ret = sock_create_kern(&init_net, AF_RXRPC, SOCK_DGRAM, PF_INET, &socket); if (ret < 0) { destroy_workqueue(afs_async_calls); _leave(" = %d [socket]", ret); diff --git a/fs/befs/btree.c b/fs/befs/btree.c index 0826e91..22c16628 100644 --- a/fs/befs/btree.c +++ b/fs/befs/btree.c @@ -137,8 +137,8 @@ static int befs_bt_read_super(struct super_block *sb, befs_data_stream * ds, befs_btree_super * sup) { - struct buffer_head *bh = NULL; - befs_disk_btree_super *od_sup = NULL; + struct buffer_head *bh; + befs_disk_btree_super *od_sup; befs_debug(sb, "---> %s", __func__); @@ -250,7 +250,7 @@ int befs_btree_find(struct super_block *sb, befs_data_stream * ds, const char *key, befs_off_t * value) { - struct befs_btree_node *this_node = NULL; + struct befs_btree_node *this_node; befs_btree_super bt_super; befs_off_t node_off; int res; diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 5046b62..6b65996 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -918,7 +918,7 @@ static int load_elf_binary(struct linux_binprm *bprm) total_size = total_mapping_size(elf_phdata, loc->elf_ex.e_phnum); if (!total_size) { - error = -EINVAL; + retval = -EINVAL; goto out_free_dentry; } } diff --git a/fs/block_dev.c b/fs/block_dev.c index 12b22dd..1982437 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -14,6 +14,7 @@ #include <linux/device_cgroup.h> #include <linux/highmem.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <linux/module.h> #include <linux/blkpg.h> #include <linux/magic.h> @@ -42,7 +43,7 @@ static inline struct bdev_inode *BDEV_I(struct inode *inode) return container_of(inode, struct bdev_inode, vfs_inode); } -inline struct block_device *I_BDEV(struct inode *inode) +struct block_device *I_BDEV(struct inode *inode) { return &BDEV_I(inode)->bdev; } @@ -379,7 +380,7 @@ int bdev_read_page(struct block_device *bdev, sector_t sector, struct page *page) { const struct block_device_operations *ops = bdev->bd_disk->fops; - if (!ops->rw_page) + if (!ops->rw_page || bdev_get_integrity(bdev)) return -EOPNOTSUPP; return ops->rw_page(bdev, sector + get_start_sect(bdev), page, READ); } @@ -410,7 +411,7 @@ int bdev_write_page(struct block_device *bdev, sector_t sector, int result; int rw = (wbc->sync_mode == WB_SYNC_ALL) ? WRITE_SYNC : WRITE; const struct block_device_operations *ops = bdev->bd_disk->fops; - if (!ops->rw_page) + if (!ops->rw_page || bdev_get_integrity(bdev)) return -EOPNOTSUPP; set_page_writeback(page); result = ops->rw_page(bdev, sector + get_start_sect(bdev), page, rw); @@ -555,7 +556,8 @@ static struct file_system_type bd_type = { .kill_sb = kill_anon_super, }; -static struct super_block *blockdev_superblock __read_mostly; +struct super_block *blockdev_superblock __read_mostly; +EXPORT_SYMBOL_GPL(blockdev_superblock); void __init bdev_cache_init(void) { @@ -696,11 +698,6 @@ static struct block_device *bd_acquire(struct inode *inode) return bdev; } -int sb_is_blkdev_sb(struct super_block *sb) -{ - return sb == blockdev_superblock; -} - /* Call when you free inode */ void bd_forget(struct inode *inode) diff --git a/fs/btrfs/async-thread.c b/fs/btrfs/async-thread.c index df9932b..1ce06c84 100644 --- a/fs/btrfs/async-thread.c +++ b/fs/btrfs/async-thread.c @@ -85,6 +85,7 @@ BTRFS_WORK_HELPER(extent_refs_helper); BTRFS_WORK_HELPER(scrub_helper); BTRFS_WORK_HELPER(scrubwrc_helper); BTRFS_WORK_HELPER(scrubnc_helper); +BTRFS_WORK_HELPER(scrubparity_helper); static struct __btrfs_workqueue * __btrfs_alloc_workqueue(const char *name, unsigned int flags, int max_active, diff --git a/fs/btrfs/async-thread.h b/fs/btrfs/async-thread.h index ec2ee47..b0b093b 100644 --- a/fs/btrfs/async-thread.h +++ b/fs/btrfs/async-thread.h @@ -64,6 +64,8 @@ BTRFS_WORK_HELPER_PROTO(extent_refs_helper); BTRFS_WORK_HELPER_PROTO(scrub_helper); BTRFS_WORK_HELPER_PROTO(scrubwrc_helper); BTRFS_WORK_HELPER_PROTO(scrubnc_helper); +BTRFS_WORK_HELPER_PROTO(scrubparity_helper); + struct btrfs_workqueue *btrfs_alloc_workqueue(const char *name, unsigned int flags, diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index 9de772e..802fabb 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -250,8 +250,12 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path, * the first item to check. But sometimes, we may enter it with * slot==nritems. In that case, go to the next leaf before we continue. */ - if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) - ret = btrfs_next_old_leaf(root, path, time_seq); + if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) { + if (time_seq == (u64)-1) + ret = btrfs_next_leaf(root, path); + else + ret = btrfs_next_old_leaf(root, path, time_seq); + } while (!ret && count < total_refs) { eb = path->nodes[0]; @@ -291,7 +295,10 @@ static int add_all_parents(struct btrfs_root *root, struct btrfs_path *path, eie = NULL; } next: - ret = btrfs_next_old_item(root, path, time_seq); + if (time_seq == (u64)-1) + ret = btrfs_next_item(root, path); + else + ret = btrfs_next_old_item(root, path, time_seq); } if (ret > 0) @@ -334,6 +341,8 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info, if (path->search_commit_root) root_level = btrfs_header_level(root->commit_root); + else if (time_seq == (u64)-1) + root_level = btrfs_header_level(root->node); else root_level = btrfs_old_root_level(root, time_seq); @@ -343,7 +352,12 @@ static int __resolve_indirect_ref(struct btrfs_fs_info *fs_info, } path->lowest_level = level; - ret = btrfs_search_old_slot(root, &ref->key_for_search, path, time_seq); + if (time_seq == (u64)-1) + ret = btrfs_search_slot(NULL, root, &ref->key_for_search, path, + 0, 0); + else + ret = btrfs_search_old_slot(root, &ref->key_for_search, path, + time_seq); /* root node has been locked, we can release @subvol_srcu safely here */ srcu_read_unlock(&fs_info->subvol_srcu, index); @@ -491,7 +505,9 @@ static int __add_missing_keys(struct btrfs_fs_info *fs_info, BUG_ON(!ref->wanted_disk_byte); eb = read_tree_block(fs_info->tree_root, ref->wanted_disk_byte, 0); - if (!eb || !extent_buffer_uptodate(eb)) { + if (IS_ERR(eb)) { + return PTR_ERR(eb); + } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); return -EIO; } @@ -507,7 +523,7 @@ static int __add_missing_keys(struct btrfs_fs_info *fs_info, } /* - * merge two lists of backrefs and adjust counts accordingly + * merge backrefs and adjust counts accordingly * * mode = 1: merge identical keys, if key is set * FIXME: if we add more keys in __add_prelim_ref, we can merge more here. @@ -535,9 +551,9 @@ static void __merge_refs(struct list_head *head, int mode) ref2 = list_entry(pos2, struct __prelim_ref, list); + if (!ref_for_same_block(ref1, ref2)) + continue; if (mode == 1) { - if (!ref_for_same_block(ref1, ref2)) - continue; if (!ref1->parent && ref2->parent) { xchg = ref1; ref1 = ref2; @@ -572,8 +588,8 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq, struct list_head *prefs, u64 *total_refs, u64 inum) { + struct btrfs_delayed_ref_node *node; struct btrfs_delayed_extent_op *extent_op = head->extent_op; - struct rb_node *n = &head->node.rb_node; struct btrfs_key key; struct btrfs_key op_key = {0}; int sgn; @@ -583,12 +599,7 @@ static int __add_delayed_refs(struct btrfs_delayed_ref_head *head, u64 seq, btrfs_disk_key_to_cpu(&op_key, &extent_op->key); spin_lock(&head->lock); - n = rb_first(&head->ref_root); - while (n) { - struct btrfs_delayed_ref_node *node; - node = rb_entry(n, struct btrfs_delayed_ref_node, - rb_node); - n = rb_next(n); + list_for_each_entry(node, &head->ref_list, list) { if (node->seq > seq) continue; @@ -880,6 +891,13 @@ static int __add_keyed_refs(struct btrfs_fs_info *fs_info, * indirect refs to their parent bytenr. * When roots are found, they're added to the roots list * + * NOTE: This can return values > 0 + * + * If time_seq is set to (u64)-1, it will not search delayed_refs, and behave + * much like trans == NULL case, the difference only lies in it will not + * commit root. + * The special case is for qgroup to search roots in commit_transaction(). + * * FIXME some caching might speed things up */ static int find_parent_nodes(struct btrfs_trans_handle *trans, @@ -918,6 +936,9 @@ static int find_parent_nodes(struct btrfs_trans_handle *trans, path->skip_locking = 1; } + if (time_seq == (u64)-1) + path->skip_locking = 1; + /* * grab both a lock on the path and a lock on the delayed ref head. * We need both to get a consistent picture of how the refs look @@ -932,9 +953,10 @@ again: BUG_ON(ret == 0); #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS - if (trans && likely(trans->type != __TRANS_DUMMY)) { + if (trans && likely(trans->type != __TRANS_DUMMY) && + time_seq != (u64)-1) { #else - if (trans) { + if (trans && time_seq != (u64)-1) { #endif /* * look if there are updates for this ref queued and lock the @@ -1032,7 +1054,10 @@ again: eb = read_tree_block(fs_info->extent_root, ref->parent, 0); - if (!eb || !extent_buffer_uptodate(eb)) { + if (IS_ERR(eb)) { + ret = PTR_ERR(eb); + goto out; + } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); ret = -EIO; goto out; @@ -1198,6 +1223,19 @@ int btrfs_find_all_roots(struct btrfs_trans_handle *trans, return ret; } +/** + * btrfs_check_shared - tell us whether an extent is shared + * + * @trans: optional trans handle + * + * btrfs_check_shared uses the backref walking code but will short + * circuit as soon as it finds a root or inode that doesn't match the + * one passed in. This provides a significant performance benefit for + * callers (such as fiemap) which want to know whether the extent is + * shared but do not need a ref count. + * + * Return: 0 if extent is not shared, 1 if it is shared, < 0 on error. + */ int btrfs_check_shared(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info, u64 root_objectid, u64 inum, u64 bytenr) @@ -1226,11 +1264,13 @@ int btrfs_check_shared(struct btrfs_trans_handle *trans, ret = find_parent_nodes(trans, fs_info, bytenr, elem.seq, tmp, roots, NULL, root_objectid, inum); if (ret == BACKREF_FOUND_SHARED) { + /* this is the only condition under which we return 1 */ ret = 1; break; } if (ret < 0 && ret != -ENOENT) break; + ret = 0; node = ulist_next(tmp, &uiter); if (!node) break; diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 0f11ebc..54114b4 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -1439,8 +1439,9 @@ get_old_root(struct btrfs_root *root, u64 time_seq) btrfs_tree_read_unlock(eb_root); free_extent_buffer(eb_root); old = read_tree_block(root, logical, 0); - if (WARN_ON(!old || !extent_buffer_uptodate(old))) { - free_extent_buffer(old); + if (WARN_ON(IS_ERR(old) || !extent_buffer_uptodate(old))) { + if (!IS_ERR(old)) + free_extent_buffer(old); btrfs_warn(root->fs_info, "failed to read tree block %llu from get_old_root", logical); } else { @@ -1685,7 +1686,9 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans, if (!cur || !uptodate) { if (!cur) { cur = read_tree_block(root, blocknr, gen); - if (!cur || !extent_buffer_uptodate(cur)) { + if (IS_ERR(cur)) { + return PTR_ERR(cur); + } else if (!extent_buffer_uptodate(cur)) { free_extent_buffer(cur); return -EIO; } @@ -1864,8 +1867,9 @@ static noinline struct extent_buffer *read_node_slot(struct btrfs_root *root, eb = read_tree_block(root, btrfs_node_blockptr(parent, slot), btrfs_node_ptr_generation(parent, slot)); - if (eb && !extent_buffer_uptodate(eb)) { - free_extent_buffer(eb); + if (IS_ERR(eb) || !extent_buffer_uptodate(eb)) { + if (!IS_ERR(eb)) + free_extent_buffer(eb); eb = NULL; } @@ -2494,7 +2498,7 @@ read_block_for_search(struct btrfs_trans_handle *trans, ret = -EAGAIN; tmp = read_tree_block(root, blocknr, 0); - if (tmp) { + if (!IS_ERR(tmp)) { /* * If the read above didn't mark this buffer up to date, * it will never end up being up to date. Set ret to EIO now diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 6f364e1..80a9aef 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -174,7 +174,7 @@ struct btrfs_ordered_sum; /* csum types */ #define BTRFS_CSUM_TYPE_CRC32 0 -static int btrfs_csum_sizes[] = { 4, 0 }; +static int btrfs_csum_sizes[] = { 4 }; /* four bytes for CRC32 */ #define BTRFS_EMPTY_DIR_SIZE 0 @@ -1619,10 +1619,7 @@ struct btrfs_fs_info { struct task_struct *cleaner_kthread; int thread_pool_size; - struct kobject super_kobj; struct kobject *space_info_kobj; - struct kobject *device_dir_kobj; - struct completion kobj_unregister; int do_barriers; int closing; int log_root_recovering; @@ -1698,6 +1695,7 @@ struct btrfs_fs_info { struct btrfs_workqueue *scrub_workers; struct btrfs_workqueue *scrub_wr_completion_workers; struct btrfs_workqueue *scrub_nocow_workers; + struct btrfs_workqueue *scrub_parity_workers; #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY u32 check_integrity_print_mask; @@ -1735,7 +1733,7 @@ struct btrfs_fs_info { /* list of dirty qgroups to be written at next commit */ struct list_head dirty_qgroups; - /* used by btrfs_qgroup_record_ref for an efficient tree traversal */ + /* used by qgroup for an efficient tree traversal */ u64 qgroup_seq; /* qgroup rescan items */ @@ -3458,6 +3456,7 @@ int btrfs_check_data_free_space(struct inode *inode, u64 bytes, u64 write_bytes) void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes); void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans, struct btrfs_root *root); +void btrfs_trans_release_chunk_metadata(struct btrfs_trans_handle *trans); int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans, struct inode *inode); void btrfs_orphan_release_metadata(struct inode *inode); @@ -3515,6 +3514,9 @@ int btrfs_delayed_refs_qgroup_accounting(struct btrfs_trans_handle *trans, int __get_raid_index(u64 flags); int btrfs_start_write_no_snapshoting(struct btrfs_root *root); void btrfs_end_write_no_snapshoting(struct btrfs_root *root); +void check_system_chunk(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + const u64 type); /* ctree.c */ int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key, int level, int *slot); @@ -4050,6 +4052,7 @@ void btrfs_printk(const struct btrfs_fs_info *fs_info, const char *fmt, ...) #ifdef CONFIG_BTRFS_ASSERT +__cold static inline void assfail(char *expr, char *file, int line) { pr_err("BTRFS: assertion failed: %s, file: %s, line: %d", @@ -4065,10 +4068,12 @@ static inline void assfail(char *expr, char *file, int line) #define btrfs_assert() __printf(5, 6) +__cold void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, unsigned int line, int errno, const char *fmt, ...); +__cold void __btrfs_abort_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root, const char *function, unsigned int line, int errno); @@ -4111,11 +4116,17 @@ static inline int __btrfs_fs_incompat(struct btrfs_fs_info *fs_info, u64 flag) * Call btrfs_abort_transaction as early as possible when an error condition is * detected, that way the exact line number is reported. */ - #define btrfs_abort_transaction(trans, root, errno) \ do { \ - __btrfs_abort_transaction(trans, root, __func__, \ - __LINE__, errno); \ + /* Report first abort since mount */ \ + if (!test_and_set_bit(BTRFS_FS_STATE_TRANS_ABORTED, \ + &((root)->fs_info->fs_state))) { \ + WARN(1, KERN_DEBUG \ + "BTRFS: Transaction aborted (error %d)\n", \ + (errno)); \ + } \ + __btrfs_abort_transaction((trans), (root), __func__, \ + __LINE__, (errno)); \ } while (0) #define btrfs_std_error(fs_info, errno) \ @@ -4132,6 +4143,7 @@ do { \ } while (0) __printf(5, 6) +__cold void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function, unsigned int line, int errno, const char *fmt, ...); diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index 8f8ed7d..ac3e81d 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -22,6 +22,7 @@ #include "ctree.h" #include "delayed-ref.h" #include "transaction.h" +#include "qgroup.h" struct kmem_cache *btrfs_delayed_ref_head_cachep; struct kmem_cache *btrfs_delayed_tree_ref_cachep; @@ -84,87 +85,6 @@ static int comp_data_refs(struct btrfs_delayed_data_ref *ref2, return 0; } -/* - * entries in the rb tree are ordered by the byte number of the extent, - * type of the delayed backrefs and content of delayed backrefs. - */ -static int comp_entry(struct btrfs_delayed_ref_node *ref2, - struct btrfs_delayed_ref_node *ref1, - bool compare_seq) -{ - if (ref1->bytenr < ref2->bytenr) - return -1; - if (ref1->bytenr > ref2->bytenr) - return 1; - if (ref1->is_head && ref2->is_head) - return 0; - if (ref2->is_head) - return -1; - if (ref1->is_head) - return 1; - if (ref1->type < ref2->type) - return -1; - if (ref1->type > ref2->type) - return 1; - if (ref1->no_quota > ref2->no_quota) - return 1; - if (ref1->no_quota < ref2->no_quota) - return -1; - /* merging of sequenced refs is not allowed */ - if (compare_seq) { - if (ref1->seq < ref2->seq) - return -1; - if (ref1->seq > ref2->seq) - return 1; - } - if (ref1->type == BTRFS_TREE_BLOCK_REF_KEY || - ref1->type == BTRFS_SHARED_BLOCK_REF_KEY) { - return comp_tree_refs(btrfs_delayed_node_to_tree_ref(ref2), - btrfs_delayed_node_to_tree_ref(ref1), - ref1->type); - } else if (ref1->type == BTRFS_EXTENT_DATA_REF_KEY || - ref1->type == BTRFS_SHARED_DATA_REF_KEY) { - return comp_data_refs(btrfs_delayed_node_to_data_ref(ref2), - btrfs_delayed_node_to_data_ref(ref1)); - } - BUG(); - return 0; -} - -/* - * insert a new ref into the rbtree. This returns any existing refs - * for the same (bytenr,parent) tuple, or NULL if the new node was properly - * inserted. - */ -static struct btrfs_delayed_ref_node *tree_insert(struct rb_root *root, - struct rb_node *node) -{ - struct rb_node **p = &root->rb_node; - struct rb_node *parent_node = NULL; - struct btrfs_delayed_ref_node *entry; - struct btrfs_delayed_ref_node *ins; - int cmp; - - ins = rb_entry(node, struct btrfs_delayed_ref_node, rb_node); - while (*p) { - parent_node = *p; - entry = rb_entry(parent_node, struct btrfs_delayed_ref_node, - rb_node); - - cmp = comp_entry(entry, ins, 1); - if (cmp < 0) - p = &(*p)->rb_left; - else if (cmp > 0) - p = &(*p)->rb_right; - else - return entry; - } - - rb_link_node(node, parent_node, p); - rb_insert_color(node, root); - return NULL; -} - /* insert a new ref to head ref rbtree */ static struct btrfs_delayed_ref_head *htree_insert(struct rb_root *root, struct rb_node *node) @@ -268,7 +188,7 @@ static inline void drop_delayed_ref(struct btrfs_trans_handle *trans, rb_erase(&head->href_node, &delayed_refs->href_root); } else { assert_spin_locked(&head->lock); - rb_erase(&ref->rb_node, &head->ref_root); + list_del(&ref->list); } ref->in_tree = 0; btrfs_put_delayed_ref(ref); @@ -277,99 +197,6 @@ static inline void drop_delayed_ref(struct btrfs_trans_handle *trans, trans->delayed_ref_updates--; } -static int merge_ref(struct btrfs_trans_handle *trans, - struct btrfs_delayed_ref_root *delayed_refs, - struct btrfs_delayed_ref_head *head, - struct btrfs_delayed_ref_node *ref, u64 seq) -{ - struct rb_node *node; - int mod = 0; - int done = 0; - - node = rb_next(&ref->rb_node); - while (!done && node) { - struct btrfs_delayed_ref_node *next; - - next = rb_entry(node, struct btrfs_delayed_ref_node, rb_node); - node = rb_next(node); - if (seq && next->seq >= seq) - break; - if (comp_entry(ref, next, 0)) - continue; - - if (ref->action == next->action) { - mod = next->ref_mod; - } else { - if (ref->ref_mod < next->ref_mod) { - struct btrfs_delayed_ref_node *tmp; - - tmp = ref; - ref = next; - next = tmp; - done = 1; - } - mod = -next->ref_mod; - } - - drop_delayed_ref(trans, delayed_refs, head, next); - ref->ref_mod += mod; - if (ref->ref_mod == 0) { - drop_delayed_ref(trans, delayed_refs, head, ref); - done = 1; - } else { - /* - * You can't have multiples of the same ref on a tree - * block. - */ - WARN_ON(ref->type == BTRFS_TREE_BLOCK_REF_KEY || - ref->type == BTRFS_SHARED_BLOCK_REF_KEY); - } - } - return done; -} - -void btrfs_merge_delayed_refs(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, - struct btrfs_delayed_ref_root *delayed_refs, - struct btrfs_delayed_ref_head *head) -{ - struct rb_node *node; - u64 seq = 0; - - assert_spin_locked(&head->lock); - /* - * We don't have too much refs to merge in the case of delayed data - * refs. - */ - if (head->is_data) - return; - - spin_lock(&fs_info->tree_mod_seq_lock); - if (!list_empty(&fs_info->tree_mod_seq_list)) { - struct seq_list *elem; - - elem = list_first_entry(&fs_info->tree_mod_seq_list, - struct seq_list, list); - seq = elem->seq; - } - spin_unlock(&fs_info->tree_mod_seq_lock); - - node = rb_first(&head->ref_root); - while (node) { - struct btrfs_delayed_ref_node *ref; - - ref = rb_entry(node, struct btrfs_delayed_ref_node, - rb_node); - /* We can't merge refs that are outside of our seq count */ - if (seq && ref->seq >= seq) - break; - if (merge_ref(trans, delayed_refs, head, ref, seq)) - node = rb_first(&head->ref_root); - else - node = rb_next(&ref->rb_node); - } -} - int btrfs_check_delayed_seq(struct btrfs_fs_info *fs_info, struct btrfs_delayed_ref_root *delayed_refs, u64 seq) @@ -443,45 +270,71 @@ again: } /* - * helper function to update an extent delayed ref in the - * rbtree. existing and update must both have the same - * bytenr and parent + * Helper to insert the ref_node to the tail or merge with tail. * - * This may free existing if the update cancels out whatever - * operation it was doing. + * Return 0 for insert. + * Return >0 for merge. */ -static noinline void -update_existing_ref(struct btrfs_trans_handle *trans, - struct btrfs_delayed_ref_root *delayed_refs, - struct btrfs_delayed_ref_head *head, - struct btrfs_delayed_ref_node *existing, - struct btrfs_delayed_ref_node *update) +static int +add_delayed_ref_tail_merge(struct btrfs_trans_handle *trans, + struct btrfs_delayed_ref_root *root, + struct btrfs_delayed_ref_head *href, + struct btrfs_delayed_ref_node *ref) { - if (update->action != existing->action) { - /* - * this is effectively undoing either an add or a - * drop. We decrement the ref_mod, and if it goes - * down to zero we just delete the entry without - * every changing the extent allocation tree. - */ - existing->ref_mod--; - if (existing->ref_mod == 0) - drop_delayed_ref(trans, delayed_refs, head, existing); - else - WARN_ON(existing->type == BTRFS_TREE_BLOCK_REF_KEY || - existing->type == BTRFS_SHARED_BLOCK_REF_KEY); + struct btrfs_delayed_ref_node *exist; + int mod; + int ret = 0; + + spin_lock(&href->lock); + /* Check whether we can merge the tail node with ref */ + if (list_empty(&href->ref_list)) + goto add_tail; + exist = list_entry(href->ref_list.prev, struct btrfs_delayed_ref_node, + list); + /* No need to compare bytenr nor is_head */ + if (exist->type != ref->type || exist->no_quota != ref->no_quota || + exist->seq != ref->seq) + goto add_tail; + + if ((exist->type == BTRFS_TREE_BLOCK_REF_KEY || + exist->type == BTRFS_SHARED_BLOCK_REF_KEY) && + comp_tree_refs(btrfs_delayed_node_to_tree_ref(exist), + btrfs_delayed_node_to_tree_ref(ref), + ref->type)) + goto add_tail; + if ((exist->type == BTRFS_EXTENT_DATA_REF_KEY || + exist->type == BTRFS_SHARED_DATA_REF_KEY) && + comp_data_refs(btrfs_delayed_node_to_data_ref(exist), + btrfs_delayed_node_to_data_ref(ref))) + goto add_tail; + + /* Now we are sure we can merge */ + ret = 1; + if (exist->action == ref->action) { + mod = ref->ref_mod; } else { - WARN_ON(existing->type == BTRFS_TREE_BLOCK_REF_KEY || - existing->type == BTRFS_SHARED_BLOCK_REF_KEY); - /* - * the action on the existing ref matches - * the action on the ref we're trying to add. - * Bump the ref_mod by one so the backref that - * is eventually added/removed has the correct - * reference count - */ - existing->ref_mod += update->ref_mod; + /* Need to change action */ + if (exist->ref_mod < ref->ref_mod) { + exist->action = ref->action; + mod = -exist->ref_mod; + exist->ref_mod = ref->ref_mod; + } else + mod = -ref->ref_mod; } + exist->ref_mod += mod; + + /* remove existing tail if its ref_mod is zero */ + if (exist->ref_mod == 0) + drop_delayed_ref(trans, root, href, exist); + spin_unlock(&href->lock); + return ret; + +add_tail: + list_add_tail(&ref->list, &href->ref_list); + atomic_inc(&root->num_entries); + trans->delayed_ref_updates++; + spin_unlock(&href->lock); + return ret; } /* @@ -568,12 +421,14 @@ update_existing_head_ref(struct btrfs_delayed_ref_root *delayed_refs, static noinline struct btrfs_delayed_ref_head * add_delayed_ref_head(struct btrfs_fs_info *fs_info, struct btrfs_trans_handle *trans, - struct btrfs_delayed_ref_node *ref, u64 bytenr, - u64 num_bytes, int action, int is_data) + struct btrfs_delayed_ref_node *ref, + struct btrfs_qgroup_extent_record *qrecord, + u64 bytenr, u64 num_bytes, int action, int is_data) { struct btrfs_delayed_ref_head *existing; struct btrfs_delayed_ref_head *head_ref = NULL; struct btrfs_delayed_ref_root *delayed_refs; + struct btrfs_qgroup_extent_record *qexisting; int count_mod = 1; int must_insert_reserved = 0; @@ -618,10 +473,22 @@ add_delayed_ref_head(struct btrfs_fs_info *fs_info, head_ref = btrfs_delayed_node_to_head(ref); head_ref->must_insert_reserved = must_insert_reserved; head_ref->is_data = is_data; - head_ref->ref_root = RB_ROOT; + INIT_LIST_HEAD(&head_ref->ref_list); head_ref->processing = 0; head_ref->total_ref_mod = count_mod; + /* Record qgroup extent info if provided */ + if (qrecord) { + qrecord->bytenr = bytenr; + qrecord->num_bytes = num_bytes; + qrecord->old_roots = NULL; + + qexisting = btrfs_qgroup_insert_dirty_extent(delayed_refs, + qrecord); + if (qexisting) + kfree(qrecord); + } + spin_lock_init(&head_ref->lock); mutex_init(&head_ref->mutex); @@ -659,10 +526,10 @@ add_delayed_tree_ref(struct btrfs_fs_info *fs_info, u64 num_bytes, u64 parent, u64 ref_root, int level, int action, int no_quota) { - struct btrfs_delayed_ref_node *existing; struct btrfs_delayed_tree_ref *full_ref; struct btrfs_delayed_ref_root *delayed_refs; u64 seq = 0; + int ret; if (action == BTRFS_ADD_DELAYED_EXTENT) action = BTRFS_ADD_DELAYED_REF; @@ -693,21 +560,14 @@ add_delayed_tree_ref(struct btrfs_fs_info *fs_info, trace_add_delayed_tree_ref(ref, full_ref, action); - spin_lock(&head_ref->lock); - existing = tree_insert(&head_ref->ref_root, &ref->rb_node); - if (existing) { - update_existing_ref(trans, delayed_refs, head_ref, existing, - ref); - /* - * we've updated the existing ref, free the newly - * allocated ref - */ + ret = add_delayed_ref_tail_merge(trans, delayed_refs, head_ref, ref); + + /* + * XXX: memory should be freed at the same level allocated. + * But bad practice is anywhere... Follow it now. Need cleanup. + */ + if (ret > 0) kmem_cache_free(btrfs_delayed_tree_ref_cachep, full_ref); - } else { - atomic_inc(&delayed_refs->num_entries); - trans->delayed_ref_updates++; - } - spin_unlock(&head_ref->lock); } /* @@ -721,10 +581,10 @@ add_delayed_data_ref(struct btrfs_fs_info *fs_info, u64 num_bytes, u64 parent, u64 ref_root, u64 owner, u64 offset, int action, int no_quota) { - struct btrfs_delayed_ref_node *existing; struct btrfs_delayed_data_ref *full_ref; struct btrfs_delayed_ref_root *delayed_refs; u64 seq = 0; + int ret; if (action == BTRFS_ADD_DELAYED_EXTENT) action = BTRFS_ADD_DELAYED_REF; @@ -758,21 +618,10 @@ add_delayed_data_ref(struct btrfs_fs_info *fs_info, trace_add_delayed_data_ref(ref, full_ref, action); - spin_lock(&head_ref->lock); - existing = tree_insert(&head_ref->ref_root, &ref->rb_node); - if (existing) { - update_existing_ref(trans, delayed_refs, head_ref, existing, - ref); - /* - * we've updated the existing ref, free the newly - * allocated ref - */ + ret = add_delayed_ref_tail_merge(trans, delayed_refs, head_ref, ref); + + if (ret > 0) kmem_cache_free(btrfs_delayed_data_ref_cachep, full_ref); - } else { - atomic_inc(&delayed_refs->num_entries); - trans->delayed_ref_updates++; - } - spin_unlock(&head_ref->lock); } /* @@ -790,6 +639,7 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, struct btrfs_delayed_tree_ref *ref; struct btrfs_delayed_ref_head *head_ref; struct btrfs_delayed_ref_root *delayed_refs; + struct btrfs_qgroup_extent_record *record = NULL; if (!is_fstree(ref_root) || !fs_info->quota_enabled) no_quota = 0; @@ -800,9 +650,13 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, return -ENOMEM; head_ref = kmem_cache_alloc(btrfs_delayed_ref_head_cachep, GFP_NOFS); - if (!head_ref) { - kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref); - return -ENOMEM; + if (!head_ref) + goto free_ref; + + if (fs_info->quota_enabled && is_fstree(ref_root)) { + record = kmalloc(sizeof(*record), GFP_NOFS); + if (!record) + goto free_head_ref; } head_ref->extent_op = extent_op; @@ -814,7 +668,7 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, * insert both the head node and the new ref without dropping * the spin lock */ - head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, + head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, record, bytenr, num_bytes, action, 0); add_delayed_tree_ref(fs_info, trans, head_ref, &ref->node, bytenr, @@ -823,6 +677,13 @@ int btrfs_add_delayed_tree_ref(struct btrfs_fs_info *fs_info, spin_unlock(&delayed_refs->lock); return 0; + +free_head_ref: + kmem_cache_free(btrfs_delayed_ref_head_cachep, head_ref); +free_ref: + kmem_cache_free(btrfs_delayed_tree_ref_cachep, ref); + + return -ENOMEM; } /* @@ -839,6 +700,7 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, struct btrfs_delayed_data_ref *ref; struct btrfs_delayed_ref_head *head_ref; struct btrfs_delayed_ref_root *delayed_refs; + struct btrfs_qgroup_extent_record *record = NULL; if (!is_fstree(ref_root) || !fs_info->quota_enabled) no_quota = 0; @@ -854,6 +716,16 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, return -ENOMEM; } + if (fs_info->quota_enabled && is_fstree(ref_root)) { + record = kmalloc(sizeof(*record), GFP_NOFS); + if (!record) { + kmem_cache_free(btrfs_delayed_data_ref_cachep, ref); + kmem_cache_free(btrfs_delayed_ref_head_cachep, + head_ref); + return -ENOMEM; + } + } + head_ref->extent_op = extent_op; delayed_refs = &trans->transaction->delayed_refs; @@ -863,7 +735,7 @@ int btrfs_add_delayed_data_ref(struct btrfs_fs_info *fs_info, * insert both the head node and the new ref without dropping * the spin lock */ - head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, + head_ref = add_delayed_ref_head(fs_info, trans, &head_ref->node, record, bytenr, num_bytes, action, 1); add_delayed_data_ref(fs_info, trans, head_ref, &ref->node, bytenr, @@ -891,9 +763,9 @@ int btrfs_add_delayed_extent_op(struct btrfs_fs_info *fs_info, delayed_refs = &trans->transaction->delayed_refs; spin_lock(&delayed_refs->lock); - add_delayed_ref_head(fs_info, trans, &head_ref->node, bytenr, - num_bytes, BTRFS_UPDATE_DELAYED_HEAD, - extent_op->is_data); + add_delayed_ref_head(fs_info, trans, &head_ref->node, NULL, bytenr, + num_bytes, BTRFS_UPDATE_DELAYED_HEAD, + extent_op->is_data); spin_unlock(&delayed_refs->lock); return 0; diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index 5eb0892..13fb5e6 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -24,9 +24,25 @@ #define BTRFS_ADD_DELAYED_EXTENT 3 /* record a full extent allocation */ #define BTRFS_UPDATE_DELAYED_HEAD 4 /* not changing ref count on head ref */ +/* + * XXX: Qu: I really hate the design that ref_head and tree/data ref shares the + * same ref_node structure. + * Ref_head is in a higher logic level than tree/data ref, and duplicated + * bytenr/num_bytes in ref_node is really a waste or memory, they should be + * referred from ref_head. + * This gets more disgusting after we use list to store tree/data ref in + * ref_head. Must clean this mess up later. + */ struct btrfs_delayed_ref_node { + /* + * ref_head use rb tree, stored in ref_root->href. + * indexed by bytenr + */ struct rb_node rb_node; + /*data/tree ref use list, stored in ref_head->ref_list. */ + struct list_head list; + /* the starting bytenr of the extent */ u64 bytenr; @@ -83,7 +99,7 @@ struct btrfs_delayed_ref_head { struct mutex mutex; spinlock_t lock; - struct rb_root ref_root; + struct list_head ref_list; struct rb_node href_node; @@ -132,6 +148,9 @@ struct btrfs_delayed_ref_root { /* head ref rbtree */ struct rb_root href_root; + /* dirty extent records */ + struct rb_root dirty_extent_root; + /* this spin lock protects the rbtree and the entries inside */ spinlock_t lock; @@ -156,6 +175,14 @@ struct btrfs_delayed_ref_root { int flushing; u64 run_delayed_start; + + /* + * To make qgroup to skip given root. + * This is for snapshot, as btrfs_qgroup_inherit() will manully + * modify counters for snapshot and its source, so we should skip + * the snapshot in new_root/old_roots or it will get calculated twice + */ + u64 qgroup_to_skip; }; extern struct kmem_cache *btrfs_delayed_ref_head_cachep; diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c index 0573848..862fbc2 100644 --- a/fs/btrfs/dev-replace.c +++ b/fs/btrfs/dev-replace.c @@ -376,6 +376,10 @@ int btrfs_dev_replace_start(struct btrfs_root *root, WARN_ON(!tgt_device); dev_replace->tgtdev = tgt_device; + ret = btrfs_kobj_add_device(tgt_device->fs_devices, tgt_device); + if (ret) + btrfs_error(root->fs_info, ret, "kobj add dev failed"); + printk_in_rcu(KERN_INFO "BTRFS: dev_replace from %s (devid %llu) to %s started\n", src_device->missing ? "<missing disk>" : @@ -583,8 +587,7 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info, mutex_unlock(&uuid_mutex); /* replace the sysfs entry */ - btrfs_kobj_rm_device(fs_info, src_device); - btrfs_kobj_add_device(fs_info, tgt_device); + btrfs_kobj_rm_device(fs_info->fs_devices, src_device); btrfs_rm_dev_replace_free_srcdev(fs_info, src_device); /* write back the superblocks */ diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 2ef9a4b..3f43bfea 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -1149,12 +1149,12 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr, buf = btrfs_find_create_tree_block(root, bytenr); if (!buf) - return NULL; + return ERR_PTR(-ENOMEM); ret = btree_read_extent_buffer_pages(root, buf, 0, parent_transid); if (ret) { free_extent_buffer(buf); - return NULL; + return ERR_PTR(ret); } return buf; @@ -1509,20 +1509,19 @@ static struct btrfs_root *btrfs_read_tree_root(struct btrfs_root *tree_root, generation = btrfs_root_generation(&root->root_item); root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item), generation); - if (!root->node) { - ret = -ENOMEM; + if (IS_ERR(root->node)) { + ret = PTR_ERR(root->node); goto find_fail; } else if (!btrfs_buffer_uptodate(root->node, generation, 0)) { ret = -EIO; - goto read_fail; + free_extent_buffer(root->node); + goto find_fail; } root->commit_root = btrfs_root_node(root); out: btrfs_free_path(path); return root; -read_fail: - free_extent_buffer(root->node); find_fail: kfree(root); alloc_fail: @@ -1745,7 +1744,7 @@ static void end_workqueue_fn(struct btrfs_work *work) bio->bi_private = end_io_wq->private; bio->bi_end_io = end_io_wq->end_io; kmem_cache_free(btrfs_end_io_wq_cache, end_io_wq); - bio_endio_nodec(bio, error); + bio_endio(bio, error); } static int cleaner_kthread(void *arg) @@ -2320,8 +2319,12 @@ static int btrfs_replay_log(struct btrfs_fs_info *fs_info, log_tree_root->node = read_tree_block(tree_root, bytenr, fs_info->generation + 1); - if (!log_tree_root->node || - !extent_buffer_uptodate(log_tree_root->node)) { + if (IS_ERR(log_tree_root->node)) { + printk(KERN_ERR "BTRFS: failed to read log tree\n"); + ret = PTR_ERR(log_tree_root->node); + kfree(log_tree_root); + return ret; + } else if (!extent_buffer_uptodate(log_tree_root->node)) { printk(KERN_ERR "BTRFS: failed to read log tree\n"); free_extent_buffer(log_tree_root->node); kfree(log_tree_root); @@ -2494,7 +2497,6 @@ int open_ctree(struct super_block *sb, seqlock_init(&fs_info->profiles_lock); init_rwsem(&fs_info->delayed_iput_sem); - init_completion(&fs_info->kobj_unregister); INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots); INIT_LIST_HEAD(&fs_info->space_info); INIT_LIST_HEAD(&fs_info->tree_mod_seq_list); @@ -2797,8 +2799,8 @@ int open_ctree(struct super_block *sb, chunk_root->node = read_tree_block(chunk_root, btrfs_super_chunk_root(disk_super), generation); - if (!chunk_root->node || - !test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) { + if (IS_ERR(chunk_root->node) || + !extent_buffer_uptodate(chunk_root->node)) { printk(KERN_ERR "BTRFS: failed to read chunk root on %s\n", sb->s_id); goto fail_tree_roots; @@ -2834,8 +2836,8 @@ retry_root_backup: tree_root->node = read_tree_block(tree_root, btrfs_super_root(disk_super), generation); - if (!tree_root->node || - !test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) { + if (IS_ERR(tree_root->node) || + !extent_buffer_uptodate(tree_root->node)) { printk(KERN_WARNING "BTRFS: failed to read tree root on %s\n", sb->s_id); @@ -2874,10 +2876,22 @@ retry_root_backup: btrfs_close_extra_devices(fs_devices, 1); + ret = btrfs_sysfs_add_fsid(fs_devices, NULL); + if (ret) { + pr_err("BTRFS: failed to init sysfs fsid interface: %d\n", ret); + goto fail_block_groups; + } + + ret = btrfs_sysfs_add_device(fs_devices); + if (ret) { + pr_err("BTRFS: failed to init sysfs device interface: %d\n", ret); + goto fail_fsdev_sysfs; + } + ret = btrfs_sysfs_add_one(fs_info); if (ret) { pr_err("BTRFS: failed to init sysfs interface: %d\n", ret); - goto fail_block_groups; + goto fail_fsdev_sysfs; } ret = btrfs_init_space_info(fs_info); @@ -3055,6 +3069,9 @@ fail_cleaner: fail_sysfs: btrfs_sysfs_remove_one(fs_info); +fail_fsdev_sysfs: + btrfs_sysfs_remove_fsid(fs_info->fs_devices); + fail_block_groups: btrfs_put_block_group_cache(fs_info); btrfs_free_block_groups(fs_info); @@ -3269,11 +3286,8 @@ static int write_dev_supers(struct btrfs_device *device, */ static void btrfs_end_empty_barrier(struct bio *bio, int err) { - if (err) { - if (err == -EOPNOTSUPP) - set_bit(BIO_EOPNOTSUPP, &bio->bi_flags); + if (err) clear_bit(BIO_UPTODATE, &bio->bi_flags); - } if (bio->bi_private) complete(bio->bi_private); bio_put(bio); @@ -3301,11 +3315,7 @@ static int write_dev_flush(struct btrfs_device *device, int wait) wait_for_completion(&device->flush_wait); - if (bio_flagged(bio, BIO_EOPNOTSUPP)) { - printk_in_rcu("BTRFS: disabling barriers on dev %s\n", - rcu_str_deref(device->name)); - device->nobarriers = 1; - } else if (!bio_flagged(bio, BIO_UPTODATE)) { + if (!bio_flagged(bio, BIO_UPTODATE)) { ret = -EIO; btrfs_dev_stat_inc_and_print(device, BTRFS_DEV_STAT_FLUSH_ERRS); @@ -3732,6 +3742,7 @@ void close_ctree(struct btrfs_root *root) } btrfs_sysfs_remove_one(fs_info); + btrfs_sysfs_remove_fsid(fs_info->fs_devices); btrfs_free_fs_roots(fs_info); @@ -4060,6 +4071,7 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, while ((node = rb_first(&delayed_refs->href_root)) != NULL) { struct btrfs_delayed_ref_head *head; + struct btrfs_delayed_ref_node *tmp; bool pin_bytes = false; head = rb_entry(node, struct btrfs_delayed_ref_head, @@ -4075,11 +4087,10 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, continue; } spin_lock(&head->lock); - while ((node = rb_first(&head->ref_root)) != NULL) { - ref = rb_entry(node, struct btrfs_delayed_ref_node, - rb_node); + list_for_each_entry_safe_reverse(ref, tmp, &head->ref_list, + list) { ref->in_tree = 0; - rb_erase(&ref->rb_node, &head->ref_root); + list_del(&ref->list); atomic_dec(&delayed_refs->num_entries); btrfs_put_delayed_ref(ref); } diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 0ec8e22..38b76cc 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -79,11 +79,10 @@ static int update_block_group(struct btrfs_trans_handle *trans, u64 num_bytes, int alloc); static int __btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, - u64 bytenr, u64 num_bytes, u64 parent, + struct btrfs_delayed_ref_node *node, u64 parent, u64 root_objectid, u64 owner_objectid, u64 owner_offset, int refs_to_drop, - struct btrfs_delayed_extent_op *extra_op, - int no_quota); + struct btrfs_delayed_extent_op *extra_op); static void __run_delayed_extent_op(struct btrfs_delayed_extent_op *extent_op, struct extent_buffer *leaf, struct btrfs_extent_item *ei); @@ -1967,10 +1966,9 @@ int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - u64 bytenr, u64 num_bytes, + struct btrfs_delayed_ref_node *node, u64 parent, u64 root_objectid, u64 owner, u64 offset, int refs_to_add, - int no_quota, struct btrfs_delayed_extent_op *extent_op) { struct btrfs_fs_info *fs_info = root->fs_info; @@ -1978,9 +1976,11 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, struct extent_buffer *leaf; struct btrfs_extent_item *item; struct btrfs_key key; + u64 bytenr = node->bytenr; + u64 num_bytes = node->num_bytes; u64 refs; int ret; - enum btrfs_qgroup_operation_type type = BTRFS_QGROUP_OPER_ADD_EXCL; + int no_quota = node->no_quota; path = btrfs_alloc_path(); if (!path) @@ -1996,26 +1996,8 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, bytenr, num_bytes, parent, root_objectid, owner, offset, refs_to_add, extent_op); - if ((ret < 0 && ret != -EAGAIN) || (!ret && no_quota)) + if ((ret < 0 && ret != -EAGAIN) || !ret) goto out; - /* - * Ok we were able to insert an inline extent and it appears to be a new - * reference, deal with the qgroup accounting. - */ - if (!ret && !no_quota) { - ASSERT(root->fs_info->quota_enabled); - leaf = path->nodes[0]; - btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - item = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_extent_item); - if (btrfs_extent_refs(leaf, item) > (u64)refs_to_add) - type = BTRFS_QGROUP_OPER_ADD_SHARED; - btrfs_release_path(path); - - ret = btrfs_qgroup_record_ref(trans, fs_info, root_objectid, - bytenr, num_bytes, type, 0); - goto out; - } /* * Ok we had -EAGAIN which means we didn't have space to insert and @@ -2026,8 +2008,6 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item); refs = btrfs_extent_refs(leaf, item); - if (refs) - type = BTRFS_QGROUP_OPER_ADD_SHARED; btrfs_set_extent_refs(leaf, item, refs + refs_to_add); if (extent_op) __run_delayed_extent_op(extent_op, leaf, item); @@ -2035,13 +2015,6 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(leaf); btrfs_release_path(path); - if (!no_quota) { - ret = btrfs_qgroup_record_ref(trans, fs_info, root_objectid, - bytenr, num_bytes, type, 0); - if (ret) - goto out; - } - path->reada = 1; path->leave_spinning = 1; /* now insert the actual backref */ @@ -2087,17 +2060,15 @@ static int run_delayed_data_ref(struct btrfs_trans_handle *trans, ref->objectid, ref->offset, &ins, node->ref_mod); } else if (node->action == BTRFS_ADD_DELAYED_REF) { - ret = __btrfs_inc_extent_ref(trans, root, node->bytenr, - node->num_bytes, parent, + ret = __btrfs_inc_extent_ref(trans, root, node, parent, ref_root, ref->objectid, ref->offset, node->ref_mod, - node->no_quota, extent_op); + extent_op); } else if (node->action == BTRFS_DROP_DELAYED_REF) { - ret = __btrfs_free_extent(trans, root, node->bytenr, - node->num_bytes, parent, + ret = __btrfs_free_extent(trans, root, node, parent, ref_root, ref->objectid, ref->offset, node->ref_mod, - extent_op, node->no_quota); + extent_op); } else { BUG(); } @@ -2255,15 +2226,14 @@ static int run_delayed_tree_ref(struct btrfs_trans_handle *trans, ref->level, &ins, node->no_quota); } else if (node->action == BTRFS_ADD_DELAYED_REF) { - ret = __btrfs_inc_extent_ref(trans, root, node->bytenr, - node->num_bytes, parent, ref_root, - ref->level, 0, 1, node->no_quota, + ret = __btrfs_inc_extent_ref(trans, root, node, + parent, ref_root, + ref->level, 0, 1, extent_op); } else if (node->action == BTRFS_DROP_DELAYED_REF) { - ret = __btrfs_free_extent(trans, root, node->bytenr, - node->num_bytes, parent, ref_root, - ref->level, 0, 1, extent_op, - node->no_quota); + ret = __btrfs_free_extent(trans, root, node, + parent, ref_root, + ref->level, 0, 1, extent_op); } else { BUG(); } @@ -2323,28 +2293,14 @@ static int run_one_delayed_ref(struct btrfs_trans_handle *trans, return ret; } -static noinline struct btrfs_delayed_ref_node * +static inline struct btrfs_delayed_ref_node * select_delayed_ref(struct btrfs_delayed_ref_head *head) { - struct rb_node *node; - struct btrfs_delayed_ref_node *ref, *last = NULL;; + if (list_empty(&head->ref_list)) + return NULL; - /* - * select delayed ref of type BTRFS_ADD_DELAYED_REF first. - * this prevents ref count from going down to zero when - * there still are pending delayed ref. - */ - node = rb_first(&head->ref_root); - while (node) { - ref = rb_entry(node, struct btrfs_delayed_ref_node, - rb_node); - if (ref->action == BTRFS_ADD_DELAYED_REF) - return ref; - else if (last == NULL) - last = ref; - node = rb_next(node); - } - return last; + return list_entry(head->ref_list.next, struct btrfs_delayed_ref_node, + list); } /* @@ -2396,16 +2352,7 @@ static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, } } - /* - * We need to try and merge add/drops of the same ref since we - * can run into issues with relocate dropping the implicit ref - * and then it being added back again before the drop can - * finish. If we merged anything we need to re-loop so we can - * get a good ref. - */ spin_lock(&locked_ref->lock); - btrfs_merge_delayed_refs(trans, fs_info, delayed_refs, - locked_ref); /* * locked_ref is the head node, so we have to go one @@ -2482,7 +2429,7 @@ static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, spin_unlock(&locked_ref->lock); spin_lock(&delayed_refs->lock); spin_lock(&locked_ref->lock); - if (rb_first(&locked_ref->ref_root) || + if (!list_empty(&locked_ref->ref_list) || locked_ref->extent_op) { spin_unlock(&locked_ref->lock); spin_unlock(&delayed_refs->lock); @@ -2496,7 +2443,7 @@ static noinline int __btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, } else { actual_count++; ref->in_tree = 0; - rb_erase(&ref->rb_node, &locked_ref->ref_root); + list_del(&ref->list); } atomic_dec(&delayed_refs->num_entries); @@ -2864,9 +2811,6 @@ again: goto again; } out: - ret = btrfs_delayed_qgroup_accounting(trans, root->fs_info); - if (ret) - return ret; assert_qgroups_uptodate(trans); return 0; } @@ -2905,7 +2849,6 @@ static noinline int check_delayed_ref(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_node *ref; struct btrfs_delayed_data_ref *data_ref; struct btrfs_delayed_ref_root *delayed_refs; - struct rb_node *node; int ret = 0; delayed_refs = &trans->transaction->delayed_refs; @@ -2934,11 +2877,7 @@ static noinline int check_delayed_ref(struct btrfs_trans_handle *trans, spin_unlock(&delayed_refs->lock); spin_lock(&head->lock); - node = rb_first(&head->ref_root); - while (node) { - ref = rb_entry(node, struct btrfs_delayed_ref_node, rb_node); - node = rb_next(node); - + list_for_each_entry(ref, &head->ref_list, list) { /* If it's a shared ref we know a cross reference exists */ if (ref->type != BTRFS_EXTENT_DATA_REF_KEY) { ret = 1; @@ -3180,8 +3119,6 @@ static int write_one_cache_group(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(leaf); fail: btrfs_release_path(path); - if (ret) - btrfs_abort_transaction(trans, root, ret); return ret; } @@ -3487,8 +3424,30 @@ again: ret = 0; } } - if (!ret) + if (!ret) { ret = write_one_cache_group(trans, root, path, cache); + /* + * Our block group might still be attached to the list + * of new block groups in the transaction handle of some + * other task (struct btrfs_trans_handle->new_bgs). This + * means its block group item isn't yet in the extent + * tree. If this happens ignore the error, as we will + * try again later in the critical section of the + * transaction commit. + */ + if (ret == -ENOENT) { + ret = 0; + spin_lock(&cur_trans->dirty_bgs_lock); + if (list_empty(&cache->dirty_list)) { + list_add_tail(&cache->dirty_list, + &cur_trans->dirty_bgs); + btrfs_get_block_group(cache); + } + spin_unlock(&cur_trans->dirty_bgs_lock); + } else if (ret) { + btrfs_abort_transaction(trans, root, ret); + } + } /* if its not on the io list, we need to put the block group */ if (should_put) @@ -3597,8 +3556,11 @@ int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, ret = 0; } } - if (!ret) + if (!ret) { ret = write_one_cache_group(trans, root, path, cache); + if (ret) + btrfs_abort_transaction(trans, root, ret); + } /* if its not on the io list, we need to put the block group */ if (should_put) @@ -3670,7 +3632,8 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, found->disk_total += total_bytes * factor; found->bytes_used += bytes_used; found->disk_used += bytes_used * factor; - found->full = 0; + if (total_bytes > 0) + found->full = 0; spin_unlock(&found->lock); *space_info = found; return 0; @@ -3698,7 +3661,10 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, found->bytes_reserved = 0; found->bytes_readonly = 0; found->bytes_may_use = 0; - found->full = 0; + if (total_bytes > 0) + found->full = 0; + else + found->full = 1; found->force_alloc = CHUNK_ALLOC_NO_FORCE; found->chunk_alloc = 0; found->flush = 0; @@ -3952,6 +3918,9 @@ commit_trans: !atomic_read(&root->fs_info->open_ioctl_trans)) { need_commit--; + if (need_commit > 0) + btrfs_wait_ordered_roots(fs_info, -1); + trans = btrfs_join_transaction(root); if (IS_ERR(trans)) return PTR_ERR(trans); @@ -4065,7 +4034,7 @@ static int should_alloc_chunk(struct btrfs_root *root, return 1; } -static u64 get_system_chunk_thresh(struct btrfs_root *root, u64 type) +static u64 get_profile_num_devs(struct btrfs_root *root, u64 type) { u64 num_dev; @@ -4079,24 +4048,43 @@ static u64 get_system_chunk_thresh(struct btrfs_root *root, u64 type) else num_dev = 1; /* DUP or single */ - /* metadata for updaing devices and chunk tree */ - return btrfs_calc_trans_metadata_size(root, num_dev + 1); + return num_dev; } -static void check_system_chunk(struct btrfs_trans_handle *trans, - struct btrfs_root *root, u64 type) +/* + * If @is_allocation is true, reserve space in the system space info necessary + * for allocating a chunk, otherwise if it's false, reserve space necessary for + * removing a chunk. + */ +void check_system_chunk(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + u64 type) { struct btrfs_space_info *info; u64 left; u64 thresh; + int ret = 0; + u64 num_devs; + + /* + * Needed because we can end up allocating a system chunk and for an + * atomic and race free space reservation in the chunk block reserve. + */ + ASSERT(mutex_is_locked(&root->fs_info->chunk_mutex)); info = __find_space_info(root->fs_info, BTRFS_BLOCK_GROUP_SYSTEM); spin_lock(&info->lock); left = info->total_bytes - info->bytes_used - info->bytes_pinned - - info->bytes_reserved - info->bytes_readonly; + info->bytes_reserved - info->bytes_readonly - + info->bytes_may_use; spin_unlock(&info->lock); - thresh = get_system_chunk_thresh(root, type); + num_devs = get_profile_num_devs(root, type); + + /* num_devs device items to update and 1 chunk item to add or remove */ + thresh = btrfs_calc_trunc_metadata_size(root, num_devs) + + btrfs_calc_trans_metadata_size(root, 1); + if (left < thresh && btrfs_test_opt(root, ENOSPC_DEBUG)) { btrfs_info(root->fs_info, "left=%llu, need=%llu, flags=%llu", left, thresh, type); @@ -4107,7 +4095,21 @@ static void check_system_chunk(struct btrfs_trans_handle *trans, u64 flags; flags = btrfs_get_alloc_profile(root->fs_info->chunk_root, 0); - btrfs_alloc_chunk(trans, root, flags); + /* + * Ignore failure to create system chunk. We might end up not + * needing it, as we might not need to COW all nodes/leafs from + * the paths we visit in the chunk tree (they were already COWed + * or created in the current transaction for example). + */ + ret = btrfs_alloc_chunk(trans, root, flags); + } + + if (!ret) { + ret = btrfs_block_rsv_add(root->fs_info->chunk_root, + &root->fs_info->chunk_block_rsv, + thresh, BTRFS_RESERVE_NO_FLUSH); + if (!ret) + trans->chunk_bytes_reserved += thresh; } } @@ -5165,6 +5167,24 @@ void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans, trans->bytes_reserved = 0; } +/* + * To be called after all the new block groups attached to the transaction + * handle have been created (btrfs_create_pending_block_groups()). + */ +void btrfs_trans_release_chunk_metadata(struct btrfs_trans_handle *trans) +{ + struct btrfs_fs_info *fs_info = trans->root->fs_info; + + if (!trans->chunk_bytes_reserved) + return; + + WARN_ON_ONCE(!list_empty(&trans->new_bgs)); + + block_rsv_release_bytes(fs_info, &fs_info->chunk_block_rsv, NULL, + trans->chunk_bytes_reserved); + trans->chunk_bytes_reserved = 0; +} + /* Can only return 0 or -ENOSPC */ int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans, struct inode *inode) @@ -6069,11 +6089,10 @@ static void add_pinned_bytes(struct btrfs_fs_info *fs_info, u64 num_bytes, static int __btrfs_free_extent(struct btrfs_trans_handle *trans, struct btrfs_root *root, - u64 bytenr, u64 num_bytes, u64 parent, + struct btrfs_delayed_ref_node *node, u64 parent, u64 root_objectid, u64 owner_objectid, u64 owner_offset, int refs_to_drop, - struct btrfs_delayed_extent_op *extent_op, - int no_quota) + struct btrfs_delayed_extent_op *extent_op) { struct btrfs_key key; struct btrfs_path *path; @@ -6087,10 +6106,12 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, int extent_slot = 0; int found_extent = 0; int num_to_del = 1; + int no_quota = node->no_quota; u32 item_size; u64 refs; + u64 bytenr = node->bytenr; + u64 num_bytes = node->num_bytes; int last_ref = 0; - enum btrfs_qgroup_operation_type type = BTRFS_QGROUP_OPER_SUB_EXCL; bool skinny_metadata = btrfs_fs_incompat(root->fs_info, SKINNY_METADATA); @@ -6271,7 +6292,6 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, refs -= refs_to_drop; if (refs > 0) { - type = BTRFS_QGROUP_OPER_SUB_SHARED; if (extent_op) __run_delayed_extent_op(extent_op, leaf, ei); /* @@ -6333,18 +6353,6 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, } btrfs_release_path(path); - /* Deal with the quota accounting */ - if (!ret && last_ref && !no_quota) { - int mod_seq = 0; - - if (owner_objectid >= BTRFS_FIRST_FREE_OBJECTID && - type == BTRFS_QGROUP_OPER_SUB_SHARED) - mod_seq = 1; - - ret = btrfs_qgroup_record_ref(trans, info, root_objectid, - bytenr, num_bytes, type, - mod_seq); - } out: btrfs_free_path(path); return ret; @@ -6370,7 +6378,7 @@ static noinline int check_ref_cleanup(struct btrfs_trans_handle *trans, goto out_delayed_unlock; spin_lock(&head->lock); - if (rb_first(&head->ref_root)) + if (!list_empty(&head->ref_list)) goto out; if (head->extent_op) { @@ -7280,13 +7288,6 @@ static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(path->nodes[0]); btrfs_free_path(path); - /* Always set parent to 0 here since its exclusive anyway. */ - ret = btrfs_qgroup_record_ref(trans, fs_info, root_objectid, - ins->objectid, ins->offset, - BTRFS_QGROUP_OPER_ADD_EXCL, 0); - if (ret) - return ret; - ret = update_block_group(trans, root, ins->objectid, ins->offset, 1); if (ret) { /* -ENOENT, logic error */ btrfs_err(fs_info, "update block group failed for %llu %llu", @@ -7368,14 +7369,6 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(leaf); btrfs_free_path(path); - if (!no_quota) { - ret = btrfs_qgroup_record_ref(trans, fs_info, root_objectid, - ins->objectid, num_bytes, - BTRFS_QGROUP_OPER_ADD_EXCL, 0); - if (ret) - return ret; - } - ret = update_block_group(trans, root, ins->objectid, root->nodesize, 1); if (ret) { /* -ENOENT, logic error */ @@ -7732,12 +7725,18 @@ reada: wc->reada_slot = slot; } +/* + * TODO: Modify related function to add related node/leaf to dirty_extent_root, + * for later qgroup accounting. + * + * Current, this function does nothing. + */ static int account_leaf_items(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *eb) { int nr = btrfs_header_nritems(eb); - int i, extent_type, ret; + int i, extent_type; struct btrfs_key key; struct btrfs_file_extent_item *fi; u64 bytenr, num_bytes; @@ -7760,13 +7759,6 @@ static int account_leaf_items(struct btrfs_trans_handle *trans, continue; num_bytes = btrfs_file_extent_disk_num_bytes(eb, fi); - - ret = btrfs_qgroup_record_ref(trans, root->fs_info, - root->objectid, - bytenr, num_bytes, - BTRFS_QGROUP_OPER_SUB_SUBTREE, 0); - if (ret) - return ret; } return 0; } @@ -7835,6 +7827,8 @@ static int adjust_slots_upwards(struct btrfs_root *root, /* * root_eb is the subtree root and is locked before this function is called. + * TODO: Modify this function to mark all (including complete shared node) + * to dirty_extent_root to allow it get accounted in qgroup. */ static int account_shared_subtree(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -7897,7 +7891,11 @@ walk_down: child_gen = btrfs_node_ptr_generation(eb, parent_slot); eb = read_tree_block(root, child_bytenr, child_gen); - if (!eb || !extent_buffer_uptodate(eb)) { + if (IS_ERR(eb)) { + ret = PTR_ERR(eb); + goto out; + } else if (!extent_buffer_uptodate(eb)) { + free_extent_buffer(eb); ret = -EIO; goto out; } @@ -7908,16 +7906,6 @@ walk_down: btrfs_tree_read_lock(eb); btrfs_set_lock_blocking_rw(eb, BTRFS_READ_LOCK); path->locks[level] = BTRFS_READ_LOCK_BLOCKING; - - ret = btrfs_qgroup_record_ref(trans, root->fs_info, - root->objectid, - child_bytenr, - root->nodesize, - BTRFS_QGROUP_OPER_SUB_SUBTREE, - 0); - if (ret) - goto out; - } if (level == 0) { @@ -8128,7 +8116,9 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans, if (reada && level == 1) reada_walk_down(trans, root, wc, path); next = read_tree_block(root, bytenr, generation); - if (!next || !extent_buffer_uptodate(next)) { + if (IS_ERR(next)) { + return PTR_ERR(next); + } else if (!extent_buffer_uptodate(next)) { free_extent_buffer(next); return -EIO; } @@ -8510,24 +8500,6 @@ int btrfs_drop_snapshot(struct btrfs_root *root, goto out_end_trans; } - /* - * Qgroup update accounting is run from - * delayed ref handling. This usually works - * out because delayed refs are normally the - * only way qgroup updates are added. However, - * we may have added updates during our tree - * walk so run qgroups here to make sure we - * don't lose any updates. - */ - ret = btrfs_delayed_qgroup_accounting(trans, - root->fs_info); - if (ret) - printk_ratelimited(KERN_ERR "BTRFS: Failure %d " - "running qgroup updates " - "during snapshot delete. " - "Quota is out of sync, " - "rescan required.\n", ret); - btrfs_end_transaction_throttle(trans, tree_root); if (!for_reloc && btrfs_need_cleaner_sleep(root)) { pr_debug("BTRFS: drop snapshot early exit\n"); @@ -8581,14 +8553,6 @@ int btrfs_drop_snapshot(struct btrfs_root *root, } root_dropped = true; out_end_trans: - ret = btrfs_delayed_qgroup_accounting(trans, tree_root->fs_info); - if (ret) - printk_ratelimited(KERN_ERR "BTRFS: Failure %d " - "running qgroup updates " - "during snapshot delete. " - "Quota is out of sync, " - "rescan required.\n", ret); - btrfs_end_transaction_throttle(trans, tree_root); out_free: kfree(wc); @@ -8806,6 +8770,24 @@ again: goto again; } + /* + * if we are changing raid levels, try to allocate a corresponding + * block group with the new raid level. + */ + alloc_flags = update_block_group_flags(root, cache->flags); + if (alloc_flags != cache->flags) { + ret = do_chunk_alloc(trans, root, alloc_flags, + CHUNK_ALLOC_FORCE); + /* + * ENOSPC is allowed here, we may have enough space + * already allocated at the new raid level to + * carry on + */ + if (ret == -ENOSPC) + ret = 0; + if (ret < 0) + goto out; + } ret = set_block_group_ro(cache, 0); if (!ret) @@ -8819,7 +8801,9 @@ again: out: if (cache->flags & BTRFS_BLOCK_GROUP_SYSTEM) { alloc_flags = update_block_group_flags(root, cache->flags); + lock_chunks(root->fs_info->chunk_root); check_system_chunk(trans, root, alloc_flags); + unlock_chunks(root->fs_info->chunk_root); } mutex_unlock(&root->fs_info->ro_block_group_mutex); @@ -9519,6 +9503,19 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, free_excluded_extents(root, cache); + /* + * Call to ensure the corresponding space_info object is created and + * assigned to our block group, but don't update its counters just yet. + * We want our bg to be added to the rbtree with its ->space_info set. + */ + ret = update_space_info(root->fs_info, cache->flags, 0, 0, + &cache->space_info); + if (ret) { + btrfs_remove_free_space_cache(cache); + btrfs_put_block_group(cache); + return ret; + } + ret = btrfs_add_block_group_cache(root->fs_info, cache); if (ret) { btrfs_remove_free_space_cache(cache); @@ -9526,6 +9523,10 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, return ret; } + /* + * Now that our block group has its ->space_info set and is inserted in + * the rbtree, update the space info's counters. + */ ret = update_space_info(root->fs_info, cache->flags, size, bytes_used, &cache->space_info); if (ret) { diff --git a/fs/btrfs/extent-tree.h b/fs/btrfs/extent-tree.h new file mode 100644 index 0000000..e69de29 --- /dev/null +++ b/fs/btrfs/extent-tree.h diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 43af5a6..02d0581 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -1277,7 +1277,12 @@ int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, unsigned bits, gfp_t mask) { - return clear_extent_bit(tree, start, end, bits, 0, 0, NULL, mask); + int wake = 0; + + if (bits & EXTENT_LOCKED) + wake = 1; + + return clear_extent_bit(tree, start, end, bits, wake, 0, NULL, mask); } int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end, @@ -2767,8 +2772,6 @@ static int __must_check submit_one_bio(int rw, struct bio *bio, else btrfsic_submit_bio(rw, bio); - if (bio_flagged(bio, BIO_EOPNOTSUPP)) - ret = -EOPNOTSUPP; bio_put(bio); return ret; } @@ -4492,6 +4495,8 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, } if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) flags |= FIEMAP_EXTENT_ENCODED; + if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) + flags |= FIEMAP_EXTENT_UNWRITTEN; free_extent_map(em); em = NULL; @@ -4772,6 +4777,25 @@ struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info, start >> PAGE_CACHE_SHIFT); if (eb && atomic_inc_not_zero(&eb->refs)) { rcu_read_unlock(); + /* + * Lock our eb's refs_lock to avoid races with + * free_extent_buffer. When we get our eb it might be flagged + * with EXTENT_BUFFER_STALE and another task running + * free_extent_buffer might have seen that flag set, + * eb->refs == 2, that the buffer isn't under IO (dirty and + * writeback flags not set) and it's still in the tree (flag + * EXTENT_BUFFER_TREE_REF set), therefore being in the process + * of decrementing the extent buffer's reference count twice. + * So here we could race and increment the eb's reference count, + * clear its stale flag, mark it as dirty and drop our reference + * before the other task finishes executing free_extent_buffer, + * which would later result in an attempt to free an extent + * buffer that is dirty. + */ + if (test_bit(EXTENT_BUFFER_STALE, &eb->bflags)) { + spin_lock(&eb->refs_lock); + spin_unlock(&eb->refs_lock); + } mark_extent_buffer_accessed(eb, NULL); return eb; } diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 86f9728..b823fac9 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1868,6 +1868,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) struct btrfs_log_ctx ctx; int ret = 0; bool full_sync = 0; + const u64 len = end - start + 1; trace_btrfs_sync_file(file, datasync); @@ -1896,7 +1897,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) * all extents are persisted and the respective file extent * items are in the fs/subvol btree. */ - ret = btrfs_wait_ordered_range(inode, start, end - start + 1); + ret = btrfs_wait_ordered_range(inode, start, len); } else { /* * Start any new ordered operations before starting to log the @@ -1968,8 +1969,10 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) */ smp_mb(); if (btrfs_inode_in_log(inode, root->fs_info->generation) || - (full_sync && BTRFS_I(inode)->last_trans <= - root->fs_info->last_trans_committed)) { + (BTRFS_I(inode)->last_trans <= + root->fs_info->last_trans_committed && + (full_sync || + !btrfs_have_ordered_extents_in_range(inode, start, len)))) { /* * We'v had everything committed since the last time we were * modified so clear this flag in case it was set for whatever diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 5e020d7..fb5a6b1 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -231,6 +231,7 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root, { int ret = 0; struct btrfs_path *path = btrfs_alloc_path(); + bool locked = false; if (!path) { ret = -ENOMEM; @@ -238,6 +239,7 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root, } if (block_group) { + locked = true; mutex_lock(&trans->transaction->cache_write_mutex); if (!list_empty(&block_group->io_list)) { list_del_init(&block_group->io_list); @@ -269,18 +271,14 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root, */ ret = btrfs_truncate_inode_items(trans, root, inode, 0, BTRFS_EXTENT_DATA_KEY); - if (ret) { - mutex_unlock(&trans->transaction->cache_write_mutex); - btrfs_abort_transaction(trans, root, ret); - return ret; - } + if (ret) + goto fail; ret = btrfs_update_inode(trans, root, inode); - if (block_group) - mutex_unlock(&trans->transaction->cache_write_mutex); - fail: + if (locked) + mutex_unlock(&trans->transaction->cache_write_mutex); if (ret) btrfs_abort_transaction(trans, root, ret); @@ -3466,6 +3464,7 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root, struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; int ret; struct btrfs_io_ctl io_ctl; + bool release_metadata = true; if (!btrfs_test_opt(root, INODE_MAP_CACHE)) return 0; @@ -3473,11 +3472,20 @@ int btrfs_write_out_ino_cache(struct btrfs_root *root, memset(&io_ctl, 0, sizeof(io_ctl)); ret = __btrfs_write_out_cache(root, inode, ctl, NULL, &io_ctl, trans, path, 0); - if (!ret) + if (!ret) { + /* + * At this point writepages() didn't error out, so our metadata + * reservation is released when the writeback finishes, at + * inode.c:btrfs_finish_ordered_io(), regardless of it finishing + * with or without an error. + */ + release_metadata = false; ret = btrfs_wait_cache_io(root, trans, NULL, &io_ctl, path, 0); + } if (ret) { - btrfs_delalloc_release_metadata(inode, inode->i_size); + if (release_metadata) + btrfs_delalloc_release_metadata(inode, inode->i_size); #ifdef DEBUG btrfs_err(root->fs_info, "failed to write free ino cache for root %llu", diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 8bb01367..855935f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -4986,24 +4986,40 @@ static void evict_inode_truncate_pages(struct inode *inode) } write_unlock(&map_tree->lock); + /* + * Keep looping until we have no more ranges in the io tree. + * We can have ongoing bios started by readpages (called from readahead) + * that didn't get their end io callbacks called yet or they are still + * in progress ((extent_io.c:end_bio_extent_readpage()). This means some + * ranges can still be locked and eviction started because before + * submitting those bios, which are executed by a separate task (work + * queue kthread), inode references (inode->i_count) were not taken + * (which would be dropped in the end io callback of each bio). + * Therefore here we effectively end up waiting for those bios and + * anyone else holding locked ranges without having bumped the inode's + * reference count - if we don't do it, when they access the inode's + * io_tree to unlock a range it may be too late, leading to an + * use-after-free issue. + */ spin_lock(&io_tree->lock); while (!RB_EMPTY_ROOT(&io_tree->state)) { struct extent_state *state; struct extent_state *cached_state = NULL; + u64 start; + u64 end; node = rb_first(&io_tree->state); state = rb_entry(node, struct extent_state, rb_node); - atomic_inc(&state->refs); + start = state->start; + end = state->end; spin_unlock(&io_tree->lock); - lock_extent_bits(io_tree, state->start, state->end, - 0, &cached_state); - clear_extent_bit(io_tree, state->start, state->end, + lock_extent_bits(io_tree, start, end, 0, &cached_state); + clear_extent_bit(io_tree, start, end, EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING | EXTENT_DEFRAG, 1, 1, &cached_state, GFP_NOFS); - free_extent_state(state); cond_resched(); spin_lock(&io_tree->lock); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 1c22c65..c86b835 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -553,8 +553,8 @@ static noinline int create_subvol(struct inode *dir, key.offset = (u64)-1; new_root = btrfs_read_fs_root_no_name(root->fs_info, &key); if (IS_ERR(new_root)) { - btrfs_abort_transaction(trans, root, PTR_ERR(new_root)); ret = PTR_ERR(new_root); + btrfs_abort_transaction(trans, root, ret); goto fail; } @@ -1318,7 +1318,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, i = range->start >> PAGE_CACHE_SHIFT; } if (!max_to_defrag) - max_to_defrag = last_index + 1; + max_to_defrag = last_index - i + 1; /* * make writeback starts from i, so the defrag range can be @@ -1368,7 +1368,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, ra_index = max(i, ra_index); btrfs_force_ra(inode->i_mapping, ra, file, ra_index, cluster); - ra_index += max_cluster; + ra_index += cluster; } mutex_lock(&inode->i_mutex); @@ -2271,10 +2271,7 @@ static noinline int btrfs_ioctl_ino_lookup(struct file *file, { struct btrfs_ioctl_ino_lookup_args *args; struct inode *inode; - int ret; - - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; + int ret = 0; args = memdup_user(argp, sizeof(*args)); if (IS_ERR(args)) @@ -2282,13 +2279,28 @@ static noinline int btrfs_ioctl_ino_lookup(struct file *file, inode = file_inode(file); + /* + * Unprivileged query to obtain the containing subvolume root id. The + * path is reset so it's consistent with btrfs_search_path_in_tree. + */ if (args->treeid == 0) args->treeid = BTRFS_I(inode)->root->root_key.objectid; + if (args->objectid == BTRFS_FIRST_FREE_OBJECTID) { + args->name[0] = 0; + goto out; + } + + if (!capable(CAP_SYS_ADMIN)) { + ret = -EPERM; + goto out; + } + ret = btrfs_search_path_in_tree(BTRFS_I(inode)->root->fs_info, args->treeid, args->objectid, args->name); +out: if (ret == 0 && copy_to_user(argp, args, sizeof(*args))) ret = -EFAULT; @@ -2413,8 +2425,6 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file, goto out_unlock_inode; } - d_invalidate(dentry); - down_write(&root->fs_info->subvol_sem); err = may_destroy_subvol(dest); @@ -2508,7 +2518,7 @@ out_up_write: out_unlock_inode: mutex_unlock(&inode->i_mutex); if (!err) { - shrink_dcache_sb(root->fs_info->sb); + d_invalidate(dentry); btrfs_invalidate_inodes(dest); d_delete(dentry); ASSERT(dest->send_in_progress == 0); @@ -2879,12 +2889,19 @@ static int btrfs_cmp_data(struct inode *src, u64 loff, struct inode *dst, return ret; } -static int extent_same_check_offsets(struct inode *inode, u64 off, u64 len) +static int extent_same_check_offsets(struct inode *inode, u64 off, u64 *plen, + u64 olen) { + u64 len = *plen; u64 bs = BTRFS_I(inode)->root->fs_info->sb->s_blocksize; - if (off + len > inode->i_size || off + len < off) + if (off + olen > inode->i_size || off + olen < off) return -EINVAL; + + /* if we extend to eof, continue to block boundary */ + if (off + len == inode->i_size) + *plen = len = ALIGN(inode->i_size, bs) - off; + /* Check that we are block aligned - btrfs_clone() requires this */ if (!IS_ALIGNED(off, bs) || !IS_ALIGNED(off + len, bs)) return -EINVAL; @@ -2892,10 +2909,11 @@ static int extent_same_check_offsets(struct inode *inode, u64 off, u64 len) return 0; } -static int btrfs_extent_same(struct inode *src, u64 loff, u64 len, +static int btrfs_extent_same(struct inode *src, u64 loff, u64 olen, struct inode *dst, u64 dst_loff) { int ret; + u64 len = olen; /* * btrfs_clone() can't handle extents in the same file @@ -2910,11 +2928,11 @@ static int btrfs_extent_same(struct inode *src, u64 loff, u64 len, btrfs_double_lock(src, loff, dst, dst_loff, len); - ret = extent_same_check_offsets(src, loff, len); + ret = extent_same_check_offsets(src, loff, &len, olen); if (ret) goto out_unlock; - ret = extent_same_check_offsets(dst, dst_loff, len); + ret = extent_same_check_offsets(dst, dst_loff, &len, olen); if (ret) goto out_unlock; @@ -2927,7 +2945,7 @@ static int btrfs_extent_same(struct inode *src, u64 loff, u64 len, ret = btrfs_cmp_data(src, loff, dst, dst_loff, len); if (ret == 0) - ret = btrfs_clone(src, dst, loff, len, len, dst_loff); + ret = btrfs_clone(src, dst, loff, olen, len, dst_loff); out_unlock: btrfs_double_unlock(src, loff, dst, dst_loff, len); diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index 157cc54..89656d7 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -198,9 +198,6 @@ static int __btrfs_add_ordered_extent(struct inode *inode, u64 file_offset, entry->file_offset = file_offset; entry->start = start; entry->len = len; - if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM) && - !(type == BTRFS_ORDERED_NOCOW)) - entry->csum_bytes_left = disk_len; entry->disk_len = disk_len; entry->bytes_left = len; entry->inode = igrab(inode); @@ -286,10 +283,6 @@ void btrfs_add_ordered_sum(struct inode *inode, tree = &BTRFS_I(inode)->ordered_tree; spin_lock_irq(&tree->lock); list_add_tail(&sum->list, &entry->list); - WARN_ON(entry->csum_bytes_left < sum->len); - entry->csum_bytes_left -= sum->len; - if (entry->csum_bytes_left == 0) - wake_up(&entry->wait); spin_unlock_irq(&tree->lock); } @@ -509,7 +502,21 @@ void btrfs_wait_logged_extents(struct btrfs_trans_handle *trans, wait_event(ordered->wait, test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags)); - list_add_tail(&ordered->trans_list, &trans->ordered); + /* + * If our ordered extent completed it means it updated the + * fs/subvol and csum trees already, so no need to make the + * current transaction's commit wait for it, as we end up + * holding memory unnecessarily and delaying the inode's iput + * until the transaction commit (we schedule an iput for the + * inode when the ordered extent's refcount drops to 0), which + * prevents it from being evictable until the transaction + * commits. + */ + if (test_bit(BTRFS_ORDERED_COMPLETE, &ordered->flags)) + btrfs_put_ordered_extent(ordered); + else + list_add_tail(&ordered->trans_list, &trans->ordered); + spin_lock_irq(&log->log_extents_lock[index]); } spin_unlock_irq(&log->log_extents_lock[index]); @@ -722,6 +729,7 @@ void btrfs_start_ordered_extent(struct inode *inode, int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len) { int ret = 0; + int ret_wb = 0; u64 end; u64 orig_end; struct btrfs_ordered_extent *ordered; @@ -741,9 +749,14 @@ int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len) if (ret) return ret; - ret = filemap_fdatawait_range(inode->i_mapping, start, orig_end); - if (ret) - return ret; + /* + * If we have a writeback error don't return immediately. Wait first + * for any ordered extents that haven't completed yet. This is to make + * sure no one can dirty the same page ranges and call writepages() + * before the ordered extents complete - to avoid failures (-EEXIST) + * when adding the new ordered extents to the ordered tree. + */ + ret_wb = filemap_fdatawait_range(inode->i_mapping, start, orig_end); end = orig_end; while (1) { @@ -767,7 +780,7 @@ int btrfs_wait_ordered_range(struct inode *inode, u64 start, u64 len) break; end--; } - return ret; + return ret_wb ? ret_wb : ret; } /* @@ -838,6 +851,20 @@ out: return entry; } +bool btrfs_have_ordered_extents_in_range(struct inode *inode, + u64 file_offset, + u64 len) +{ + struct btrfs_ordered_extent *oe; + + oe = btrfs_lookup_ordered_range(inode, file_offset, len); + if (oe) { + btrfs_put_ordered_extent(oe); + return true; + } + return false; +} + /* * lookup and return any extent before 'file_offset'. NULL is returned * if none is found diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h index e96cd4c..7176cc0 100644 --- a/fs/btrfs/ordered-data.h +++ b/fs/btrfs/ordered-data.h @@ -89,9 +89,6 @@ struct btrfs_ordered_extent { /* number of bytes that still need writing */ u64 bytes_left; - /* number of bytes that still need csumming */ - u64 csum_bytes_left; - /* * the end of the ordered extent which is behind it but * didn't update disk_i_size. Please see the comment of @@ -191,6 +188,9 @@ btrfs_lookup_first_ordered_extent(struct inode * inode, u64 file_offset); struct btrfs_ordered_extent *btrfs_lookup_ordered_range(struct inode *inode, u64 file_offset, u64 len); +bool btrfs_have_ordered_extents_in_range(struct inode *inode, + u64 file_offset, + u64 len); int btrfs_ordered_update_i_size(struct inode *inode, u64 offset, struct btrfs_ordered_extent *ordered); int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr, diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 3d65465..d5f1f03 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -34,6 +34,7 @@ #include "extent_io.h" #include "qgroup.h" + /* TODO XXX FIXME * - subvol delete -> delete when ref goes to 0? delete limits also? * - reorganize keys @@ -84,11 +85,42 @@ struct btrfs_qgroup { /* * temp variables for accounting operations + * Refer to qgroup_shared_accouting() for details. */ u64 old_refcnt; u64 new_refcnt; }; +static void btrfs_qgroup_update_old_refcnt(struct btrfs_qgroup *qg, u64 seq, + int mod) +{ + if (qg->old_refcnt < seq) + qg->old_refcnt = seq; + qg->old_refcnt += mod; +} + +static void btrfs_qgroup_update_new_refcnt(struct btrfs_qgroup *qg, u64 seq, + int mod) +{ + if (qg->new_refcnt < seq) + qg->new_refcnt = seq; + qg->new_refcnt += mod; +} + +static inline u64 btrfs_qgroup_get_old_refcnt(struct btrfs_qgroup *qg, u64 seq) +{ + if (qg->old_refcnt < seq) + return 0; + return qg->old_refcnt - seq; +} + +static inline u64 btrfs_qgroup_get_new_refcnt(struct btrfs_qgroup *qg, u64 seq) +{ + if (qg->new_refcnt < seq) + return 0; + return qg->new_refcnt - seq; +} + /* * glue structure to represent the relations between qgroups. */ @@ -1115,14 +1147,14 @@ int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans, struct ulist *tmp; int ret = 0; - tmp = ulist_alloc(GFP_NOFS); - if (!tmp) - return -ENOMEM; - /* Check the level of src and dst first */ if (btrfs_qgroup_level(src) >= btrfs_qgroup_level(dst)) return -EINVAL; + tmp = ulist_alloc(GFP_NOFS); + if (!tmp) + return -ENOMEM; + mutex_lock(&fs_info->qgroup_ioctl_lock); quota_root = fs_info->quota_root; if (!quota_root) { @@ -1356,239 +1388,86 @@ out: return ret; } -static int comp_oper_exist(struct btrfs_qgroup_operation *oper1, - struct btrfs_qgroup_operation *oper2) +int btrfs_qgroup_prepare_account_extents(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info) { - /* - * Ignore seq and type here, we're looking for any operation - * at all related to this extent on that root. - */ - if (oper1->bytenr < oper2->bytenr) - return -1; - if (oper1->bytenr > oper2->bytenr) - return 1; - if (oper1->ref_root < oper2->ref_root) - return -1; - if (oper1->ref_root > oper2->ref_root) - return 1; - return 0; -} + struct btrfs_qgroup_extent_record *record; + struct btrfs_delayed_ref_root *delayed_refs; + struct rb_node *node; + u64 qgroup_to_skip; + int ret = 0; -static int qgroup_oper_exists(struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper) -{ - struct rb_node *n; - struct btrfs_qgroup_operation *cur; - int cmp; + delayed_refs = &trans->transaction->delayed_refs; + qgroup_to_skip = delayed_refs->qgroup_to_skip; - spin_lock(&fs_info->qgroup_op_lock); - n = fs_info->qgroup_op_tree.rb_node; - while (n) { - cur = rb_entry(n, struct btrfs_qgroup_operation, n); - cmp = comp_oper_exist(cur, oper); - if (cmp < 0) { - n = n->rb_right; - } else if (cmp) { - n = n->rb_left; - } else { - spin_unlock(&fs_info->qgroup_op_lock); - return -EEXIST; - } + /* + * No need to do lock, since this function will only be called in + * btrfs_commmit_transaction(). + */ + node = rb_first(&delayed_refs->dirty_extent_root); + while (node) { + record = rb_entry(node, struct btrfs_qgroup_extent_record, + node); + ret = btrfs_find_all_roots(NULL, fs_info, record->bytenr, 0, + &record->old_roots); + if (ret < 0) + break; + if (qgroup_to_skip) + ulist_del(record->old_roots, qgroup_to_skip, 0); + node = rb_next(node); } - spin_unlock(&fs_info->qgroup_op_lock); - return 0; -} - -static int comp_oper(struct btrfs_qgroup_operation *oper1, - struct btrfs_qgroup_operation *oper2) -{ - if (oper1->bytenr < oper2->bytenr) - return -1; - if (oper1->bytenr > oper2->bytenr) - return 1; - if (oper1->ref_root < oper2->ref_root) - return -1; - if (oper1->ref_root > oper2->ref_root) - return 1; - if (oper1->seq < oper2->seq) - return -1; - if (oper1->seq > oper2->seq) - return 1; - if (oper1->type < oper2->type) - return -1; - if (oper1->type > oper2->type) - return 1; - return 0; + return ret; } -static int insert_qgroup_oper(struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper) +struct btrfs_qgroup_extent_record +*btrfs_qgroup_insert_dirty_extent(struct btrfs_delayed_ref_root *delayed_refs, + struct btrfs_qgroup_extent_record *record) { - struct rb_node **p; - struct rb_node *parent = NULL; - struct btrfs_qgroup_operation *cur; - int cmp; + struct rb_node **p = &delayed_refs->dirty_extent_root.rb_node; + struct rb_node *parent_node = NULL; + struct btrfs_qgroup_extent_record *entry; + u64 bytenr = record->bytenr; - spin_lock(&fs_info->qgroup_op_lock); - p = &fs_info->qgroup_op_tree.rb_node; while (*p) { - parent = *p; - cur = rb_entry(parent, struct btrfs_qgroup_operation, n); - cmp = comp_oper(cur, oper); - if (cmp < 0) { - p = &(*p)->rb_right; - } else if (cmp) { + parent_node = *p; + entry = rb_entry(parent_node, struct btrfs_qgroup_extent_record, + node); + if (bytenr < entry->bytenr) p = &(*p)->rb_left; - } else { - spin_unlock(&fs_info->qgroup_op_lock); - return -EEXIST; - } - } - rb_link_node(&oper->n, parent, p); - rb_insert_color(&oper->n, &fs_info->qgroup_op_tree); - spin_unlock(&fs_info->qgroup_op_lock); - return 0; -} - -/* - * Record a quota operation for processing later on. - * @trans: the transaction we are adding the delayed op to. - * @fs_info: the fs_info for this fs. - * @ref_root: the root of the reference we are acting on, - * @bytenr: the bytenr we are acting on. - * @num_bytes: the number of bytes in the reference. - * @type: the type of operation this is. - * @mod_seq: do we need to get a sequence number for looking up roots. - * - * We just add it to our trans qgroup_ref_list and carry on and process these - * operations in order at some later point. If the reference root isn't a fs - * root then we don't bother with doing anything. - * - * MUST BE HOLDING THE REF LOCK. - */ -int btrfs_qgroup_record_ref(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, u64 ref_root, - u64 bytenr, u64 num_bytes, - enum btrfs_qgroup_operation_type type, int mod_seq) -{ - struct btrfs_qgroup_operation *oper; - int ret; - - if (!is_fstree(ref_root) || !fs_info->quota_enabled) - return 0; - - oper = kmalloc(sizeof(*oper), GFP_NOFS); - if (!oper) - return -ENOMEM; - - oper->ref_root = ref_root; - oper->bytenr = bytenr; - oper->num_bytes = num_bytes; - oper->type = type; - oper->seq = atomic_inc_return(&fs_info->qgroup_op_seq); - INIT_LIST_HEAD(&oper->elem.list); - oper->elem.seq = 0; - - trace_btrfs_qgroup_record_ref(oper); - - if (type == BTRFS_QGROUP_OPER_SUB_SUBTREE) { - /* - * If any operation for this bytenr/ref_root combo - * exists, then we know it's not exclusively owned and - * shouldn't be queued up. - * - * This also catches the case where we have a cloned - * extent that gets queued up multiple times during - * drop snapshot. - */ - if (qgroup_oper_exists(fs_info, oper)) { - kfree(oper); - return 0; - } - } - - ret = insert_qgroup_oper(fs_info, oper); - if (ret) { - /* Shouldn't happen so have an assert for developers */ - ASSERT(0); - kfree(oper); - return ret; + else if (bytenr > entry->bytenr) + p = &(*p)->rb_right; + else + return entry; } - list_add_tail(&oper->list, &trans->qgroup_ref_list); - if (mod_seq) - btrfs_get_tree_mod_seq(fs_info, &oper->elem); - - return 0; -} - -static int qgroup_excl_accounting(struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper) -{ - struct ulist *tmp; - int sign = 0; - int ret = 0; - - tmp = ulist_alloc(GFP_NOFS); - if (!tmp) - return -ENOMEM; - - spin_lock(&fs_info->qgroup_lock); - if (!fs_info->quota_root) - goto out; - - switch (oper->type) { - case BTRFS_QGROUP_OPER_ADD_EXCL: - sign = 1; - break; - case BTRFS_QGROUP_OPER_SUB_EXCL: - sign = -1; - break; - default: - ASSERT(0); - } - ret = __qgroup_excl_accounting(fs_info, tmp, oper->ref_root, - oper->num_bytes, sign); -out: - spin_unlock(&fs_info->qgroup_lock); - ulist_free(tmp); - return ret; + rb_link_node(&record->node, parent_node, p); + rb_insert_color(&record->node, &delayed_refs->dirty_extent_root); + return NULL; } +#define UPDATE_NEW 0 +#define UPDATE_OLD 1 /* - * Walk all of the roots that pointed to our bytenr and adjust their refcnts as - * properly. + * Walk all of the roots that points to the bytenr and adjust their refcnts. */ -static int qgroup_calc_old_refcnt(struct btrfs_fs_info *fs_info, - u64 root_to_skip, struct ulist *tmp, - struct ulist *roots, struct ulist *qgroups, - u64 seq, int *old_roots, int rescan) +static int qgroup_update_refcnt(struct btrfs_fs_info *fs_info, + struct ulist *roots, struct ulist *tmp, + struct ulist *qgroups, u64 seq, int update_old) { struct ulist_node *unode; struct ulist_iterator uiter; struct ulist_node *tmp_unode; struct ulist_iterator tmp_uiter; struct btrfs_qgroup *qg; - int ret; + int ret = 0; + if (!roots) + return 0; ULIST_ITER_INIT(&uiter); while ((unode = ulist_next(roots, &uiter))) { - /* We don't count our current root here */ - if (unode->val == root_to_skip) - continue; qg = find_qgroup_rb(fs_info, unode->val); if (!qg) continue; - /* - * We could have a pending removal of this same ref so we may - * not have actually found our ref root when doing - * btrfs_find_all_roots, so we need to keep track of how many - * old roots we find in case we removed ours and added a - * different one at the same time. I don't think this could - * happen in practice but that sort of thinking leads to pain - * and suffering and to the dark side. - */ - (*old_roots)++; ulist_reinit(tmp); ret = ulist_add(qgroups, qg->qgroupid, ptr_to_u64(qg), @@ -1603,29 +1482,10 @@ static int qgroup_calc_old_refcnt(struct btrfs_fs_info *fs_info, struct btrfs_qgroup_list *glist; qg = u64_to_ptr(tmp_unode->aux); - /* - * We use this sequence number to keep from having to - * run the whole list and 0 out the refcnt every time. - * We basically use sequnce as the known 0 count and - * then add 1 everytime we see a qgroup. This is how we - * get how many of the roots actually point up to the - * upper level qgroups in order to determine exclusive - * counts. - * - * For rescan we want to set old_refcnt to seq so our - * exclusive calculations end up correct. - */ - if (rescan) - qg->old_refcnt = seq; - else if (qg->old_refcnt < seq) - qg->old_refcnt = seq + 1; + if (update_old) + btrfs_qgroup_update_old_refcnt(qg, seq, 1); else - qg->old_refcnt++; - - if (qg->new_refcnt < seq) - qg->new_refcnt = seq + 1; - else - qg->new_refcnt++; + btrfs_qgroup_update_new_refcnt(qg, seq, 1); list_for_each_entry(glist, &qg->groups, next_group) { ret = ulist_add(qgroups, glist->group->qgroupid, ptr_to_u64(glist->group), @@ -1644,161 +1504,46 @@ static int qgroup_calc_old_refcnt(struct btrfs_fs_info *fs_info, } /* - * We need to walk forward in our operation tree and account for any roots that - * were deleted after we made this operation. - */ -static int qgroup_account_deleted_refs(struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper, - struct ulist *tmp, - struct ulist *qgroups, u64 seq, - int *old_roots) -{ - struct ulist_node *unode; - struct ulist_iterator uiter; - struct btrfs_qgroup *qg; - struct btrfs_qgroup_operation *tmp_oper; - struct rb_node *n; - int ret; - - ulist_reinit(tmp); - - /* - * We only walk forward in the tree since we're only interested in - * removals that happened _after_ our operation. - */ - spin_lock(&fs_info->qgroup_op_lock); - n = rb_next(&oper->n); - spin_unlock(&fs_info->qgroup_op_lock); - if (!n) - return 0; - tmp_oper = rb_entry(n, struct btrfs_qgroup_operation, n); - while (tmp_oper->bytenr == oper->bytenr) { - /* - * If it's not a removal we don't care, additions work out - * properly with our refcnt tracking. - */ - if (tmp_oper->type != BTRFS_QGROUP_OPER_SUB_SHARED && - tmp_oper->type != BTRFS_QGROUP_OPER_SUB_EXCL) - goto next; - qg = find_qgroup_rb(fs_info, tmp_oper->ref_root); - if (!qg) - goto next; - ret = ulist_add(qgroups, qg->qgroupid, ptr_to_u64(qg), - GFP_ATOMIC); - if (ret) { - if (ret < 0) - return ret; - /* - * We only want to increase old_roots if this qgroup is - * not already in the list of qgroups. If it is already - * there then that means it must have been re-added or - * the delete will be discarded because we had an - * existing ref that we haven't looked up yet. In this - * case we don't want to increase old_roots. So if ret - * == 1 then we know that this is the first time we've - * seen this qgroup and we can bump the old_roots. - */ - (*old_roots)++; - ret = ulist_add(tmp, qg->qgroupid, ptr_to_u64(qg), - GFP_ATOMIC); - if (ret < 0) - return ret; - } -next: - spin_lock(&fs_info->qgroup_op_lock); - n = rb_next(&tmp_oper->n); - spin_unlock(&fs_info->qgroup_op_lock); - if (!n) - break; - tmp_oper = rb_entry(n, struct btrfs_qgroup_operation, n); - } - - /* Ok now process the qgroups we found */ - ULIST_ITER_INIT(&uiter); - while ((unode = ulist_next(tmp, &uiter))) { - struct btrfs_qgroup_list *glist; - - qg = u64_to_ptr(unode->aux); - if (qg->old_refcnt < seq) - qg->old_refcnt = seq + 1; - else - qg->old_refcnt++; - if (qg->new_refcnt < seq) - qg->new_refcnt = seq + 1; - else - qg->new_refcnt++; - list_for_each_entry(glist, &qg->groups, next_group) { - ret = ulist_add(qgroups, glist->group->qgroupid, - ptr_to_u64(glist->group), GFP_ATOMIC); - if (ret < 0) - return ret; - ret = ulist_add(tmp, glist->group->qgroupid, - ptr_to_u64(glist->group), GFP_ATOMIC); - if (ret < 0) - return ret; - } - } - return 0; -} - -/* Add refcnt for the newly added reference. */ -static int qgroup_calc_new_refcnt(struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper, - struct btrfs_qgroup *qgroup, - struct ulist *tmp, struct ulist *qgroups, - u64 seq) -{ - struct ulist_node *unode; - struct ulist_iterator uiter; - struct btrfs_qgroup *qg; - int ret; - - ulist_reinit(tmp); - ret = ulist_add(qgroups, qgroup->qgroupid, ptr_to_u64(qgroup), - GFP_ATOMIC); - if (ret < 0) - return ret; - ret = ulist_add(tmp, qgroup->qgroupid, ptr_to_u64(qgroup), - GFP_ATOMIC); - if (ret < 0) - return ret; - ULIST_ITER_INIT(&uiter); - while ((unode = ulist_next(tmp, &uiter))) { - struct btrfs_qgroup_list *glist; - - qg = u64_to_ptr(unode->aux); - if (oper->type == BTRFS_QGROUP_OPER_ADD_SHARED) { - if (qg->new_refcnt < seq) - qg->new_refcnt = seq + 1; - else - qg->new_refcnt++; - } else { - if (qg->old_refcnt < seq) - qg->old_refcnt = seq + 1; - else - qg->old_refcnt++; - } - list_for_each_entry(glist, &qg->groups, next_group) { - ret = ulist_add(tmp, glist->group->qgroupid, - ptr_to_u64(glist->group), GFP_ATOMIC); - if (ret < 0) - return ret; - ret = ulist_add(qgroups, glist->group->qgroupid, - ptr_to_u64(glist->group), GFP_ATOMIC); - if (ret < 0) - return ret; - } - } - return 0; -} - -/* - * This adjusts the counters for all referenced qgroups if need be. + * Update qgroup rfer/excl counters. + * Rfer update is easy, codes can explain themselves. + * + * Excl update is tricky, the update is split into 2 part. + * Part 1: Possible exclusive <-> sharing detect: + * | A | !A | + * ------------------------------------- + * B | * | - | + * ------------------------------------- + * !B | + | ** | + * ------------------------------------- + * + * Conditions: + * A: cur_old_roots < nr_old_roots (not exclusive before) + * !A: cur_old_roots == nr_old_roots (possible exclusive before) + * B: cur_new_roots < nr_new_roots (not exclusive now) + * !B: cur_new_roots == nr_new_roots (possible exclsuive now) + * + * Results: + * +: Possible sharing -> exclusive -: Possible exclusive -> sharing + * *: Definitely not changed. **: Possible unchanged. + * + * For !A and !B condition, the exception is cur_old/new_roots == 0 case. + * + * To make the logic clear, we first use condition A and B to split + * combination into 4 results. + * + * Then, for result "+" and "-", check old/new_roots == 0 case, as in them + * only on variant maybe 0. + * + * Lastly, check result **, since there are 2 variants maybe 0, split them + * again(2x2). + * But this time we don't need to consider other things, the codes and logic + * is easy to understand now. */ -static int qgroup_adjust_counters(struct btrfs_fs_info *fs_info, - u64 root_to_skip, u64 num_bytes, - struct ulist *qgroups, u64 seq, - int old_roots, int new_roots, int rescan) +static int qgroup_update_counters(struct btrfs_fs_info *fs_info, + struct ulist *qgroups, + u64 nr_old_roots, + u64 nr_new_roots, + u64 num_bytes, u64 seq) { struct ulist_node *unode; struct ulist_iterator uiter; @@ -1810,423 +1555,191 @@ static int qgroup_adjust_counters(struct btrfs_fs_info *fs_info, bool dirty = false; qg = u64_to_ptr(unode->aux); - /* - * Wasn't referenced before but is now, add to the reference - * counters. - */ - if (qg->old_refcnt <= seq && qg->new_refcnt > seq) { + cur_old_count = btrfs_qgroup_get_old_refcnt(qg, seq); + cur_new_count = btrfs_qgroup_get_new_refcnt(qg, seq); + + /* Rfer update part */ + if (cur_old_count == 0 && cur_new_count > 0) { qg->rfer += num_bytes; qg->rfer_cmpr += num_bytes; dirty = true; } - - /* - * Was referenced before but isn't now, subtract from the - * reference counters. - */ - if (qg->old_refcnt > seq && qg->new_refcnt <= seq) { + if (cur_old_count > 0 && cur_new_count == 0) { qg->rfer -= num_bytes; qg->rfer_cmpr -= num_bytes; dirty = true; } - if (qg->old_refcnt < seq) - cur_old_count = 0; - else - cur_old_count = qg->old_refcnt - seq; - if (qg->new_refcnt < seq) - cur_new_count = 0; - else - cur_new_count = qg->new_refcnt - seq; - - /* - * If our refcount was the same as the roots previously but our - * new count isn't the same as the number of roots now then we - * went from having a exclusive reference on this range to not. - */ - if (old_roots && cur_old_count == old_roots && - (cur_new_count != new_roots || new_roots == 0)) { - WARN_ON(cur_new_count != new_roots && new_roots == 0); - qg->excl -= num_bytes; - qg->excl_cmpr -= num_bytes; - dirty = true; + /* Excl update part */ + /* Exclusive/none -> shared case */ + if (cur_old_count == nr_old_roots && + cur_new_count < nr_new_roots) { + /* Exclusive -> shared */ + if (cur_old_count != 0) { + qg->excl -= num_bytes; + qg->excl_cmpr -= num_bytes; + dirty = true; + } } - /* - * If we didn't reference all the roots before but now we do we - * have an exclusive reference to this range. - */ - if ((!old_roots || (old_roots && cur_old_count != old_roots)) - && cur_new_count == new_roots) { - qg->excl += num_bytes; - qg->excl_cmpr += num_bytes; - dirty = true; + /* Shared -> exclusive/none case */ + if (cur_old_count < nr_old_roots && + cur_new_count == nr_new_roots) { + /* Shared->exclusive */ + if (cur_new_count != 0) { + qg->excl += num_bytes; + qg->excl_cmpr += num_bytes; + dirty = true; + } } + /* Exclusive/none -> exclusive/none case */ + if (cur_old_count == nr_old_roots && + cur_new_count == nr_new_roots) { + if (cur_old_count == 0) { + /* None -> exclusive/none */ + + if (cur_new_count != 0) { + /* None -> exclusive */ + qg->excl += num_bytes; + qg->excl_cmpr += num_bytes; + dirty = true; + } + /* None -> none, nothing changed */ + } else { + /* Exclusive -> exclusive/none */ + + if (cur_new_count == 0) { + /* Exclusive -> none */ + qg->excl -= num_bytes; + qg->excl_cmpr -= num_bytes; + dirty = true; + } + /* Exclusive -> exclusive, nothing changed */ + } + } if (dirty) qgroup_dirty(fs_info, qg); } return 0; } -/* - * If we removed a data extent and there were other references for that bytenr - * then we need to lookup all referenced roots to make sure we still don't - * reference this bytenr. If we do then we can just discard this operation. - */ -static int check_existing_refs(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper) -{ - struct ulist *roots = NULL; - struct ulist_node *unode; - struct ulist_iterator uiter; - int ret = 0; - - ret = btrfs_find_all_roots(trans, fs_info, oper->bytenr, - oper->elem.seq, &roots); - if (ret < 0) - return ret; - ret = 0; - - ULIST_ITER_INIT(&uiter); - while ((unode = ulist_next(roots, &uiter))) { - if (unode->val == oper->ref_root) { - ret = 1; - break; - } - } - ulist_free(roots); - btrfs_put_tree_mod_seq(fs_info, &oper->elem); - - return ret; -} - -/* - * If we share a reference across multiple roots then we may need to adjust - * various qgroups referenced and exclusive counters. The basic premise is this - * - * 1) We have seq to represent a 0 count. Instead of looping through all of the - * qgroups and resetting their refcount to 0 we just constantly bump this - * sequence number to act as the base reference count. This means that if - * anybody is equal to or below this sequence they were never referenced. We - * jack this sequence up by the number of roots we found each time in order to - * make sure we don't have any overlap. - * - * 2) We first search all the roots that reference the area _except_ the root - * we're acting on currently. This makes up the old_refcnt of all the qgroups - * before. - * - * 3) We walk all of the qgroups referenced by the root we are currently acting - * on, and will either adjust old_refcnt in the case of a removal or the - * new_refcnt in the case of an addition. - * - * 4) Finally we walk all the qgroups that are referenced by this range - * including the root we are acting on currently. We will adjust the counters - * based on the number of roots we had and will have after this operation. - * - * Take this example as an illustration - * - * [qgroup 1/0] - * / | \ - * [qg 0/0] [qg 0/1] [qg 0/2] - * \ | / - * [ extent ] - * - * Say we are adding a reference that is covered by qg 0/0. The first step - * would give a refcnt of 1 to qg 0/1 and 0/2 and a refcnt of 2 to qg 1/0 with - * old_roots being 2. Because it is adding new_roots will be 1. We then go - * through qg 0/0 which will get the new_refcnt set to 1 and add 1 to qg 1/0's - * new_refcnt, bringing it to 3. We then walk through all of the qgroups, we - * notice that the old refcnt for qg 0/0 < the new refcnt, so we added a - * reference and thus must add the size to the referenced bytes. Everything - * else is the same so nothing else changes. - */ -static int qgroup_shared_accounting(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper) +int +btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, + u64 bytenr, u64 num_bytes, + struct ulist *old_roots, struct ulist *new_roots) { - struct ulist *roots = NULL; - struct ulist *qgroups, *tmp; - struct btrfs_qgroup *qgroup; - struct seq_list elem = SEQ_LIST_INIT(elem); + struct ulist *qgroups = NULL; + struct ulist *tmp = NULL; u64 seq; - int old_roots = 0; - int new_roots = 0; + u64 nr_new_roots = 0; + u64 nr_old_roots = 0; int ret = 0; - if (oper->elem.seq) { - ret = check_existing_refs(trans, fs_info, oper); - if (ret < 0) - return ret; - if (ret) - return 0; - } + if (new_roots) + nr_new_roots = new_roots->nnodes; + if (old_roots) + nr_old_roots = old_roots->nnodes; - qgroups = ulist_alloc(GFP_NOFS); - if (!qgroups) - return -ENOMEM; + if (!fs_info->quota_enabled) + goto out_free; + BUG_ON(!fs_info->quota_root); + qgroups = ulist_alloc(GFP_NOFS); + if (!qgroups) { + ret = -ENOMEM; + goto out_free; + } tmp = ulist_alloc(GFP_NOFS); if (!tmp) { - ulist_free(qgroups); - return -ENOMEM; + ret = -ENOMEM; + goto out_free; } - btrfs_get_tree_mod_seq(fs_info, &elem); - ret = btrfs_find_all_roots(trans, fs_info, oper->bytenr, elem.seq, - &roots); - btrfs_put_tree_mod_seq(fs_info, &elem); - if (ret < 0) { - ulist_free(qgroups); - ulist_free(tmp); - return ret; + mutex_lock(&fs_info->qgroup_rescan_lock); + if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) { + if (fs_info->qgroup_rescan_progress.objectid <= bytenr) { + mutex_unlock(&fs_info->qgroup_rescan_lock); + ret = 0; + goto out_free; + } } + mutex_unlock(&fs_info->qgroup_rescan_lock); + spin_lock(&fs_info->qgroup_lock); - qgroup = find_qgroup_rb(fs_info, oper->ref_root); - if (!qgroup) - goto out; seq = fs_info->qgroup_seq; - /* - * So roots is the list of all the roots currently pointing at the - * bytenr, including the ref we are adding if we are adding, or not if - * we are removing a ref. So we pass in the ref_root to skip that root - * in our calculations. We set old_refnct and new_refcnt cause who the - * hell knows what everything looked like before, and it doesn't matter - * except... - */ - ret = qgroup_calc_old_refcnt(fs_info, oper->ref_root, tmp, roots, qgroups, - seq, &old_roots, 0); + /* Update old refcnts using old_roots */ + ret = qgroup_update_refcnt(fs_info, old_roots, tmp, qgroups, seq, + UPDATE_OLD); if (ret < 0) goto out; - /* - * Now adjust the refcounts of the qgroups that care about this - * reference, either the old_count in the case of removal or new_count - * in the case of an addition. - */ - ret = qgroup_calc_new_refcnt(fs_info, oper, qgroup, tmp, qgroups, - seq); + /* Update new refcnts using new_roots */ + ret = qgroup_update_refcnt(fs_info, new_roots, tmp, qgroups, seq, + UPDATE_NEW); if (ret < 0) goto out; - /* - * ...in the case of removals. If we had a removal before we got around - * to processing this operation then we need to find that guy and count - * his references as if they really existed so we don't end up screwing - * up the exclusive counts. Then whenever we go to process the delete - * everything will be grand and we can account for whatever exclusive - * changes need to be made there. We also have to pass in old_roots so - * we have an accurate count of the roots as it pertains to this - * operations view of the world. - */ - ret = qgroup_account_deleted_refs(fs_info, oper, tmp, qgroups, seq, - &old_roots); - if (ret < 0) - goto out; + qgroup_update_counters(fs_info, qgroups, nr_old_roots, nr_new_roots, + num_bytes, seq); /* - * We are adding our root, need to adjust up the number of roots, - * otherwise old_roots is the number of roots we want. + * Bump qgroup_seq to avoid seq overlap */ - if (oper->type == BTRFS_QGROUP_OPER_ADD_SHARED) { - new_roots = old_roots + 1; - } else { - new_roots = old_roots; - old_roots++; - } - fs_info->qgroup_seq += old_roots + 1; - - - /* - * And now the magic happens, bless Arne for having a pretty elegant - * solution for this. - */ - qgroup_adjust_counters(fs_info, oper->ref_root, oper->num_bytes, - qgroups, seq, old_roots, new_roots, 0); + fs_info->qgroup_seq += max(nr_old_roots, nr_new_roots) + 1; out: spin_unlock(&fs_info->qgroup_lock); - ulist_free(qgroups); - ulist_free(roots); +out_free: ulist_free(tmp); + ulist_free(qgroups); + ulist_free(old_roots); + ulist_free(new_roots); return ret; } -/* - * Process a reference to a shared subtree. This type of operation is - * queued during snapshot removal when we encounter extents which are - * shared between more than one root. - */ -static int qgroup_subtree_accounting(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper) -{ - struct ulist *roots = NULL; - struct ulist_node *unode; - struct ulist_iterator uiter; - struct btrfs_qgroup_list *glist; - struct ulist *parents; - int ret = 0; - int err; - struct btrfs_qgroup *qg; - u64 root_obj = 0; - struct seq_list elem = SEQ_LIST_INIT(elem); - - parents = ulist_alloc(GFP_NOFS); - if (!parents) - return -ENOMEM; - - btrfs_get_tree_mod_seq(fs_info, &elem); - ret = btrfs_find_all_roots(trans, fs_info, oper->bytenr, - elem.seq, &roots); - btrfs_put_tree_mod_seq(fs_info, &elem); - if (ret < 0) - goto out; - - if (roots->nnodes != 1) - goto out; - - ULIST_ITER_INIT(&uiter); - unode = ulist_next(roots, &uiter); /* Only want 1 so no need to loop */ - /* - * If we find our ref root then that means all refs - * this extent has to the root have not yet been - * deleted. In that case, we do nothing and let the - * last ref for this bytenr drive our update. - * - * This can happen for example if an extent is - * referenced multiple times in a snapshot (clone, - * etc). If we are in the middle of snapshot removal, - * queued updates for such an extent will find the - * root if we have not yet finished removing the - * snapshot. - */ - if (unode->val == oper->ref_root) - goto out; - - root_obj = unode->val; - BUG_ON(!root_obj); - - spin_lock(&fs_info->qgroup_lock); - qg = find_qgroup_rb(fs_info, root_obj); - if (!qg) - goto out_unlock; - - qg->excl += oper->num_bytes; - qg->excl_cmpr += oper->num_bytes; - qgroup_dirty(fs_info, qg); - - /* - * Adjust counts for parent groups. First we find all - * parents, then in the 2nd loop we do the adjustment - * while adding parents of the parents to our ulist. - */ - list_for_each_entry(glist, &qg->groups, next_group) { - err = ulist_add(parents, glist->group->qgroupid, - ptr_to_u64(glist->group), GFP_ATOMIC); - if (err < 0) { - ret = err; - goto out_unlock; - } - } - - ULIST_ITER_INIT(&uiter); - while ((unode = ulist_next(parents, &uiter))) { - qg = u64_to_ptr(unode->aux); - qg->excl += oper->num_bytes; - qg->excl_cmpr += oper->num_bytes; - qgroup_dirty(fs_info, qg); - - /* Add any parents of the parents */ - list_for_each_entry(glist, &qg->groups, next_group) { - err = ulist_add(parents, glist->group->qgroupid, - ptr_to_u64(glist->group), GFP_ATOMIC); - if (err < 0) { - ret = err; - goto out_unlock; - } - } - } - -out_unlock: - spin_unlock(&fs_info->qgroup_lock); - -out: - ulist_free(roots); - ulist_free(parents); - return ret; -} - -/* - * btrfs_qgroup_account_ref is called for every ref that is added to or deleted - * from the fs. First, all roots referencing the extent are searched, and - * then the space is accounted accordingly to the different roots. The - * accounting algorithm works in 3 steps documented inline. - */ -static int btrfs_qgroup_account(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper) +int btrfs_qgroup_account_extents(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info) { + struct btrfs_qgroup_extent_record *record; + struct btrfs_delayed_ref_root *delayed_refs; + struct ulist *new_roots = NULL; + struct rb_node *node; + u64 qgroup_to_skip; int ret = 0; - if (!fs_info->quota_enabled) - return 0; - - BUG_ON(!fs_info->quota_root); + delayed_refs = &trans->transaction->delayed_refs; + qgroup_to_skip = delayed_refs->qgroup_to_skip; + while ((node = rb_first(&delayed_refs->dirty_extent_root))) { + record = rb_entry(node, struct btrfs_qgroup_extent_record, + node); - mutex_lock(&fs_info->qgroup_rescan_lock); - if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) { - if (fs_info->qgroup_rescan_progress.objectid <= oper->bytenr) { - mutex_unlock(&fs_info->qgroup_rescan_lock); - return 0; + if (!ret) { + /* + * Use (u64)-1 as time_seq to do special search, which + * doesn't lock tree or delayed_refs and search current + * root. It's safe inside commit_transaction(). + */ + ret = btrfs_find_all_roots(trans, fs_info, + record->bytenr, (u64)-1, &new_roots); + if (ret < 0) + goto cleanup; + if (qgroup_to_skip) + ulist_del(new_roots, qgroup_to_skip, 0); + ret = btrfs_qgroup_account_extent(trans, fs_info, + record->bytenr, record->num_bytes, + record->old_roots, new_roots); + record->old_roots = NULL; + new_roots = NULL; } - } - mutex_unlock(&fs_info->qgroup_rescan_lock); +cleanup: + ulist_free(record->old_roots); + ulist_free(new_roots); + new_roots = NULL; + rb_erase(node, &delayed_refs->dirty_extent_root); + kfree(record); - ASSERT(is_fstree(oper->ref_root)); - - trace_btrfs_qgroup_account(oper); - - switch (oper->type) { - case BTRFS_QGROUP_OPER_ADD_EXCL: - case BTRFS_QGROUP_OPER_SUB_EXCL: - ret = qgroup_excl_accounting(fs_info, oper); - break; - case BTRFS_QGROUP_OPER_ADD_SHARED: - case BTRFS_QGROUP_OPER_SUB_SHARED: - ret = qgroup_shared_accounting(trans, fs_info, oper); - break; - case BTRFS_QGROUP_OPER_SUB_SUBTREE: - ret = qgroup_subtree_accounting(trans, fs_info, oper); - break; - default: - ASSERT(0); - } - return ret; -} - -/* - * Needs to be called everytime we run delayed refs, even if there is an error - * in order to cleanup outstanding operations. - */ -int btrfs_delayed_qgroup_accounting(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info) -{ - struct btrfs_qgroup_operation *oper; - int ret = 0; - - while (!list_empty(&trans->qgroup_ref_list)) { - oper = list_first_entry(&trans->qgroup_ref_list, - struct btrfs_qgroup_operation, list); - list_del_init(&oper->list); - if (!ret || !trans->aborted) - ret = btrfs_qgroup_account(trans, fs_info, oper); - spin_lock(&fs_info->qgroup_op_lock); - rb_erase(&oper->n, &fs_info->qgroup_op_tree); - spin_unlock(&fs_info->qgroup_op_lock); - btrfs_put_tree_mod_seq(fs_info, &oper->elem); - kfree(oper); } return ret; } @@ -2637,15 +2150,13 @@ void assert_qgroups_uptodate(struct btrfs_trans_handle *trans) */ static int qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path, - struct btrfs_trans_handle *trans, struct ulist *qgroups, - struct ulist *tmp, struct extent_buffer *scratch_leaf) + struct btrfs_trans_handle *trans, + struct extent_buffer *scratch_leaf) { struct btrfs_key found; struct ulist *roots = NULL; struct seq_list tree_mod_seq_elem = SEQ_LIST_INIT(tree_mod_seq_elem); u64 num_bytes; - u64 seq; - int new_roots; int slot; int ret; @@ -2695,33 +2206,15 @@ qgroup_rescan_leaf(struct btrfs_fs_info *fs_info, struct btrfs_path *path, else num_bytes = found.offset; - ulist_reinit(qgroups); ret = btrfs_find_all_roots(NULL, fs_info, found.objectid, 0, &roots); if (ret < 0) goto out; - spin_lock(&fs_info->qgroup_lock); - seq = fs_info->qgroup_seq; - fs_info->qgroup_seq += roots->nnodes + 1; /* max refcnt */ - - new_roots = 0; - ret = qgroup_calc_old_refcnt(fs_info, 0, tmp, roots, qgroups, - seq, &new_roots, 1); - if (ret < 0) { - spin_unlock(&fs_info->qgroup_lock); - ulist_free(roots); - goto out; - } - - ret = qgroup_adjust_counters(fs_info, 0, num_bytes, qgroups, - seq, 0, new_roots, 1); - if (ret < 0) { - spin_unlock(&fs_info->qgroup_lock); - ulist_free(roots); + /* For rescan, just pass old_roots as NULL */ + ret = btrfs_qgroup_account_extent(trans, fs_info, + found.objectid, num_bytes, NULL, roots); + if (ret < 0) goto out; - } - spin_unlock(&fs_info->qgroup_lock); - ulist_free(roots); } out: btrfs_put_tree_mod_seq(fs_info, &tree_mod_seq_elem); @@ -2735,7 +2228,6 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work) qgroup_rescan_work); struct btrfs_path *path; struct btrfs_trans_handle *trans = NULL; - struct ulist *tmp = NULL, *qgroups = NULL; struct extent_buffer *scratch_leaf = NULL; int err = -ENOMEM; int ret = 0; @@ -2743,12 +2235,6 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work) path = btrfs_alloc_path(); if (!path) goto out; - qgroups = ulist_alloc(GFP_NOFS); - if (!qgroups) - goto out; - tmp = ulist_alloc(GFP_NOFS); - if (!tmp) - goto out; scratch_leaf = kmalloc(sizeof(*scratch_leaf), GFP_NOFS); if (!scratch_leaf) goto out; @@ -2764,7 +2250,7 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work) err = -EINTR; } else { err = qgroup_rescan_leaf(fs_info, path, trans, - qgroups, tmp, scratch_leaf); + scratch_leaf); } if (err > 0) btrfs_commit_transaction(trans, fs_info->fs_root); @@ -2774,8 +2260,6 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work) out: kfree(scratch_leaf); - ulist_free(qgroups); - ulist_free(tmp); btrfs_free_path(path); mutex_lock(&fs_info->qgroup_rescan_lock); diff --git a/fs/btrfs/qgroup.h b/fs/btrfs/qgroup.h index c5242aa..6387dcf 100644 --- a/fs/btrfs/qgroup.h +++ b/fs/btrfs/qgroup.h @@ -19,43 +19,18 @@ #ifndef __BTRFS_QGROUP__ #define __BTRFS_QGROUP__ +#include "ulist.h" +#include "delayed-ref.h" + /* - * A description of the operations, all of these operations only happen when we - * are adding the 1st reference for that subvolume in the case of adding space - * or on the last reference delete in the case of subtraction. The only - * exception is the last one, which is added for confusion. - * - * BTRFS_QGROUP_OPER_ADD_EXCL: adding bytes where this subvolume is the only - * one pointing at the bytes we are adding. This is called on the first - * allocation. - * - * BTRFS_QGROUP_OPER_ADD_SHARED: adding bytes where this bytenr is going to be - * shared between subvols. This is called on the creation of a ref that already - * has refs from a different subvolume, so basically reflink. - * - * BTRFS_QGROUP_OPER_SUB_EXCL: removing bytes where this subvolume is the only - * one referencing the range. - * - * BTRFS_QGROUP_OPER_SUB_SHARED: removing bytes where this subvolume shares with - * refs with other subvolumes. + * Record a dirty extent, and info qgroup to update quota on it + * TODO: Use kmem cache to alloc it. */ -enum btrfs_qgroup_operation_type { - BTRFS_QGROUP_OPER_ADD_EXCL, - BTRFS_QGROUP_OPER_ADD_SHARED, - BTRFS_QGROUP_OPER_SUB_EXCL, - BTRFS_QGROUP_OPER_SUB_SHARED, - BTRFS_QGROUP_OPER_SUB_SUBTREE, -}; - -struct btrfs_qgroup_operation { - u64 ref_root; +struct btrfs_qgroup_extent_record { + struct rb_node node; u64 bytenr; u64 num_bytes; - u64 seq; - enum btrfs_qgroup_operation_type type; - struct seq_list elem; - struct rb_node n; - struct list_head list; + struct ulist *old_roots; }; int btrfs_quota_enable(struct btrfs_trans_handle *trans, @@ -79,16 +54,18 @@ int btrfs_limit_qgroup(struct btrfs_trans_handle *trans, int btrfs_read_qgroup_config(struct btrfs_fs_info *fs_info); void btrfs_free_qgroup_config(struct btrfs_fs_info *fs_info); struct btrfs_delayed_extent_op; -int btrfs_qgroup_record_ref(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, u64 ref_root, +int btrfs_qgroup_prepare_account_extents(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info); +struct btrfs_qgroup_extent_record +*btrfs_qgroup_insert_dirty_extent(struct btrfs_delayed_ref_root *delayed_refs, + struct btrfs_qgroup_extent_record *record); +int +btrfs_qgroup_account_extent(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info, u64 bytenr, u64 num_bytes, - enum btrfs_qgroup_operation_type type, - int mod_seq); -int btrfs_delayed_qgroup_accounting(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info); -void btrfs_remove_qgroup_operation(struct btrfs_trans_handle *trans, - struct btrfs_fs_info *fs_info, - struct btrfs_qgroup_operation *oper); + struct ulist *old_roots, struct ulist *new_roots); +int btrfs_qgroup_account_extents(struct btrfs_trans_handle *trans, + struct btrfs_fs_info *fs_info); int btrfs_run_qgroups(struct btrfs_trans_handle *trans, struct btrfs_fs_info *fs_info); int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 74b24b0..827951f 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -1847,8 +1847,10 @@ again: } eb = read_tree_block(dest, old_bytenr, old_ptr_gen); - if (!eb || !extent_buffer_uptodate(eb)) { - ret = (!eb) ? -ENOMEM : -EIO; + if (IS_ERR(eb)) { + ret = PTR_ERR(eb); + } else if (!extent_buffer_uptodate(eb)) { + ret = -EIO; free_extent_buffer(eb); break; } @@ -2002,7 +2004,9 @@ int walk_down_reloc_tree(struct btrfs_root *root, struct btrfs_path *path, bytenr = btrfs_node_blockptr(eb, path->slots[i]); eb = read_tree_block(root, bytenr, ptr_gen); - if (!eb || !extent_buffer_uptodate(eb)) { + if (IS_ERR(eb)) { + return PTR_ERR(eb); + } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); return -EIO; } @@ -2710,7 +2714,10 @@ static int do_relocation(struct btrfs_trans_handle *trans, blocksize = root->nodesize; generation = btrfs_node_ptr_generation(upper->eb, slot); eb = read_tree_block(root, bytenr, generation); - if (!eb || !extent_buffer_uptodate(eb)) { + if (IS_ERR(eb)) { + err = PTR_ERR(eb); + goto next; + } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); err = -EIO; goto next; @@ -2873,7 +2880,9 @@ static int get_tree_block_key(struct reloc_control *rc, BUG_ON(block->key_ready); eb = read_tree_block(rc->extent_root, block->bytenr, block->key.offset); - if (!eb || !extent_buffer_uptodate(eb)) { + if (IS_ERR(eb)) { + return PTR_ERR(eb); + } else if (!extent_buffer_uptodate(eb)) { free_extent_buffer(eb); return -EIO; } diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index ab58115..9f2feab 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -2662,18 +2662,30 @@ static void scrub_free_parity(struct scrub_parity *sparity) kfree(sparity); } +static void scrub_parity_bio_endio_worker(struct btrfs_work *work) +{ + struct scrub_parity *sparity = container_of(work, struct scrub_parity, + work); + struct scrub_ctx *sctx = sparity->sctx; + + scrub_free_parity(sparity); + scrub_pending_bio_dec(sctx); +} + static void scrub_parity_bio_endio(struct bio *bio, int error) { struct scrub_parity *sparity = (struct scrub_parity *)bio->bi_private; - struct scrub_ctx *sctx = sparity->sctx; if (error) bitmap_or(sparity->ebitmap, sparity->ebitmap, sparity->dbitmap, sparity->nsectors); - scrub_free_parity(sparity); - scrub_pending_bio_dec(sctx); bio_put(bio); + + btrfs_init_work(&sparity->work, btrfs_scrubparity_helper, + scrub_parity_bio_endio_worker, NULL, NULL); + btrfs_queue_work(sparity->sctx->dev_root->fs_info->scrub_parity_workers, + &sparity->work); } static void scrub_parity_check_and_repair(struct scrub_parity *sparity) @@ -3589,6 +3601,13 @@ static noinline_for_stack int scrub_workers_get(struct btrfs_fs_info *fs_info, ret = -ENOMEM; goto out; } + fs_info->scrub_parity_workers = + btrfs_alloc_workqueue("btrfs-scrubparity", flags, + max_active, 2); + if (!fs_info->scrub_parity_workers) { + ret = -ENOMEM; + goto out; + } } ++fs_info->scrub_workers_refcnt; out: @@ -3601,6 +3620,7 @@ static noinline_for_stack void scrub_workers_put(struct btrfs_fs_info *fs_info) btrfs_destroy_workqueue(fs_info->scrub_workers); btrfs_destroy_workqueue(fs_info->scrub_wr_completion_workers); btrfs_destroy_workqueue(fs_info->scrub_nocow_workers); + btrfs_destroy_workqueue(fs_info->scrub_parity_workers); } WARN_ON(fs_info->scrub_workers_refcnt < 0); } diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index a1216f9..aa72bfd 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -243,6 +243,7 @@ struct waiting_dir_move { * after this directory is moved, we can try to rmdir the ino rmdir_ino. */ u64 rmdir_ino; + bool orphanized; }; struct orphan_dir_info { @@ -1158,6 +1159,9 @@ struct backref_ctx { /* may be truncated in case it's the last extent in a file */ u64 extent_len; + /* data offset in the file extent item */ + u64 data_offset; + /* Just to check for bugs in backref resolving */ int found_itself; }; @@ -1221,7 +1225,7 @@ static int __iterate_backrefs(u64 ino, u64 offset, u64 root, void *ctx_) if (ret < 0) return ret; - if (offset + bctx->extent_len > i_size) + if (offset + bctx->data_offset + bctx->extent_len > i_size) return 0; /* @@ -1363,6 +1367,19 @@ static int find_extent_clone(struct send_ctx *sctx, backref_ctx->cur_offset = data_offset; backref_ctx->found_itself = 0; backref_ctx->extent_len = num_bytes; + /* + * For non-compressed extents iterate_extent_inodes() gives us extent + * offsets that already take into account the data offset, but not for + * compressed extents, since the offset is logical and not relative to + * the physical extent locations. We must take this into account to + * avoid sending clone offsets that go beyond the source file's size, + * which would result in the clone ioctl failing with -EINVAL on the + * receiving end. + */ + if (compressed == BTRFS_COMPRESS_NONE) + backref_ctx->data_offset = 0; + else + backref_ctx->data_offset = btrfs_file_extent_offset(eb, fi); /* * The last extent of a file may be too large due to page alignment. @@ -1900,8 +1917,13 @@ static int did_overwrite_ref(struct send_ctx *sctx, goto out; } - /* we know that it is or will be overwritten. check this now */ - if (ow_inode < sctx->send_progress) + /* + * We know that it is or will be overwritten. Check this now. + * The current inode being processed might have been the one that caused + * inode 'ino' to be orphanized, therefore ow_inode can actually be the + * same as sctx->send_progress. + */ + if (ow_inode <= sctx->send_progress) ret = 1; else ret = 0; @@ -2223,6 +2245,8 @@ static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen, fs_path_reset(dest); while (!stop && ino != BTRFS_FIRST_FREE_OBJECTID) { + struct waiting_dir_move *wdm; + fs_path_reset(name); if (is_waiting_for_rm(sctx, ino)) { @@ -2233,7 +2257,11 @@ static int get_cur_path(struct send_ctx *sctx, u64 ino, u64 gen, break; } - if (is_waiting_for_move(sctx, ino)) { + wdm = get_waiting_dir_move(sctx, ino); + if (wdm && wdm->orphanized) { + ret = gen_unique_name(sctx, ino, gen, name); + stop = 1; + } else if (wdm) { ret = get_first_ref(sctx->parent_root, ino, &parent_inode, &parent_gen, name); } else { @@ -2328,8 +2356,12 @@ static int send_subvol_begin(struct send_ctx *sctx) TLV_PUT_U64(sctx, BTRFS_SEND_A_CTRANSID, le64_to_cpu(sctx->send_root->root_item.ctransid)); if (parent_root) { - TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, - sctx->parent_root->root_item.uuid); + if (!btrfs_is_empty_uuid(parent_root->root_item.received_uuid)) + TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, + parent_root->root_item.received_uuid); + else + TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, + parent_root->root_item.uuid); TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID, le64_to_cpu(sctx->parent_root->root_item.ctransid)); } @@ -2923,7 +2955,7 @@ static int is_waiting_for_move(struct send_ctx *sctx, u64 ino) return entry != NULL; } -static int add_waiting_dir_move(struct send_ctx *sctx, u64 ino) +static int add_waiting_dir_move(struct send_ctx *sctx, u64 ino, bool orphanized) { struct rb_node **p = &sctx->waiting_dir_moves.rb_node; struct rb_node *parent = NULL; @@ -2934,6 +2966,7 @@ static int add_waiting_dir_move(struct send_ctx *sctx, u64 ino) return -ENOMEM; dm->ino = ino; dm->rmdir_ino = 0; + dm->orphanized = orphanized; while (*p) { parent = *p; @@ -3030,7 +3063,7 @@ static int add_pending_dir_move(struct send_ctx *sctx, goto out; } - ret = add_waiting_dir_move(sctx, pm->ino); + ret = add_waiting_dir_move(sctx, pm->ino, is_orphan); if (ret) goto out; @@ -3353,8 +3386,40 @@ out: return ret; } +/* + * Check if ino ino1 is an ancestor of inode ino2 in the given root. + * Return 1 if true, 0 if false and < 0 on error. + */ +static int is_ancestor(struct btrfs_root *root, + const u64 ino1, + const u64 ino1_gen, + const u64 ino2, + struct fs_path *fs_path) +{ + u64 ino = ino2; + + while (ino > BTRFS_FIRST_FREE_OBJECTID) { + int ret; + u64 parent; + u64 parent_gen; + + fs_path_reset(fs_path); + ret = get_first_ref(root, ino, &parent, &parent_gen, fs_path); + if (ret < 0) { + if (ret == -ENOENT && ino == ino2) + ret = 0; + return ret; + } + if (parent == ino1) + return parent_gen == ino1_gen ? 1 : 0; + ino = parent; + } + return 0; +} + static int wait_for_parent_move(struct send_ctx *sctx, - struct recorded_ref *parent_ref) + struct recorded_ref *parent_ref, + const bool is_orphan) { int ret = 0; u64 ino = parent_ref->dir; @@ -3374,11 +3439,24 @@ static int wait_for_parent_move(struct send_ctx *sctx, * Our current directory inode may not yet be renamed/moved because some * ancestor (immediate or not) has to be renamed/moved first. So find if * such ancestor exists and make sure our own rename/move happens after - * that ancestor is processed. + * that ancestor is processed to avoid path build infinite loops (done + * at get_cur_path()). */ while (ino > BTRFS_FIRST_FREE_OBJECTID) { if (is_waiting_for_move(sctx, ino)) { - ret = 1; + /* + * If the current inode is an ancestor of ino in the + * parent root, we need to delay the rename of the + * current inode, otherwise don't delayed the rename + * because we can end up with a circular dependency + * of renames, resulting in some directories never + * getting the respective rename operations issued in + * the send stream or getting into infinite path build + * loops. + */ + ret = is_ancestor(sctx->parent_root, + sctx->cur_ino, sctx->cur_inode_gen, + ino, path_before); break; } @@ -3420,7 +3498,7 @@ out: ino, &sctx->new_refs, &sctx->deleted_refs, - false); + is_orphan); if (!ret) ret = 1; } @@ -3589,6 +3667,17 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino); } } + if (S_ISDIR(sctx->cur_inode_mode) && sctx->parent_root && + can_rename) { + ret = wait_for_parent_move(sctx, cur, is_orphan); + if (ret < 0) + goto out; + if (ret == 1) { + can_rename = false; + *pending_move = 1; + } + } + /* * link/move the ref to the new place. If we have an orphan * inode, move it and update valid_path. If not, link or move @@ -3609,18 +3698,11 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino); * dirs, we always have one new and one deleted * ref. The deleted ref is ignored later. */ - ret = wait_for_parent_move(sctx, cur); - if (ret < 0) - goto out; - if (ret) { - *pending_move = 1; - } else { - ret = send_rename(sctx, valid_path, - cur->full_path); - if (!ret) - ret = fs_path_copy(valid_path, - cur->full_path); - } + ret = send_rename(sctx, valid_path, + cur->full_path); + if (!ret) + ret = fs_path_copy(valid_path, + cur->full_path); if (ret < 0) goto out; } else { @@ -4508,8 +4590,21 @@ verbose_printk("btrfs: send_clone offset=%llu, len=%d, clone_root=%llu, " if (ret < 0) goto out; - TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, - clone_root->root->root_item.uuid); + /* + * If the parent we're using has a received_uuid set then use that as + * our clone source as that is what we will look for when doing a + * receive. + * + * This covers the case that we create a snapshot off of a received + * subvolume and then use that as the parent and try to receive on a + * different host. + */ + if (!btrfs_is_empty_uuid(clone_root->root->root_item.received_uuid)) + TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, + clone_root->root->root_item.received_uuid); + else + TLV_PUT_UUID(sctx, BTRFS_SEND_A_CLONE_UUID, + clone_root->root->root_item.uuid); TLV_PUT_U64(sctx, BTRFS_SEND_A_CLONE_CTRANSID, le64_to_cpu(clone_root->root->root_item.ctransid)); TLV_PUT_PATH(sctx, BTRFS_SEND_A_CLONE_PATH, p); diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 9e66f5e..cd7ef34 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -135,6 +135,7 @@ static void btrfs_handle_error(struct btrfs_fs_info *fs_info) * __btrfs_std_error decodes expected errors from the caller and * invokes the approciate error response. */ +__cold void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, unsigned int line, int errno, const char *fmt, ...) { @@ -247,18 +248,11 @@ void __btrfs_std_error(struct btrfs_fs_info *fs_info, const char *function, * We'll complete the cleanup in btrfs_end_transaction and * btrfs_commit_transaction. */ +__cold void __btrfs_abort_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root, const char *function, unsigned int line, int errno) { - /* - * Report first abort since mount - */ - if (!test_and_set_bit(BTRFS_FS_STATE_TRANS_ABORTED, - &root->fs_info->fs_state)) { - WARN(1, KERN_DEBUG "BTRFS: Transaction aborted (error %d)\n", - errno); - } trans->aborted = errno; /* Nothing used. The other threads that have joined this * transaction may be able to continue. */ @@ -281,6 +275,7 @@ void __btrfs_abort_transaction(struct btrfs_trans_handle *trans, * __btrfs_panic decodes unexpected, fatal errors from the caller, * issues an alert, and either panics or BUGs, depending on mount options. */ +__cold void __btrfs_panic(struct btrfs_fs_info *fs_info, const char *function, unsigned int line, int errno, const char *fmt, ...) { @@ -841,33 +836,153 @@ out: return error; } -static struct dentry *get_default_root(struct super_block *sb, - u64 subvol_objectid) +static char *get_subvol_name_from_objectid(struct btrfs_fs_info *fs_info, + u64 subvol_objectid) { - struct btrfs_fs_info *fs_info = btrfs_sb(sb); struct btrfs_root *root = fs_info->tree_root; - struct btrfs_root *new_root; - struct btrfs_dir_item *di; - struct btrfs_path *path; - struct btrfs_key location; - struct inode *inode; - u64 dir_id; - int new = 0; + struct btrfs_root *fs_root; + struct btrfs_root_ref *root_ref; + struct btrfs_inode_ref *inode_ref; + struct btrfs_key key; + struct btrfs_path *path = NULL; + char *name = NULL, *ptr; + u64 dirid; + int len; + int ret; + + path = btrfs_alloc_path(); + if (!path) { + ret = -ENOMEM; + goto err; + } + path->leave_spinning = 1; + + name = kmalloc(PATH_MAX, GFP_NOFS); + if (!name) { + ret = -ENOMEM; + goto err; + } + ptr = name + PATH_MAX - 1; + ptr[0] = '\0'; /* - * We have a specific subvol we want to mount, just setup location and - * go look up the root. + * Walk up the subvolume trees in the tree of tree roots by root + * backrefs until we hit the top-level subvolume. */ - if (subvol_objectid) { - location.objectid = subvol_objectid; - location.type = BTRFS_ROOT_ITEM_KEY; - location.offset = (u64)-1; - goto find_root; + while (subvol_objectid != BTRFS_FS_TREE_OBJECTID) { + key.objectid = subvol_objectid; + key.type = BTRFS_ROOT_BACKREF_KEY; + key.offset = (u64)-1; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) { + goto err; + } else if (ret > 0) { + ret = btrfs_previous_item(root, path, subvol_objectid, + BTRFS_ROOT_BACKREF_KEY); + if (ret < 0) { + goto err; + } else if (ret > 0) { + ret = -ENOENT; + goto err; + } + } + + btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); + subvol_objectid = key.offset; + + root_ref = btrfs_item_ptr(path->nodes[0], path->slots[0], + struct btrfs_root_ref); + len = btrfs_root_ref_name_len(path->nodes[0], root_ref); + ptr -= len + 1; + if (ptr < name) { + ret = -ENAMETOOLONG; + goto err; + } + read_extent_buffer(path->nodes[0], ptr + 1, + (unsigned long)(root_ref + 1), len); + ptr[0] = '/'; + dirid = btrfs_root_ref_dirid(path->nodes[0], root_ref); + btrfs_release_path(path); + + key.objectid = subvol_objectid; + key.type = BTRFS_ROOT_ITEM_KEY; + key.offset = (u64)-1; + fs_root = btrfs_read_fs_root_no_name(fs_info, &key); + if (IS_ERR(fs_root)) { + ret = PTR_ERR(fs_root); + goto err; + } + + /* + * Walk up the filesystem tree by inode refs until we hit the + * root directory. + */ + while (dirid != BTRFS_FIRST_FREE_OBJECTID) { + key.objectid = dirid; + key.type = BTRFS_INODE_REF_KEY; + key.offset = (u64)-1; + + ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0); + if (ret < 0) { + goto err; + } else if (ret > 0) { + ret = btrfs_previous_item(fs_root, path, dirid, + BTRFS_INODE_REF_KEY); + if (ret < 0) { + goto err; + } else if (ret > 0) { + ret = -ENOENT; + goto err; + } + } + + btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); + dirid = key.offset; + + inode_ref = btrfs_item_ptr(path->nodes[0], + path->slots[0], + struct btrfs_inode_ref); + len = btrfs_inode_ref_name_len(path->nodes[0], + inode_ref); + ptr -= len + 1; + if (ptr < name) { + ret = -ENAMETOOLONG; + goto err; + } + read_extent_buffer(path->nodes[0], ptr + 1, + (unsigned long)(inode_ref + 1), len); + ptr[0] = '/'; + btrfs_release_path(path); + } } + btrfs_free_path(path); + if (ptr == name + PATH_MAX - 1) { + name[0] = '/'; + name[1] = '\0'; + } else { + memmove(name, ptr, name + PATH_MAX - ptr); + } + return name; + +err: + btrfs_free_path(path); + kfree(name); + return ERR_PTR(ret); +} + +static int get_default_subvol_objectid(struct btrfs_fs_info *fs_info, u64 *objectid) +{ + struct btrfs_root *root = fs_info->tree_root; + struct btrfs_dir_item *di; + struct btrfs_path *path; + struct btrfs_key location; + u64 dir_id; + path = btrfs_alloc_path(); if (!path) - return ERR_PTR(-ENOMEM); + return -ENOMEM; path->leave_spinning = 1; /* @@ -879,58 +994,23 @@ static struct dentry *get_default_root(struct super_block *sb, di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0); if (IS_ERR(di)) { btrfs_free_path(path); - return ERR_CAST(di); + return PTR_ERR(di); } if (!di) { /* * Ok the default dir item isn't there. This is weird since * it's always been there, but don't freak out, just try and - * mount to root most subvolume. + * mount the top-level subvolume. */ btrfs_free_path(path); - dir_id = BTRFS_FIRST_FREE_OBJECTID; - new_root = fs_info->fs_root; - goto setup_root; + *objectid = BTRFS_FS_TREE_OBJECTID; + return 0; } btrfs_dir_item_key_to_cpu(path->nodes[0], di, &location); btrfs_free_path(path); - -find_root: - new_root = btrfs_read_fs_root_no_name(fs_info, &location); - if (IS_ERR(new_root)) - return ERR_CAST(new_root); - - if (!(sb->s_flags & MS_RDONLY)) { - int ret; - down_read(&fs_info->cleanup_work_sem); - ret = btrfs_orphan_cleanup(new_root); - up_read(&fs_info->cleanup_work_sem); - if (ret) - return ERR_PTR(ret); - } - - dir_id = btrfs_root_dirid(&new_root->root_item); -setup_root: - location.objectid = dir_id; - location.type = BTRFS_INODE_ITEM_KEY; - location.offset = 0; - - inode = btrfs_iget(sb, &location, new_root, &new); - if (IS_ERR(inode)) - return ERR_CAST(inode); - - /* - * If we're just mounting the root most subvol put the inode and return - * a reference to the dentry. We will have already gotten a reference - * to the inode in btrfs_fill_super so we're good to go. - */ - if (!new && d_inode(sb->s_root) == inode) { - iput(inode); - return dget(sb->s_root); - } - - return d_obtain_root(inode); + *objectid = location.objectid; + return 0; } static int btrfs_fill_super(struct super_block *sb, @@ -1108,6 +1188,10 @@ static int btrfs_show_options(struct seq_file *seq, struct dentry *dentry) seq_puts(seq, ",fatal_errors=panic"); if (info->commit_interval != BTRFS_DEFAULT_COMMIT_INTERVAL) seq_printf(seq, ",commit=%d", info->commit_interval); + seq_printf(seq, ",subvolid=%llu", + BTRFS_I(d_inode(dentry))->root->root_key.objectid); + seq_puts(seq, ",subvol="); + seq_dentry(seq, dentry, " \t\n\\"); return 0; } @@ -1138,107 +1222,139 @@ static inline int is_subvolume_inode(struct inode *inode) } /* - * This will strip out the subvol=%s argument for an argument string and add - * subvolid=0 to make sure we get the actual tree root for path walking to the - * subvol we want. + * This will add subvolid=0 to the argument string while removing any subvol= + * and subvolid= arguments to make sure we get the top-level root for path + * walking to the subvol we want. */ static char *setup_root_args(char *args) { - unsigned len = strlen(args) + 2 + 1; - char *src, *dst, *buf; + char *buf, *dst, *sep; - /* - * We need the same args as before, but with this substitution: - * s!subvol=[^,]+!subvolid=0! - * - * Since the replacement string is up to 2 bytes longer than the - * original, allocate strlen(args) + 2 + 1 bytes. - */ + if (!args) + return kstrdup("subvolid=0", GFP_NOFS); - src = strstr(args, "subvol="); - /* This shouldn't happen, but just in case.. */ - if (!src) - return NULL; - - buf = dst = kmalloc(len, GFP_NOFS); + /* The worst case is that we add ",subvolid=0" to the end. */ + buf = dst = kmalloc(strlen(args) + strlen(",subvolid=0") + 1, GFP_NOFS); if (!buf) return NULL; - /* - * If the subvol= arg is not at the start of the string, - * copy whatever precedes it into buf. - */ - if (src != args) { - *src++ = '\0'; - strcpy(buf, args); - dst += strlen(args); + while (1) { + sep = strchrnul(args, ','); + if (!strstarts(args, "subvol=") && + !strstarts(args, "subvolid=")) { + memcpy(dst, args, sep - args); + dst += sep - args; + *dst++ = ','; + } + if (*sep) + args = sep + 1; + else + break; } - strcpy(dst, "subvolid=0"); - dst += strlen("subvolid=0"); - - /* - * If there is a "," after the original subvol=... string, - * copy that suffix into our buffer. Otherwise, we're done. - */ - src = strchr(src, ','); - if (src) - strcpy(dst, src); return buf; } -static struct dentry *mount_subvol(const char *subvol_name, int flags, - const char *device_name, char *data) +static struct dentry *mount_subvol(const char *subvol_name, u64 subvol_objectid, + int flags, const char *device_name, + char *data) { struct dentry *root; - struct vfsmount *mnt; + struct vfsmount *mnt = NULL; char *newargs; + int ret; newargs = setup_root_args(data); - if (!newargs) - return ERR_PTR(-ENOMEM); - mnt = vfs_kern_mount(&btrfs_fs_type, flags, device_name, - newargs); + if (!newargs) { + root = ERR_PTR(-ENOMEM); + goto out; + } - if (PTR_RET(mnt) == -EBUSY) { + mnt = vfs_kern_mount(&btrfs_fs_type, flags, device_name, newargs); + if (PTR_ERR_OR_ZERO(mnt) == -EBUSY) { if (flags & MS_RDONLY) { - mnt = vfs_kern_mount(&btrfs_fs_type, flags & ~MS_RDONLY, device_name, - newargs); + mnt = vfs_kern_mount(&btrfs_fs_type, flags & ~MS_RDONLY, + device_name, newargs); } else { - int r; - mnt = vfs_kern_mount(&btrfs_fs_type, flags | MS_RDONLY, device_name, - newargs); + mnt = vfs_kern_mount(&btrfs_fs_type, flags | MS_RDONLY, + device_name, newargs); if (IS_ERR(mnt)) { - kfree(newargs); - return ERR_CAST(mnt); + root = ERR_CAST(mnt); + mnt = NULL; + goto out; } - r = btrfs_remount(mnt->mnt_sb, &flags, NULL); - if (r < 0) { - /* FIXME: release vfsmount mnt ??*/ - kfree(newargs); - return ERR_PTR(r); + down_write(&mnt->mnt_sb->s_umount); + ret = btrfs_remount(mnt->mnt_sb, &flags, NULL); + up_write(&mnt->mnt_sb->s_umount); + if (ret < 0) { + root = ERR_PTR(ret); + goto out; } } } + if (IS_ERR(mnt)) { + root = ERR_CAST(mnt); + mnt = NULL; + goto out; + } - kfree(newargs); + if (!subvol_name) { + if (!subvol_objectid) { + ret = get_default_subvol_objectid(btrfs_sb(mnt->mnt_sb), + &subvol_objectid); + if (ret) { + root = ERR_PTR(ret); + goto out; + } + } + subvol_name = get_subvol_name_from_objectid(btrfs_sb(mnt->mnt_sb), + subvol_objectid); + if (IS_ERR(subvol_name)) { + root = ERR_CAST(subvol_name); + subvol_name = NULL; + goto out; + } - if (IS_ERR(mnt)) - return ERR_CAST(mnt); + } root = mount_subtree(mnt, subvol_name); + /* mount_subtree() drops our reference on the vfsmount. */ + mnt = NULL; - if (!IS_ERR(root) && !is_subvolume_inode(d_inode(root))) { + if (!IS_ERR(root)) { struct super_block *s = root->d_sb; - dput(root); - root = ERR_PTR(-EINVAL); - deactivate_locked_super(s); - printk(KERN_ERR "BTRFS: '%s' is not a valid subvolume\n", - subvol_name); + struct inode *root_inode = d_inode(root); + u64 root_objectid = BTRFS_I(root_inode)->root->root_key.objectid; + + ret = 0; + if (!is_subvolume_inode(root_inode)) { + pr_err("BTRFS: '%s' is not a valid subvolume\n", + subvol_name); + ret = -EINVAL; + } + if (subvol_objectid && root_objectid != subvol_objectid) { + /* + * This will also catch a race condition where a + * subvolume which was passed by ID is renamed and + * another subvolume is renamed over the old location. + */ + pr_err("BTRFS: subvol '%s' does not match subvolid %llu\n", + subvol_name, subvol_objectid); + ret = -EINVAL; + } + if (ret) { + dput(root); + root = ERR_PTR(ret); + deactivate_locked_super(s); + } } +out: + mntput(mnt); + kfree(newargs); + kfree(subvol_name); return root; } @@ -1303,7 +1419,6 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, { struct block_device *bdev = NULL; struct super_block *s; - struct dentry *root; struct btrfs_fs_devices *fs_devices = NULL; struct btrfs_fs_info *fs_info = NULL; struct security_mnt_opts new_sec_opts; @@ -1323,10 +1438,10 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, return ERR_PTR(error); } - if (subvol_name) { - root = mount_subvol(subvol_name, flags, device_name, data); - kfree(subvol_name); - return root; + if (subvol_name || subvol_objectid != BTRFS_FS_TREE_OBJECTID) { + /* mount_subvol() will free subvol_name. */ + return mount_subvol(subvol_name, subvol_objectid, flags, + device_name, data); } security_init_mnt_opts(&new_sec_opts); @@ -1392,23 +1507,19 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, error = btrfs_fill_super(s, fs_devices, data, flags & MS_SILENT ? 1 : 0); } - - root = !error ? get_default_root(s, subvol_objectid) : ERR_PTR(error); - if (IS_ERR(root)) { + if (error) { deactivate_locked_super(s); - error = PTR_ERR(root); goto error_sec_opts; } fs_info = btrfs_sb(s); error = setup_security_options(fs_info, s, &new_sec_opts); if (error) { - dput(root); deactivate_locked_super(s); goto error_sec_opts; } - return root; + return dget(s->s_root); error_close_devices: btrfs_close_devices(fs_devices); diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c index e8a4c86..603b0cc 100644 --- a/fs/btrfs/sysfs.c +++ b/fs/btrfs/sysfs.c @@ -33,6 +33,7 @@ #include "volumes.h" static inline struct btrfs_fs_info *to_fs_info(struct kobject *kobj); +static inline struct btrfs_fs_devices *to_fs_devs(struct kobject *kobj); static u64 get_features(struct btrfs_fs_info *fs_info, enum btrfs_feature_set set) @@ -428,7 +429,7 @@ static ssize_t btrfs_clone_alignment_show(struct kobject *kobj, BTRFS_ATTR(clone_alignment, btrfs_clone_alignment_show); -static struct attribute *btrfs_attrs[] = { +static const struct attribute *btrfs_attrs[] = { BTRFS_ATTR_PTR(label), BTRFS_ATTR_PTR(nodesize), BTRFS_ATTR_PTR(sectorsize), @@ -438,21 +439,29 @@ static struct attribute *btrfs_attrs[] = { static void btrfs_release_super_kobj(struct kobject *kobj) { - struct btrfs_fs_info *fs_info = to_fs_info(kobj); - complete(&fs_info->kobj_unregister); + struct btrfs_fs_devices *fs_devs = to_fs_devs(kobj); + + memset(&fs_devs->super_kobj, 0, sizeof(struct kobject)); + complete(&fs_devs->kobj_unregister); } static struct kobj_type btrfs_ktype = { .sysfs_ops = &kobj_sysfs_ops, .release = btrfs_release_super_kobj, - .default_attrs = btrfs_attrs, }; +static inline struct btrfs_fs_devices *to_fs_devs(struct kobject *kobj) +{ + if (kobj->ktype != &btrfs_ktype) + return NULL; + return container_of(kobj, struct btrfs_fs_devices, super_kobj); +} + static inline struct btrfs_fs_info *to_fs_info(struct kobject *kobj) { if (kobj->ktype != &btrfs_ktype) return NULL; - return container_of(kobj, struct btrfs_fs_info, super_kobj); + return to_fs_devs(kobj)->fs_info; } #define NUM_FEATURE_BITS 64 @@ -493,12 +502,12 @@ static int addrm_unknown_feature_attrs(struct btrfs_fs_info *fs_info, bool add) attrs[0] = &fa->kobj_attr.attr; if (add) { int ret; - ret = sysfs_merge_group(&fs_info->super_kobj, + ret = sysfs_merge_group(&fs_info->fs_devices->super_kobj, &agroup); if (ret) return ret; } else - sysfs_unmerge_group(&fs_info->super_kobj, + sysfs_unmerge_group(&fs_info->fs_devices->super_kobj, &agroup); } @@ -506,25 +515,49 @@ static int addrm_unknown_feature_attrs(struct btrfs_fs_info *fs_info, bool add) return 0; } -static void __btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info) +static void __btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs) +{ + if (fs_devs->device_dir_kobj) { + kobject_del(fs_devs->device_dir_kobj); + kobject_put(fs_devs->device_dir_kobj); + fs_devs->device_dir_kobj = NULL; + } + + if (fs_devs->super_kobj.state_initialized) { + kobject_del(&fs_devs->super_kobj); + kobject_put(&fs_devs->super_kobj); + wait_for_completion(&fs_devs->kobj_unregister); + } +} + +/* when fs_devs is NULL it will remove all fsid kobject */ +void btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs) { - kobject_del(&fs_info->super_kobj); - kobject_put(&fs_info->super_kobj); - wait_for_completion(&fs_info->kobj_unregister); + struct list_head *fs_uuids = btrfs_get_fs_uuids(); + + if (fs_devs) { + __btrfs_sysfs_remove_fsid(fs_devs); + return; + } + + list_for_each_entry(fs_devs, fs_uuids, list) { + __btrfs_sysfs_remove_fsid(fs_devs); + } } void btrfs_sysfs_remove_one(struct btrfs_fs_info *fs_info) { + btrfs_reset_fs_info_ptr(fs_info); + if (fs_info->space_info_kobj) { sysfs_remove_files(fs_info->space_info_kobj, allocation_attrs); kobject_del(fs_info->space_info_kobj); kobject_put(fs_info->space_info_kobj); } - kobject_del(fs_info->device_dir_kobj); - kobject_put(fs_info->device_dir_kobj); addrm_unknown_feature_attrs(fs_info, false); - sysfs_remove_group(&fs_info->super_kobj, &btrfs_feature_attr_group); - __btrfs_sysfs_remove_one(fs_info); + sysfs_remove_group(&fs_info->fs_devices->super_kobj, &btrfs_feature_attr_group); + sysfs_remove_files(&fs_info->fs_devices->super_kobj, btrfs_attrs); + btrfs_kobj_rm_device(fs_info->fs_devices, NULL); } const char * const btrfs_feature_set_names[3] = { @@ -602,40 +635,60 @@ static void init_feature_attrs(void) } } -int btrfs_kobj_rm_device(struct btrfs_fs_info *fs_info, +/* when one_device is NULL, it removes all device links */ + +int btrfs_kobj_rm_device(struct btrfs_fs_devices *fs_devices, struct btrfs_device *one_device) { struct hd_struct *disk; struct kobject *disk_kobj; - if (!fs_info->device_dir_kobj) + if (!fs_devices->device_dir_kobj) return -EINVAL; if (one_device && one_device->bdev) { disk = one_device->bdev->bd_part; disk_kobj = &part_to_dev(disk)->kobj; - sysfs_remove_link(fs_info->device_dir_kobj, + sysfs_remove_link(fs_devices->device_dir_kobj, + disk_kobj->name); + } + + if (one_device) + return 0; + + list_for_each_entry(one_device, + &fs_devices->devices, dev_list) { + if (!one_device->bdev) + continue; + disk = one_device->bdev->bd_part; + disk_kobj = &part_to_dev(disk)->kobj; + + sysfs_remove_link(fs_devices->device_dir_kobj, disk_kobj->name); } return 0; } -int btrfs_kobj_add_device(struct btrfs_fs_info *fs_info, - struct btrfs_device *one_device) +int btrfs_sysfs_add_device(struct btrfs_fs_devices *fs_devs) { - int error = 0; - struct btrfs_fs_devices *fs_devices = fs_info->fs_devices; - struct btrfs_device *dev; - - if (!fs_info->device_dir_kobj) - fs_info->device_dir_kobj = kobject_create_and_add("devices", - &fs_info->super_kobj); + if (!fs_devs->device_dir_kobj) + fs_devs->device_dir_kobj = kobject_create_and_add("devices", + &fs_devs->super_kobj); - if (!fs_info->device_dir_kobj) + if (!fs_devs->device_dir_kobj) return -ENOMEM; + return 0; +} + +int btrfs_kobj_add_device(struct btrfs_fs_devices *fs_devices, + struct btrfs_device *one_device) +{ + int error = 0; + struct btrfs_device *dev; + list_for_each_entry(dev, &fs_devices->devices, dev_list) { struct hd_struct *disk; struct kobject *disk_kobj; @@ -649,7 +702,7 @@ int btrfs_kobj_add_device(struct btrfs_fs_info *fs_info, disk = dev->bdev->bd_part; disk_kobj = &part_to_dev(disk)->kobj; - error = sysfs_create_link(fs_info->device_dir_kobj, + error = sysfs_create_link(fs_devices->device_dir_kobj, disk_kobj, disk_kobj->name); if (error) break; @@ -667,34 +720,51 @@ static struct dentry *btrfs_debugfs_root_dentry; /* Debugging tunables and exported data */ u64 btrfs_debugfs_test; +/* + * Can be called by the device discovery thread. + * And parent can be specified for seed device + */ +int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs, + struct kobject *parent) +{ + int error; + + init_completion(&fs_devs->kobj_unregister); + fs_devs->super_kobj.kset = btrfs_kset; + error = kobject_init_and_add(&fs_devs->super_kobj, + &btrfs_ktype, parent, "%pU", fs_devs->fsid); + return error; +} + int btrfs_sysfs_add_one(struct btrfs_fs_info *fs_info) { int error; + struct btrfs_fs_devices *fs_devs = fs_info->fs_devices; + struct kobject *super_kobj = &fs_devs->super_kobj; + + btrfs_set_fs_info_ptr(fs_info); - init_completion(&fs_info->kobj_unregister); - fs_info->super_kobj.kset = btrfs_kset; - error = kobject_init_and_add(&fs_info->super_kobj, &btrfs_ktype, NULL, - "%pU", fs_info->fsid); + error = btrfs_kobj_add_device(fs_devs, NULL); if (error) return error; - error = sysfs_create_group(&fs_info->super_kobj, - &btrfs_feature_attr_group); + error = sysfs_create_files(super_kobj, btrfs_attrs); if (error) { - __btrfs_sysfs_remove_one(fs_info); + btrfs_kobj_rm_device(fs_devs, NULL); return error; } - error = addrm_unknown_feature_attrs(fs_info, true); + error = sysfs_create_group(super_kobj, + &btrfs_feature_attr_group); if (error) goto failure; - error = btrfs_kobj_add_device(fs_info, NULL); + error = addrm_unknown_feature_attrs(fs_info, true); if (error) goto failure; fs_info->space_info_kobj = kobject_create_and_add("allocation", - &fs_info->super_kobj); + super_kobj); if (!fs_info->space_info_kobj) { error = -ENOMEM; goto failure; diff --git a/fs/btrfs/sysfs.h b/fs/btrfs/sysfs.h index 3a4bbed..6392527 100644 --- a/fs/btrfs/sysfs.h +++ b/fs/btrfs/sysfs.h @@ -82,8 +82,12 @@ char *btrfs_printable_features(enum btrfs_feature_set set, u64 flags); extern const char * const btrfs_feature_set_names[3]; extern struct kobj_type space_info_ktype; extern struct kobj_type btrfs_raid_ktype; -int btrfs_kobj_add_device(struct btrfs_fs_info *fs_info, +int btrfs_kobj_add_device(struct btrfs_fs_devices *fs_devices, struct btrfs_device *one_device); -int btrfs_kobj_rm_device(struct btrfs_fs_info *fs_info, +int btrfs_kobj_rm_device(struct btrfs_fs_devices *fs_devices, struct btrfs_device *one_device); +int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs, + struct kobject *parent); +int btrfs_sysfs_add_device(struct btrfs_fs_devices *fs_devs); +void btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs); #endif /* _BTRFS_SYSFS_H_ */ diff --git a/fs/btrfs/tests/qgroup-tests.c b/fs/btrfs/tests/qgroup-tests.c index c32a7ba..846d277 100644 --- a/fs/btrfs/tests/qgroup-tests.c +++ b/fs/btrfs/tests/qgroup-tests.c @@ -21,6 +21,7 @@ #include "../transaction.h" #include "../disk-io.h" #include "../qgroup.h" +#include "../backref.h" static void init_dummy_trans(struct btrfs_trans_handle *trans) { @@ -227,6 +228,8 @@ static int test_no_shared_qgroup(struct btrfs_root *root) { struct btrfs_trans_handle trans; struct btrfs_fs_info *fs_info = root->fs_info; + struct ulist *old_roots = NULL; + struct ulist *new_roots = NULL; int ret; init_dummy_trans(&trans); @@ -238,10 +241,15 @@ static int test_no_shared_qgroup(struct btrfs_root *root) return ret; } - ret = btrfs_qgroup_record_ref(&trans, fs_info, 5, 4096, 4096, - BTRFS_QGROUP_OPER_ADD_EXCL, 0); + /* + * Since the test trans doesn't havee the complicated delayed refs, + * we can only call btrfs_qgroup_account_extent() directly to test + * quota. + */ + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &old_roots); if (ret) { - test_msg("Couldn't add space to a qgroup %d\n", ret); + ulist_free(old_roots); + test_msg("Couldn't find old roots: %d\n", ret); return ret; } @@ -249,9 +257,18 @@ static int test_no_shared_qgroup(struct btrfs_root *root) if (ret) return ret; - ret = btrfs_delayed_qgroup_accounting(&trans, fs_info); + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &new_roots); + if (ret) { + ulist_free(old_roots); + ulist_free(new_roots); + test_msg("Couldn't find old roots: %d\n", ret); + return ret; + } + + ret = btrfs_qgroup_account_extent(&trans, fs_info, 4096, 4096, + old_roots, new_roots); if (ret) { - test_msg("Delayed qgroup accounting failed %d\n", ret); + test_msg("Couldn't account space for a qgroup %d\n", ret); return ret; } @@ -259,21 +276,32 @@ static int test_no_shared_qgroup(struct btrfs_root *root) test_msg("Qgroup counts didn't match expected values\n"); return -EINVAL; } + old_roots = NULL; + new_roots = NULL; + + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &old_roots); + if (ret) { + ulist_free(old_roots); + test_msg("Couldn't find old roots: %d\n", ret); + return ret; + } ret = remove_extent_item(root, 4096, 4096); if (ret) return -EINVAL; - ret = btrfs_qgroup_record_ref(&trans, fs_info, 5, 4096, 4096, - BTRFS_QGROUP_OPER_SUB_EXCL, 0); + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &new_roots); if (ret) { - test_msg("Couldn't remove space from the qgroup %d\n", ret); - return -EINVAL; + ulist_free(old_roots); + ulist_free(new_roots); + test_msg("Couldn't find old roots: %d\n", ret); + return ret; } - ret = btrfs_delayed_qgroup_accounting(&trans, fs_info); + ret = btrfs_qgroup_account_extent(&trans, fs_info, 4096, 4096, + old_roots, new_roots); if (ret) { - test_msg("Qgroup accounting failed %d\n", ret); + test_msg("Couldn't account space for a qgroup %d\n", ret); return -EINVAL; } @@ -294,6 +322,8 @@ static int test_multiple_refs(struct btrfs_root *root) { struct btrfs_trans_handle trans; struct btrfs_fs_info *fs_info = root->fs_info; + struct ulist *old_roots = NULL; + struct ulist *new_roots = NULL; int ret; init_dummy_trans(&trans); @@ -307,20 +337,29 @@ static int test_multiple_refs(struct btrfs_root *root) return ret; } + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &old_roots); + if (ret) { + ulist_free(old_roots); + test_msg("Couldn't find old roots: %d\n", ret); + return ret; + } + ret = insert_normal_tree_ref(root, 4096, 4096, 0, 5); if (ret) return ret; - ret = btrfs_qgroup_record_ref(&trans, fs_info, 5, 4096, 4096, - BTRFS_QGROUP_OPER_ADD_EXCL, 0); + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &new_roots); if (ret) { - test_msg("Couldn't add space to a qgroup %d\n", ret); + ulist_free(old_roots); + ulist_free(new_roots); + test_msg("Couldn't find old roots: %d\n", ret); return ret; } - ret = btrfs_delayed_qgroup_accounting(&trans, fs_info); + ret = btrfs_qgroup_account_extent(&trans, fs_info, 4096, 4096, + old_roots, new_roots); if (ret) { - test_msg("Delayed qgroup accounting failed %d\n", ret); + test_msg("Couldn't account space for a qgroup %d\n", ret); return ret; } @@ -329,20 +368,29 @@ static int test_multiple_refs(struct btrfs_root *root) return -EINVAL; } + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &old_roots); + if (ret) { + ulist_free(old_roots); + test_msg("Couldn't find old roots: %d\n", ret); + return ret; + } + ret = add_tree_ref(root, 4096, 4096, 0, 256); if (ret) return ret; - ret = btrfs_qgroup_record_ref(&trans, fs_info, 256, 4096, 4096, - BTRFS_QGROUP_OPER_ADD_SHARED, 0); + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &new_roots); if (ret) { - test_msg("Qgroup record ref failed %d\n", ret); + ulist_free(old_roots); + ulist_free(new_roots); + test_msg("Couldn't find old roots: %d\n", ret); return ret; } - ret = btrfs_delayed_qgroup_accounting(&trans, fs_info); + ret = btrfs_qgroup_account_extent(&trans, fs_info, 4096, 4096, + old_roots, new_roots); if (ret) { - test_msg("Qgroup accounting failed %d\n", ret); + test_msg("Couldn't account space for a qgroup %d\n", ret); return ret; } @@ -356,20 +404,29 @@ static int test_multiple_refs(struct btrfs_root *root) return -EINVAL; } + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &old_roots); + if (ret) { + ulist_free(old_roots); + test_msg("Couldn't find old roots: %d\n", ret); + return ret; + } + ret = remove_extent_ref(root, 4096, 4096, 0, 256); if (ret) return ret; - ret = btrfs_qgroup_record_ref(&trans, fs_info, 256, 4096, 4096, - BTRFS_QGROUP_OPER_SUB_SHARED, 0); + ret = btrfs_find_all_roots(&trans, fs_info, 4096, 0, &new_roots); if (ret) { - test_msg("Qgroup record ref failed %d\n", ret); + ulist_free(old_roots); + ulist_free(new_roots); + test_msg("Couldn't find old roots: %d\n", ret); return ret; } - ret = btrfs_delayed_qgroup_accounting(&trans, fs_info); + ret = btrfs_qgroup_account_extent(&trans, fs_info, 4096, 4096, + old_roots, new_roots); if (ret) { - test_msg("Qgroup accounting failed %d\n", ret); + test_msg("Couldn't account space for a qgroup %d\n", ret); return ret; } diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 5628e25..c0f18e7 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -225,12 +225,14 @@ loop: cur_trans->dirty_bg_run = 0; cur_trans->delayed_refs.href_root = RB_ROOT; + cur_trans->delayed_refs.dirty_extent_root = RB_ROOT; atomic_set(&cur_trans->delayed_refs.num_entries, 0); cur_trans->delayed_refs.num_heads_ready = 0; cur_trans->delayed_refs.pending_csums = 0; cur_trans->delayed_refs.num_heads = 0; cur_trans->delayed_refs.flushing = 0; cur_trans->delayed_refs.run_delayed_start = 0; + cur_trans->delayed_refs.qgroup_to_skip = 0; /* * although the tree mod log is per file system and not per transaction, @@ -509,6 +511,7 @@ again: h->transaction = cur_trans; h->blocks_used = 0; h->bytes_reserved = 0; + h->chunk_bytes_reserved = 0; h->root = root; h->delayed_ref_updates = 0; h->use_count = 1; @@ -792,6 +795,8 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, if (!list_empty(&trans->new_bgs)) btrfs_create_pending_block_groups(trans, root); + btrfs_trans_release_chunk_metadata(trans); + if (lock && !atomic_read(&root->fs_info->open_ioctl_trans) && should_end_transaction(trans, root) && ACCESS_ONCE(cur_trans->state) == TRANS_STATE_RUNNING) { @@ -1290,6 +1295,12 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, if (pending->error) goto no_free_objectid; + /* + * Make qgroup to skip current new snapshot's qgroupid, as it is + * accounted by later btrfs_qgroup_inherit(). + */ + btrfs_set_skip_qgroup(trans, objectid); + btrfs_reloc_pre_snapshot(trans, pending, &to_reserve); if (to_reserve > 0) { @@ -1298,7 +1309,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, to_reserve, BTRFS_RESERVE_NO_FLUSH); if (pending->error) - goto no_free_objectid; + goto clear_skip_qgroup; } key.objectid = objectid; @@ -1396,25 +1407,6 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, btrfs_abort_transaction(trans, root, ret); goto fail; } - - /* - * We need to flush delayed refs in order to make sure all of our quota - * operations have been done before we call btrfs_qgroup_inherit. - */ - ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); - if (ret) { - btrfs_abort_transaction(trans, root, ret); - goto fail; - } - - ret = btrfs_qgroup_inherit(trans, fs_info, - root->root_key.objectid, - objectid, pending->inherit); - if (ret) { - btrfs_abort_transaction(trans, root, ret); - goto fail; - } - /* see comments in should_cow_block() */ set_bit(BTRFS_ROOT_FORCE_COW, &root->state); smp_wmb(); @@ -1497,11 +1489,37 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, goto fail; } } + + ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); + if (ret) { + btrfs_abort_transaction(trans, root, ret); + goto fail; + } + + /* + * account qgroup counters before qgroup_inherit() + */ + ret = btrfs_qgroup_prepare_account_extents(trans, fs_info); + if (ret) + goto fail; + ret = btrfs_qgroup_account_extents(trans, fs_info); + if (ret) + goto fail; + ret = btrfs_qgroup_inherit(trans, fs_info, + root->root_key.objectid, + objectid, pending->inherit); + if (ret) { + btrfs_abort_transaction(trans, root, ret); + goto fail; + } + fail: pending->error = ret; dir_item_existed: trans->block_rsv = rsv; trans->bytes_reserved = 0; +clear_skip_qgroup: + btrfs_clear_skip_qgroup(trans); no_free_objectid: kfree(new_root_item); root_item_alloc_fail: @@ -1963,6 +1981,13 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, goto scrub_continue; } + /* Reocrd old roots for later qgroup accounting */ + ret = btrfs_qgroup_prepare_account_extents(trans, root->fs_info); + if (ret) { + mutex_unlock(&root->fs_info->reloc_mutex); + goto scrub_continue; + } + /* * make sure none of the code above managed to slip in a * delayed item @@ -2004,6 +2029,17 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, */ btrfs_free_log_root_tree(trans, root->fs_info); + /* + * Since fs roots are all committed, we can get a quite accurate + * new_roots. So let's do quota accounting. + */ + ret = btrfs_qgroup_account_extents(trans, root->fs_info); + if (ret < 0) { + mutex_unlock(&root->fs_info->tree_log_mutex); + mutex_unlock(&root->fs_info->reloc_mutex); + goto scrub_continue; + } + ret = commit_cowonly_roots(trans, root); if (ret) { mutex_unlock(&root->fs_info->tree_log_mutex); @@ -2054,6 +2090,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, clear_bit(BTRFS_INODE_BTREE_LOG1_ERR, &btree_ino->runtime_flags); clear_bit(BTRFS_INODE_BTREE_LOG2_ERR, &btree_ino->runtime_flags); + btrfs_trans_release_chunk_metadata(trans); + spin_lock(&root->fs_info->trans_lock); cur_trans->state = TRANS_STATE_UNBLOCKED; root->fs_info->running_transaction = NULL; @@ -2123,6 +2161,7 @@ scrub_continue: btrfs_scrub_continue(root); cleanup_transaction: btrfs_trans_release_metadata(trans, root); + btrfs_trans_release_chunk_metadata(trans); trans->block_rsv = NULL; if (trans->qgroup_reserved) { btrfs_qgroup_free(root, trans->qgroup_reserved); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 0b24755..eb09c20 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -102,6 +102,7 @@ struct btrfs_transaction { struct btrfs_trans_handle { u64 transid; u64 bytes_reserved; + u64 chunk_bytes_reserved; u64 qgroup_reserved; unsigned long use_count; unsigned long blocks_reserved; @@ -153,6 +154,29 @@ static inline void btrfs_set_inode_last_trans(struct btrfs_trans_handle *trans, spin_unlock(&BTRFS_I(inode)->lock); } +/* + * Make qgroup codes to skip given qgroupid, means the old/new_roots for + * qgroup won't contain the qgroupid in it. + */ +static inline void btrfs_set_skip_qgroup(struct btrfs_trans_handle *trans, + u64 qgroupid) +{ + struct btrfs_delayed_ref_root *delayed_refs; + + delayed_refs = &trans->transaction->delayed_refs; + WARN_ON(delayed_refs->qgroup_to_skip); + delayed_refs->qgroup_to_skip = qgroupid; +} + +static inline void btrfs_clear_skip_qgroup(struct btrfs_trans_handle *trans) +{ + struct btrfs_delayed_ref_root *delayed_refs; + + delayed_refs = &trans->transaction->delayed_refs; + WARN_ON(!delayed_refs->qgroup_to_skip); + delayed_refs->qgroup_to_skip = 0; +} + int btrfs_end_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root); struct btrfs_trans_handle *btrfs_start_transaction(struct btrfs_root *root, diff --git a/fs/btrfs/tree-defrag.c b/fs/btrfs/tree-defrag.c index a63719c..a4b9c8b 100644 --- a/fs/btrfs/tree-defrag.c +++ b/fs/btrfs/tree-defrag.c @@ -52,9 +52,6 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, if (!test_bit(BTRFS_ROOT_REF_COWS, &root->state)) goto out; - if (btrfs_test_opt(root, SSD)) - goto out; - path = btrfs_alloc_path(); if (!path) return -ENOMEM; diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index d049683..1ce80c1 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -3881,12 +3881,6 @@ static int wait_ordered_extents(struct btrfs_trans_handle *trans, &ordered->flags)) continue; - if (ordered->csum_bytes_left) { - btrfs_start_ordered_extent(inode, ordered, 0); - wait_event(ordered->wait, - ordered->csum_bytes_left == 0); - } - list_for_each_entry(sum, &ordered->list, list) { ret = btrfs_csum_file_blocks(trans, log, sum); if (ret) diff --git a/fs/btrfs/ulist.c b/fs/btrfs/ulist.c index 840a38b..91feb2b 100644 --- a/fs/btrfs/ulist.c +++ b/fs/btrfs/ulist.c @@ -132,6 +132,15 @@ static struct ulist_node *ulist_rbtree_search(struct ulist *ulist, u64 val) return NULL; } +static void ulist_rbtree_erase(struct ulist *ulist, struct ulist_node *node) +{ + rb_erase(&node->rb_node, &ulist->root); + list_del(&node->list); + kfree(node); + BUG_ON(ulist->nnodes == 0); + ulist->nnodes--; +} + static int ulist_rbtree_insert(struct ulist *ulist, struct ulist_node *ins) { struct rb_node **p = &ulist->root.rb_node; @@ -197,9 +206,6 @@ int ulist_add_merge(struct ulist *ulist, u64 val, u64 aux, node->val = val; node->aux = aux; -#ifdef CONFIG_BTRFS_DEBUG - node->seqnum = ulist->nnodes; -#endif ret = ulist_rbtree_insert(ulist, node); ASSERT(!ret); @@ -209,6 +215,33 @@ int ulist_add_merge(struct ulist *ulist, u64 val, u64 aux, return 1; } +/* + * ulist_del - delete one node from ulist + * @ulist: ulist to remove node from + * @val: value to delete + * @aux: aux to delete + * + * The deletion will only be done when *BOTH* val and aux matches. + * Return 0 for successful delete. + * Return > 0 for not found. + */ +int ulist_del(struct ulist *ulist, u64 val, u64 aux) +{ + struct ulist_node *node; + + node = ulist_rbtree_search(ulist, val); + /* Not found */ + if (!node) + return 1; + + if (node->aux != aux) + return 1; + + /* Found and delete */ + ulist_rbtree_erase(ulist, node); + return 0; +} + /** * ulist_next - iterate ulist * @ulist: ulist to iterate @@ -237,15 +270,7 @@ struct ulist_node *ulist_next(struct ulist *ulist, struct ulist_iterator *uiter) uiter->cur_list = uiter->cur_list->next; } else { uiter->cur_list = ulist->nodes.next; -#ifdef CONFIG_BTRFS_DEBUG - uiter->i = 0; -#endif } node = list_entry(uiter->cur_list, struct ulist_node, list); -#ifdef CONFIG_BTRFS_DEBUG - ASSERT(node->seqnum == uiter->i); - ASSERT(uiter->i >= 0 && uiter->i < ulist->nnodes); - uiter->i++; -#endif return node; } diff --git a/fs/btrfs/ulist.h b/fs/btrfs/ulist.h index 4c29db6..a01a2c4 100644 --- a/fs/btrfs/ulist.h +++ b/fs/btrfs/ulist.h @@ -57,6 +57,7 @@ void ulist_free(struct ulist *ulist); int ulist_add(struct ulist *ulist, u64 val, u64 aux, gfp_t gfp_mask); int ulist_add_merge(struct ulist *ulist, u64 val, u64 aux, u64 *old_aux, gfp_t gfp_mask); +int ulist_del(struct ulist *ulist, u64 val, u64 aux); /* just like ulist_add_merge() but take a pointer for the aux data */ static inline int ulist_add_merge_ptr(struct ulist *ulist, u64 val, void *aux, diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 96aebf3..4b438b4 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -52,6 +52,10 @@ static void btrfs_dev_stat_print_on_load(struct btrfs_device *device); DEFINE_MUTEX(uuid_mutex); static LIST_HEAD(fs_uuids); +struct list_head *btrfs_get_fs_uuids(void) +{ + return &fs_uuids; +} static struct btrfs_fs_devices *__alloc_fs_devices(void) { @@ -345,7 +349,7 @@ loop_lock: waitqueue_active(&fs_info->async_submit_wait)) wake_up(&fs_info->async_submit_wait); - BUG_ON(atomic_read(&cur->bi_cnt) == 0); + BUG_ON(atomic_read(&cur->__bi_cnt) == 0); /* * if we're doing the sync list, record that our @@ -441,6 +445,61 @@ static void pending_bios_fn(struct btrfs_work *work) run_scheduled_bios(device); } + +void btrfs_free_stale_device(struct btrfs_device *cur_dev) +{ + struct btrfs_fs_devices *fs_devs; + struct btrfs_device *dev; + + if (!cur_dev->name) + return; + + list_for_each_entry(fs_devs, &fs_uuids, list) { + int del = 1; + + if (fs_devs->opened) + continue; + if (fs_devs->seeding) + continue; + + list_for_each_entry(dev, &fs_devs->devices, dev_list) { + + if (dev == cur_dev) + continue; + if (!dev->name) + continue; + + /* + * Todo: This won't be enough. What if the same device + * comes back (with new uuid and) with its mapper path? + * But for now, this does help as mostly an admin will + * either use mapper or non mapper path throughout. + */ + rcu_read_lock(); + del = strcmp(rcu_str_deref(dev->name), + rcu_str_deref(cur_dev->name)); + rcu_read_unlock(); + if (!del) + break; + } + + if (!del) { + /* delete the stale device */ + if (fs_devs->num_devices == 1) { + btrfs_sysfs_remove_fsid(fs_devs); + list_del(&fs_devs->list); + free_fs_devices(fs_devs); + } else { + fs_devs->num_devices--; + list_del(&dev->dev_list); + rcu_string_free(dev->name); + kfree(dev); + } + break; + } + } +} + /* * Add new device to list of registered devices * @@ -556,6 +615,12 @@ static noinline int device_list_add(const char *path, if (!fs_devices->opened) device->generation = found_transid; + /* + * if there is new btrfs on an already registered device, + * then remove the stale device entry. + */ + btrfs_free_stale_device(device); + *fs_devices_ret = fs_devices; return ret; @@ -693,13 +758,13 @@ static void free_device(struct rcu_head *head) static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices) { - struct btrfs_device *device; + struct btrfs_device *device, *tmp; if (--fs_devices->opened > 0) return 0; mutex_lock(&fs_devices->device_list_mutex); - list_for_each_entry(device, &fs_devices->devices, dev_list) { + list_for_each_entry_safe(device, tmp, &fs_devices->devices, dev_list) { struct btrfs_device *new_device; struct rcu_string *name; @@ -1067,15 +1132,31 @@ again: map = (struct map_lookup *)em->bdev; for (i = 0; i < map->num_stripes; i++) { + u64 end; + if (map->stripes[i].dev != device) continue; if (map->stripes[i].physical >= physical_start + len || map->stripes[i].physical + em->orig_block_len <= physical_start) continue; - *start = map->stripes[i].physical + - em->orig_block_len; - ret = 1; + /* + * Make sure that while processing the pinned list we do + * not override our *start with a lower value, because + * we can have pinned chunks that fall within this + * device hole and that have lower physical addresses + * than the pending chunks we processed before. If we + * do not take this special care we can end up getting + * 2 pending chunks that start at the same physical + * device offsets because the end offset of a pinned + * chunk can be equal to the start offset of some + * pending chunk. + */ + end = map->stripes[i].physical + em->orig_block_len; + if (end > *start) { + *start = end; + ret = 1; + } } } if (search_list == &trans->transaction->pending_chunks) { @@ -1706,7 +1787,7 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) if (device->bdev) { device->fs_devices->open_devices--; /* remove sysfs entry */ - btrfs_kobj_rm_device(root->fs_info, device); + btrfs_kobj_rm_device(root->fs_info->fs_devices, device); } call_rcu(&device->rcu, free_device); @@ -1875,6 +1956,9 @@ void btrfs_destroy_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, mutex_lock(&uuid_mutex); WARN_ON(!tgtdev); mutex_lock(&fs_info->fs_devices->device_list_mutex); + + btrfs_kobj_rm_device(fs_info->fs_devices, tgtdev); + if (tgtdev->bdev) { btrfs_scratch_superblock(tgtdev); fs_info->fs_devices->open_devices--; @@ -2211,7 +2295,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) tmp + 1); /* add sysfs device entry */ - btrfs_kobj_add_device(root->fs_info, device); + btrfs_kobj_add_device(root->fs_info->fs_devices, device); /* * we've got more storage, clear any full flags on the space @@ -2252,8 +2336,9 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) */ snprintf(fsid_buf, BTRFS_UUID_UNPARSED_SIZE, "%pU", root->fs_info->fsid); - if (kobject_rename(&root->fs_info->super_kobj, fsid_buf)) - goto error_trans; + if (kobject_rename(&root->fs_info->fs_devices->super_kobj, + fsid_buf)) + pr_warn("BTRFS: sysfs: failed to create fsid for sprout\n"); } root->fs_info->num_tolerated_disk_barrier_failures = @@ -2289,7 +2374,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) error_trans: btrfs_end_transaction(trans, root); rcu_string_free(device->name); - btrfs_kobj_rm_device(root->fs_info, device); + btrfs_kobj_rm_device(root->fs_info->fs_devices, device); kfree(device); error: blkdev_put(bdev, FMODE_EXCL); @@ -2609,6 +2694,9 @@ int btrfs_remove_chunk(struct btrfs_trans_handle *trans, return -EINVAL; } map = (struct map_lookup *)em->bdev; + lock_chunks(root->fs_info->chunk_root); + check_system_chunk(trans, extent_root, map->type); + unlock_chunks(root->fs_info->chunk_root); for (i = 0; i < map->num_stripes; i++) { struct btrfs_device *device = map->stripes[i].dev; @@ -3908,9 +3996,9 @@ int btrfs_create_uuid_tree(struct btrfs_fs_info *fs_info) uuid_root = btrfs_create_tree(trans, fs_info, BTRFS_UUID_TREE_OBJECTID); if (IS_ERR(uuid_root)) { - btrfs_abort_transaction(trans, tree_root, - PTR_ERR(uuid_root)); - return PTR_ERR(uuid_root); + ret = PTR_ERR(uuid_root); + btrfs_abort_transaction(trans, tree_root, ret); + return ret; } fs_info->uuid_root = uuid_root; @@ -3965,6 +4053,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size) int slot; int failed = 0; bool retried = false; + bool checked_pending_chunks = false; struct extent_buffer *l; struct btrfs_key key; struct btrfs_super_block *super_copy = root->fs_info->super_copy; @@ -4045,15 +4134,6 @@ again: goto again; } else if (failed && retried) { ret = -ENOSPC; - lock_chunks(root); - - btrfs_device_set_total_bytes(device, old_size); - if (device->writeable) - device->fs_devices->total_rw_bytes += diff; - spin_lock(&root->fs_info->free_chunk_lock); - root->fs_info->free_chunk_space += diff; - spin_unlock(&root->fs_info->free_chunk_lock); - unlock_chunks(root); goto done; } @@ -4065,6 +4145,35 @@ again: } lock_chunks(root); + + /* + * We checked in the above loop all device extents that were already in + * the device tree. However before we have updated the device's + * total_bytes to the new size, we might have had chunk allocations that + * have not complete yet (new block groups attached to transaction + * handles), and therefore their device extents were not yet in the + * device tree and we missed them in the loop above. So if we have any + * pending chunk using a device extent that overlaps the device range + * that we can not use anymore, commit the current transaction and + * repeat the search on the device tree - this way we guarantee we will + * not have chunks using device extents that end beyond 'new_size'. + */ + if (!checked_pending_chunks) { + u64 start = new_size; + u64 len = old_size - new_size; + + if (contains_pending_extent(trans, device, &start, len)) { + unlock_chunks(root); + checked_pending_chunks = true; + failed = 0; + retried = false; + ret = btrfs_commit_transaction(trans, root); + if (ret) + goto done; + goto again; + } + } + btrfs_device_set_disk_total_bytes(device, new_size); if (list_empty(&device->resized_list)) list_add_tail(&device->resized_list, @@ -4079,6 +4188,16 @@ again: btrfs_end_transaction(trans, root); done: btrfs_free_path(path); + if (ret) { + lock_chunks(root); + btrfs_device_set_total_bytes(device, old_size); + if (device->writeable) + device->fs_devices->total_rw_bytes += diff; + spin_lock(&root->fs_info->free_chunk_lock); + root->fs_info->free_chunk_space += diff; + spin_unlock(&root->fs_info->free_chunk_lock); + unlock_chunks(root); + } return ret; } @@ -4625,6 +4744,7 @@ int btrfs_alloc_chunk(struct btrfs_trans_handle *trans, { u64 chunk_offset; + ASSERT(mutex_is_locked(&extent_root->fs_info->chunk_mutex)); chunk_offset = find_next_chunk(extent_root->fs_info); return __btrfs_alloc_chunk(trans, extent_root, chunk_offset, type); } @@ -5585,10 +5705,10 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, static inline void btrfs_end_bbio(struct btrfs_bio *bbio, struct bio *bio, int err) { - if (likely(bbio->flags & BTRFS_BIO_ORIG_BIO_SUBMITTED)) - bio_endio_nodec(bio, err); - else - bio_endio(bio, err); + bio->bi_private = bbio->private; + bio->bi_end_io = bbio->end_io; + bio_endio(bio, err); + btrfs_put_bbio(bbio); } @@ -5632,8 +5752,6 @@ static void btrfs_end_bio(struct bio *bio, int err) bio = bbio->orig_bio; } - bio->bi_private = bbio->private; - bio->bi_end_io = bbio->end_io; btrfs_io_bio(bio)->mirror_num = bbio->mirror_num; /* only send an error to the higher layers if it is * beyond the tolerance of the btrfs bio @@ -5815,8 +5933,6 @@ static void bbio_error(struct btrfs_bio *bbio, struct bio *bio, u64 logical) /* Shoud be the original bio. */ WARN_ON(bio != bbio->orig_bio); - bio->bi_private = bbio->private; - bio->bi_end_io = bbio->end_io; btrfs_io_bio(bio)->mirror_num = bbio->mirror_num; bio->bi_iter.bi_sector = logical >> 9; @@ -5897,10 +6013,8 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, if (dev_nr < total_devs - 1) { bio = btrfs_bio_clone(first_bio, GFP_NOFS); BUG_ON(!bio); /* -ENOMEM */ - } else { + } else bio = first_bio; - bbio->flags |= BTRFS_BIO_ORIG_BIO_SUBMITTED; - } submit_stripe_bio(root, bbio, bio, bbio->stripes[dev_nr].physical, dev_nr, rw, @@ -6077,6 +6191,8 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key, free_extent_map(em); return -EIO; } + btrfs_warn(root->fs_info, "devid %llu uuid %pU is missing", + devid, uuid); } map->stripes[i].dev->in_fs_metadata = 1; } @@ -6196,10 +6312,11 @@ static int read_one_dev(struct btrfs_root *root, if (!btrfs_test_opt(root, DEGRADED)) return -EIO; - btrfs_warn(root->fs_info, "devid %llu missing", devid); device = add_missing_dev(root, fs_devices, devid, dev_uuid); if (!device) return -ENOMEM; + btrfs_warn(root->fs_info, "devid %llu uuid %pU missing", + devid, dev_uuid); } else { if (!device->bdev && !btrfs_test_opt(root, DEGRADED)) return -EIO; @@ -6727,3 +6844,21 @@ void btrfs_update_commit_device_bytes_used(struct btrfs_root *root, } unlock_chunks(root); } + +void btrfs_set_fs_info_ptr(struct btrfs_fs_info *fs_info) +{ + struct btrfs_fs_devices *fs_devices = fs_info->fs_devices; + while (fs_devices) { + fs_devices->fs_info = fs_info; + fs_devices = fs_devices->seed; + } +} + +void btrfs_reset_fs_info_ptr(struct btrfs_fs_info *fs_info) +{ + struct btrfs_fs_devices *fs_devices = fs_info->fs_devices; + while (fs_devices) { + fs_devices->fs_info = NULL; + fs_devices = fs_devices->seed; + } +} diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index ebc3133..95842a9 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -253,6 +253,12 @@ struct btrfs_fs_devices { * nonrot flag set */ int rotating; + + struct btrfs_fs_info *fs_info; + /* sysfs kobjects */ + struct kobject super_kobj; + struct kobject *device_dir_kobj; + struct completion kobj_unregister; }; #define BTRFS_BIO_INLINE_CSUM_SIZE 64 @@ -292,8 +298,6 @@ struct btrfs_bio_stripe { struct btrfs_bio; typedef void (btrfs_bio_end_io_t) (struct btrfs_bio *bio, int err); -#define BTRFS_BIO_ORIG_BIO_SUBMITTED (1 << 0) - struct btrfs_bio { atomic_t refs; atomic_t stripes_pending; @@ -537,5 +541,8 @@ static inline void unlock_chunks(struct btrfs_root *root) mutex_unlock(&root->fs_info->chunk_mutex); } +struct list_head *btrfs_get_fs_uuids(void); +void btrfs_set_fs_info_ptr(struct btrfs_fs_info *fs_info); +void btrfs_reset_fs_info_ptr(struct btrfs_fs_info *fs_info); #endif diff --git a/fs/buffer.c b/fs/buffer.c index c7a5602..1cf7a53 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -30,6 +30,7 @@ #include <linux/quotaops.h> #include <linux/highmem.h> #include <linux/export.h> +#include <linux/backing-dev.h> #include <linux/writeback.h> #include <linux/hash.h> #include <linux/suspend.h> @@ -44,6 +45,9 @@ #include <trace/events/block.h> static int fsync_buffers_list(spinlock_t *lock, struct list_head *list); +static int submit_bh_wbc(int rw, struct buffer_head *bh, + unsigned long bio_flags, + struct writeback_control *wbc); #define BH_ENTRY(list) list_entry((list), struct buffer_head, b_assoc_buffers) @@ -623,21 +627,22 @@ EXPORT_SYMBOL(mark_buffer_dirty_inode); * * If warn is true, then emit a warning if the page is not uptodate and has * not been truncated. + * + * The caller must hold mem_cgroup_begin_page_stat() lock. */ -static void __set_page_dirty(struct page *page, - struct address_space *mapping, int warn) +static void __set_page_dirty(struct page *page, struct address_space *mapping, + struct mem_cgroup *memcg, int warn) { unsigned long flags; spin_lock_irqsave(&mapping->tree_lock, flags); if (page->mapping) { /* Race with truncate? */ WARN_ON_ONCE(warn && !PageUptodate(page)); - account_page_dirtied(page, mapping); + account_page_dirtied(page, mapping, memcg); radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); } spin_unlock_irqrestore(&mapping->tree_lock, flags); - __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); } /* @@ -668,6 +673,7 @@ static void __set_page_dirty(struct page *page, int __set_page_dirty_buffers(struct page *page) { int newly_dirty; + struct mem_cgroup *memcg; struct address_space *mapping = page_mapping(page); if (unlikely(!mapping)) @@ -683,11 +689,22 @@ int __set_page_dirty_buffers(struct page *page) bh = bh->b_this_page; } while (bh != head); } + /* + * Use mem_group_begin_page_stat() to keep PageDirty synchronized with + * per-memcg dirty page counters. + */ + memcg = mem_cgroup_begin_page_stat(page); newly_dirty = !TestSetPageDirty(page); spin_unlock(&mapping->private_lock); if (newly_dirty) - __set_page_dirty(page, mapping, 1); + __set_page_dirty(page, mapping, memcg, 1); + + mem_cgroup_end_page_stat(memcg); + + if (newly_dirty) + __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); + return newly_dirty; } EXPORT_SYMBOL(__set_page_dirty_buffers); @@ -1158,11 +1175,18 @@ void mark_buffer_dirty(struct buffer_head *bh) if (!test_set_buffer_dirty(bh)) { struct page *page = bh->b_page; + struct address_space *mapping = NULL; + struct mem_cgroup *memcg; + + memcg = mem_cgroup_begin_page_stat(page); if (!TestSetPageDirty(page)) { - struct address_space *mapping = page_mapping(page); + mapping = page_mapping(page); if (mapping) - __set_page_dirty(page, mapping, 0); + __set_page_dirty(page, mapping, memcg, 0); } + mem_cgroup_end_page_stat(memcg); + if (mapping) + __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); } } EXPORT_SYMBOL(mark_buffer_dirty); @@ -1684,8 +1708,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page, struct buffer_head *bh, *head; unsigned int blocksize, bbits; int nr_underway = 0; - int write_op = (wbc->sync_mode == WB_SYNC_ALL ? - WRITE_SYNC : WRITE); + int write_op = (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE); head = create_page_buffers(page, inode, (1 << BH_Dirty)|(1 << BH_Uptodate)); @@ -1774,7 +1797,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page, do { struct buffer_head *next = bh->b_this_page; if (buffer_async_write(bh)) { - submit_bh(write_op, bh); + submit_bh_wbc(write_op, bh, 0, wbc); nr_underway++; } bh = next; @@ -1828,7 +1851,7 @@ recover: struct buffer_head *next = bh->b_this_page; if (buffer_async_write(bh)) { clear_buffer_dirty(bh); - submit_bh(write_op, bh); + submit_bh_wbc(write_op, bh, 0, wbc); nr_underway++; } bh = next; @@ -2938,10 +2961,6 @@ static void end_bio_bh_io_sync(struct bio *bio, int err) { struct buffer_head *bh = bio->bi_private; - if (err == -EOPNOTSUPP) { - set_bit(BIO_EOPNOTSUPP, &bio->bi_flags); - } - if (unlikely (test_bit(BIO_QUIET,&bio->bi_flags))) set_bit(BH_Quiet, &bh->b_state); @@ -2997,10 +3016,10 @@ void guard_bio_eod(int rw, struct bio *bio) } } -int _submit_bh(int rw, struct buffer_head *bh, unsigned long bio_flags) +static int submit_bh_wbc(int rw, struct buffer_head *bh, + unsigned long bio_flags, struct writeback_control *wbc) { struct bio *bio; - int ret = 0; BUG_ON(!buffer_locked(bh)); BUG_ON(!buffer_mapped(bh)); @@ -3020,6 +3039,11 @@ int _submit_bh(int rw, struct buffer_head *bh, unsigned long bio_flags) */ bio = bio_alloc(GFP_NOIO, 1); + if (wbc) { + wbc_init_bio(wbc, bio); + wbc_account_io(wbc, bh->b_page, bh->b_size); + } + bio->bi_iter.bi_sector = bh->b_blocknr * (bh->b_size >> 9); bio->bi_bdev = bh->b_bdev; bio->bi_io_vec[0].bv_page = bh->b_page; @@ -3041,20 +3065,19 @@ int _submit_bh(int rw, struct buffer_head *bh, unsigned long bio_flags) if (buffer_prio(bh)) rw |= REQ_PRIO; - bio_get(bio); submit_bio(rw, bio); + return 0; +} - if (bio_flagged(bio, BIO_EOPNOTSUPP)) - ret = -EOPNOTSUPP; - - bio_put(bio); - return ret; +int _submit_bh(int rw, struct buffer_head *bh, unsigned long bio_flags) +{ + return submit_bh_wbc(rw, bh, bio_flags, NULL); } EXPORT_SYMBOL_GPL(_submit_bh); int submit_bh(int rw, struct buffer_head *bh) { - return _submit_bh(rw, bh, 0); + return submit_bh_wbc(rw, bh, 0, NULL); } EXPORT_SYMBOL(submit_bh); @@ -3243,8 +3266,8 @@ int try_to_free_buffers(struct page *page) * to synchronise against __set_page_dirty_buffers and prevent the * dirty bit from being lost. */ - if (ret && TestClearPageDirty(page)) - account_page_cleaned(page, mapping); + if (ret) + cancel_dirty_page(page); spin_unlock(&mapping->private_lock); out: if (buffers_to_free) { diff --git a/fs/ceph/acl.c b/fs/ceph/acl.c index 64fa248..8f84646 100644 --- a/fs/ceph/acl.c +++ b/fs/ceph/acl.c @@ -187,10 +187,10 @@ int ceph_pre_init_acls(struct inode *dir, umode_t *mode, val_size2 = posix_acl_xattr_size(default_acl->a_count); err = -ENOMEM; - tmp_buf = kmalloc(max(val_size1, val_size2), GFP_NOFS); + tmp_buf = kmalloc(max(val_size1, val_size2), GFP_KERNEL); if (!tmp_buf) goto out_err; - pagelist = kmalloc(sizeof(struct ceph_pagelist), GFP_NOFS); + pagelist = kmalloc(sizeof(struct ceph_pagelist), GFP_KERNEL); if (!pagelist) goto out_err; ceph_pagelist_init(pagelist); diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index e162bcd..890c509 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -87,17 +87,21 @@ static int ceph_set_page_dirty(struct page *page) inode = mapping->host; ci = ceph_inode(inode); - /* - * Note that we're grabbing a snapc ref here without holding - * any locks! - */ - snapc = ceph_get_snap_context(ci->i_snap_realm->cached_context); - /* dirty the head */ spin_lock(&ci->i_ceph_lock); - if (ci->i_head_snapc == NULL) - ci->i_head_snapc = ceph_get_snap_context(snapc); - ++ci->i_wrbuffer_ref_head; + BUG_ON(ci->i_wr_ref == 0); // caller should hold Fw reference + if (__ceph_have_pending_cap_snap(ci)) { + struct ceph_cap_snap *capsnap = + list_last_entry(&ci->i_cap_snaps, + struct ceph_cap_snap, + ci_item); + snapc = ceph_get_snap_context(capsnap->context); + capsnap->dirty_pages++; + } else { + BUG_ON(!ci->i_head_snapc); + snapc = ceph_get_snap_context(ci->i_head_snapc); + ++ci->i_wrbuffer_ref_head; + } if (ci->i_wrbuffer_ref == 0) ihold(inode); ++ci->i_wrbuffer_ref; @@ -346,7 +350,7 @@ static int start_read(struct inode *inode, struct list_head *page_list, int max) /* build page vector */ nr_pages = calc_pages_for(0, len); - pages = kmalloc(sizeof(*pages) * nr_pages, GFP_NOFS); + pages = kmalloc(sizeof(*pages) * nr_pages, GFP_KERNEL); ret = -ENOMEM; if (!pages) goto out; @@ -358,7 +362,7 @@ static int start_read(struct inode *inode, struct list_head *page_list, int max) dout("start_read %p adding %p idx %lu\n", inode, page, page->index); if (add_to_page_cache_lru(page, &inode->i_data, page->index, - GFP_NOFS)) { + GFP_KERNEL)) { ceph_fscache_uncache_page(inode, page); page_cache_release(page); dout("start_read %p add_to_page_cache failed %p\n", @@ -436,7 +440,7 @@ out: * only snap context we are allowed to write back. */ static struct ceph_snap_context *get_oldest_context(struct inode *inode, - u64 *snap_size) + loff_t *snap_size) { struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_snap_context *snapc = NULL; @@ -476,8 +480,9 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) struct ceph_osd_client *osdc; struct ceph_snap_context *snapc, *oldest; loff_t page_off = page_offset(page); + loff_t snap_size = -1; long writeback_stat; - u64 truncate_size, snap_size = 0; + u64 truncate_size; u32 truncate_seq; int err = 0, len = PAGE_CACHE_SIZE; @@ -512,7 +517,7 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) spin_lock(&ci->i_ceph_lock); truncate_seq = ci->i_truncate_seq; truncate_size = ci->i_truncate_size; - if (!snap_size) + if (snap_size == -1) snap_size = i_size_read(inode); spin_unlock(&ci->i_ceph_lock); @@ -695,7 +700,8 @@ static int ceph_writepages_start(struct address_space *mapping, unsigned wsize = 1 << inode->i_blkbits; struct ceph_osd_request *req = NULL; int do_sync = 0; - u64 truncate_size, snap_size; + loff_t snap_size, i_size; + u64 truncate_size; u32 truncate_seq; /* @@ -741,7 +747,7 @@ static int ceph_writepages_start(struct address_space *mapping, retry: /* find oldest snap context with dirty data */ ceph_put_snap_context(snapc); - snap_size = 0; + snap_size = -1; snapc = get_oldest_context(inode, &snap_size); if (!snapc) { /* hmm, why does writepages get called when there @@ -749,16 +755,13 @@ retry: dout(" no snap context with dirty data?\n"); goto out; } - if (snap_size == 0) - snap_size = i_size_read(inode); dout(" oldest snapc is %p seq %lld (%d snaps)\n", snapc, snapc->seq, snapc->num_snaps); spin_lock(&ci->i_ceph_lock); truncate_seq = ci->i_truncate_seq; truncate_size = ci->i_truncate_size; - if (!snap_size) - snap_size = i_size_read(inode); + i_size = i_size_read(inode); spin_unlock(&ci->i_ceph_lock); if (last_snapc && snapc != last_snapc) { @@ -828,8 +831,10 @@ get_more_pages: dout("waiting on writeback %p\n", page); wait_on_page_writeback(page); } - if (page_offset(page) >= snap_size) { - dout("%p page eof %llu\n", page, snap_size); + if (page_offset(page) >= + (snap_size == -1 ? i_size : snap_size)) { + dout("%p page eof %llu\n", page, + (snap_size == -1 ? i_size : snap_size)); done = 1; unlock_page(page); break; @@ -884,7 +889,8 @@ get_more_pages: } if (do_sync) - osd_req_op_init(req, 1, CEPH_OSD_OP_STARTSYNC); + osd_req_op_init(req, 1, + CEPH_OSD_OP_STARTSYNC, 0); req->r_callback = writepages_finish; req->r_inode = inode; @@ -944,10 +950,18 @@ get_more_pages: } /* Format the osd request message and submit the write */ - offset = page_offset(pages[0]); - len = min(snap_size - offset, - (u64)locked_pages << PAGE_CACHE_SHIFT); + len = (u64)locked_pages << PAGE_CACHE_SHIFT; + if (snap_size == -1) { + len = min(len, (u64)i_size_read(inode) - offset); + /* writepages_finish() clears writeback pages + * according to the data length, so make sure + * data length covers all locked pages */ + len = max(len, 1 + + ((u64)(locked_pages - 1) << PAGE_CACHE_SHIFT)); + } else { + len = min(len, snap_size - offset); + } dout("writepages got %d pages at %llu~%llu\n", locked_pages, offset, len); @@ -1032,7 +1046,6 @@ static int ceph_update_writeable_page(struct file *file, { struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); - struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; loff_t page_off = pos & PAGE_CACHE_MASK; int pos_in_page = pos & ~PAGE_CACHE_MASK; int end_in_page = pos_in_page + len; @@ -1044,10 +1057,6 @@ retry_locked: /* writepages currently holds page lock, but if we change that later, */ wait_on_page_writeback(page); - /* check snap context */ - BUG_ON(!ci->i_snap_realm); - down_read(&mdsc->snap_rwsem); - BUG_ON(!ci->i_snap_realm->cached_context); snapc = page_snap_context(page); if (snapc && snapc != ci->i_head_snapc) { /* @@ -1055,7 +1064,6 @@ retry_locked: * context! is it writeable now? */ oldest = get_oldest_context(inode, NULL); - up_read(&mdsc->snap_rwsem); if (snapc->seq > oldest->seq) { ceph_put_snap_context(oldest); @@ -1112,7 +1120,6 @@ retry_locked: } /* we need to read it. */ - up_read(&mdsc->snap_rwsem); r = readpage_nounlock(file, page); if (r < 0) goto fail_nosnap; @@ -1157,16 +1164,13 @@ static int ceph_write_begin(struct file *file, struct address_space *mapping, /* * we don't do anything in here that simple_write_end doesn't do - * except adjust dirty page accounting and drop read lock on - * mdsc->snap_rwsem. + * except adjust dirty page accounting */ static int ceph_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { struct inode *inode = file_inode(file); - struct ceph_fs_client *fsc = ceph_inode_to_client(inode); - struct ceph_mds_client *mdsc = fsc->mdsc; unsigned from = pos & (PAGE_CACHE_SIZE - 1); int check_cap = 0; @@ -1188,7 +1192,6 @@ static int ceph_write_end(struct file *file, struct address_space *mapping, set_page_dirty(page); unlock_page(page); - up_read(&mdsc->snap_rwsem); page_cache_release(page); if (check_cap) @@ -1314,13 +1317,17 @@ static int ceph_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) struct inode *inode = file_inode(vma->vm_file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_file_info *fi = vma->vm_file->private_data; - struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; + struct ceph_cap_flush *prealloc_cf; struct page *page = vmf->page; loff_t off = page_offset(page); loff_t size = i_size_read(inode); size_t len; int want, got, ret; + prealloc_cf = ceph_alloc_cap_flush(); + if (!prealloc_cf) + return VM_FAULT_SIGBUS; + if (ci->i_inline_version != CEPH_INLINE_NONE) { struct page *locked_page = NULL; if (off == 0) { @@ -1330,8 +1337,10 @@ static int ceph_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) ret = ceph_uninline_data(vma->vm_file, locked_page); if (locked_page) unlock_page(locked_page); - if (ret < 0) - return VM_FAULT_SIGBUS; + if (ret < 0) { + ret = VM_FAULT_SIGBUS; + goto out_free; + } } if (off + PAGE_CACHE_SIZE <= size) @@ -1353,7 +1362,8 @@ static int ceph_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) break; if (ret != -ERESTARTSYS) { WARN_ON(1); - return VM_FAULT_SIGBUS; + ret = VM_FAULT_SIGBUS; + goto out_free; } } dout("page_mkwrite %p %llu~%zd got cap refs on %s\n", @@ -1373,7 +1383,6 @@ static int ceph_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) if (ret == 0) { /* success. we'll keep the page locked. */ set_page_dirty(page); - up_read(&mdsc->snap_rwsem); ret = VM_FAULT_LOCKED; } else { if (ret == -ENOMEM) @@ -1389,7 +1398,8 @@ out: int dirty; spin_lock(&ci->i_ceph_lock); ci->i_inline_version = CEPH_INLINE_NONE; - dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR, + &prealloc_cf); spin_unlock(&ci->i_ceph_lock); if (dirty) __mark_inode_dirty(inode, dirty); @@ -1398,6 +1408,8 @@ out: dout("page_mkwrite %p %llu~%zd dropping cap refs on %s ret %d\n", inode, off, len, ceph_cap_string(got), ret); ceph_put_cap_refs(ci, got); +out_free: + ceph_free_cap_flush(prealloc_cf); return ret; } @@ -1509,8 +1521,7 @@ int ceph_uninline_data(struct file *filp, struct page *locked_page) ceph_vino(inode), 0, &len, 0, 1, CEPH_OSD_OP_CREATE, CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE, - ci->i_snap_realm->cached_context, - 0, 0, false); + ceph_empty_snapc, 0, 0, false); if (IS_ERR(req)) { err = PTR_ERR(req); goto out; @@ -1528,7 +1539,7 @@ int ceph_uninline_data(struct file *filp, struct page *locked_page) ceph_vino(inode), 0, &len, 1, 3, CEPH_OSD_OP_WRITE, CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE, - ci->i_snap_realm->cached_context, + ceph_empty_snapc, ci->i_truncate_seq, ci->i_truncate_size, false); if (IS_ERR(req)) { @@ -1597,3 +1608,206 @@ int ceph_mmap(struct file *file, struct vm_area_struct *vma) vma->vm_ops = &ceph_vmops; return 0; } + +enum { + POOL_READ = 1, + POOL_WRITE = 2, +}; + +static int __ceph_pool_perm_get(struct ceph_inode_info *ci, u32 pool) +{ + struct ceph_fs_client *fsc = ceph_inode_to_client(&ci->vfs_inode); + struct ceph_mds_client *mdsc = fsc->mdsc; + struct ceph_osd_request *rd_req = NULL, *wr_req = NULL; + struct rb_node **p, *parent; + struct ceph_pool_perm *perm; + struct page **pages; + int err = 0, err2 = 0, have = 0; + + down_read(&mdsc->pool_perm_rwsem); + p = &mdsc->pool_perm_tree.rb_node; + while (*p) { + perm = rb_entry(*p, struct ceph_pool_perm, node); + if (pool < perm->pool) + p = &(*p)->rb_left; + else if (pool > perm->pool) + p = &(*p)->rb_right; + else { + have = perm->perm; + break; + } + } + up_read(&mdsc->pool_perm_rwsem); + if (*p) + goto out; + + dout("__ceph_pool_perm_get pool %u no perm cached\n", pool); + + down_write(&mdsc->pool_perm_rwsem); + parent = NULL; + while (*p) { + parent = *p; + perm = rb_entry(parent, struct ceph_pool_perm, node); + if (pool < perm->pool) + p = &(*p)->rb_left; + else if (pool > perm->pool) + p = &(*p)->rb_right; + else { + have = perm->perm; + break; + } + } + if (*p) { + up_write(&mdsc->pool_perm_rwsem); + goto out; + } + + rd_req = ceph_osdc_alloc_request(&fsc->client->osdc, + ceph_empty_snapc, + 1, false, GFP_NOFS); + if (!rd_req) { + err = -ENOMEM; + goto out_unlock; + } + + rd_req->r_flags = CEPH_OSD_FLAG_READ; + osd_req_op_init(rd_req, 0, CEPH_OSD_OP_STAT, 0); + rd_req->r_base_oloc.pool = pool; + snprintf(rd_req->r_base_oid.name, sizeof(rd_req->r_base_oid.name), + "%llx.00000000", ci->i_vino.ino); + rd_req->r_base_oid.name_len = strlen(rd_req->r_base_oid.name); + + wr_req = ceph_osdc_alloc_request(&fsc->client->osdc, + ceph_empty_snapc, + 1, false, GFP_NOFS); + if (!wr_req) { + err = -ENOMEM; + goto out_unlock; + } + + wr_req->r_flags = CEPH_OSD_FLAG_WRITE | + CEPH_OSD_FLAG_ACK | CEPH_OSD_FLAG_ONDISK; + osd_req_op_init(wr_req, 0, CEPH_OSD_OP_CREATE, CEPH_OSD_OP_FLAG_EXCL); + wr_req->r_base_oloc.pool = pool; + wr_req->r_base_oid = rd_req->r_base_oid; + + /* one page should be large enough for STAT data */ + pages = ceph_alloc_page_vector(1, GFP_KERNEL); + if (IS_ERR(pages)) { + err = PTR_ERR(pages); + goto out_unlock; + } + + osd_req_op_raw_data_in_pages(rd_req, 0, pages, PAGE_SIZE, + 0, false, true); + ceph_osdc_build_request(rd_req, 0, NULL, CEPH_NOSNAP, + &ci->vfs_inode.i_mtime); + err = ceph_osdc_start_request(&fsc->client->osdc, rd_req, false); + + ceph_osdc_build_request(wr_req, 0, NULL, CEPH_NOSNAP, + &ci->vfs_inode.i_mtime); + err2 = ceph_osdc_start_request(&fsc->client->osdc, wr_req, false); + + if (!err) + err = ceph_osdc_wait_request(&fsc->client->osdc, rd_req); + if (!err2) + err2 = ceph_osdc_wait_request(&fsc->client->osdc, wr_req); + + if (err >= 0 || err == -ENOENT) + have |= POOL_READ; + else if (err != -EPERM) + goto out_unlock; + + if (err2 == 0 || err2 == -EEXIST) + have |= POOL_WRITE; + else if (err2 != -EPERM) { + err = err2; + goto out_unlock; + } + + perm = kmalloc(sizeof(*perm), GFP_NOFS); + if (!perm) { + err = -ENOMEM; + goto out_unlock; + } + + perm->pool = pool; + perm->perm = have; + rb_link_node(&perm->node, parent, p); + rb_insert_color(&perm->node, &mdsc->pool_perm_tree); + err = 0; +out_unlock: + up_write(&mdsc->pool_perm_rwsem); + + if (rd_req) + ceph_osdc_put_request(rd_req); + if (wr_req) + ceph_osdc_put_request(wr_req); +out: + if (!err) + err = have; + dout("__ceph_pool_perm_get pool %u result = %d\n", pool, err); + return err; +} + +int ceph_pool_perm_check(struct ceph_inode_info *ci, int need) +{ + u32 pool; + int ret, flags; + + if (ceph_test_mount_opt(ceph_inode_to_client(&ci->vfs_inode), + NOPOOLPERM)) + return 0; + + spin_lock(&ci->i_ceph_lock); + flags = ci->i_ceph_flags; + pool = ceph_file_layout_pg_pool(ci->i_layout); + spin_unlock(&ci->i_ceph_lock); +check: + if (flags & CEPH_I_POOL_PERM) { + if ((need & CEPH_CAP_FILE_RD) && !(flags & CEPH_I_POOL_RD)) { + dout("ceph_pool_perm_check pool %u no read perm\n", + pool); + return -EPERM; + } + if ((need & CEPH_CAP_FILE_WR) && !(flags & CEPH_I_POOL_WR)) { + dout("ceph_pool_perm_check pool %u no write perm\n", + pool); + return -EPERM; + } + return 0; + } + + ret = __ceph_pool_perm_get(ci, pool); + if (ret < 0) + return ret; + + flags = CEPH_I_POOL_PERM; + if (ret & POOL_READ) + flags |= CEPH_I_POOL_RD; + if (ret & POOL_WRITE) + flags |= CEPH_I_POOL_WR; + + spin_lock(&ci->i_ceph_lock); + if (pool == ceph_file_layout_pg_pool(ci->i_layout)) { + ci->i_ceph_flags = flags; + } else { + pool = ceph_file_layout_pg_pool(ci->i_layout); + flags = ci->i_ceph_flags; + } + spin_unlock(&ci->i_ceph_lock); + goto check; +} + +void ceph_pool_perm_destroy(struct ceph_mds_client *mdsc) +{ + struct ceph_pool_perm *perm; + struct rb_node *n; + + while (!RB_EMPTY_ROOT(&mdsc->pool_perm_tree)) { + n = rb_first(&mdsc->pool_perm_tree); + perm = rb_entry(n, struct ceph_pool_perm, node); + rb_erase(n, &mdsc->pool_perm_tree); + kfree(perm); + } +} diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index be5ea6a..dc10c9d 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -833,7 +833,9 @@ int __ceph_caps_used(struct ceph_inode_info *ci) used |= CEPH_CAP_PIN; if (ci->i_rd_ref) used |= CEPH_CAP_FILE_RD; - if (ci->i_rdcache_ref || ci->vfs_inode.i_data.nrpages) + if (ci->i_rdcache_ref || + (!S_ISDIR(ci->vfs_inode.i_mode) && /* ignore readdir cache */ + ci->vfs_inode.i_data.nrpages)) used |= CEPH_CAP_FILE_CACHE; if (ci->i_wr_ref) used |= CEPH_CAP_FILE_WR; @@ -926,16 +928,6 @@ void __ceph_remove_cap(struct ceph_cap *cap, bool queue_release) /* remove from session list */ spin_lock(&session->s_cap_lock); - /* - * s_cap_reconnect is protected by s_cap_lock. no one changes - * s_cap_gen while session is in the reconnect state. - */ - if (queue_release && - (!session->s_cap_reconnect || - cap->cap_gen == session->s_cap_gen)) - __queue_cap_release(session, ci->i_vino.ino, cap->cap_id, - cap->mseq, cap->issue_seq); - if (session->s_cap_iterator == cap) { /* not yet, we are iterating over this very cap */ dout("__ceph_remove_cap delaying %p removal from session %p\n", @@ -948,6 +940,25 @@ void __ceph_remove_cap(struct ceph_cap *cap, bool queue_release) } /* protect backpointer with s_cap_lock: see iterate_session_caps */ cap->ci = NULL; + + /* + * s_cap_reconnect is protected by s_cap_lock. no one changes + * s_cap_gen while session is in the reconnect state. + */ + if (queue_release && + (!session->s_cap_reconnect || cap->cap_gen == session->s_cap_gen)) { + cap->queue_release = 1; + if (removed) { + list_add_tail(&cap->session_caps, + &session->s_cap_releases); + session->s_num_cap_releases++; + removed = 0; + } + } else { + cap->queue_release = 0; + } + cap->cap_ino = ci->i_vino.ino; + spin_unlock(&session->s_cap_lock); /* remove from inode list */ @@ -977,8 +988,8 @@ void __ceph_remove_cap(struct ceph_cap *cap, bool queue_release) static int send_cap_msg(struct ceph_mds_session *session, u64 ino, u64 cid, int op, int caps, int wanted, int dirty, - u32 seq, u64 flush_tid, u32 issue_seq, u32 mseq, - u64 size, u64 max_size, + u32 seq, u64 flush_tid, u64 oldest_flush_tid, + u32 issue_seq, u32 mseq, u64 size, u64 max_size, struct timespec *mtime, struct timespec *atime, u64 time_warp_seq, kuid_t uid, kgid_t gid, umode_t mode, @@ -992,20 +1003,23 @@ static int send_cap_msg(struct ceph_mds_session *session, size_t extra_len; dout("send_cap_msg %s %llx %llx caps %s wanted %s dirty %s" - " seq %u/%u mseq %u follows %lld size %llu/%llu" + " seq %u/%u tid %llu/%llu mseq %u follows %lld size %llu/%llu" " xattr_ver %llu xattr_len %d\n", ceph_cap_op_name(op), cid, ino, ceph_cap_string(caps), ceph_cap_string(wanted), ceph_cap_string(dirty), - seq, issue_seq, mseq, follows, size, max_size, + seq, issue_seq, flush_tid, oldest_flush_tid, + mseq, follows, size, max_size, xattr_version, xattrs_buf ? (int)xattrs_buf->vec.iov_len : 0); - /* flock buffer size + inline version + inline data size */ - extra_len = 4 + 8 + 4; + /* flock buffer size + inline version + inline data size + + * osd_epoch_barrier + oldest_flush_tid */ + extra_len = 4 + 8 + 4 + 4 + 8; msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPS, sizeof(*fc) + extra_len, GFP_NOFS, false); if (!msg) return -ENOMEM; + msg->hdr.version = cpu_to_le16(6); msg->hdr.tid = cpu_to_le64(flush_tid); fc = msg->front.iov_base; @@ -1041,6 +1055,10 @@ static int send_cap_msg(struct ceph_mds_session *session, ceph_encode_64(&p, inline_data ? 0 : CEPH_INLINE_NONE); /* inline data size */ ceph_encode_32(&p, 0); + /* osd_epoch_barrier */ + ceph_encode_32(&p, 0); + /* oldest_flush_tid */ + ceph_encode_64(&p, oldest_flush_tid); fc->xattr_version = cpu_to_le64(xattr_version); if (xattrs_buf) { @@ -1053,44 +1071,6 @@ static int send_cap_msg(struct ceph_mds_session *session, return 0; } -void __queue_cap_release(struct ceph_mds_session *session, - u64 ino, u64 cap_id, u32 migrate_seq, - u32 issue_seq) -{ - struct ceph_msg *msg; - struct ceph_mds_cap_release *head; - struct ceph_mds_cap_item *item; - - BUG_ON(!session->s_num_cap_releases); - msg = list_first_entry(&session->s_cap_releases, - struct ceph_msg, list_head); - - dout(" adding %llx release to mds%d msg %p (%d left)\n", - ino, session->s_mds, msg, session->s_num_cap_releases); - - BUG_ON(msg->front.iov_len + sizeof(*item) > PAGE_CACHE_SIZE); - head = msg->front.iov_base; - le32_add_cpu(&head->num, 1); - item = msg->front.iov_base + msg->front.iov_len; - item->ino = cpu_to_le64(ino); - item->cap_id = cpu_to_le64(cap_id); - item->migrate_seq = cpu_to_le32(migrate_seq); - item->seq = cpu_to_le32(issue_seq); - - session->s_num_cap_releases--; - - msg->front.iov_len += sizeof(*item); - if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) { - dout(" release msg %p full\n", msg); - list_move_tail(&msg->list_head, &session->s_cap_releases_done); - } else { - dout(" release msg %p at %d/%d (%d)\n", msg, - (int)le32_to_cpu(head->num), - (int)CEPH_CAPS_PER_RELEASE, - (int)msg->front.iov_len); - } -} - /* * Queue cap releases when an inode is dropped from our cache. Since * inode is about to be destroyed, there is no need for i_ceph_lock. @@ -1127,7 +1107,7 @@ void ceph_queue_caps_release(struct inode *inode) */ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, int op, int used, int want, int retain, int flushing, - unsigned *pflush_tid) + u64 flush_tid, u64 oldest_flush_tid) __releases(cap->ci->i_ceph_lock) { struct ceph_inode_info *ci = cap->ci; @@ -1145,8 +1125,6 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, u64 xattr_version = 0; struct ceph_buffer *xattr_blob = NULL; int delayed = 0; - u64 flush_tid = 0; - int i; int ret; bool inline_data; @@ -1190,26 +1168,7 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, cap->implemented &= cap->issued | used; cap->mds_wanted = want; - if (flushing) { - /* - * assign a tid for flush operations so we can avoid - * flush1 -> dirty1 -> flush2 -> flushack1 -> mark - * clean type races. track latest tid for every bit - * so we can handle flush AxFw, flush Fw, and have the - * first ack clean Ax. - */ - flush_tid = ++ci->i_cap_flush_last_tid; - if (pflush_tid) - *pflush_tid = flush_tid; - dout(" cap_flush_tid %d\n", (int)flush_tid); - for (i = 0; i < CEPH_CAP_BITS; i++) - if (flushing & (1 << i)) - ci->i_cap_flush_tid[i] = flush_tid; - - follows = ci->i_head_snapc->seq; - } else { - follows = 0; - } + follows = flushing ? ci->i_head_snapc->seq : 0; keep = cap->implemented; seq = cap->seq; @@ -1237,7 +1196,8 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, spin_unlock(&ci->i_ceph_lock); ret = send_cap_msg(session, ceph_vino(inode).ino, cap_id, - op, keep, want, flushing, seq, flush_tid, issue_seq, mseq, + op, keep, want, flushing, seq, + flush_tid, oldest_flush_tid, issue_seq, mseq, size, max_size, &mtime, &atime, time_warp_seq, uid, gid, mode, xattr_version, xattr_blob, follows, inline_data); @@ -1259,14 +1219,14 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, * asynchronously back to the MDS once sync writes complete and dirty * data is written out. * - * Unless @again is true, skip cap_snaps that were already sent to + * Unless @kick is true, skip cap_snaps that were already sent to * the MDS (i.e., during this session). * * Called under i_ceph_lock. Takes s_mutex as needed. */ void __ceph_flush_snaps(struct ceph_inode_info *ci, struct ceph_mds_session **psession, - int again) + int kick) __releases(ci->i_ceph_lock) __acquires(ci->i_ceph_lock) { @@ -1297,11 +1257,8 @@ retry: if (capsnap->dirty_pages || capsnap->writing) break; - /* - * if cap writeback already occurred, we should have dropped - * the capsnap in ceph_put_wrbuffer_cap_refs. - */ - BUG_ON(capsnap->dirty == 0); + /* should be removed by ceph_try_drop_cap_snap() */ + BUG_ON(!capsnap->need_flush); /* pick mds, take s_mutex */ if (ci->i_auth_cap == NULL) { @@ -1310,7 +1267,7 @@ retry: } /* only flush each capsnap once */ - if (!again && !list_empty(&capsnap->flushing_item)) { + if (!kick && !list_empty(&capsnap->flushing_item)) { dout("already flushed %p, skipping\n", capsnap); continue; } @@ -1320,6 +1277,9 @@ retry: if (session && session->s_mds != mds) { dout("oops, wrong session %p mutex\n", session); + if (kick) + goto out; + mutex_unlock(&session->s_mutex); ceph_put_mds_session(session); session = NULL; @@ -1343,20 +1303,22 @@ retry: goto retry; } - capsnap->flush_tid = ++ci->i_cap_flush_last_tid; + spin_lock(&mdsc->cap_dirty_lock); + capsnap->flush_tid = ++mdsc->last_cap_flush_tid; + spin_unlock(&mdsc->cap_dirty_lock); + atomic_inc(&capsnap->nref); - if (!list_empty(&capsnap->flushing_item)) - list_del_init(&capsnap->flushing_item); - list_add_tail(&capsnap->flushing_item, - &session->s_cap_snaps_flushing); + if (list_empty(&capsnap->flushing_item)) + list_add_tail(&capsnap->flushing_item, + &session->s_cap_snaps_flushing); spin_unlock(&ci->i_ceph_lock); dout("flush_snaps %p cap_snap %p follows %lld tid %llu\n", inode, capsnap, capsnap->follows, capsnap->flush_tid); send_cap_msg(session, ceph_vino(inode).ino, 0, CEPH_CAP_OP_FLUSHSNAP, capsnap->issued, 0, - capsnap->dirty, 0, capsnap->flush_tid, 0, mseq, - capsnap->size, 0, + capsnap->dirty, 0, capsnap->flush_tid, 0, + 0, mseq, capsnap->size, 0, &capsnap->mtime, &capsnap->atime, capsnap->time_warp_seq, capsnap->uid, capsnap->gid, capsnap->mode, @@ -1396,7 +1358,8 @@ static void ceph_flush_snaps(struct ceph_inode_info *ci) * Caller is then responsible for calling __mark_inode_dirty with the * returned flags value. */ -int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) +int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask, + struct ceph_cap_flush **pcf) { struct ceph_mds_client *mdsc = ceph_sb_to_client(ci->vfs_inode.i_sb)->mdsc; @@ -1416,9 +1379,14 @@ int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) ceph_cap_string(was | mask)); ci->i_dirty_caps |= mask; if (was == 0) { - if (!ci->i_head_snapc) + WARN_ON_ONCE(ci->i_prealloc_cap_flush); + swap(ci->i_prealloc_cap_flush, *pcf); + + if (!ci->i_head_snapc) { + WARN_ON_ONCE(!rwsem_is_locked(&mdsc->snap_rwsem)); ci->i_head_snapc = ceph_get_snap_context( ci->i_snap_realm->cached_context); + } dout(" inode %p now dirty snapc %p auth cap %p\n", &ci->vfs_inode, ci->i_head_snapc, ci->i_auth_cap); BUG_ON(!list_empty(&ci->i_dirty_item)); @@ -1429,6 +1397,8 @@ int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) ihold(inode); dirty |= I_DIRTY_SYNC; } + } else { + WARN_ON_ONCE(!ci->i_prealloc_cap_flush); } BUG_ON(list_empty(&ci->i_dirty_item)); if (((was | ci->i_flushing_caps) & CEPH_CAP_FILE_BUFFER) && @@ -1438,6 +1408,74 @@ int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) return dirty; } +static void __add_cap_flushing_to_inode(struct ceph_inode_info *ci, + struct ceph_cap_flush *cf) +{ + struct rb_node **p = &ci->i_cap_flush_tree.rb_node; + struct rb_node *parent = NULL; + struct ceph_cap_flush *other = NULL; + + while (*p) { + parent = *p; + other = rb_entry(parent, struct ceph_cap_flush, i_node); + + if (cf->tid < other->tid) + p = &(*p)->rb_left; + else if (cf->tid > other->tid) + p = &(*p)->rb_right; + else + BUG(); + } + + rb_link_node(&cf->i_node, parent, p); + rb_insert_color(&cf->i_node, &ci->i_cap_flush_tree); +} + +static void __add_cap_flushing_to_mdsc(struct ceph_mds_client *mdsc, + struct ceph_cap_flush *cf) +{ + struct rb_node **p = &mdsc->cap_flush_tree.rb_node; + struct rb_node *parent = NULL; + struct ceph_cap_flush *other = NULL; + + while (*p) { + parent = *p; + other = rb_entry(parent, struct ceph_cap_flush, g_node); + + if (cf->tid < other->tid) + p = &(*p)->rb_left; + else if (cf->tid > other->tid) + p = &(*p)->rb_right; + else + BUG(); + } + + rb_link_node(&cf->g_node, parent, p); + rb_insert_color(&cf->g_node, &mdsc->cap_flush_tree); +} + +struct ceph_cap_flush *ceph_alloc_cap_flush(void) +{ + return kmem_cache_alloc(ceph_cap_flush_cachep, GFP_KERNEL); +} + +void ceph_free_cap_flush(struct ceph_cap_flush *cf) +{ + if (cf) + kmem_cache_free(ceph_cap_flush_cachep, cf); +} + +static u64 __get_oldest_flush_tid(struct ceph_mds_client *mdsc) +{ + struct rb_node *n = rb_first(&mdsc->cap_flush_tree); + if (n) { + struct ceph_cap_flush *cf = + rb_entry(n, struct ceph_cap_flush, g_node); + return cf->tid; + } + return 0; +} + /* * Add dirty inode to the flushing list. Assigned a seq number so we * can wait for caps to flush without starving. @@ -1445,14 +1483,17 @@ int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) * Called under i_ceph_lock. */ static int __mark_caps_flushing(struct inode *inode, - struct ceph_mds_session *session) + struct ceph_mds_session *session, + u64 *flush_tid, u64 *oldest_flush_tid) { struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; struct ceph_inode_info *ci = ceph_inode(inode); + struct ceph_cap_flush *cf = NULL; int flushing; BUG_ON(ci->i_dirty_caps == 0); BUG_ON(list_empty(&ci->i_dirty_item)); + BUG_ON(!ci->i_prealloc_cap_flush); flushing = ci->i_dirty_caps; dout("__mark_caps_flushing flushing %s, flushing_caps %s -> %s\n", @@ -1463,22 +1504,31 @@ static int __mark_caps_flushing(struct inode *inode, ci->i_dirty_caps = 0; dout(" inode %p now !dirty\n", inode); + swap(cf, ci->i_prealloc_cap_flush); + cf->caps = flushing; + cf->kick = false; + spin_lock(&mdsc->cap_dirty_lock); list_del_init(&ci->i_dirty_item); + cf->tid = ++mdsc->last_cap_flush_tid; + __add_cap_flushing_to_mdsc(mdsc, cf); + *oldest_flush_tid = __get_oldest_flush_tid(mdsc); + if (list_empty(&ci->i_flushing_item)) { - ci->i_cap_flush_seq = ++mdsc->cap_flush_seq; list_add_tail(&ci->i_flushing_item, &session->s_cap_flushing); mdsc->num_cap_flushing++; - dout(" inode %p now flushing seq %lld\n", inode, - ci->i_cap_flush_seq); + dout(" inode %p now flushing tid %llu\n", inode, cf->tid); } else { list_move_tail(&ci->i_flushing_item, &session->s_cap_flushing); - dout(" inode %p now flushing (more) seq %lld\n", inode, - ci->i_cap_flush_seq); + dout(" inode %p now flushing (more) tid %llu\n", + inode, cf->tid); } spin_unlock(&mdsc->cap_dirty_lock); + __add_cap_flushing_to_inode(ci, cf); + + *flush_tid = cf->tid; return flushing; } @@ -1524,6 +1574,7 @@ void ceph_check_caps(struct ceph_inode_info *ci, int flags, struct ceph_mds_client *mdsc = fsc->mdsc; struct inode *inode = &ci->vfs_inode; struct ceph_cap *cap; + u64 flush_tid, oldest_flush_tid; int file_wanted, used, cap_used; int took_snap_rwsem = 0; /* true if mdsc->snap_rwsem held */ int issued, implemented, want, retain, revoking, flushing = 0; @@ -1553,13 +1604,13 @@ retry: retry_locked: file_wanted = __ceph_caps_file_wanted(ci); used = __ceph_caps_used(ci); - want = file_wanted | used; issued = __ceph_caps_issued(ci, &implemented); revoking = implemented & ~issued; - retain = want | CEPH_CAP_PIN; + want = file_wanted; + retain = file_wanted | used | CEPH_CAP_PIN; if (!mdsc->stopping && inode->i_nlink > 0) { - if (want) { + if (file_wanted) { retain |= CEPH_CAP_ANY; /* be greedy */ } else if (S_ISDIR(inode->i_mode) && (issued & CEPH_CAP_FILE_SHARED) && @@ -1602,9 +1653,10 @@ retry_locked: * If we fail, it's because pages are locked.... try again later. */ if ((!is_delayed || mdsc->stopping) && - ci->i_wrbuffer_ref == 0 && /* no dirty pages... */ - inode->i_data.nrpages && /* have cached pages */ - (file_wanted == 0 || /* no open files */ + !S_ISDIR(inode->i_mode) && /* ignore readdir cache */ + ci->i_wrbuffer_ref == 0 && /* no dirty pages... */ + inode->i_data.nrpages && /* have cached pages */ + (file_wanted == 0 || /* no open files */ (revoking & (CEPH_CAP_FILE_CACHE| CEPH_CAP_FILE_LAZYIO))) && /* or revoking cache */ !tried_invalidate) { @@ -1742,17 +1794,25 @@ ack: took_snap_rwsem = 1; } - if (cap == ci->i_auth_cap && ci->i_dirty_caps) - flushing = __mark_caps_flushing(inode, session); - else + if (cap == ci->i_auth_cap && ci->i_dirty_caps) { + flushing = __mark_caps_flushing(inode, session, + &flush_tid, + &oldest_flush_tid); + } else { flushing = 0; + flush_tid = 0; + spin_lock(&mdsc->cap_dirty_lock); + oldest_flush_tid = __get_oldest_flush_tid(mdsc); + spin_unlock(&mdsc->cap_dirty_lock); + } mds = cap->mds; /* remember mds, so we don't repeat */ sent++; /* __send_cap drops i_ceph_lock */ delayed += __send_cap(mdsc, cap, CEPH_CAP_OP_UPDATE, cap_used, - want, retain, flushing, NULL); + want, retain, flushing, + flush_tid, oldest_flush_tid); goto retry; /* retake i_ceph_lock and restart our cap scan. */ } @@ -1781,12 +1841,13 @@ ack: /* * Try to flush dirty caps back to the auth mds. */ -static int try_flush_caps(struct inode *inode, unsigned *flush_tid) +static int try_flush_caps(struct inode *inode, u64 *ptid) { struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; struct ceph_inode_info *ci = ceph_inode(inode); - int flushing = 0; struct ceph_mds_session *session = NULL; + int flushing = 0; + u64 flush_tid = 0, oldest_flush_tid = 0; retry: spin_lock(&ci->i_ceph_lock); @@ -1811,42 +1872,54 @@ retry: if (cap->session->s_state < CEPH_MDS_SESSION_OPEN) goto out; - flushing = __mark_caps_flushing(inode, session); + flushing = __mark_caps_flushing(inode, session, &flush_tid, + &oldest_flush_tid); /* __send_cap drops i_ceph_lock */ delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, used, want, - cap->issued | cap->implemented, flushing, - flush_tid); - if (!delayed) - goto out_unlocked; + (cap->issued | cap->implemented), + flushing, flush_tid, oldest_flush_tid); - spin_lock(&ci->i_ceph_lock); - __cap_delay_requeue(mdsc, ci); + if (delayed) { + spin_lock(&ci->i_ceph_lock); + __cap_delay_requeue(mdsc, ci); + spin_unlock(&ci->i_ceph_lock); + } + } else { + struct rb_node *n = rb_last(&ci->i_cap_flush_tree); + if (n) { + struct ceph_cap_flush *cf = + rb_entry(n, struct ceph_cap_flush, i_node); + flush_tid = cf->tid; + } + flushing = ci->i_flushing_caps; + spin_unlock(&ci->i_ceph_lock); } out: - spin_unlock(&ci->i_ceph_lock); -out_unlocked: if (session) mutex_unlock(&session->s_mutex); + + *ptid = flush_tid; return flushing; } /* * Return true if we've flushed caps through the given flush_tid. */ -static int caps_are_flushed(struct inode *inode, unsigned tid) +static int caps_are_flushed(struct inode *inode, u64 flush_tid) { struct ceph_inode_info *ci = ceph_inode(inode); - int i, ret = 1; + struct ceph_cap_flush *cf; + struct rb_node *n; + int ret = 1; spin_lock(&ci->i_ceph_lock); - for (i = 0; i < CEPH_CAP_BITS; i++) - if ((ci->i_flushing_caps & (1 << i)) && - ci->i_cap_flush_tid[i] <= tid) { - /* still flushing this bit */ + n = rb_first(&ci->i_cap_flush_tree); + if (n) { + cf = rb_entry(n, struct ceph_cap_flush, i_node); + if (cf->tid <= flush_tid) ret = 0; - break; - } + } spin_unlock(&ci->i_ceph_lock); return ret; } @@ -1864,13 +1937,16 @@ static void sync_write_wait(struct inode *inode) struct ceph_osd_request *req; u64 last_tid; + if (!S_ISREG(inode->i_mode)) + return; + spin_lock(&ci->i_unsafe_lock); if (list_empty(head)) goto out; /* set upper bound as _last_ entry in chain */ - req = list_entry(head->prev, struct ceph_osd_request, - r_unsafe_item); + req = list_last_entry(head, struct ceph_osd_request, + r_unsafe_item); last_tid = req->r_tid; do { @@ -1888,18 +1964,64 @@ static void sync_write_wait(struct inode *inode) */ if (list_empty(head)) break; - req = list_entry(head->next, struct ceph_osd_request, - r_unsafe_item); + req = list_first_entry(head, struct ceph_osd_request, + r_unsafe_item); } while (req->r_tid < last_tid); out: spin_unlock(&ci->i_unsafe_lock); } +/* + * wait for any uncommitted directory operations to commit. + */ +static int unsafe_dirop_wait(struct inode *inode) +{ + struct ceph_inode_info *ci = ceph_inode(inode); + struct list_head *head = &ci->i_unsafe_dirops; + struct ceph_mds_request *req; + u64 last_tid; + int ret = 0; + + if (!S_ISDIR(inode->i_mode)) + return 0; + + spin_lock(&ci->i_unsafe_lock); + if (list_empty(head)) + goto out; + + req = list_last_entry(head, struct ceph_mds_request, + r_unsafe_dir_item); + last_tid = req->r_tid; + + do { + ceph_mdsc_get_request(req); + spin_unlock(&ci->i_unsafe_lock); + + dout("unsafe_dirop_wait %p wait on tid %llu (until %llu)\n", + inode, req->r_tid, last_tid); + ret = !wait_for_completion_timeout(&req->r_safe_completion, + ceph_timeout_jiffies(req->r_timeout)); + if (ret) + ret = -EIO; /* timed out */ + + ceph_mdsc_put_request(req); + + spin_lock(&ci->i_unsafe_lock); + if (ret || list_empty(head)) + break; + req = list_first_entry(head, struct ceph_mds_request, + r_unsafe_dir_item); + } while (req->r_tid < last_tid); +out: + spin_unlock(&ci->i_unsafe_lock); + return ret; +} + int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync) { struct inode *inode = file->f_mapping->host; struct ceph_inode_info *ci = ceph_inode(inode); - unsigned flush_tid; + u64 flush_tid; int ret; int dirty; @@ -1908,25 +2030,30 @@ int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync) ret = filemap_write_and_wait_range(inode->i_mapping, start, end); if (ret < 0) - return ret; + goto out; + + if (datasync) + goto out; + mutex_lock(&inode->i_mutex); dirty = try_flush_caps(inode, &flush_tid); dout("fsync dirty caps are %s\n", ceph_cap_string(dirty)); + ret = unsafe_dirop_wait(inode); + /* * only wait on non-file metadata writeback (the mds * can recover size and mtime, so we don't need to * wait for that) */ - if (!datasync && (dirty & ~CEPH_CAP_ANY_FILE_WR)) { - dout("fsync waiting for flush_tid %u\n", flush_tid); + if (!ret && (dirty & ~CEPH_CAP_ANY_FILE_WR)) { ret = wait_event_interruptible(ci->i_cap_wq, - caps_are_flushed(inode, flush_tid)); + caps_are_flushed(inode, flush_tid)); } - - dout("fsync %p%s done\n", inode, datasync ? " datasync" : ""); mutex_unlock(&inode->i_mutex); +out: + dout("fsync %p%s result=%d\n", inode, datasync ? " datasync" : "", ret); return ret; } @@ -1939,7 +2066,7 @@ int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync) int ceph_write_inode(struct inode *inode, struct writeback_control *wbc) { struct ceph_inode_info *ci = ceph_inode(inode); - unsigned flush_tid; + u64 flush_tid; int err = 0; int dirty; int wait = wbc->sync_mode == WB_SYNC_ALL; @@ -1994,6 +2121,104 @@ static void kick_flushing_capsnaps(struct ceph_mds_client *mdsc, } } +static int __kick_flushing_caps(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session, + struct ceph_inode_info *ci, + bool kick_all) +{ + struct inode *inode = &ci->vfs_inode; + struct ceph_cap *cap; + struct ceph_cap_flush *cf; + struct rb_node *n; + int delayed = 0; + u64 first_tid = 0; + u64 oldest_flush_tid; + + spin_lock(&mdsc->cap_dirty_lock); + oldest_flush_tid = __get_oldest_flush_tid(mdsc); + spin_unlock(&mdsc->cap_dirty_lock); + + while (true) { + spin_lock(&ci->i_ceph_lock); + cap = ci->i_auth_cap; + if (!(cap && cap->session == session)) { + pr_err("%p auth cap %p not mds%d ???\n", inode, + cap, session->s_mds); + spin_unlock(&ci->i_ceph_lock); + break; + } + + for (n = rb_first(&ci->i_cap_flush_tree); n; n = rb_next(n)) { + cf = rb_entry(n, struct ceph_cap_flush, i_node); + if (cf->tid < first_tid) + continue; + if (kick_all || cf->kick) + break; + } + if (!n) { + spin_unlock(&ci->i_ceph_lock); + break; + } + + cf = rb_entry(n, struct ceph_cap_flush, i_node); + cf->kick = false; + + first_tid = cf->tid + 1; + + dout("kick_flushing_caps %p cap %p tid %llu %s\n", inode, + cap, cf->tid, ceph_cap_string(cf->caps)); + delayed |= __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, + __ceph_caps_used(ci), + __ceph_caps_wanted(ci), + cap->issued | cap->implemented, + cf->caps, cf->tid, oldest_flush_tid); + } + return delayed; +} + +void ceph_early_kick_flushing_caps(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session) +{ + struct ceph_inode_info *ci; + struct ceph_cap *cap; + struct ceph_cap_flush *cf; + struct rb_node *n; + + dout("early_kick_flushing_caps mds%d\n", session->s_mds); + list_for_each_entry(ci, &session->s_cap_flushing, i_flushing_item) { + spin_lock(&ci->i_ceph_lock); + cap = ci->i_auth_cap; + if (!(cap && cap->session == session)) { + pr_err("%p auth cap %p not mds%d ???\n", + &ci->vfs_inode, cap, session->s_mds); + spin_unlock(&ci->i_ceph_lock); + continue; + } + + + /* + * if flushing caps were revoked, we re-send the cap flush + * in client reconnect stage. This guarantees MDS * processes + * the cap flush message before issuing the flushing caps to + * other client. + */ + if ((cap->issued & ci->i_flushing_caps) != + ci->i_flushing_caps) { + spin_unlock(&ci->i_ceph_lock); + if (!__kick_flushing_caps(mdsc, session, ci, true)) + continue; + spin_lock(&ci->i_ceph_lock); + } + + for (n = rb_first(&ci->i_cap_flush_tree); n; n = rb_next(n)) { + cf = rb_entry(n, struct ceph_cap_flush, i_node); + cf->kick = true; + } + + spin_unlock(&ci->i_ceph_lock); + } +} + void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc, struct ceph_mds_session *session) { @@ -2003,28 +2228,10 @@ void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc, dout("kick_flushing_caps mds%d\n", session->s_mds); list_for_each_entry(ci, &session->s_cap_flushing, i_flushing_item) { - struct inode *inode = &ci->vfs_inode; - struct ceph_cap *cap; - int delayed = 0; - - spin_lock(&ci->i_ceph_lock); - cap = ci->i_auth_cap; - if (cap && cap->session == session) { - dout("kick_flushing_caps %p cap %p %s\n", inode, - cap, ceph_cap_string(ci->i_flushing_caps)); - delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, - __ceph_caps_used(ci), - __ceph_caps_wanted(ci), - cap->issued | cap->implemented, - ci->i_flushing_caps, NULL); - if (delayed) { - spin_lock(&ci->i_ceph_lock); - __cap_delay_requeue(mdsc, ci); - spin_unlock(&ci->i_ceph_lock); - } - } else { - pr_err("%p auth cap %p not mds%d ???\n", inode, - cap, session->s_mds); + int delayed = __kick_flushing_caps(mdsc, session, ci, false); + if (delayed) { + spin_lock(&ci->i_ceph_lock); + __cap_delay_requeue(mdsc, ci); spin_unlock(&ci->i_ceph_lock); } } @@ -2036,26 +2243,25 @@ static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc, { struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_cap *cap; - int delayed = 0; spin_lock(&ci->i_ceph_lock); cap = ci->i_auth_cap; - dout("kick_flushing_inode_caps %p flushing %s flush_seq %lld\n", inode, - ceph_cap_string(ci->i_flushing_caps), ci->i_cap_flush_seq); + dout("kick_flushing_inode_caps %p flushing %s\n", inode, + ceph_cap_string(ci->i_flushing_caps)); __ceph_flush_snaps(ci, &session, 1); if (ci->i_flushing_caps) { + int delayed; + spin_lock(&mdsc->cap_dirty_lock); list_move_tail(&ci->i_flushing_item, &cap->session->s_cap_flushing); spin_unlock(&mdsc->cap_dirty_lock); - delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, - __ceph_caps_used(ci), - __ceph_caps_wanted(ci), - cap->issued | cap->implemented, - ci->i_flushing_caps, NULL); + spin_unlock(&ci->i_ceph_lock); + + delayed = __kick_flushing_caps(mdsc, session, ci, true); if (delayed) { spin_lock(&ci->i_ceph_lock); __cap_delay_requeue(mdsc, ci); @@ -2073,7 +2279,8 @@ static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc, * * Protected by i_ceph_lock. */ -static void __take_cap_refs(struct ceph_inode_info *ci, int got) +static void __take_cap_refs(struct ceph_inode_info *ci, int got, + bool snap_rwsem_locked) { if (got & CEPH_CAP_PIN) ci->i_pin_ref++; @@ -2081,8 +2288,14 @@ static void __take_cap_refs(struct ceph_inode_info *ci, int got) ci->i_rd_ref++; if (got & CEPH_CAP_FILE_CACHE) ci->i_rdcache_ref++; - if (got & CEPH_CAP_FILE_WR) + if (got & CEPH_CAP_FILE_WR) { + if (ci->i_wr_ref == 0 && !ci->i_head_snapc) { + BUG_ON(!snap_rwsem_locked); + ci->i_head_snapc = ceph_get_snap_context( + ci->i_snap_realm->cached_context); + } ci->i_wr_ref++; + } if (got & CEPH_CAP_FILE_BUFFER) { if (ci->i_wb_ref == 0) ihold(&ci->vfs_inode); @@ -2100,16 +2313,19 @@ static void __take_cap_refs(struct ceph_inode_info *ci, int got) * requested from the MDS. */ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want, - loff_t endoff, int *got, int *check_max, int *err) + loff_t endoff, bool nonblock, int *got, int *err) { struct inode *inode = &ci->vfs_inode; + struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; int ret = 0; int have, implemented; int file_wanted; + bool snap_rwsem_locked = false; dout("get_cap_refs %p need %s want %s\n", inode, ceph_cap_string(need), ceph_cap_string(want)); +again: spin_lock(&ci->i_ceph_lock); /* make sure file is actually open */ @@ -2125,6 +2341,10 @@ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want, /* finish pending truncate */ while (ci->i_truncate_pending) { spin_unlock(&ci->i_ceph_lock); + if (snap_rwsem_locked) { + up_read(&mdsc->snap_rwsem); + snap_rwsem_locked = false; + } __ceph_do_pending_vmtruncate(inode); spin_lock(&ci->i_ceph_lock); } @@ -2136,7 +2356,7 @@ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want, dout("get_cap_refs %p endoff %llu > maxsize %llu\n", inode, endoff, ci->i_max_size); if (endoff > ci->i_requested_max_size) { - *check_max = 1; + *err = -EAGAIN; ret = 1; } goto out_unlock; @@ -2164,8 +2384,29 @@ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want, inode, ceph_cap_string(have), ceph_cap_string(not), ceph_cap_string(revoking)); if ((revoking & not) == 0) { + if (!snap_rwsem_locked && + !ci->i_head_snapc && + (need & CEPH_CAP_FILE_WR)) { + if (!down_read_trylock(&mdsc->snap_rwsem)) { + /* + * we can not call down_read() when + * task isn't in TASK_RUNNING state + */ + if (nonblock) { + *err = -EAGAIN; + ret = 1; + goto out_unlock; + } + + spin_unlock(&ci->i_ceph_lock); + down_read(&mdsc->snap_rwsem); + snap_rwsem_locked = true; + goto again; + } + snap_rwsem_locked = true; + } *got = need | (have & want); - __take_cap_refs(ci, *got); + __take_cap_refs(ci, *got, true); ret = 1; } } else { @@ -2189,6 +2430,8 @@ static int try_get_cap_refs(struct ceph_inode_info *ci, int need, int want, } out_unlock: spin_unlock(&ci->i_ceph_lock); + if (snap_rwsem_locked) + up_read(&mdsc->snap_rwsem); dout("get_cap_refs %p ret %d got %s\n", inode, ret, ceph_cap_string(*got)); @@ -2231,50 +2474,70 @@ static void check_max_size(struct inode *inode, loff_t endoff) int ceph_get_caps(struct ceph_inode_info *ci, int need, int want, loff_t endoff, int *got, struct page **pinned_page) { - int _got, check_max, ret, err = 0; + int _got, ret, err = 0; -retry: - if (endoff > 0) - check_max_size(&ci->vfs_inode, endoff); - _got = 0; - check_max = 0; - ret = wait_event_interruptible(ci->i_cap_wq, - try_get_cap_refs(ci, need, want, endoff, - &_got, &check_max, &err)); - if (err) - ret = err; + ret = ceph_pool_perm_check(ci, need); if (ret < 0) return ret; - if (check_max) - goto retry; + while (true) { + if (endoff > 0) + check_max_size(&ci->vfs_inode, endoff); - if (ci->i_inline_version != CEPH_INLINE_NONE && - (_got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) && - i_size_read(&ci->vfs_inode) > 0) { - struct page *page = find_get_page(ci->vfs_inode.i_mapping, 0); - if (page) { - if (PageUptodate(page)) { - *pinned_page = page; - goto out; - } - page_cache_release(page); - } - /* - * drop cap refs first because getattr while holding - * caps refs can cause deadlock. - */ - ceph_put_cap_refs(ci, _got); + err = 0; _got = 0; + ret = try_get_cap_refs(ci, need, want, endoff, + false, &_got, &err); + if (ret) { + if (err == -EAGAIN) + continue; + if (err < 0) + return err; + } else { + ret = wait_event_interruptible(ci->i_cap_wq, + try_get_cap_refs(ci, need, want, endoff, + true, &_got, &err)); + if (err == -EAGAIN) + continue; + if (err < 0) + ret = err; + if (ret < 0) + return ret; + } - /* getattr request will bring inline data into page cache */ - ret = __ceph_do_getattr(&ci->vfs_inode, NULL, - CEPH_STAT_CAP_INLINE_DATA, true); - if (ret < 0) - return ret; - goto retry; + if (ci->i_inline_version != CEPH_INLINE_NONE && + (_got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) && + i_size_read(&ci->vfs_inode) > 0) { + struct page *page = + find_get_page(ci->vfs_inode.i_mapping, 0); + if (page) { + if (PageUptodate(page)) { + *pinned_page = page; + break; + } + page_cache_release(page); + } + /* + * drop cap refs first because getattr while + * holding * caps refs can cause deadlock. + */ + ceph_put_cap_refs(ci, _got); + _got = 0; + + /* + * getattr request will bring inline data into + * page cache + */ + ret = __ceph_do_getattr(&ci->vfs_inode, NULL, + CEPH_STAT_CAP_INLINE_DATA, + true); + if (ret < 0) + return ret; + continue; + } + break; } -out: + *got = _got; return 0; } @@ -2286,10 +2549,31 @@ out: void ceph_get_cap_refs(struct ceph_inode_info *ci, int caps) { spin_lock(&ci->i_ceph_lock); - __take_cap_refs(ci, caps); + __take_cap_refs(ci, caps, false); spin_unlock(&ci->i_ceph_lock); } + +/* + * drop cap_snap that is not associated with any snapshot. + * we don't need to send FLUSHSNAP message for it. + */ +static int ceph_try_drop_cap_snap(struct ceph_cap_snap *capsnap) +{ + if (!capsnap->need_flush && + !capsnap->writing && !capsnap->dirty_pages) { + + dout("dropping cap_snap %p follows %llu\n", + capsnap, capsnap->follows); + ceph_put_snap_context(capsnap->context); + list_del(&capsnap->ci_item); + list_del(&capsnap->flushing_item); + ceph_put_cap_snap(capsnap); + return 1; + } + return 0; +} + /* * Release cap refs. * @@ -2303,7 +2587,6 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had) { struct inode *inode = &ci->vfs_inode; int last = 0, put = 0, flushsnaps = 0, wake = 0; - struct ceph_cap_snap *capsnap; spin_lock(&ci->i_ceph_lock); if (had & CEPH_CAP_PIN) @@ -2325,17 +2608,24 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had) if (had & CEPH_CAP_FILE_WR) if (--ci->i_wr_ref == 0) { last++; - if (!list_empty(&ci->i_cap_snaps)) { - capsnap = list_first_entry(&ci->i_cap_snaps, - struct ceph_cap_snap, - ci_item); - if (capsnap->writing) { - capsnap->writing = 0; - flushsnaps = - __ceph_finish_cap_snap(ci, - capsnap); - wake = 1; - } + if (__ceph_have_pending_cap_snap(ci)) { + struct ceph_cap_snap *capsnap = + list_last_entry(&ci->i_cap_snaps, + struct ceph_cap_snap, + ci_item); + capsnap->writing = 0; + if (ceph_try_drop_cap_snap(capsnap)) + put++; + else if (__ceph_finish_cap_snap(ci, capsnap)) + flushsnaps = 1; + wake = 1; + } + if (ci->i_wrbuffer_ref_head == 0 && + ci->i_dirty_caps == 0 && + ci->i_flushing_caps == 0) { + BUG_ON(!ci->i_head_snapc); + ceph_put_snap_context(ci->i_head_snapc); + ci->i_head_snapc = NULL; } /* see comment in __ceph_remove_cap() */ if (!__ceph_is_any_caps(ci) && ci->i_snap_realm) @@ -2352,7 +2642,7 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had) ceph_flush_snaps(ci); if (wake) wake_up_all(&ci->i_cap_wq); - if (put) + while (put-- > 0) iput(inode); } @@ -2380,7 +2670,9 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, if (ci->i_head_snapc == snapc) { ci->i_wrbuffer_ref_head -= nr; if (ci->i_wrbuffer_ref_head == 0 && - ci->i_dirty_caps == 0 && ci->i_flushing_caps == 0) { + ci->i_wr_ref == 0 && + ci->i_dirty_caps == 0 && + ci->i_flushing_caps == 0) { BUG_ON(!ci->i_head_snapc); ceph_put_snap_context(ci->i_head_snapc); ci->i_head_snapc = NULL; @@ -2401,25 +2693,15 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, capsnap->dirty_pages -= nr; if (capsnap->dirty_pages == 0) { complete_capsnap = 1; - if (capsnap->dirty == 0) - /* cap writeback completed before we created - * the cap_snap; no FLUSHSNAP is needed */ - drop_capsnap = 1; + drop_capsnap = ceph_try_drop_cap_snap(capsnap); } dout("put_wrbuffer_cap_refs on %p cap_snap %p " - " snap %lld %d/%d -> %d/%d %s%s%s\n", + " snap %lld %d/%d -> %d/%d %s%s\n", inode, capsnap, capsnap->context->seq, ci->i_wrbuffer_ref+nr, capsnap->dirty_pages + nr, ci->i_wrbuffer_ref, capsnap->dirty_pages, last ? " (wrbuffer last)" : "", - complete_capsnap ? " (complete capsnap)" : "", - drop_capsnap ? " (drop capsnap)" : ""); - if (drop_capsnap) { - ceph_put_snap_context(capsnap->context); - list_del(&capsnap->ci_item); - list_del(&capsnap->flushing_item); - ceph_put_cap_snap(capsnap); - } + complete_capsnap ? " (complete capsnap)" : ""); } spin_unlock(&ci->i_ceph_lock); @@ -2526,7 +2808,8 @@ static void handle_cap_grant(struct ceph_mds_client *mdsc, * try to invalidate (once). (If there are dirty buffers, we * will invalidate _after_ writeback.) */ - if (((cap->issued & ~newcaps) & CEPH_CAP_FILE_CACHE) && + if (!S_ISDIR(inode->i_mode) && /* don't invalidate readdir cache */ + ((cap->issued & ~newcaps) & CEPH_CAP_FILE_CACHE) && (newcaps & CEPH_CAP_FILE_LAZYIO) == 0 && !ci->i_wrbuffer_ref) { if (try_nonblocking_invalidate(inode)) { @@ -2732,16 +3015,29 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid, { struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; + struct ceph_cap_flush *cf; + struct rb_node *n; + LIST_HEAD(to_remove); unsigned seq = le32_to_cpu(m->seq); int dirty = le32_to_cpu(m->dirty); int cleaned = 0; int drop = 0; - int i; - for (i = 0; i < CEPH_CAP_BITS; i++) - if ((dirty & (1 << i)) && - (u16)flush_tid == ci->i_cap_flush_tid[i]) - cleaned |= 1 << i; + n = rb_first(&ci->i_cap_flush_tree); + while (n) { + cf = rb_entry(n, struct ceph_cap_flush, i_node); + n = rb_next(&cf->i_node); + if (cf->tid == flush_tid) + cleaned = cf->caps; + if (cf->tid <= flush_tid) { + rb_erase(&cf->i_node, &ci->i_cap_flush_tree); + list_add_tail(&cf->list, &to_remove); + } else { + cleaned &= ~cf->caps; + if (!cleaned) + break; + } + } dout("handle_cap_flush_ack inode %p mds%d seq %d on %s cleaned %s," " flushing %s -> %s\n", @@ -2749,12 +3045,23 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid, ceph_cap_string(cleaned), ceph_cap_string(ci->i_flushing_caps), ceph_cap_string(ci->i_flushing_caps & ~cleaned)); - if (ci->i_flushing_caps == (ci->i_flushing_caps & ~cleaned)) + if (list_empty(&to_remove) && !cleaned) goto out; ci->i_flushing_caps &= ~cleaned; spin_lock(&mdsc->cap_dirty_lock); + + if (!list_empty(&to_remove)) { + list_for_each_entry(cf, &to_remove, list) + rb_erase(&cf->g_node, &mdsc->cap_flush_tree); + + n = rb_first(&mdsc->cap_flush_tree); + cf = n ? rb_entry(n, struct ceph_cap_flush, g_node) : NULL; + if (!cf || cf->tid > flush_tid) + wake_up_all(&mdsc->cap_flushing_wq); + } + if (ci->i_flushing_caps == 0) { list_del_init(&ci->i_flushing_item); if (!list_empty(&session->s_cap_flushing)) @@ -2764,14 +3071,14 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid, struct ceph_inode_info, i_flushing_item)->vfs_inode); mdsc->num_cap_flushing--; - wake_up_all(&mdsc->cap_flushing_wq); dout(" inode %p now !flushing\n", inode); if (ci->i_dirty_caps == 0) { dout(" inode %p now clean\n", inode); BUG_ON(!list_empty(&ci->i_dirty_item)); drop = 1; - if (ci->i_wrbuffer_ref_head == 0) { + if (ci->i_wr_ref == 0 && + ci->i_wrbuffer_ref_head == 0) { BUG_ON(!ci->i_head_snapc); ceph_put_snap_context(ci->i_head_snapc); ci->i_head_snapc = NULL; @@ -2785,6 +3092,13 @@ static void handle_cap_flush_ack(struct inode *inode, u64 flush_tid, out: spin_unlock(&ci->i_ceph_lock); + + while (!list_empty(&to_remove)) { + cf = list_first_entry(&to_remove, + struct ceph_cap_flush, list); + list_del(&cf->list); + ceph_free_cap_flush(cf); + } if (drop) iput(inode); } @@ -2800,6 +3114,7 @@ static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid, struct ceph_mds_session *session) { struct ceph_inode_info *ci = ceph_inode(inode); + struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; u64 follows = le64_to_cpu(m->snap_follows); struct ceph_cap_snap *capsnap; int drop = 0; @@ -2823,6 +3138,7 @@ static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid, list_del(&capsnap->ci_item); list_del(&capsnap->flushing_item); ceph_put_cap_snap(capsnap); + wake_up_all(&mdsc->cap_flushing_wq); drop = 1; break; } else { @@ -2971,7 +3287,6 @@ retry: mutex_lock_nested(&session->s_mutex, SINGLE_DEPTH_NESTING); } - ceph_add_cap_releases(mdsc, tsession); new_cap = ceph_get_cap(mdsc, NULL); } else { WARN_ON(1); @@ -3167,16 +3482,20 @@ void ceph_handle_caps(struct ceph_mds_session *session, dout(" mds%d seq %lld cap seq %u\n", session->s_mds, session->s_seq, (unsigned)seq); - if (op == CEPH_CAP_OP_IMPORT) - ceph_add_cap_releases(mdsc, session); - if (!inode) { dout(" i don't have ino %llx\n", vino.ino); if (op == CEPH_CAP_OP_IMPORT) { + cap = ceph_get_cap(mdsc, NULL); + cap->cap_ino = vino.ino; + cap->queue_release = 1; + cap->cap_id = cap_id; + cap->mseq = mseq; + cap->seq = seq; spin_lock(&session->s_cap_lock); - __queue_cap_release(session, vino.ino, cap_id, - mseq, seq); + list_add_tail(&cap->session_caps, + &session->s_cap_releases); + session->s_num_cap_releases++; spin_unlock(&session->s_cap_lock); } goto flush_cap_releases; @@ -3252,11 +3571,10 @@ void ceph_handle_caps(struct ceph_mds_session *session, flush_cap_releases: /* - * send any full release message to try to move things + * send any cap release message to try to move things * along for the mds (who clearly thinks we still have this * cap). */ - ceph_add_cap_releases(mdsc, session); ceph_send_cap_releases(mdsc, session); done: diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index edbb8da..9314b4e 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -38,7 +38,7 @@ int ceph_init_dentry(struct dentry *dentry) if (dentry->d_fsdata) return 0; - di = kmem_cache_alloc(ceph_dentry_cachep, GFP_NOFS | __GFP_ZERO); + di = kmem_cache_alloc(ceph_dentry_cachep, GFP_KERNEL | __GFP_ZERO); if (!di) return -ENOMEM; /* oh well */ @@ -107,6 +107,27 @@ static int fpos_cmp(loff_t l, loff_t r) } /* + * make note of the last dentry we read, so we can + * continue at the same lexicographical point, + * regardless of what dir changes take place on the + * server. + */ +static int note_last_dentry(struct ceph_file_info *fi, const char *name, + int len, unsigned next_offset) +{ + char *buf = kmalloc(len+1, GFP_KERNEL); + if (!buf) + return -ENOMEM; + kfree(fi->last_name); + fi->last_name = buf; + memcpy(fi->last_name, name, len); + fi->last_name[len] = 0; + fi->next_offset = next_offset; + dout("note_last_dentry '%s'\n", fi->last_name); + return 0; +} + +/* * When possible, we try to satisfy a readdir by peeking at the * dcache. We make this work by carefully ordering dentries on * d_child when we initially get results back from the MDS, and @@ -123,123 +144,113 @@ static int __dcache_readdir(struct file *file, struct dir_context *ctx, struct ceph_file_info *fi = file->private_data; struct dentry *parent = file->f_path.dentry; struct inode *dir = d_inode(parent); - struct list_head *p; - struct dentry *dentry, *last; + struct dentry *dentry, *last = NULL; struct ceph_dentry_info *di; + unsigned nsize = PAGE_CACHE_SIZE / sizeof(struct dentry *); int err = 0; + loff_t ptr_pos = 0; + struct ceph_readdir_cache_control cache_ctl = {}; - /* claim ref on last dentry we returned */ - last = fi->dentry; - fi->dentry = NULL; - - dout("__dcache_readdir %p v%u at %llu (last %p)\n", - dir, shared_gen, ctx->pos, last); + dout("__dcache_readdir %p v%u at %llu\n", dir, shared_gen, ctx->pos); - spin_lock(&parent->d_lock); - - /* start at beginning? */ - if (ctx->pos == 2 || last == NULL || - fpos_cmp(ctx->pos, ceph_dentry(last)->offset) < 0) { - if (list_empty(&parent->d_subdirs)) - goto out_unlock; - p = parent->d_subdirs.prev; - dout(" initial p %p/%p\n", p->prev, p->next); - } else { - p = last->d_child.prev; + /* we can calculate cache index for the first dirfrag */ + if (ceph_frag_is_leftmost(fpos_frag(ctx->pos))) { + cache_ctl.index = fpos_off(ctx->pos) - 2; + BUG_ON(cache_ctl.index < 0); + ptr_pos = cache_ctl.index * sizeof(struct dentry *); } -more: - dentry = list_entry(p, struct dentry, d_child); - di = ceph_dentry(dentry); - while (1) { - dout(" p %p/%p %s d_subdirs %p/%p\n", p->prev, p->next, - d_unhashed(dentry) ? "!hashed" : "hashed", - parent->d_subdirs.prev, parent->d_subdirs.next); - if (p == &parent->d_subdirs) { + while (true) { + pgoff_t pgoff; + bool emit_dentry; + + if (ptr_pos >= i_size_read(dir)) { fi->flags |= CEPH_F_ATEND; - goto out_unlock; + err = 0; + break; + } + + err = -EAGAIN; + pgoff = ptr_pos >> PAGE_CACHE_SHIFT; + if (!cache_ctl.page || pgoff != page_index(cache_ctl.page)) { + ceph_readdir_cache_release(&cache_ctl); + cache_ctl.page = find_lock_page(&dir->i_data, pgoff); + if (!cache_ctl.page) { + dout(" page %lu not found\n", pgoff); + break; + } + /* reading/filling the cache are serialized by + * i_mutex, no need to use page lock */ + unlock_page(cache_ctl.page); + cache_ctl.dentries = kmap(cache_ctl.page); } - spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); + + rcu_read_lock(); + spin_lock(&parent->d_lock); + /* check i_size again here, because empty directory can be + * marked as complete while not holding the i_mutex. */ + if (ceph_dir_is_complete_ordered(dir) && + ptr_pos < i_size_read(dir)) + dentry = cache_ctl.dentries[cache_ctl.index % nsize]; + else + dentry = NULL; + spin_unlock(&parent->d_lock); + if (dentry && !lockref_get_not_dead(&dentry->d_lockref)) + dentry = NULL; + rcu_read_unlock(); + if (!dentry) + break; + + emit_dentry = false; + di = ceph_dentry(dentry); + spin_lock(&dentry->d_lock); if (di->lease_shared_gen == shared_gen && - simple_positive(dentry) && + d_really_is_positive(dentry) && ceph_snap(d_inode(dentry)) != CEPH_SNAPDIR && ceph_ino(d_inode(dentry)) != CEPH_INO_CEPH && - fpos_cmp(ctx->pos, di->offset) <= 0) - break; - dout(" skipping %p %pd at %llu (%llu)%s%s\n", dentry, - dentry, di->offset, - ctx->pos, d_unhashed(dentry) ? " unhashed" : "", - !d_inode(dentry) ? " null" : ""); + fpos_cmp(ctx->pos, di->offset) <= 0) { + emit_dentry = true; + } spin_unlock(&dentry->d_lock); - p = p->prev; - dentry = list_entry(p, struct dentry, d_child); - di = ceph_dentry(dentry); - } - - dget_dlock(dentry); - spin_unlock(&dentry->d_lock); - spin_unlock(&parent->d_lock); - /* make sure a dentry wasn't dropped while we didn't have parent lock */ - if (!ceph_dir_is_complete_ordered(dir)) { - dout(" lost dir complete on %p; falling back to mds\n", dir); - dput(dentry); - err = -EAGAIN; - goto out; - } + if (emit_dentry) { + dout(" %llu (%llu) dentry %p %pd %p\n", di->offset, ctx->pos, + dentry, dentry, d_inode(dentry)); + ctx->pos = di->offset; + if (!dir_emit(ctx, dentry->d_name.name, + dentry->d_name.len, + ceph_translate_ino(dentry->d_sb, + d_inode(dentry)->i_ino), + d_inode(dentry)->i_mode >> 12)) { + dput(dentry); + err = 0; + break; + } + ctx->pos++; - dout(" %llu (%llu) dentry %p %pd %p\n", di->offset, ctx->pos, - dentry, dentry, d_inode(dentry)); - if (!dir_emit(ctx, dentry->d_name.name, - dentry->d_name.len, - ceph_translate_ino(dentry->d_sb, d_inode(dentry)->i_ino), - d_inode(dentry)->i_mode >> 12)) { - if (last) { - /* remember our position */ - fi->dentry = last; - fi->next_offset = fpos_off(di->offset); + if (last) + dput(last); + last = dentry; + } else { + dput(dentry); } - dput(dentry); - return 0; - } - - ctx->pos = di->offset + 1; - if (last) - dput(last); - last = dentry; - - spin_lock(&parent->d_lock); - p = p->prev; /* advance to next dentry */ - goto more; - -out_unlock: - spin_unlock(&parent->d_lock); -out: - if (last) + cache_ctl.index++; + ptr_pos += sizeof(struct dentry *); + } + ceph_readdir_cache_release(&cache_ctl); + if (last) { + int ret; + di = ceph_dentry(last); + ret = note_last_dentry(fi, last->d_name.name, last->d_name.len, + fpos_off(di->offset) + 1); + if (ret < 0) + err = ret; dput(last); + } return err; } -/* - * make note of the last dentry we read, so we can - * continue at the same lexicographical point, - * regardless of what dir changes take place on the - * server. - */ -static int note_last_dentry(struct ceph_file_info *fi, const char *name, - int len) -{ - kfree(fi->last_name); - fi->last_name = kmalloc(len+1, GFP_NOFS); - if (!fi->last_name) - return -ENOMEM; - memcpy(fi->last_name, name, len); - fi->last_name[len] = 0; - dout("note_last_dentry '%s'\n", fi->last_name); - return 0; -} - static int ceph_readdir(struct file *file, struct dir_context *ctx) { struct ceph_file_info *fi = file->private_data; @@ -280,8 +291,7 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx) /* can we use the dcache? */ spin_lock(&ci->i_ceph_lock); - if ((ctx->pos == 2 || fi->dentry) && - ceph_test_mount_opt(fsc, DCACHE) && + if (ceph_test_mount_opt(fsc, DCACHE) && !ceph_test_mount_opt(fsc, NOASYNCREADDIR) && ceph_snap(inode) != CEPH_SNAPDIR && __ceph_dir_is_complete_ordered(ci) && @@ -296,24 +306,8 @@ static int ceph_readdir(struct file *file, struct dir_context *ctx) } else { spin_unlock(&ci->i_ceph_lock); } - if (fi->dentry) { - err = note_last_dentry(fi, fi->dentry->d_name.name, - fi->dentry->d_name.len); - if (err) - return err; - dput(fi->dentry); - fi->dentry = NULL; - } /* proceed with a normal readdir */ - - if (ctx->pos == 2) { - /* note dir version at start of readdir so we can tell - * if any dentries get dropped */ - fi->dir_release_count = atomic_read(&ci->i_release_count); - fi->dir_ordered_count = ci->i_ordered_count; - } - more: /* do we have the correct frag content buffered? */ if (fi->frag != frag || fi->last_readdir == NULL) { @@ -342,12 +336,15 @@ more: req->r_direct_hash = ceph_frag_value(frag); req->r_direct_is_hash = true; if (fi->last_name) { - req->r_path2 = kstrdup(fi->last_name, GFP_NOFS); + req->r_path2 = kstrdup(fi->last_name, GFP_KERNEL); if (!req->r_path2) { ceph_mdsc_put_request(req); return -ENOMEM; } } + req->r_dir_release_cnt = fi->dir_release_count; + req->r_dir_ordered_cnt = fi->dir_ordered_count; + req->r_readdir_cache_idx = fi->readdir_cache_idx; req->r_readdir_offset = fi->next_offset; req->r_args.readdir.frag = cpu_to_le32(frag); @@ -364,26 +361,38 @@ more: (int)req->r_reply_info.dir_end, (int)req->r_reply_info.dir_complete); - if (!req->r_did_prepopulate) { - dout("readdir !did_prepopulate"); - /* preclude from marking dir complete */ - fi->dir_release_count--; - } /* note next offset and last dentry name */ rinfo = &req->r_reply_info; if (le32_to_cpu(rinfo->dir_dir->frag) != frag) { frag = le32_to_cpu(rinfo->dir_dir->frag); - if (ceph_frag_is_leftmost(frag)) - fi->next_offset = 2; - else - fi->next_offset = 0; - off = fi->next_offset; + off = req->r_readdir_offset; + fi->next_offset = off; } + fi->frag = frag; fi->offset = fi->next_offset; fi->last_readdir = req; + if (req->r_did_prepopulate) { + fi->readdir_cache_idx = req->r_readdir_cache_idx; + if (fi->readdir_cache_idx < 0) { + /* preclude from marking dir ordered */ + fi->dir_ordered_count = 0; + } else if (ceph_frag_is_leftmost(frag) && off == 2) { + /* note dir version at start of readdir so + * we can tell if any dentries get dropped */ + fi->dir_release_count = req->r_dir_release_cnt; + fi->dir_ordered_count = req->r_dir_ordered_cnt; + } + } else { + dout("readdir !did_prepopulate"); + /* disable readdir cache */ + fi->readdir_cache_idx = -1; + /* preclude from marking dir complete */ + fi->dir_release_count = 0; + } + if (req->r_reply_info.dir_end) { kfree(fi->last_name); fi->last_name = NULL; @@ -394,10 +403,10 @@ more: } else { err = note_last_dentry(fi, rinfo->dir_dname[rinfo->dir_nr-1], - rinfo->dir_dname_len[rinfo->dir_nr-1]); + rinfo->dir_dname_len[rinfo->dir_nr-1], + fi->next_offset + rinfo->dir_nr); if (err) return err; - fi->next_offset += rinfo->dir_nr; } } @@ -453,16 +462,22 @@ more: * were released during the whole readdir, and we should have * the complete dir contents in our cache. */ - spin_lock(&ci->i_ceph_lock); - if (atomic_read(&ci->i_release_count) == fi->dir_release_count) { - if (ci->i_ordered_count == fi->dir_ordered_count) + if (atomic64_read(&ci->i_release_count) == fi->dir_release_count) { + spin_lock(&ci->i_ceph_lock); + if (fi->dir_ordered_count == atomic64_read(&ci->i_ordered_count)) { dout(" marking %p complete and ordered\n", inode); - else + /* use i_size to track number of entries in + * readdir cache */ + BUG_ON(fi->readdir_cache_idx < 0); + i_size_write(inode, fi->readdir_cache_idx * + sizeof(struct dentry*)); + } else { dout(" marking %p complete\n", inode); + } __ceph_dir_set_complete(ci, fi->dir_release_count, fi->dir_ordered_count); + spin_unlock(&ci->i_ceph_lock); } - spin_unlock(&ci->i_ceph_lock); dout("readdir %p file %p done.\n", inode, file); return 0; @@ -476,14 +491,12 @@ static void reset_readdir(struct ceph_file_info *fi, unsigned frag) } kfree(fi->last_name); fi->last_name = NULL; + fi->dir_release_count = 0; + fi->readdir_cache_idx = -1; if (ceph_frag_is_leftmost(frag)) fi->next_offset = 2; /* compensate for . and .. */ else fi->next_offset = 0; - if (fi->dentry) { - dput(fi->dentry); - fi->dentry = NULL; - } fi->flags &= ~CEPH_F_ATEND; } @@ -497,13 +510,12 @@ static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int whence) mutex_lock(&inode->i_mutex); retval = -EINVAL; switch (whence) { - case SEEK_END: - offset += inode->i_size + 2; /* FIXME */ - break; case SEEK_CUR: offset += file->f_pos; case SEEK_SET: break; + case SEEK_END: + retval = -EOPNOTSUPP; default: goto out; } @@ -516,20 +528,18 @@ static loff_t ceph_dir_llseek(struct file *file, loff_t offset, int whence) } retval = offset; - /* - * discard buffered readdir content on seekdir(0), or - * seek to new frag, or seek prior to current chunk. - */ if (offset == 0 || fpos_frag(offset) != fi->frag || fpos_off(offset) < fi->offset) { + /* discard buffered readdir content on seekdir(0), or + * seek to new frag, or seek prior to current chunk */ dout("dir_llseek dropping %p content\n", file); reset_readdir(fi, fpos_frag(offset)); + } else if (fpos_cmp(offset, old_offset) > 0) { + /* reset dir_release_count if we did a forward seek */ + fi->dir_release_count = 0; + fi->readdir_cache_idx = -1; } - - /* bump dir_release_count if we did a forward seek */ - if (fpos_cmp(offset, old_offset) > 0) - fi->dir_release_count--; } out: mutex_unlock(&inode->i_mutex); @@ -764,7 +774,7 @@ static int ceph_symlink(struct inode *dir, struct dentry *dentry, err = PTR_ERR(req); goto out; } - req->r_path2 = kstrdup(dest, GFP_NOFS); + req->r_path2 = kstrdup(dest, GFP_KERNEL); if (!req->r_path2) { err = -ENOMEM; ceph_mdsc_put_request(req); @@ -985,16 +995,15 @@ static int ceph_rename(struct inode *old_dir, struct dentry *old_dentry, * to do it here. */ + /* d_move screws up sibling dentries' offsets */ + ceph_dir_clear_complete(old_dir); + ceph_dir_clear_complete(new_dir); + d_move(old_dentry, new_dentry); /* ensure target dentry is invalidated, despite rehashing bug in vfs_rename_dir */ ceph_invalidate_dentry_lease(new_dentry); - - /* d_move screws up sibling dentries' offsets */ - ceph_dir_clear_complete(old_dir); - ceph_dir_clear_complete(new_dir); - } ceph_mdsc_put_request(req); return err; @@ -1189,7 +1198,7 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size, return -EISDIR; if (!cf->dir_info) { - cf->dir_info = kmalloc(bufsize, GFP_NOFS); + cf->dir_info = kmalloc(bufsize, GFP_KERNEL); if (!cf->dir_info) return -ENOMEM; cf->dir_info_len = @@ -1224,66 +1233,6 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size, } /* - * an fsync() on a dir will wait for any uncommitted directory - * operations to commit. - */ -static int ceph_dir_fsync(struct file *file, loff_t start, loff_t end, - int datasync) -{ - struct inode *inode = file_inode(file); - struct ceph_inode_info *ci = ceph_inode(inode); - struct list_head *head = &ci->i_unsafe_dirops; - struct ceph_mds_request *req; - u64 last_tid; - int ret = 0; - - dout("dir_fsync %p\n", inode); - ret = filemap_write_and_wait_range(inode->i_mapping, start, end); - if (ret) - return ret; - mutex_lock(&inode->i_mutex); - - spin_lock(&ci->i_unsafe_lock); - if (list_empty(head)) - goto out; - - req = list_entry(head->prev, - struct ceph_mds_request, r_unsafe_dir_item); - last_tid = req->r_tid; - - do { - ceph_mdsc_get_request(req); - spin_unlock(&ci->i_unsafe_lock); - - dout("dir_fsync %p wait on tid %llu (until %llu)\n", - inode, req->r_tid, last_tid); - if (req->r_timeout) { - unsigned long time_left = wait_for_completion_timeout( - &req->r_safe_completion, - req->r_timeout); - if (time_left > 0) - ret = 0; - else - ret = -EIO; /* timed out */ - } else { - wait_for_completion(&req->r_safe_completion); - } - ceph_mdsc_put_request(req); - - spin_lock(&ci->i_unsafe_lock); - if (ret || list_empty(head)) - break; - req = list_entry(head->next, - struct ceph_mds_request, r_unsafe_dir_item); - } while (req->r_tid < last_tid); -out: - spin_unlock(&ci->i_unsafe_lock); - mutex_unlock(&inode->i_mutex); - - return ret; -} - -/* * We maintain a private dentry LRU. * * FIXME: this needs to be changed to a per-mds lru to be useful. @@ -1353,7 +1302,7 @@ const struct file_operations ceph_dir_fops = { .open = ceph_open, .release = ceph_release, .unlocked_ioctl = ceph_ioctl, - .fsync = ceph_dir_fsync, + .fsync = ceph_fsync, }; const struct file_operations ceph_snapdir_fops = { diff --git a/fs/ceph/file.c b/fs/ceph/file.c index e55fe32..8b79d87 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -89,13 +89,14 @@ static int ceph_init_file(struct inode *inode, struct file *file, int fmode) case S_IFDIR: dout("init_file %p %p 0%o (regular)\n", inode, file, inode->i_mode); - cf = kmem_cache_alloc(ceph_file_cachep, GFP_NOFS | __GFP_ZERO); + cf = kmem_cache_alloc(ceph_file_cachep, GFP_KERNEL | __GFP_ZERO); if (cf == NULL) { ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */ return -ENOMEM; } cf->fmode = fmode; cf->next_offset = 2; + cf->readdir_cache_idx = -1; file->private_data = cf; BUG_ON(inode->i_fop->release != ceph_release); break; @@ -324,7 +325,6 @@ int ceph_release(struct inode *inode, struct file *file) ceph_mdsc_put_request(cf->last_readdir); kfree(cf->last_name); kfree(cf->dir_info); - dput(cf->dentry); kmem_cache_free(ceph_file_cachep, cf); /* wake up anyone waiting for caps on this inode */ @@ -483,7 +483,7 @@ static ssize_t ceph_sync_read(struct kiocb *iocb, struct iov_iter *i, } } else { num_pages = calc_pages_for(off, len); - pages = ceph_alloc_page_vector(num_pages, GFP_NOFS); + pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); if (IS_ERR(pages)) return PTR_ERR(pages); ret = striped_read(inode, off, len, pages, @@ -557,13 +557,13 @@ static void ceph_sync_write_unsafe(struct ceph_osd_request *req, bool unsafe) * objects, rollback on failure, etc.) */ static ssize_t -ceph_sync_direct_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos) +ceph_sync_direct_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos, + struct ceph_snap_context *snapc) { struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_fs_client *fsc = ceph_inode_to_client(inode); - struct ceph_snap_context *snapc; struct ceph_vino vino; struct ceph_osd_request *req; struct page **pages; @@ -600,7 +600,6 @@ ceph_sync_direct_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos) size_t start; ssize_t n; - snapc = ci->i_snap_realm->cached_context; vino = ceph_vino(inode); req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout, vino, pos, &len, 0, @@ -614,7 +613,7 @@ ceph_sync_direct_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos) break; } - osd_req_op_init(req, 1, CEPH_OSD_OP_STARTSYNC); + osd_req_op_init(req, 1, CEPH_OSD_OP_STARTSYNC, 0); n = iov_iter_get_pages_alloc(from, &pages, len, &start); if (unlikely(n < 0)) { @@ -674,13 +673,13 @@ ceph_sync_direct_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos) * objects, rollback on failure, etc.) */ static ssize_t -ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos) +ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos, + struct ceph_snap_context *snapc) { struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_fs_client *fsc = ceph_inode_to_client(inode); - struct ceph_snap_context *snapc; struct ceph_vino vino; struct ceph_osd_request *req; struct page **pages; @@ -717,7 +716,6 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos) size_t left; int n; - snapc = ci->i_snap_realm->cached_context; vino = ceph_vino(inode); req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout, vino, pos, &len, 0, 1, @@ -736,7 +734,7 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos) */ num_pages = (len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - pages = ceph_alloc_page_vector(num_pages, GFP_NOFS); + pages = ceph_alloc_page_vector(num_pages, GFP_KERNEL); if (IS_ERR(pages)) { ret = PTR_ERR(pages); goto out; @@ -860,7 +858,7 @@ again: struct page *page = NULL; loff_t i_size; if (retry_op == READ_INLINE) { - page = __page_cache_alloc(GFP_NOFS); + page = __page_cache_alloc(GFP_KERNEL); if (!page) return -ENOMEM; } @@ -941,6 +939,7 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from) struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_osd_client *osdc = &ceph_sb_to_client(inode->i_sb)->client->osdc; + struct ceph_cap_flush *prealloc_cf; ssize_t count, written = 0; int err, want, got; loff_t pos; @@ -948,6 +947,10 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from) if (ceph_snap(inode) != CEPH_NOSNAP) return -EROFS; + prealloc_cf = ceph_alloc_cap_flush(); + if (!prealloc_cf) + return -ENOMEM; + mutex_lock(&inode->i_mutex); /* We can write back this queue in page reclaim */ @@ -996,14 +999,30 @@ retry_snap: if ((got & (CEPH_CAP_FILE_BUFFER|CEPH_CAP_FILE_LAZYIO)) == 0 || (iocb->ki_flags & IOCB_DIRECT) || (fi->flags & CEPH_F_SYNC)) { + struct ceph_snap_context *snapc; struct iov_iter data; mutex_unlock(&inode->i_mutex); + + spin_lock(&ci->i_ceph_lock); + if (__ceph_have_pending_cap_snap(ci)) { + struct ceph_cap_snap *capsnap = + list_last_entry(&ci->i_cap_snaps, + struct ceph_cap_snap, + ci_item); + snapc = ceph_get_snap_context(capsnap->context); + } else { + BUG_ON(!ci->i_head_snapc); + snapc = ceph_get_snap_context(ci->i_head_snapc); + } + spin_unlock(&ci->i_ceph_lock); + /* we might need to revert back to that point */ data = *from; if (iocb->ki_flags & IOCB_DIRECT) - written = ceph_sync_direct_write(iocb, &data, pos); + written = ceph_sync_direct_write(iocb, &data, pos, + snapc); else - written = ceph_sync_write(iocb, &data, pos); + written = ceph_sync_write(iocb, &data, pos, snapc); if (written == -EOLDSNAPC) { dout("aio_write %p %llx.%llx %llu~%u" "got EOLDSNAPC, retrying\n", @@ -1014,6 +1033,7 @@ retry_snap: } if (written > 0) iov_iter_advance(from, written); + ceph_put_snap_context(snapc); } else { loff_t old_size = inode->i_size; /* @@ -1035,7 +1055,8 @@ retry_snap: int dirty; spin_lock(&ci->i_ceph_lock); ci->i_inline_version = CEPH_INLINE_NONE; - dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR, + &prealloc_cf); spin_unlock(&ci->i_ceph_lock); if (dirty) __mark_inode_dirty(inode, dirty); @@ -1059,6 +1080,7 @@ retry_snap: out: mutex_unlock(&inode->i_mutex); out_unlocked: + ceph_free_cap_flush(prealloc_cf); current->backing_dev_info = NULL; return written ? written : err; } @@ -1255,6 +1277,7 @@ static long ceph_fallocate(struct file *file, int mode, struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_osd_client *osdc = &ceph_inode_to_client(inode)->client->osdc; + struct ceph_cap_flush *prealloc_cf; int want, got = 0; int dirty; int ret = 0; @@ -1267,6 +1290,10 @@ static long ceph_fallocate(struct file *file, int mode, if (!S_ISREG(inode->i_mode)) return -EOPNOTSUPP; + prealloc_cf = ceph_alloc_cap_flush(); + if (!prealloc_cf) + return -ENOMEM; + mutex_lock(&inode->i_mutex); if (ceph_snap(inode) != CEPH_NOSNAP) { @@ -1313,7 +1340,8 @@ static long ceph_fallocate(struct file *file, int mode, if (!ret) { spin_lock(&ci->i_ceph_lock); ci->i_inline_version = CEPH_INLINE_NONE; - dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR, + &prealloc_cf); spin_unlock(&ci->i_ceph_lock); if (dirty) __mark_inode_dirty(inode, dirty); @@ -1322,6 +1350,7 @@ static long ceph_fallocate(struct file *file, int mode, ceph_put_cap_refs(ci, got); unlock: mutex_unlock(&inode->i_mutex); + ceph_free_cap_flush(prealloc_cf); return ret; } diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index 571acd8..96d2bd8 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -389,9 +389,10 @@ struct inode *ceph_alloc_inode(struct super_block *sb) ci->i_inline_version = 0; ci->i_time_warp_seq = 0; ci->i_ceph_flags = 0; - ci->i_ordered_count = 0; - atomic_set(&ci->i_release_count, 1); - atomic_set(&ci->i_complete_count, 0); + atomic64_set(&ci->i_ordered_count, 1); + atomic64_set(&ci->i_release_count, 1); + atomic64_set(&ci->i_complete_seq[0], 0); + atomic64_set(&ci->i_complete_seq[1], 0); ci->i_symlink = NULL; memset(&ci->i_dir_layout, 0, sizeof(ci->i_dir_layout)); @@ -415,9 +416,8 @@ struct inode *ceph_alloc_inode(struct super_block *sb) ci->i_flushing_caps = 0; INIT_LIST_HEAD(&ci->i_dirty_item); INIT_LIST_HEAD(&ci->i_flushing_item); - ci->i_cap_flush_seq = 0; - ci->i_cap_flush_last_tid = 0; - memset(&ci->i_cap_flush_tid, 0, sizeof(ci->i_cap_flush_tid)); + ci->i_prealloc_cap_flush = NULL; + ci->i_cap_flush_tree = RB_ROOT; init_waitqueue_head(&ci->i_cap_wq); ci->i_hold_caps_min = 0; ci->i_hold_caps_max = 0; @@ -752,7 +752,10 @@ static int fill_inode(struct inode *inode, struct page *locked_page, if (new_version || (new_issued & (CEPH_CAP_ANY_FILE_RD | CEPH_CAP_ANY_FILE_WR))) { + if (ci->i_layout.fl_pg_pool != info->layout.fl_pg_pool) + ci->i_ceph_flags &= ~CEPH_I_POOL_PERM; ci->i_layout = info->layout; + queue_trunc = ceph_fill_file_size(inode, issued, le32_to_cpu(info->truncate_seq), le64_to_cpu(info->truncate_size), @@ -858,9 +861,10 @@ static int fill_inode(struct inode *inode, struct page *locked_page, (issued & CEPH_CAP_FILE_EXCL) == 0 && !__ceph_dir_is_complete(ci)) { dout(" marking %p complete (empty)\n", inode); + i_size_write(inode, 0); __ceph_dir_set_complete(ci, - atomic_read(&ci->i_release_count), - ci->i_ordered_count); + atomic64_read(&ci->i_release_count), + atomic64_read(&ci->i_ordered_count)); } wake = true; @@ -1212,6 +1216,10 @@ retry_lookup: dout("fill_trace doing d_move %p -> %p\n", req->r_old_dentry, dn); + /* d_move screws up sibling dentries' offsets */ + ceph_dir_clear_ordered(dir); + ceph_dir_clear_ordered(olddir); + d_move(req->r_old_dentry, dn); dout(" src %p '%pd' dst %p '%pd'\n", req->r_old_dentry, @@ -1222,10 +1230,6 @@ retry_lookup: rehashing bug in vfs_rename_dir */ ceph_invalidate_dentry_lease(dn); - /* d_move screws up sibling dentries' offsets */ - ceph_dir_clear_ordered(dir); - ceph_dir_clear_ordered(olddir); - dout("dn %p gets new offset %lld\n", req->r_old_dentry, ceph_dentry(req->r_old_dentry)->offset); @@ -1333,6 +1337,49 @@ static int readdir_prepopulate_inodes_only(struct ceph_mds_request *req, return err; } +void ceph_readdir_cache_release(struct ceph_readdir_cache_control *ctl) +{ + if (ctl->page) { + kunmap(ctl->page); + page_cache_release(ctl->page); + ctl->page = NULL; + } +} + +static int fill_readdir_cache(struct inode *dir, struct dentry *dn, + struct ceph_readdir_cache_control *ctl, + struct ceph_mds_request *req) +{ + struct ceph_inode_info *ci = ceph_inode(dir); + unsigned nsize = PAGE_CACHE_SIZE / sizeof(struct dentry*); + unsigned idx = ctl->index % nsize; + pgoff_t pgoff = ctl->index / nsize; + + if (!ctl->page || pgoff != page_index(ctl->page)) { + ceph_readdir_cache_release(ctl); + ctl->page = grab_cache_page(&dir->i_data, pgoff); + if (!ctl->page) { + ctl->index = -1; + return -ENOMEM; + } + /* reading/filling the cache are serialized by + * i_mutex, no need to use page lock */ + unlock_page(ctl->page); + ctl->dentries = kmap(ctl->page); + } + + if (req->r_dir_release_cnt == atomic64_read(&ci->i_release_count) && + req->r_dir_ordered_cnt == atomic64_read(&ci->i_ordered_count)) { + dout("readdir cache dn %p idx %d\n", dn, ctl->index); + ctl->dentries[idx] = dn; + ctl->index++; + } else { + dout("disable readdir cache\n"); + ctl->index = -1; + } + return 0; +} + int ceph_readdir_prepopulate(struct ceph_mds_request *req, struct ceph_mds_session *session) { @@ -1345,8 +1392,11 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req, struct inode *snapdir = NULL; struct ceph_mds_request_head *rhead = req->r_request->front.iov_base; struct ceph_dentry_info *di; - u64 r_readdir_offset = req->r_readdir_offset; u32 frag = le32_to_cpu(rhead->args.readdir.frag); + struct ceph_readdir_cache_control cache_ctl = {}; + + if (req->r_aborted) + return readdir_prepopulate_inodes_only(req, session); if (rinfo->dir_dir && le32_to_cpu(rinfo->dir_dir->frag) != frag) { @@ -1354,14 +1404,11 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req, frag, le32_to_cpu(rinfo->dir_dir->frag)); frag = le32_to_cpu(rinfo->dir_dir->frag); if (ceph_frag_is_leftmost(frag)) - r_readdir_offset = 2; + req->r_readdir_offset = 2; else - r_readdir_offset = 0; + req->r_readdir_offset = 0; } - if (req->r_aborted) - return readdir_prepopulate_inodes_only(req, session); - if (le32_to_cpu(rinfo->head->op) == CEPH_MDS_OP_LSSNAP) { snapdir = ceph_get_snapdir(d_inode(parent)); parent = d_find_alias(snapdir); @@ -1374,6 +1421,17 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req, ceph_fill_dirfrag(d_inode(parent), rinfo->dir_dir); } + if (ceph_frag_is_leftmost(frag) && req->r_readdir_offset == 2) { + /* note dir version at start of readdir so we can tell + * if any dentries get dropped */ + struct ceph_inode_info *ci = ceph_inode(d_inode(parent)); + req->r_dir_release_cnt = atomic64_read(&ci->i_release_count); + req->r_dir_ordered_cnt = atomic64_read(&ci->i_ordered_count); + req->r_readdir_cache_idx = 0; + } + + cache_ctl.index = req->r_readdir_cache_idx; + /* FIXME: release caps/leases if error occurs */ for (i = 0; i < rinfo->dir_nr; i++) { struct ceph_vino vino; @@ -1413,13 +1471,6 @@ retry_lookup: d_delete(dn); dput(dn); goto retry_lookup; - } else { - /* reorder parent's d_subdirs */ - spin_lock(&parent->d_lock); - spin_lock_nested(&dn->d_lock, DENTRY_D_LOCK_NESTED); - list_move(&dn->d_child, &parent->d_subdirs); - spin_unlock(&dn->d_lock); - spin_unlock(&parent->d_lock); } /* inode */ @@ -1436,13 +1487,15 @@ retry_lookup: } } - if (fill_inode(in, NULL, &rinfo->dir_in[i], NULL, session, - req->r_request_started, -1, - &req->r_caps_reservation) < 0) { + ret = fill_inode(in, NULL, &rinfo->dir_in[i], NULL, session, + req->r_request_started, -1, + &req->r_caps_reservation); + if (ret < 0) { pr_err("fill_inode badness on %p\n", in); if (d_really_is_negative(dn)) iput(in); d_drop(dn); + err = ret; goto next_item; } @@ -1458,19 +1511,28 @@ retry_lookup: } di = dn->d_fsdata; - di->offset = ceph_make_fpos(frag, i + r_readdir_offset); + di->offset = ceph_make_fpos(frag, i + req->r_readdir_offset); update_dentry_lease(dn, rinfo->dir_dlease[i], req->r_session, req->r_request_started); + + if (err == 0 && cache_ctl.index >= 0) { + ret = fill_readdir_cache(d_inode(parent), dn, + &cache_ctl, req); + if (ret < 0) + err = ret; + } next_item: if (dn) dput(dn); } - if (err == 0) - req->r_did_prepopulate = true; - out: + if (err == 0) { + req->r_did_prepopulate = true; + req->r_readdir_cache_idx = cache_ctl.index; + } + ceph_readdir_cache_release(&cache_ctl); if (snapdir) { iput(snapdir); dput(parent); @@ -1712,11 +1774,13 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) const unsigned int ia_valid = attr->ia_valid; struct ceph_mds_request *req; struct ceph_mds_client *mdsc = ceph_sb_to_client(dentry->d_sb)->mdsc; + struct ceph_cap_flush *prealloc_cf; int issued; int release = 0, dirtied = 0; int mask = 0; int err = 0; int inode_dirty_flags = 0; + bool lock_snap_rwsem = false; if (ceph_snap(inode) != CEPH_NOSNAP) return -EROFS; @@ -1725,13 +1789,31 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) if (err != 0) return err; + prealloc_cf = ceph_alloc_cap_flush(); + if (!prealloc_cf) + return -ENOMEM; + req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_SETATTR, USE_AUTH_MDS); - if (IS_ERR(req)) + if (IS_ERR(req)) { + ceph_free_cap_flush(prealloc_cf); return PTR_ERR(req); + } spin_lock(&ci->i_ceph_lock); issued = __ceph_caps_issued(ci, NULL); + + if (!ci->i_head_snapc && + (issued & (CEPH_CAP_ANY_EXCL | CEPH_CAP_FILE_WR))) { + lock_snap_rwsem = true; + if (!down_read_trylock(&mdsc->snap_rwsem)) { + spin_unlock(&ci->i_ceph_lock); + down_read(&mdsc->snap_rwsem); + spin_lock(&ci->i_ceph_lock); + issued = __ceph_caps_issued(ci, NULL); + } + } + dout("setattr %p issued %s\n", inode, ceph_cap_string(issued)); if (ia_valid & ATTR_UID) { @@ -1874,12 +1956,15 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) dout("setattr %p ATTR_FILE ... hrm!\n", inode); if (dirtied) { - inode_dirty_flags = __ceph_mark_dirty_caps(ci, dirtied); + inode_dirty_flags = __ceph_mark_dirty_caps(ci, dirtied, + &prealloc_cf); inode->i_ctime = CURRENT_TIME; } release &= issued; spin_unlock(&ci->i_ceph_lock); + if (lock_snap_rwsem) + up_read(&mdsc->snap_rwsem); if (inode_dirty_flags) __mark_inode_dirty(inode, inode_dirty_flags); @@ -1904,9 +1989,11 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) ceph_mdsc_put_request(req); if (mask & CEPH_SETATTR_SIZE) __ceph_do_pending_vmtruncate(inode); + ceph_free_cap_flush(prealloc_cf); return err; out_put: ceph_mdsc_put_request(req); + ceph_free_cap_flush(prealloc_cf); return err; } diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 84f37f3..6aa07af 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -8,6 +8,7 @@ #include <linux/debugfs.h> #include <linux/seq_file.h> #include <linux/utsname.h> +#include <linux/ratelimit.h> #include "super.h" #include "mds_client.h" @@ -458,7 +459,6 @@ static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc, s->s_cap_reconnect = 0; s->s_cap_iterator = NULL; INIT_LIST_HEAD(&s->s_cap_releases); - INIT_LIST_HEAD(&s->s_cap_releases_done); INIT_LIST_HEAD(&s->s_cap_flushing); INIT_LIST_HEAD(&s->s_cap_snaps_flushing); @@ -629,6 +629,9 @@ static void __register_request(struct ceph_mds_client *mdsc, req->r_uid = current_fsuid(); req->r_gid = current_fsgid(); + if (mdsc->oldest_tid == 0 && req->r_op != CEPH_MDS_OP_SETFILELOCK) + mdsc->oldest_tid = req->r_tid; + if (dir) { struct ceph_inode_info *ci = ceph_inode(dir); @@ -644,6 +647,21 @@ static void __unregister_request(struct ceph_mds_client *mdsc, struct ceph_mds_request *req) { dout("__unregister_request %p tid %lld\n", req, req->r_tid); + + if (req->r_tid == mdsc->oldest_tid) { + struct rb_node *p = rb_next(&req->r_node); + mdsc->oldest_tid = 0; + while (p) { + struct ceph_mds_request *next_req = + rb_entry(p, struct ceph_mds_request, r_node); + if (next_req->r_op != CEPH_MDS_OP_SETFILELOCK) { + mdsc->oldest_tid = next_req->r_tid; + break; + } + p = rb_next(p); + } + } + rb_erase(&req->r_node, &mdsc->request_tree); RB_CLEAR_NODE(&req->r_node); @@ -998,27 +1016,25 @@ void ceph_mdsc_open_export_target_sessions(struct ceph_mds_client *mdsc, * session caps */ -/* - * Free preallocated cap messages assigned to this session - */ -static void cleanup_cap_releases(struct ceph_mds_session *session) +/* caller holds s_cap_lock, we drop it */ +static void cleanup_cap_releases(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session) + __releases(session->s_cap_lock) { - struct ceph_msg *msg; + LIST_HEAD(tmp_list); + list_splice_init(&session->s_cap_releases, &tmp_list); + session->s_num_cap_releases = 0; + spin_unlock(&session->s_cap_lock); - spin_lock(&session->s_cap_lock); - while (!list_empty(&session->s_cap_releases)) { - msg = list_first_entry(&session->s_cap_releases, - struct ceph_msg, list_head); - list_del_init(&msg->list_head); - ceph_msg_put(msg); - } - while (!list_empty(&session->s_cap_releases_done)) { - msg = list_first_entry(&session->s_cap_releases_done, - struct ceph_msg, list_head); - list_del_init(&msg->list_head); - ceph_msg_put(msg); + dout("cleanup_cap_releases mds%d\n", session->s_mds); + while (!list_empty(&tmp_list)) { + struct ceph_cap *cap; + /* zero out the in-progress message */ + cap = list_first_entry(&tmp_list, + struct ceph_cap, session_caps); + list_del(&cap->session_caps); + ceph_put_cap(mdsc, cap); } - spin_unlock(&session->s_cap_lock); } static void cleanup_session_requests(struct ceph_mds_client *mdsc, @@ -1033,7 +1049,8 @@ static void cleanup_session_requests(struct ceph_mds_client *mdsc, req = list_first_entry(&session->s_unsafe, struct ceph_mds_request, r_unsafe_item); list_del_init(&req->r_unsafe_item); - pr_info(" dropping unsafe request %llu\n", req->r_tid); + pr_warn_ratelimited(" dropping unsafe request %llu\n", + req->r_tid); __unregister_request(mdsc, req); } /* zero r_attempts, so kick_requests() will re-send requests */ @@ -1095,10 +1112,16 @@ static int iterate_session_caps(struct ceph_mds_session *session, dout("iterate_session_caps finishing cap %p removal\n", cap); BUG_ON(cap->session != session); + cap->session = NULL; list_del_init(&cap->session_caps); session->s_nr_caps--; - cap->session = NULL; - old_cap = cap; /* put_cap it w/o locks held */ + if (cap->queue_release) { + list_add_tail(&cap->session_caps, + &session->s_cap_releases); + session->s_num_cap_releases++; + } else { + old_cap = cap; /* put_cap it w/o locks held */ + } } if (ret < 0) goto out; @@ -1119,6 +1142,7 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg) { struct ceph_inode_info *ci = ceph_inode(inode); + LIST_HEAD(to_remove); int drop = 0; dout("removing cap %p, ci is %p, inode is %p\n", @@ -1126,12 +1150,27 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, spin_lock(&ci->i_ceph_lock); __ceph_remove_cap(cap, false); if (!ci->i_auth_cap) { + struct ceph_cap_flush *cf; struct ceph_mds_client *mdsc = ceph_sb_to_client(inode->i_sb)->mdsc; + while (true) { + struct rb_node *n = rb_first(&ci->i_cap_flush_tree); + if (!n) + break; + cf = rb_entry(n, struct ceph_cap_flush, i_node); + rb_erase(&cf->i_node, &ci->i_cap_flush_tree); + list_add(&cf->list, &to_remove); + } + spin_lock(&mdsc->cap_dirty_lock); + + list_for_each_entry(cf, &to_remove, list) + rb_erase(&cf->g_node, &mdsc->cap_flush_tree); + if (!list_empty(&ci->i_dirty_item)) { - pr_info(" dropping dirty %s state for %p %lld\n", + pr_warn_ratelimited( + " dropping dirty %s state for %p %lld\n", ceph_cap_string(ci->i_dirty_caps), inode, ceph_ino(inode)); ci->i_dirty_caps = 0; @@ -1139,7 +1178,8 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, drop = 1; } if (!list_empty(&ci->i_flushing_item)) { - pr_info(" dropping dirty+flushing %s state for %p %lld\n", + pr_warn_ratelimited( + " dropping dirty+flushing %s state for %p %lld\n", ceph_cap_string(ci->i_flushing_caps), inode, ceph_ino(inode)); ci->i_flushing_caps = 0; @@ -1148,8 +1188,20 @@ static int remove_session_caps_cb(struct inode *inode, struct ceph_cap *cap, drop = 1; } spin_unlock(&mdsc->cap_dirty_lock); + + if (!ci->i_dirty_caps && ci->i_prealloc_cap_flush) { + list_add(&ci->i_prealloc_cap_flush->list, &to_remove); + ci->i_prealloc_cap_flush = NULL; + } } spin_unlock(&ci->i_ceph_lock); + while (!list_empty(&to_remove)) { + struct ceph_cap_flush *cf; + cf = list_first_entry(&to_remove, + struct ceph_cap_flush, list); + list_del(&cf->list); + ceph_free_cap_flush(cf); + } while (drop--) iput(inode); return 0; @@ -1191,11 +1243,12 @@ static void remove_session_caps(struct ceph_mds_session *session) spin_lock(&session->s_cap_lock); } } - spin_unlock(&session->s_cap_lock); + + // drop cap expires and unlock s_cap_lock + cleanup_cap_releases(session->s_mdsc, session); BUG_ON(session->s_nr_caps > 0); BUG_ON(!list_empty(&session->s_cap_flushing)); - cleanup_cap_releases(session); } /* @@ -1371,7 +1424,8 @@ static int trim_caps_cb(struct inode *inode, struct ceph_cap *cap, void *arg) inode, cap, ceph_cap_string(mine), ceph_cap_string(oissued), ceph_cap_string(used), ceph_cap_string(wanted)); if (cap == ci->i_auth_cap) { - if (ci->i_dirty_caps | ci->i_flushing_caps) + if (ci->i_dirty_caps || ci->i_flushing_caps || + !list_empty(&ci->i_cap_snaps)) goto out; if ((used | wanted) & CEPH_CAP_ANY_WR) goto out; @@ -1417,121 +1471,80 @@ static int trim_caps(struct ceph_mds_client *mdsc, session->s_trim_caps = 0; } - ceph_add_cap_releases(mdsc, session); ceph_send_cap_releases(mdsc, session); return 0; } -/* - * Allocate cap_release messages. If there is a partially full message - * in the queue, try to allocate enough to cover it's remainder, so that - * we can send it immediately. - * - * Called under s_mutex. - */ -int ceph_add_cap_releases(struct ceph_mds_client *mdsc, - struct ceph_mds_session *session) +static int check_capsnap_flush(struct ceph_inode_info *ci, + u64 want_snap_seq) { - struct ceph_msg *msg, *partial = NULL; - struct ceph_mds_cap_release *head; - int err = -ENOMEM; - int extra = mdsc->fsc->mount_options->cap_release_safety; - int num; - - dout("add_cap_releases %p mds%d extra %d\n", session, session->s_mds, - extra); - - spin_lock(&session->s_cap_lock); - - if (!list_empty(&session->s_cap_releases)) { - msg = list_first_entry(&session->s_cap_releases, - struct ceph_msg, - list_head); - head = msg->front.iov_base; - num = le32_to_cpu(head->num); - if (num) { - dout(" partial %p with (%d/%d)\n", msg, num, - (int)CEPH_CAPS_PER_RELEASE); - extra += CEPH_CAPS_PER_RELEASE - num; - partial = msg; - } - } - while (session->s_num_cap_releases < session->s_nr_caps + extra) { - spin_unlock(&session->s_cap_lock); - msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE, PAGE_CACHE_SIZE, - GFP_NOFS, false); - if (!msg) - goto out_unlocked; - dout("add_cap_releases %p msg %p now %d\n", session, msg, - (int)msg->front.iov_len); - head = msg->front.iov_base; - head->num = cpu_to_le32(0); - msg->front.iov_len = sizeof(*head); - spin_lock(&session->s_cap_lock); - list_add(&msg->list_head, &session->s_cap_releases); - session->s_num_cap_releases += CEPH_CAPS_PER_RELEASE; - } - - if (partial) { - head = partial->front.iov_base; - num = le32_to_cpu(head->num); - dout(" queueing partial %p with %d/%d\n", partial, num, - (int)CEPH_CAPS_PER_RELEASE); - list_move_tail(&partial->list_head, - &session->s_cap_releases_done); - session->s_num_cap_releases -= CEPH_CAPS_PER_RELEASE - num; + int ret = 1; + spin_lock(&ci->i_ceph_lock); + if (want_snap_seq > 0 && !list_empty(&ci->i_cap_snaps)) { + struct ceph_cap_snap *capsnap = + list_first_entry(&ci->i_cap_snaps, + struct ceph_cap_snap, ci_item); + ret = capsnap->follows >= want_snap_seq; } - err = 0; - spin_unlock(&session->s_cap_lock); -out_unlocked: - return err; + spin_unlock(&ci->i_ceph_lock); + return ret; } -static int check_cap_flush(struct inode *inode, u64 want_flush_seq) +static int check_caps_flush(struct ceph_mds_client *mdsc, + u64 want_flush_tid) { - struct ceph_inode_info *ci = ceph_inode(inode); - int ret; - spin_lock(&ci->i_ceph_lock); - if (ci->i_flushing_caps) - ret = ci->i_cap_flush_seq >= want_flush_seq; - else - ret = 1; - spin_unlock(&ci->i_ceph_lock); + struct rb_node *n; + struct ceph_cap_flush *cf; + int ret = 1; + + spin_lock(&mdsc->cap_dirty_lock); + n = rb_first(&mdsc->cap_flush_tree); + cf = n ? rb_entry(n, struct ceph_cap_flush, g_node) : NULL; + if (cf && cf->tid <= want_flush_tid) { + dout("check_caps_flush still flushing tid %llu <= %llu\n", + cf->tid, want_flush_tid); + ret = 0; + } + spin_unlock(&mdsc->cap_dirty_lock); return ret; } /* * flush all dirty inode data to disk. * - * returns true if we've flushed through want_flush_seq + * returns true if we've flushed through want_flush_tid */ -static void wait_caps_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq) +static void wait_caps_flush(struct ceph_mds_client *mdsc, + u64 want_flush_tid, u64 want_snap_seq) { int mds; - dout("check_cap_flush want %lld\n", want_flush_seq); + dout("check_caps_flush want %llu snap want %llu\n", + want_flush_tid, want_snap_seq); mutex_lock(&mdsc->mutex); - for (mds = 0; mds < mdsc->max_sessions; mds++) { + for (mds = 0; mds < mdsc->max_sessions; ) { struct ceph_mds_session *session = mdsc->sessions[mds]; struct inode *inode = NULL; - if (!session) + if (!session) { + mds++; continue; + } get_session(session); mutex_unlock(&mdsc->mutex); mutex_lock(&session->s_mutex); - if (!list_empty(&session->s_cap_flushing)) { - struct ceph_inode_info *ci = - list_entry(session->s_cap_flushing.next, - struct ceph_inode_info, - i_flushing_item); - - if (!check_cap_flush(&ci->vfs_inode, want_flush_seq)) { - dout("check_cap_flush still flushing %p " - "seq %lld <= %lld to mds%d\n", - &ci->vfs_inode, ci->i_cap_flush_seq, - want_flush_seq, session->s_mds); + if (!list_empty(&session->s_cap_snaps_flushing)) { + struct ceph_cap_snap *capsnap = + list_first_entry(&session->s_cap_snaps_flushing, + struct ceph_cap_snap, + flushing_item); + struct ceph_inode_info *ci = capsnap->ci; + if (!check_capsnap_flush(ci, want_snap_seq)) { + dout("check_cap_flush still flushing snap %p " + "follows %lld <= %lld to mds%d\n", + &ci->vfs_inode, capsnap->follows, + want_snap_seq, mds); inode = igrab(&ci->vfs_inode); } } @@ -1540,15 +1553,21 @@ static void wait_caps_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq) if (inode) { wait_event(mdsc->cap_flushing_wq, - check_cap_flush(inode, want_flush_seq)); + check_capsnap_flush(ceph_inode(inode), + want_snap_seq)); iput(inode); + } else { + mds++; } mutex_lock(&mdsc->mutex); } - mutex_unlock(&mdsc->mutex); - dout("check_cap_flush ok, flushed thru %lld\n", want_flush_seq); + + wait_event(mdsc->cap_flushing_wq, + check_caps_flush(mdsc, want_flush_tid)); + + dout("check_caps_flush ok, flushed thru %llu\n", want_flush_tid); } /* @@ -1557,60 +1576,74 @@ static void wait_caps_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq) void ceph_send_cap_releases(struct ceph_mds_client *mdsc, struct ceph_mds_session *session) { - struct ceph_msg *msg; + struct ceph_msg *msg = NULL; + struct ceph_mds_cap_release *head; + struct ceph_mds_cap_item *item; + struct ceph_cap *cap; + LIST_HEAD(tmp_list); + int num_cap_releases; - dout("send_cap_releases mds%d\n", session->s_mds); spin_lock(&session->s_cap_lock); - while (!list_empty(&session->s_cap_releases_done)) { - msg = list_first_entry(&session->s_cap_releases_done, - struct ceph_msg, list_head); - list_del_init(&msg->list_head); - spin_unlock(&session->s_cap_lock); - msg->hdr.front_len = cpu_to_le32(msg->front.iov_len); - dout("send_cap_releases mds%d %p\n", session->s_mds, msg); - ceph_con_send(&session->s_con, msg); - spin_lock(&session->s_cap_lock); - } +again: + list_splice_init(&session->s_cap_releases, &tmp_list); + num_cap_releases = session->s_num_cap_releases; + session->s_num_cap_releases = 0; spin_unlock(&session->s_cap_lock); -} -static void discard_cap_releases(struct ceph_mds_client *mdsc, - struct ceph_mds_session *session) -{ - struct ceph_msg *msg; - struct ceph_mds_cap_release *head; - unsigned num; - - dout("discard_cap_releases mds%d\n", session->s_mds); + while (!list_empty(&tmp_list)) { + if (!msg) { + msg = ceph_msg_new(CEPH_MSG_CLIENT_CAPRELEASE, + PAGE_CACHE_SIZE, GFP_NOFS, false); + if (!msg) + goto out_err; + head = msg->front.iov_base; + head->num = cpu_to_le32(0); + msg->front.iov_len = sizeof(*head); + } + cap = list_first_entry(&tmp_list, struct ceph_cap, + session_caps); + list_del(&cap->session_caps); + num_cap_releases--; - if (!list_empty(&session->s_cap_releases)) { - /* zero out the in-progress message */ - msg = list_first_entry(&session->s_cap_releases, - struct ceph_msg, list_head); head = msg->front.iov_base; - num = le32_to_cpu(head->num); - dout("discard_cap_releases mds%d %p %u\n", - session->s_mds, msg, num); - head->num = cpu_to_le32(0); - msg->front.iov_len = sizeof(*head); - session->s_num_cap_releases += num; + le32_add_cpu(&head->num, 1); + item = msg->front.iov_base + msg->front.iov_len; + item->ino = cpu_to_le64(cap->cap_ino); + item->cap_id = cpu_to_le64(cap->cap_id); + item->migrate_seq = cpu_to_le32(cap->mseq); + item->seq = cpu_to_le32(cap->issue_seq); + msg->front.iov_len += sizeof(*item); + + ceph_put_cap(mdsc, cap); + + if (le32_to_cpu(head->num) == CEPH_CAPS_PER_RELEASE) { + msg->hdr.front_len = cpu_to_le32(msg->front.iov_len); + dout("send_cap_releases mds%d %p\n", session->s_mds, msg); + ceph_con_send(&session->s_con, msg); + msg = NULL; + } } - /* requeue completed messages */ - while (!list_empty(&session->s_cap_releases_done)) { - msg = list_first_entry(&session->s_cap_releases_done, - struct ceph_msg, list_head); - list_del_init(&msg->list_head); + BUG_ON(num_cap_releases != 0); - head = msg->front.iov_base; - num = le32_to_cpu(head->num); - dout("discard_cap_releases mds%d %p %u\n", session->s_mds, msg, - num); - session->s_num_cap_releases += num; - head->num = cpu_to_le32(0); - msg->front.iov_len = sizeof(*head); - list_add(&msg->list_head, &session->s_cap_releases); + spin_lock(&session->s_cap_lock); + if (!list_empty(&session->s_cap_releases)) + goto again; + spin_unlock(&session->s_cap_lock); + + if (msg) { + msg->hdr.front_len = cpu_to_le32(msg->front.iov_len); + dout("send_cap_releases mds%d %p\n", session->s_mds, msg); + ceph_con_send(&session->s_con, msg); } + return; +out_err: + pr_err("send_cap_releases mds%d, failed to allocate message\n", + session->s_mds); + spin_lock(&session->s_cap_lock); + list_splice(&tmp_list, &session->s_cap_releases); + session->s_num_cap_releases += num_cap_releases; + spin_unlock(&session->s_cap_lock); } /* @@ -1635,7 +1668,8 @@ int ceph_alloc_readdir_reply_buffer(struct ceph_mds_request *req, order = get_order(size * num_entries); while (order >= 0) { - rinfo->dir_in = (void*)__get_free_pages(GFP_NOFS | __GFP_NOWARN, + rinfo->dir_in = (void*)__get_free_pages(GFP_KERNEL | + __GFP_NOWARN, order); if (rinfo->dir_in) break; @@ -1697,13 +1731,9 @@ static struct ceph_mds_request *__get_oldest_req(struct ceph_mds_client *mdsc) struct ceph_mds_request, r_node); } -static u64 __get_oldest_tid(struct ceph_mds_client *mdsc) +static inline u64 __get_oldest_tid(struct ceph_mds_client *mdsc) { - struct ceph_mds_request *req = __get_oldest_req(mdsc); - - if (req) - return req->r_tid; - return 0; + return mdsc->oldest_tid; } /* @@ -2267,15 +2297,18 @@ int ceph_mdsc_do_request(struct ceph_mds_client *mdsc, /* wait */ mutex_unlock(&mdsc->mutex); dout("do_request waiting\n"); - if (req->r_timeout) { - err = (long)wait_for_completion_killable_timeout( - &req->r_completion, req->r_timeout); - if (err == 0) - err = -EIO; - } else if (req->r_wait_for_completion) { + if (!req->r_timeout && req->r_wait_for_completion) { err = req->r_wait_for_completion(mdsc, req); } else { - err = wait_for_completion_killable(&req->r_completion); + long timeleft = wait_for_completion_killable_timeout( + &req->r_completion, + ceph_timeout_jiffies(req->r_timeout)); + if (timeleft > 0) + err = 0; + else if (!timeleft) + err = -EIO; /* timed out */ + else + err = timeleft; /* killed */ } dout("do_request waited, got %d\n", err); mutex_lock(&mdsc->mutex); @@ -2496,7 +2529,6 @@ out_err: } mutex_unlock(&mdsc->mutex); - ceph_add_cap_releases(mdsc, req->r_session); mutex_unlock(&session->s_mutex); /* kick calling process */ @@ -2888,8 +2920,7 @@ static void send_mds_reconnect(struct ceph_mds_client *mdsc, */ session->s_cap_reconnect = 1; /* drop old cap expires; we're about to reestablish that state */ - discard_cap_releases(mdsc, session); - spin_unlock(&session->s_cap_lock); + cleanup_cap_releases(mdsc, session); /* trim unused caps to reduce MDS's cache rejoin time */ if (mdsc->fsc->sb->s_root) @@ -2956,6 +2987,9 @@ static void send_mds_reconnect(struct ceph_mds_client *mdsc, reply->hdr.data_len = cpu_to_le32(pagelist->length); ceph_msg_data_add_pagelist(reply, pagelist); + + ceph_early_kick_flushing_caps(mdsc, session); + ceph_con_send(&session->s_con, reply); mutex_unlock(&session->s_mutex); @@ -3352,7 +3386,6 @@ static void delayed_work(struct work_struct *work) send_renew_caps(mdsc, s); else ceph_con_keepalive(&s->s_con); - ceph_add_cap_releases(mdsc, s); if (s->s_state == CEPH_MDS_SESSION_OPEN || s->s_state == CEPH_MDS_SESSION_HUNG) ceph_send_cap_releases(mdsc, s); @@ -3390,11 +3423,13 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) atomic_set(&mdsc->num_sessions, 0); mdsc->max_sessions = 0; mdsc->stopping = 0; + mdsc->last_snap_seq = 0; init_rwsem(&mdsc->snap_rwsem); mdsc->snap_realms = RB_ROOT; INIT_LIST_HEAD(&mdsc->snap_empty); spin_lock_init(&mdsc->snap_empty_lock); mdsc->last_tid = 0; + mdsc->oldest_tid = 0; mdsc->request_tree = RB_ROOT; INIT_DELAYED_WORK(&mdsc->delayed_work, delayed_work); mdsc->last_renew_caps = jiffies; @@ -3402,7 +3437,8 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) spin_lock_init(&mdsc->cap_delay_lock); INIT_LIST_HEAD(&mdsc->snap_flush_list); spin_lock_init(&mdsc->snap_flush_lock); - mdsc->cap_flush_seq = 0; + mdsc->last_cap_flush_tid = 1; + mdsc->cap_flush_tree = RB_ROOT; INIT_LIST_HEAD(&mdsc->cap_dirty); INIT_LIST_HEAD(&mdsc->cap_dirty_migrating); mdsc->num_cap_flushing = 0; @@ -3414,6 +3450,9 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) ceph_caps_init(mdsc); ceph_adjust_min_caps(mdsc, fsc->min_caps); + init_rwsem(&mdsc->pool_perm_rwsem); + mdsc->pool_perm_tree = RB_ROOT; + return 0; } @@ -3423,8 +3462,8 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) */ static void wait_requests(struct ceph_mds_client *mdsc) { + struct ceph_options *opts = mdsc->fsc->client->options; struct ceph_mds_request *req; - struct ceph_fs_client *fsc = mdsc->fsc; mutex_lock(&mdsc->mutex); if (__get_oldest_req(mdsc)) { @@ -3432,7 +3471,7 @@ static void wait_requests(struct ceph_mds_client *mdsc) dout("wait_requests waiting for requests\n"); wait_for_completion_timeout(&mdsc->safe_umount_waiters, - fsc->client->options->mount_timeout * HZ); + ceph_timeout_jiffies(opts->mount_timeout)); /* tear down remaining requests */ mutex_lock(&mdsc->mutex); @@ -3485,7 +3524,8 @@ restart: nextreq = rb_entry(n, struct ceph_mds_request, r_node); else nextreq = NULL; - if ((req->r_op & CEPH_MDS_OP_WRITE)) { + if (req->r_op != CEPH_MDS_OP_SETFILELOCK && + (req->r_op & CEPH_MDS_OP_WRITE)) { /* write op */ ceph_mdsc_get_request(req); if (nextreq) @@ -3513,7 +3553,7 @@ restart: void ceph_mdsc_sync(struct ceph_mds_client *mdsc) { - u64 want_tid, want_flush; + u64 want_tid, want_flush, want_snap; if (mdsc->fsc->mount_state == CEPH_MOUNT_SHUTDOWN) return; @@ -3525,13 +3565,18 @@ void ceph_mdsc_sync(struct ceph_mds_client *mdsc) ceph_flush_dirty_caps(mdsc); spin_lock(&mdsc->cap_dirty_lock); - want_flush = mdsc->cap_flush_seq; + want_flush = mdsc->last_cap_flush_tid; spin_unlock(&mdsc->cap_dirty_lock); - dout("sync want tid %lld flush_seq %lld\n", want_tid, want_flush); + down_read(&mdsc->snap_rwsem); + want_snap = mdsc->last_snap_seq; + up_read(&mdsc->snap_rwsem); + + dout("sync want tid %lld flush_seq %lld snap_seq %lld\n", + want_tid, want_flush, want_snap); wait_unsafe_requests(mdsc, want_tid); - wait_caps_flush(mdsc, want_flush); + wait_caps_flush(mdsc, want_flush, want_snap); } /* @@ -3549,10 +3594,9 @@ static bool done_closing_sessions(struct ceph_mds_client *mdsc) */ void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc) { + struct ceph_options *opts = mdsc->fsc->client->options; struct ceph_mds_session *session; int i; - struct ceph_fs_client *fsc = mdsc->fsc; - unsigned long timeout = fsc->client->options->mount_timeout * HZ; dout("close_sessions\n"); @@ -3573,7 +3617,7 @@ void ceph_mdsc_close_sessions(struct ceph_mds_client *mdsc) dout("waiting for sessions to close\n"); wait_event_timeout(mdsc->session_close_wq, done_closing_sessions(mdsc), - timeout); + ceph_timeout_jiffies(opts->mount_timeout)); /* tear down remaining sessions */ mutex_lock(&mdsc->mutex); @@ -3607,6 +3651,7 @@ static void ceph_mdsc_stop(struct ceph_mds_client *mdsc) ceph_mdsmap_destroy(mdsc->mdsmap); kfree(mdsc->sessions); ceph_caps_finalize(mdsc); + ceph_pool_perm_destroy(mdsc); } void ceph_mdsc_destroy(struct ceph_fs_client *fsc) diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index 1875b5d..762757e 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -139,7 +139,6 @@ struct ceph_mds_session { int s_cap_reconnect; int s_readonly; struct list_head s_cap_releases; /* waiting cap_release messages */ - struct list_head s_cap_releases_done; /* ready to send */ struct ceph_cap *s_cap_iterator; /* protected by mutex */ @@ -228,7 +227,7 @@ struct ceph_mds_request { int r_err; bool r_aborted; - unsigned long r_timeout; /* optional. jiffies */ + unsigned long r_timeout; /* optional. jiffies, 0 is "wait forever" */ unsigned long r_started; /* start time to measure timeout against */ unsigned long r_request_started; /* start time for mds request only, used to measure lease durations */ @@ -254,12 +253,21 @@ struct ceph_mds_request { bool r_got_unsafe, r_got_safe, r_got_result; bool r_did_prepopulate; + long long r_dir_release_cnt; + long long r_dir_ordered_cnt; + int r_readdir_cache_idx; u32 r_readdir_offset; struct ceph_cap_reservation r_caps_reservation; int r_num_caps; }; +struct ceph_pool_perm { + struct rb_node node; + u32 pool; + int perm; +}; + /* * mds client state */ @@ -284,12 +292,15 @@ struct ceph_mds_client { * references (implying they contain no inodes with caps) that * should be destroyed. */ + u64 last_snap_seq; struct rw_semaphore snap_rwsem; struct rb_root snap_realms; struct list_head snap_empty; spinlock_t snap_empty_lock; /* protect snap_empty */ u64 last_tid; /* most recent mds request */ + u64 oldest_tid; /* oldest incomplete mds request, + excluding setfilelock requests */ struct rb_root request_tree; /* pending mds requests */ struct delayed_work delayed_work; /* delayed work */ unsigned long last_renew_caps; /* last time we renewed our caps */ @@ -298,7 +309,8 @@ struct ceph_mds_client { struct list_head snap_flush_list; /* cap_snaps ready to flush */ spinlock_t snap_flush_lock; - u64 cap_flush_seq; + u64 last_cap_flush_tid; + struct rb_root cap_flush_tree; struct list_head cap_dirty; /* inodes with dirty caps */ struct list_head cap_dirty_migrating; /* ...that are migration... */ int num_cap_flushing; /* # caps we are flushing */ @@ -328,6 +340,9 @@ struct ceph_mds_client { spinlock_t dentry_lru_lock; struct list_head dentry_lru; int num_dentry; + + struct rw_semaphore pool_perm_rwsem; + struct rb_root pool_perm_tree; }; extern const char *ceph_mds_op_name(int op); @@ -379,8 +394,6 @@ static inline void ceph_mdsc_put_request(struct ceph_mds_request *req) kref_put(&req->r_kref, ceph_mdsc_release_request); } -extern int ceph_add_cap_releases(struct ceph_mds_client *mdsc, - struct ceph_mds_session *session); extern void ceph_send_cap_releases(struct ceph_mds_client *mdsc, struct ceph_mds_session *session); diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index a97e39f..233d906a 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -296,7 +296,7 @@ static int cmpu64_rev(const void *a, const void *b) } -static struct ceph_snap_context *empty_snapc; +struct ceph_snap_context *ceph_empty_snapc; /* * build the snap context for a given realm. @@ -338,9 +338,9 @@ static int build_snap_context(struct ceph_snap_realm *realm) return 0; } - if (num == 0 && realm->seq == empty_snapc->seq) { - ceph_get_snap_context(empty_snapc); - snapc = empty_snapc; + if (num == 0 && realm->seq == ceph_empty_snapc->seq) { + ceph_get_snap_context(ceph_empty_snapc); + snapc = ceph_empty_snapc; goto done; } @@ -436,6 +436,14 @@ static int dup_array(u64 **dst, __le64 *src, u32 num) return 0; } +static bool has_new_snaps(struct ceph_snap_context *o, + struct ceph_snap_context *n) +{ + if (n->num_snaps == 0) + return false; + /* snaps are in descending order */ + return n->snaps[0] > o->seq; +} /* * When a snapshot is applied, the size/mtime inode metadata is queued @@ -455,6 +463,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci) { struct inode *inode = &ci->vfs_inode; struct ceph_cap_snap *capsnap; + struct ceph_snap_context *old_snapc, *new_snapc; int used, dirty; capsnap = kzalloc(sizeof(*capsnap), GFP_NOFS); @@ -467,6 +476,9 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci) used = __ceph_caps_used(ci); dirty = __ceph_caps_dirty(ci); + old_snapc = ci->i_head_snapc; + new_snapc = ci->i_snap_realm->cached_context; + /* * If there is a write in progress, treat that as a dirty Fw, * even though it hasn't completed yet; by the time we finish @@ -481,76 +493,95 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci) writes in progress now were started before the previous cap_snap. lucky us. */ dout("queue_cap_snap %p already pending\n", inode); - kfree(capsnap); - } else if (ci->i_snap_realm->cached_context == empty_snapc) { - dout("queue_cap_snap %p empty snapc\n", inode); - kfree(capsnap); - } else if (dirty & (CEPH_CAP_AUTH_EXCL|CEPH_CAP_XATTR_EXCL| - CEPH_CAP_FILE_EXCL|CEPH_CAP_FILE_WR)) { - struct ceph_snap_context *snapc = ci->i_head_snapc; - - /* - * if we are a sync write, we may need to go to the snaprealm - * to get the current snapc. - */ - if (!snapc) - snapc = ci->i_snap_realm->cached_context; + goto update_snapc; + } + if (ci->i_wrbuffer_ref_head == 0 && + !(dirty & (CEPH_CAP_ANY_EXCL|CEPH_CAP_FILE_WR))) { + dout("queue_cap_snap %p nothing dirty|writing\n", inode); + goto update_snapc; + } - dout("queue_cap_snap %p cap_snap %p queuing under %p %s\n", - inode, capsnap, snapc, ceph_cap_string(dirty)); - ihold(inode); + BUG_ON(!old_snapc); - atomic_set(&capsnap->nref, 1); - capsnap->ci = ci; - INIT_LIST_HEAD(&capsnap->ci_item); - INIT_LIST_HEAD(&capsnap->flushing_item); - - capsnap->follows = snapc->seq; - capsnap->issued = __ceph_caps_issued(ci, NULL); - capsnap->dirty = dirty; - - capsnap->mode = inode->i_mode; - capsnap->uid = inode->i_uid; - capsnap->gid = inode->i_gid; - - if (dirty & CEPH_CAP_XATTR_EXCL) { - __ceph_build_xattrs_blob(ci); - capsnap->xattr_blob = - ceph_buffer_get(ci->i_xattrs.blob); - capsnap->xattr_version = ci->i_xattrs.version; - } else { - capsnap->xattr_blob = NULL; - capsnap->xattr_version = 0; + /* + * There is no need to send FLUSHSNAP message to MDS if there is + * no new snapshot. But when there is dirty pages or on-going + * writes, we still need to create cap_snap. cap_snap is needed + * by the write path and page writeback path. + * + * also see ceph_try_drop_cap_snap() + */ + if (has_new_snaps(old_snapc, new_snapc)) { + if (dirty & (CEPH_CAP_ANY_EXCL|CEPH_CAP_FILE_WR)) + capsnap->need_flush = true; + } else { + if (!(used & CEPH_CAP_FILE_WR) && + ci->i_wrbuffer_ref_head == 0) { + dout("queue_cap_snap %p " + "no new_snap|dirty_page|writing\n", inode); + goto update_snapc; } + } - capsnap->inline_data = ci->i_inline_version != CEPH_INLINE_NONE; - - /* dirty page count moved from _head to this cap_snap; - all subsequent writes page dirties occur _after_ this - snapshot. */ - capsnap->dirty_pages = ci->i_wrbuffer_ref_head; - ci->i_wrbuffer_ref_head = 0; - capsnap->context = snapc; - ci->i_head_snapc = - ceph_get_snap_context(ci->i_snap_realm->cached_context); - dout(" new snapc is %p\n", ci->i_head_snapc); - list_add_tail(&capsnap->ci_item, &ci->i_cap_snaps); - - if (used & CEPH_CAP_FILE_WR) { - dout("queue_cap_snap %p cap_snap %p snapc %p" - " seq %llu used WR, now pending\n", inode, - capsnap, snapc, snapc->seq); - capsnap->writing = 1; - } else { - /* note mtime, size NOW. */ - __ceph_finish_cap_snap(ci, capsnap); - } + dout("queue_cap_snap %p cap_snap %p queuing under %p %s %s\n", + inode, capsnap, old_snapc, ceph_cap_string(dirty), + capsnap->need_flush ? "" : "no_flush"); + ihold(inode); + + atomic_set(&capsnap->nref, 1); + capsnap->ci = ci; + INIT_LIST_HEAD(&capsnap->ci_item); + INIT_LIST_HEAD(&capsnap->flushing_item); + + capsnap->follows = old_snapc->seq; + capsnap->issued = __ceph_caps_issued(ci, NULL); + capsnap->dirty = dirty; + + capsnap->mode = inode->i_mode; + capsnap->uid = inode->i_uid; + capsnap->gid = inode->i_gid; + + if (dirty & CEPH_CAP_XATTR_EXCL) { + __ceph_build_xattrs_blob(ci); + capsnap->xattr_blob = + ceph_buffer_get(ci->i_xattrs.blob); + capsnap->xattr_version = ci->i_xattrs.version; } else { - dout("queue_cap_snap %p nothing dirty|writing\n", inode); - kfree(capsnap); + capsnap->xattr_blob = NULL; + capsnap->xattr_version = 0; } + capsnap->inline_data = ci->i_inline_version != CEPH_INLINE_NONE; + + /* dirty page count moved from _head to this cap_snap; + all subsequent writes page dirties occur _after_ this + snapshot. */ + capsnap->dirty_pages = ci->i_wrbuffer_ref_head; + ci->i_wrbuffer_ref_head = 0; + capsnap->context = old_snapc; + list_add_tail(&capsnap->ci_item, &ci->i_cap_snaps); + old_snapc = NULL; + + if (used & CEPH_CAP_FILE_WR) { + dout("queue_cap_snap %p cap_snap %p snapc %p" + " seq %llu used WR, now pending\n", inode, + capsnap, old_snapc, old_snapc->seq); + capsnap->writing = 1; + } else { + /* note mtime, size NOW. */ + __ceph_finish_cap_snap(ci, capsnap); + } + capsnap = NULL; + +update_snapc: + if (ci->i_head_snapc) { + ci->i_head_snapc = ceph_get_snap_context(new_snapc); + dout(" new snapc is %p\n", new_snapc); + } spin_unlock(&ci->i_ceph_lock); + + kfree(capsnap); + ceph_put_snap_context(old_snapc); } /* @@ -699,6 +730,8 @@ more: /* queue realm for cap_snap creation */ list_add(&realm->dirty_item, &dirty_realms); + if (realm->seq > mdsc->last_snap_seq) + mdsc->last_snap_seq = realm->seq; invalidate = 1; } else if (!realm->cached_context) { @@ -964,14 +997,14 @@ out: int __init ceph_snap_init(void) { - empty_snapc = ceph_create_snap_context(0, GFP_NOFS); - if (!empty_snapc) + ceph_empty_snapc = ceph_create_snap_context(0, GFP_NOFS); + if (!ceph_empty_snapc) return -ENOMEM; - empty_snapc->seq = 1; + ceph_empty_snapc->seq = 1; return 0; } void ceph_snap_exit(void) { - ceph_put_snap_context(empty_snapc); + ceph_put_snap_context(ceph_empty_snapc); } diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 4e99053..d1c833c 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -134,10 +134,12 @@ enum { Opt_noino32, Opt_fscache, Opt_nofscache, + Opt_poolperm, + Opt_nopoolperm, #ifdef CONFIG_CEPH_FS_POSIX_ACL Opt_acl, #endif - Opt_noacl + Opt_noacl, }; static match_table_t fsopt_tokens = { @@ -165,6 +167,8 @@ static match_table_t fsopt_tokens = { {Opt_noino32, "noino32"}, {Opt_fscache, "fsc"}, {Opt_nofscache, "nofsc"}, + {Opt_poolperm, "poolperm"}, + {Opt_nopoolperm, "nopoolperm"}, #ifdef CONFIG_CEPH_FS_POSIX_ACL {Opt_acl, "acl"}, #endif @@ -268,6 +272,13 @@ static int parse_fsopt_token(char *c, void *private) case Opt_nofscache: fsopt->flags &= ~CEPH_MOUNT_OPT_FSCACHE; break; + case Opt_poolperm: + fsopt->flags &= ~CEPH_MOUNT_OPT_NOPOOLPERM; + printk ("pool perm"); + break; + case Opt_nopoolperm: + fsopt->flags |= CEPH_MOUNT_OPT_NOPOOLPERM; + break; #ifdef CONFIG_CEPH_FS_POSIX_ACL case Opt_acl: fsopt->sb_flags |= MS_POSIXACL; @@ -436,6 +447,8 @@ static int ceph_show_options(struct seq_file *m, struct dentry *root) seq_puts(m, ",nodcache"); if (fsopt->flags & CEPH_MOUNT_OPT_FSCACHE) seq_puts(m, ",fsc"); + if (fsopt->flags & CEPH_MOUNT_OPT_NOPOOLPERM) + seq_puts(m, ",nopoolperm"); #ifdef CONFIG_CEPH_FS_POSIX_ACL if (fsopt->sb_flags & MS_POSIXACL) @@ -609,6 +622,7 @@ static void destroy_fs_client(struct ceph_fs_client *fsc) */ struct kmem_cache *ceph_inode_cachep; struct kmem_cache *ceph_cap_cachep; +struct kmem_cache *ceph_cap_flush_cachep; struct kmem_cache *ceph_dentry_cachep; struct kmem_cache *ceph_file_cachep; @@ -634,6 +648,10 @@ static int __init init_caches(void) SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD); if (ceph_cap_cachep == NULL) goto bad_cap; + ceph_cap_flush_cachep = KMEM_CACHE(ceph_cap_flush, + SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD); + if (ceph_cap_flush_cachep == NULL) + goto bad_cap_flush; ceph_dentry_cachep = KMEM_CACHE(ceph_dentry_info, SLAB_RECLAIM_ACCOUNT|SLAB_MEM_SPREAD); @@ -652,6 +670,8 @@ static int __init init_caches(void) bad_file: kmem_cache_destroy(ceph_dentry_cachep); bad_dentry: + kmem_cache_destroy(ceph_cap_flush_cachep); +bad_cap_flush: kmem_cache_destroy(ceph_cap_cachep); bad_cap: kmem_cache_destroy(ceph_inode_cachep); @@ -668,6 +688,7 @@ static void destroy_caches(void) kmem_cache_destroy(ceph_inode_cachep); kmem_cache_destroy(ceph_cap_cachep); + kmem_cache_destroy(ceph_cap_flush_cachep); kmem_cache_destroy(ceph_dentry_cachep); kmem_cache_destroy(ceph_file_cachep); @@ -729,7 +750,7 @@ static struct dentry *open_root_dentry(struct ceph_fs_client *fsc, req->r_ino1.ino = CEPH_INO_ROOT; req->r_ino1.snap = CEPH_NOSNAP; req->r_started = started; - req->r_timeout = fsc->client->options->mount_timeout * HZ; + req->r_timeout = fsc->client->options->mount_timeout; req->r_args.getattr.mask = cpu_to_le32(CEPH_STAT_CAP_INODE); req->r_num_caps = 2; err = ceph_mdsc_do_request(mdsc, NULL, req); diff --git a/fs/ceph/super.h b/fs/ceph/super.h index fa20e13..860cc016 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -35,6 +35,7 @@ #define CEPH_MOUNT_OPT_INO32 (1<<8) /* 32 bit inos */ #define CEPH_MOUNT_OPT_DCACHE (1<<9) /* use dcache for readdir etc */ #define CEPH_MOUNT_OPT_FSCACHE (1<<10) /* use fscache */ +#define CEPH_MOUNT_OPT_NOPOOLPERM (1<<11) /* no pool permission check */ #define CEPH_MOUNT_OPT_DEFAULT (CEPH_MOUNT_OPT_RBYTES | \ CEPH_MOUNT_OPT_DCACHE) @@ -121,11 +122,21 @@ struct ceph_cap { struct rb_node ci_node; /* per-ci cap tree */ struct ceph_mds_session *session; struct list_head session_caps; /* per-session caplist */ - int mds; u64 cap_id; /* unique cap id (mds provided) */ - int issued; /* latest, from the mds */ - int implemented; /* implemented superset of issued (for revocation) */ - int mds_wanted; + union { + /* in-use caps */ + struct { + int issued; /* latest, from the mds */ + int implemented; /* implemented superset of + issued (for revocation) */ + int mds, mds_wanted; + }; + /* caps to release */ + struct { + u64 cap_ino; + int queue_release; + }; + }; u32 seq, issue_seq, mseq; u32 cap_gen; /* active/stale cycle */ unsigned long last_used; @@ -163,6 +174,7 @@ struct ceph_cap_snap { int writing; /* a sync write is still in progress */ int dirty_pages; /* dirty pages awaiting writeback */ bool inline_data; + bool need_flush; }; static inline void ceph_put_cap_snap(struct ceph_cap_snap *capsnap) @@ -174,6 +186,17 @@ static inline void ceph_put_cap_snap(struct ceph_cap_snap *capsnap) } } +struct ceph_cap_flush { + u64 tid; + int caps; + bool kick; + struct rb_node g_node; // global + union { + struct rb_node i_node; // inode + struct list_head list; + }; +}; + /* * The frag tree describes how a directory is fragmented, potentially across * multiple metadata servers. It is also used to indicate points where @@ -259,9 +282,9 @@ struct ceph_inode_info { u32 i_time_warp_seq; unsigned i_ceph_flags; - int i_ordered_count; - atomic_t i_release_count; - atomic_t i_complete_count; + atomic64_t i_release_count; + atomic64_t i_ordered_count; + atomic64_t i_complete_seq[2]; struct ceph_dir_layout i_dir_layout; struct ceph_file_layout i_layout; @@ -283,11 +306,11 @@ struct ceph_inode_info { struct ceph_cap *i_auth_cap; /* authoritative cap, if any */ unsigned i_dirty_caps, i_flushing_caps; /* mask of dirtied fields */ struct list_head i_dirty_item, i_flushing_item; - u64 i_cap_flush_seq; /* we need to track cap writeback on a per-cap-bit basis, to allow * overlapping, pipelined cap flushes to the mds. we can probably * reduce the tid to 8 bits if we're concerned about inode size. */ - u16 i_cap_flush_last_tid, i_cap_flush_tid[CEPH_CAP_BITS]; + struct ceph_cap_flush *i_prealloc_cap_flush; + struct rb_root i_cap_flush_tree; wait_queue_head_t i_cap_wq; /* threads waiting on a capability */ unsigned long i_hold_caps_min; /* jiffies */ unsigned long i_hold_caps_max; /* jiffies */ @@ -438,36 +461,46 @@ static inline struct inode *ceph_find_inode(struct super_block *sb, /* * Ceph inode. */ -#define CEPH_I_DIR_ORDERED 1 /* dentries in dir are ordered */ -#define CEPH_I_NODELAY 4 /* do not delay cap release */ -#define CEPH_I_FLUSH 8 /* do not delay flush of dirty metadata */ -#define CEPH_I_NOFLUSH 16 /* do not flush dirty caps */ +#define CEPH_I_DIR_ORDERED (1 << 0) /* dentries in dir are ordered */ +#define CEPH_I_NODELAY (1 << 1) /* do not delay cap release */ +#define CEPH_I_FLUSH (1 << 2) /* do not delay flush of dirty metadata */ +#define CEPH_I_NOFLUSH (1 << 3) /* do not flush dirty caps */ +#define CEPH_I_POOL_PERM (1 << 4) /* pool rd/wr bits are valid */ +#define CEPH_I_POOL_RD (1 << 5) /* can read from pool */ +#define CEPH_I_POOL_WR (1 << 6) /* can write to pool */ + static inline void __ceph_dir_set_complete(struct ceph_inode_info *ci, - int release_count, int ordered_count) + long long release_count, + long long ordered_count) { - atomic_set(&ci->i_complete_count, release_count); - if (ci->i_ordered_count == ordered_count) - ci->i_ceph_flags |= CEPH_I_DIR_ORDERED; - else - ci->i_ceph_flags &= ~CEPH_I_DIR_ORDERED; + smp_mb__before_atomic(); + atomic64_set(&ci->i_complete_seq[0], release_count); + atomic64_set(&ci->i_complete_seq[1], ordered_count); } static inline void __ceph_dir_clear_complete(struct ceph_inode_info *ci) { - atomic_inc(&ci->i_release_count); + atomic64_inc(&ci->i_release_count); +} + +static inline void __ceph_dir_clear_ordered(struct ceph_inode_info *ci) +{ + atomic64_inc(&ci->i_ordered_count); } static inline bool __ceph_dir_is_complete(struct ceph_inode_info *ci) { - return atomic_read(&ci->i_complete_count) == - atomic_read(&ci->i_release_count); + return atomic64_read(&ci->i_complete_seq[0]) == + atomic64_read(&ci->i_release_count); } static inline bool __ceph_dir_is_complete_ordered(struct ceph_inode_info *ci) { - return __ceph_dir_is_complete(ci) && - (ci->i_ceph_flags & CEPH_I_DIR_ORDERED); + return atomic64_read(&ci->i_complete_seq[0]) == + atomic64_read(&ci->i_release_count) && + atomic64_read(&ci->i_complete_seq[1]) == + atomic64_read(&ci->i_ordered_count); } static inline void ceph_dir_clear_complete(struct inode *inode) @@ -477,20 +510,13 @@ static inline void ceph_dir_clear_complete(struct inode *inode) static inline void ceph_dir_clear_ordered(struct inode *inode) { - struct ceph_inode_info *ci = ceph_inode(inode); - spin_lock(&ci->i_ceph_lock); - ci->i_ordered_count++; - ci->i_ceph_flags &= ~CEPH_I_DIR_ORDERED; - spin_unlock(&ci->i_ceph_lock); + __ceph_dir_clear_ordered(ceph_inode(inode)); } static inline bool ceph_dir_is_complete_ordered(struct inode *inode) { - struct ceph_inode_info *ci = ceph_inode(inode); - bool ret; - spin_lock(&ci->i_ceph_lock); - ret = __ceph_dir_is_complete_ordered(ci); - spin_unlock(&ci->i_ceph_lock); + bool ret = __ceph_dir_is_complete_ordered(ceph_inode(inode)); + smp_rmb(); return ret; } @@ -552,7 +578,10 @@ static inline int __ceph_caps_dirty(struct ceph_inode_info *ci) { return ci->i_dirty_caps | ci->i_flushing_caps; } -extern int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask); +extern struct ceph_cap_flush *ceph_alloc_cap_flush(void); +extern void ceph_free_cap_flush(struct ceph_cap_flush *cf); +extern int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask, + struct ceph_cap_flush **pcf); extern int __ceph_caps_revoking_other(struct ceph_inode_info *ci, struct ceph_cap *ocap, int mask); @@ -606,16 +635,20 @@ struct ceph_file_info { unsigned offset; /* offset of last chunk, adjusted for . and .. */ unsigned next_offset; /* offset of next chunk (last_name's + 1) */ char *last_name; /* last entry in previous chunk */ - struct dentry *dentry; /* next dentry (for dcache readdir) */ - int dir_release_count; - int dir_ordered_count; + long long dir_release_count; + long long dir_ordered_count; + int readdir_cache_idx; /* used for -o dirstat read() on directory thing */ char *dir_info; int dir_info_len; }; - +struct ceph_readdir_cache_control { + struct page *page; + struct dentry **dentries; + int index; +}; /* * A "snap realm" describes a subset of the file hierarchy sharing @@ -687,6 +720,7 @@ static inline int default_congestion_kb(void) /* snap.c */ +extern struct ceph_snap_context *ceph_empty_snapc; struct ceph_snap_realm *ceph_lookup_snap_realm(struct ceph_mds_client *mdsc, u64 ino); extern void ceph_get_snap_realm(struct ceph_mds_client *mdsc, @@ -713,8 +747,8 @@ extern void ceph_snap_exit(void); static inline bool __ceph_have_pending_cap_snap(struct ceph_inode_info *ci) { return !list_empty(&ci->i_cap_snaps) && - list_entry(ci->i_cap_snaps.prev, struct ceph_cap_snap, - ci_item)->writing; + list_last_entry(&ci->i_cap_snaps, struct ceph_cap_snap, + ci_item)->writing; } /* inode.c */ @@ -838,12 +872,12 @@ extern void ceph_put_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap); extern int ceph_is_any_caps(struct inode *inode); -extern void __queue_cap_release(struct ceph_mds_session *session, u64 ino, - u64 cap_id, u32 migrate_seq, u32 issue_seq); extern void ceph_queue_caps_release(struct inode *inode); extern int ceph_write_inode(struct inode *inode, struct writeback_control *wbc); extern int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync); +extern void ceph_early_kick_flushing_caps(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session); extern void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc, struct ceph_mds_session *session); extern struct ceph_cap *ceph_get_cap_for_mds(struct ceph_inode_info *ci, @@ -879,6 +913,9 @@ extern void ceph_put_fmode(struct ceph_inode_info *ci, int mode); /* addr.c */ extern const struct address_space_operations ceph_aops; extern int ceph_mmap(struct file *file, struct vm_area_struct *vma); +extern int ceph_uninline_data(struct file *filp, struct page *locked_page); +extern int ceph_pool_perm_check(struct ceph_inode_info *ci, int need); +extern void ceph_pool_perm_destroy(struct ceph_mds_client* mdsc); /* file.c */ extern const struct file_operations ceph_file_fops; @@ -890,7 +927,6 @@ extern int ceph_atomic_open(struct inode *dir, struct dentry *dentry, extern int ceph_release(struct inode *inode, struct file *filp); extern void ceph_fill_inline_data(struct inode *inode, struct page *locked_page, char *data, size_t len); -int ceph_uninline_data(struct file *filp, struct page *locked_page); /* dir.c */ extern const struct file_operations ceph_dir_fops; extern const struct file_operations ceph_snapdir_fops; @@ -911,6 +947,7 @@ extern void ceph_dentry_lru_del(struct dentry *dn); extern void ceph_invalidate_dentry_lease(struct dentry *dentry); extern unsigned ceph_dentry_hash(struct inode *dir, struct dentry *dn); extern struct inode *ceph_get_dentry_parent_inode(struct dentry *dentry); +extern void ceph_readdir_cache_release(struct ceph_readdir_cache_control *ctl); /* * our d_ops vary depending on whether the inode is live, diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index cd7ffad4..819163d 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -911,6 +911,8 @@ int __ceph_setxattr(struct dentry *dentry, const char *name, struct inode *inode = d_inode(dentry); struct ceph_vxattr *vxattr; struct ceph_inode_info *ci = ceph_inode(inode); + struct ceph_mds_client *mdsc = ceph_sb_to_client(dentry->d_sb)->mdsc; + struct ceph_cap_flush *prealloc_cf = NULL; int issued; int err; int dirty = 0; @@ -920,6 +922,7 @@ int __ceph_setxattr(struct dentry *dentry, const char *name, char *newval = NULL; struct ceph_inode_xattr *xattr = NULL; int required_blob_size; + bool lock_snap_rwsem = false; if (!ceph_is_valid_xattr(name)) return -EOPNOTSUPP; @@ -948,12 +951,27 @@ int __ceph_setxattr(struct dentry *dentry, const char *name, if (!xattr) goto out; + prealloc_cf = ceph_alloc_cap_flush(); + if (!prealloc_cf) + goto out; + spin_lock(&ci->i_ceph_lock); retry: issued = __ceph_caps_issued(ci, NULL); - dout("setxattr %p issued %s\n", inode, ceph_cap_string(issued)); if (ci->i_xattrs.version == 0 || !(issued & CEPH_CAP_XATTR_EXCL)) goto do_sync; + + if (!lock_snap_rwsem && !ci->i_head_snapc) { + lock_snap_rwsem = true; + if (!down_read_trylock(&mdsc->snap_rwsem)) { + spin_unlock(&ci->i_ceph_lock); + down_read(&mdsc->snap_rwsem); + spin_lock(&ci->i_ceph_lock); + goto retry; + } + } + + dout("setxattr %p issued %s\n", inode, ceph_cap_string(issued)); __build_xattrs(inode); required_blob_size = __get_required_blob_size(ci, name_len, val_len); @@ -966,7 +984,7 @@ retry: dout(" preaallocating new blob size=%d\n", required_blob_size); blob = ceph_buffer_new(required_blob_size, GFP_NOFS); if (!blob) - goto out; + goto do_sync_unlocked; spin_lock(&ci->i_ceph_lock); if (ci->i_xattrs.prealloc_blob) ceph_buffer_put(ci->i_xattrs.prealloc_blob); @@ -978,21 +996,28 @@ retry: flags, value ? 1 : -1, &xattr); if (!err) { - dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL, + &prealloc_cf); ci->i_xattrs.dirty = true; inode->i_ctime = CURRENT_TIME; } spin_unlock(&ci->i_ceph_lock); + if (lock_snap_rwsem) + up_read(&mdsc->snap_rwsem); if (dirty) __mark_inode_dirty(inode, dirty); + ceph_free_cap_flush(prealloc_cf); return err; do_sync: spin_unlock(&ci->i_ceph_lock); do_sync_unlocked: + if (lock_snap_rwsem) + up_read(&mdsc->snap_rwsem); err = ceph_sync_setxattr(dentry, name, value, size, flags); out: + ceph_free_cap_flush(prealloc_cf); kfree(newname); kfree(newval); kfree(xattr); @@ -1044,10 +1069,13 @@ int __ceph_removexattr(struct dentry *dentry, const char *name) struct inode *inode = d_inode(dentry); struct ceph_vxattr *vxattr; struct ceph_inode_info *ci = ceph_inode(inode); + struct ceph_mds_client *mdsc = ceph_sb_to_client(dentry->d_sb)->mdsc; + struct ceph_cap_flush *prealloc_cf = NULL; int issued; int err; int required_blob_size; int dirty; + bool lock_snap_rwsem = false; if (!ceph_is_valid_xattr(name)) return -EOPNOTSUPP; @@ -1060,14 +1088,29 @@ int __ceph_removexattr(struct dentry *dentry, const char *name) if (!strncmp(name, XATTR_CEPH_PREFIX, XATTR_CEPH_PREFIX_LEN)) goto do_sync_unlocked; + prealloc_cf = ceph_alloc_cap_flush(); + if (!prealloc_cf) + return -ENOMEM; + err = -ENOMEM; spin_lock(&ci->i_ceph_lock); retry: issued = __ceph_caps_issued(ci, NULL); - dout("removexattr %p issued %s\n", inode, ceph_cap_string(issued)); - if (ci->i_xattrs.version == 0 || !(issued & CEPH_CAP_XATTR_EXCL)) goto do_sync; + + if (!lock_snap_rwsem && !ci->i_head_snapc) { + lock_snap_rwsem = true; + if (!down_read_trylock(&mdsc->snap_rwsem)) { + spin_unlock(&ci->i_ceph_lock); + down_read(&mdsc->snap_rwsem); + spin_lock(&ci->i_ceph_lock); + goto retry; + } + } + + dout("removexattr %p issued %s\n", inode, ceph_cap_string(issued)); + __build_xattrs(inode); required_blob_size = __get_required_blob_size(ci, 0, 0); @@ -1080,7 +1123,7 @@ retry: dout(" preaallocating new blob size=%d\n", required_blob_size); blob = ceph_buffer_new(required_blob_size, GFP_NOFS); if (!blob) - goto out; + goto do_sync_unlocked; spin_lock(&ci->i_ceph_lock); if (ci->i_xattrs.prealloc_blob) ceph_buffer_put(ci->i_xattrs.prealloc_blob); @@ -1090,18 +1133,24 @@ retry: err = __remove_xattr_by_name(ceph_inode(inode), name); - dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL, + &prealloc_cf); ci->i_xattrs.dirty = true; inode->i_ctime = CURRENT_TIME; spin_unlock(&ci->i_ceph_lock); + if (lock_snap_rwsem) + up_read(&mdsc->snap_rwsem); if (dirty) __mark_inode_dirty(inode, dirty); + ceph_free_cap_flush(prealloc_cf); return err; do_sync: spin_unlock(&ci->i_ceph_lock); do_sync_unlocked: + if (lock_snap_rwsem) + up_read(&mdsc->snap_rwsem); + ceph_free_cap_flush(prealloc_cf); err = ceph_send_removexattr(dentry, name); -out: return err; } diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index a2172f3..e7b478b 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -192,6 +192,15 @@ config CIFS_SMB2 options are also slightly simpler (compared to CIFS) due to protocol improvements. +config CIFS_SMB311 + bool "SMB3.1.1 network file system support (Experimental)" + depends on CIFS_SMB2 && INET + + help + This enables experimental support for the newest, SMB3.1.1, dialect. + This dialect includes improved security negotiation features. + If unsure, say N + config CIFS_FSCACHE bool "Provide CIFS client caching support" depends on CIFS=m && FSCACHE || CIFS=y && FSCACHE=y diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c index 430e034..7dc886c 100644 --- a/fs/cifs/cifs_dfs_ref.c +++ b/fs/cifs/cifs_dfs_ref.c @@ -24,6 +24,7 @@ #include "cifsfs.h" #include "dns_resolve.h" #include "cifs_debug.h" +#include "cifs_unicode.h" static LIST_HEAD(cifs_dfs_automount_list); @@ -312,7 +313,7 @@ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt) xid = get_xid(); rc = get_dfs_path(xid, ses, full_path + 1, cifs_sb->local_nls, &num_referrals, &referrals, - cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + cifs_remap(cifs_sb)); free_xid(xid); cifs_put_tlink(tlink); diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index 0303c67..5a53ac6 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -27,41 +27,6 @@ #include "cifsglob.h" #include "cifs_debug.h" -/* - * cifs_utf16_bytes - how long will a string be after conversion? - * @utf16 - pointer to input string - * @maxbytes - don't go past this many bytes of input string - * @codepage - destination codepage - * - * Walk a utf16le string and return the number of bytes that the string will - * be after being converted to the given charset, not including any null - * termination required. Don't walk past maxbytes in the source buffer. - */ -int -cifs_utf16_bytes(const __le16 *from, int maxbytes, - const struct nls_table *codepage) -{ - int i; - int charlen, outlen = 0; - int maxwords = maxbytes / 2; - char tmp[NLS_MAX_CHARSET_SIZE]; - __u16 ftmp; - - for (i = 0; i < maxwords; i++) { - ftmp = get_unaligned_le16(&from[i]); - if (ftmp == 0) - break; - - charlen = codepage->uni2char(ftmp, tmp, NLS_MAX_CHARSET_SIZE); - if (charlen > 0) - outlen += charlen; - else - outlen++; - } - - return outlen; -} - int cifs_remap(struct cifs_sb_info *cifs_sb) { int map_type; @@ -155,10 +120,13 @@ convert_sfm_char(const __u16 src_char, char *target) * enough to hold the result of the conversion (at least NLS_MAX_CHARSET_SIZE). */ static int -cifs_mapchar(char *target, const __u16 src_char, const struct nls_table *cp, +cifs_mapchar(char *target, const __u16 *from, const struct nls_table *cp, int maptype) { int len = 1; + __u16 src_char; + + src_char = *from; if ((maptype == SFM_MAP_UNI_RSVD) && convert_sfm_char(src_char, target)) return len; @@ -168,10 +136,23 @@ cifs_mapchar(char *target, const __u16 src_char, const struct nls_table *cp, /* if character not one of seven in special remap set */ len = cp->uni2char(src_char, target, NLS_MAX_CHARSET_SIZE); - if (len <= 0) { - *target = '?'; - len = 1; - } + if (len <= 0) + goto surrogate_pair; + + return len; + +surrogate_pair: + /* convert SURROGATE_PAIR and IVS */ + if (strcmp(cp->charset, "utf8")) + goto unknown; + len = utf16s_to_utf8s(from, 3, UTF16_LITTLE_ENDIAN, target, 6); + if (len <= 0) + goto unknown; + return len; + +unknown: + *target = '?'; + len = 1; return len; } @@ -206,7 +187,7 @@ cifs_from_utf16(char *to, const __le16 *from, int tolen, int fromlen, int nullsize = nls_nullsize(codepage); int fromwords = fromlen / 2; char tmp[NLS_MAX_CHARSET_SIZE]; - __u16 ftmp; + __u16 ftmp[3]; /* ftmp[3] = 3array x 2bytes = 6bytes UTF-16 */ /* * because the chars can be of varying widths, we need to take care @@ -217,9 +198,17 @@ cifs_from_utf16(char *to, const __le16 *from, int tolen, int fromlen, safelen = tolen - (NLS_MAX_CHARSET_SIZE + nullsize); for (i = 0; i < fromwords; i++) { - ftmp = get_unaligned_le16(&from[i]); - if (ftmp == 0) + ftmp[0] = get_unaligned_le16(&from[i]); + if (ftmp[0] == 0) break; + if (i + 1 < fromwords) + ftmp[1] = get_unaligned_le16(&from[i + 1]); + else + ftmp[1] = 0; + if (i + 2 < fromwords) + ftmp[2] = get_unaligned_le16(&from[i + 2]); + else + ftmp[2] = 0; /* * check to see if converting this character might make the @@ -234,6 +223,17 @@ cifs_from_utf16(char *to, const __le16 *from, int tolen, int fromlen, /* put converted char into 'to' buffer */ charlen = cifs_mapchar(&to[outlen], ftmp, codepage, map_type); outlen += charlen; + + /* charlen (=bytes of UTF-8 for 1 character) + * 4bytes UTF-8(surrogate pair) is charlen=4 + * (4bytes UTF-16 code) + * 7-8bytes UTF-8(IVS) is charlen=3+4 or 4+4 + * (2 UTF-8 pairs divided to 2 UTF-16 pairs) */ + if (charlen == 4) + i++; + else if (charlen >= 5) + /* 5-6bytes UTF-8 */ + i += 2; } /* properly null-terminate string */ @@ -296,6 +296,46 @@ success: } /* + * cifs_utf16_bytes - how long will a string be after conversion? + * @utf16 - pointer to input string + * @maxbytes - don't go past this many bytes of input string + * @codepage - destination codepage + * + * Walk a utf16le string and return the number of bytes that the string will + * be after being converted to the given charset, not including any null + * termination required. Don't walk past maxbytes in the source buffer. + */ +int +cifs_utf16_bytes(const __le16 *from, int maxbytes, + const struct nls_table *codepage) +{ + int i; + int charlen, outlen = 0; + int maxwords = maxbytes / 2; + char tmp[NLS_MAX_CHARSET_SIZE]; + __u16 ftmp[3]; + + for (i = 0; i < maxwords; i++) { + ftmp[0] = get_unaligned_le16(&from[i]); + if (ftmp[0] == 0) + break; + if (i + 1 < maxwords) + ftmp[1] = get_unaligned_le16(&from[i + 1]); + else + ftmp[1] = 0; + if (i + 2 < maxwords) + ftmp[2] = get_unaligned_le16(&from[i + 2]); + else + ftmp[2] = 0; + + charlen = cifs_mapchar(tmp, ftmp, codepage, NO_MAP_UNI_RSVD); + outlen += charlen; + } + + return outlen; +} + +/* * cifs_strndup_from_utf16 - copy a string from wire format to the local * codepage * @src - source string @@ -409,10 +449,15 @@ cifsConvertToUTF16(__le16 *target, const char *source, int srclen, char src_char; __le16 dst_char; wchar_t tmp; + wchar_t *wchar_to; /* UTF-16 */ + int ret; + unicode_t u; if (map_chars == NO_MAP_UNI_RSVD) return cifs_strtoUTF16(target, source, PATH_MAX, cp); + wchar_to = kzalloc(6, GFP_KERNEL); + for (i = 0; i < srclen; j++) { src_char = source[i]; charlen = 1; @@ -441,11 +486,55 @@ cifsConvertToUTF16(__le16 *target, const char *source, int srclen, * if no match, use question mark, which at least in * some cases serves as wild card */ - if (charlen < 1) { - dst_char = cpu_to_le16(0x003f); - charlen = 1; + if (charlen > 0) + goto ctoUTF16; + + /* convert SURROGATE_PAIR */ + if (strcmp(cp->charset, "utf8") || !wchar_to) + goto unknown; + if (*(source + i) & 0x80) { + charlen = utf8_to_utf32(source + i, 6, &u); + if (charlen < 0) + goto unknown; + } else + goto unknown; + ret = utf8s_to_utf16s(source + i, charlen, + UTF16_LITTLE_ENDIAN, + wchar_to, 6); + if (ret < 0) + goto unknown; + + i += charlen; + dst_char = cpu_to_le16(*wchar_to); + if (charlen <= 3) + /* 1-3bytes UTF-8 to 2bytes UTF-16 */ + put_unaligned(dst_char, &target[j]); + else if (charlen == 4) { + /* 4bytes UTF-8(surrogate pair) to 4bytes UTF-16 + * 7-8bytes UTF-8(IVS) divided to 2 UTF-16 + * (charlen=3+4 or 4+4) */ + put_unaligned(dst_char, &target[j]); + dst_char = cpu_to_le16(*(wchar_to + 1)); + j++; + put_unaligned(dst_char, &target[j]); + } else if (charlen >= 5) { + /* 5-6bytes UTF-8 to 6bytes UTF-16 */ + put_unaligned(dst_char, &target[j]); + dst_char = cpu_to_le16(*(wchar_to + 1)); + j++; + put_unaligned(dst_char, &target[j]); + dst_char = cpu_to_le16(*(wchar_to + 2)); + j++; + put_unaligned(dst_char, &target[j]); } + continue; + +unknown: + dst_char = cpu_to_le16(0x003f); + charlen = 1; } + +ctoUTF16: /* * character may take more than one byte in the source string, * but will take exactly two bytes in the target string @@ -456,6 +545,7 @@ cifsConvertToUTF16(__le16 *target, const char *source, int srclen, ctoUTF16_out: put_unaligned(0, &target[j]); /* Null terminate target unicode string */ + kfree(wchar_to); return j; } diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index f5089bd..0a9fb6b 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -469,6 +469,8 @@ cifs_show_options(struct seq_file *s, struct dentry *root) seq_puts(s, ",nouser_xattr"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR) seq_puts(s, ",mapchars"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SFM_CHR) + seq_puts(s, ",mapposix"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL) seq_puts(s, ",sfu"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 22b289a..b406a32 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -171,6 +171,10 @@ enum smb_version { Smb_21, Smb_30, Smb_302, +#ifdef CONFIG_CIFS_SMB311 + Smb_311, +#endif /* SMB311 */ + Smb_version_err }; struct mid_q_entry; @@ -368,6 +372,8 @@ struct smb_version_operations { void (*new_lease_key)(struct cifs_fid *); int (*generate_signingkey)(struct cifs_ses *); int (*calc_signature)(struct smb_rqst *, struct TCP_Server_Info *); + int (*set_integrity)(const unsigned int, struct cifs_tcon *tcon, + struct cifsFileInfo *src_file); int (*query_mf_symlink)(unsigned int, struct cifs_tcon *, struct cifs_sb_info *, const unsigned char *, char *, unsigned int *); @@ -386,6 +392,9 @@ struct smb_version_operations { int (*clone_range)(const unsigned int, struct cifsFileInfo *src_file, struct cifsFileInfo *target_file, u64 src_off, u64 len, u64 dest_off); + int (*duplicate_extents)(const unsigned int, struct cifsFileInfo *src, + struct cifsFileInfo *target_file, u64 src_off, u64 len, + u64 dest_off); int (*validate_negotiate)(const unsigned int, struct cifs_tcon *); ssize_t (*query_all_EAs)(const unsigned int, struct cifs_tcon *, const unsigned char *, const unsigned char *, char *, @@ -1617,4 +1626,8 @@ extern struct smb_version_values smb30_values; #define SMB302_VERSION_STRING "3.02" /*extern struct smb_version_operations smb302_operations;*/ /* not needed yet */ extern struct smb_version_values smb302_values; +#define SMB311_VERSION_STRING "3.1.1" +#define ALT_SMB311_VERSION_STRING "3.11" +extern struct smb_version_operations smb311_operations; +extern struct smb_version_values smb311_values; #endif /* _CIFS_GLOB_H */ diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index 5f9822ac..47b030d 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -2255,6 +2255,8 @@ typedef struct { /* List of FileSystemAttributes - see 2.5.1 of MS-FSCC */ +#define FILE_SUPPORTS_SPARSE_VDL 0x10000000 /* faster nonsparse extend */ +#define FILE_SUPPORTS_BLOCK_REFCOUNTING 0x08000000 /* allow ioctl dup extents */ #define FILE_SUPPORT_INTEGRITY_STREAMS 0x04000000 #define FILE_SUPPORTS_USN_JOURNAL 0x02000000 #define FILE_SUPPORTS_OPEN_BY_FILE_ID 0x01000000 @@ -2310,6 +2312,16 @@ typedef struct { /* data block encoding of response to level 263 QPathInfo */ char FileName[1]; } __attribute__((packed)) FILE_ALL_INFO; /* level 0x107 QPathInfo */ +typedef struct { + __le64 AllocationSize; + __le64 EndOfFile; /* size ie offset to first free byte in file */ + __le32 NumberOfLinks; /* hard links */ + __u8 DeletePending; + __u8 Directory; + __u16 Pad; +} __attribute__((packed)) FILE_STANDARD_INFO; /* level 0x102 QPathInfo */ + + /* defines for enumerating possible values of the Unix type field below */ #define UNIX_FILE 0 #define UNIX_DIR 1 diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index c31ce98..c63fd1d 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -361,11 +361,11 @@ extern int CIFSUnixCreateHardLink(const unsigned int xid, extern int CIFSUnixCreateSymLink(const unsigned int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, - const struct nls_table *nls_codepage); + const struct nls_table *nls_codepage, int remap); extern int CIFSSMBUnixQuerySymLink(const unsigned int xid, struct cifs_tcon *tcon, const unsigned char *searchName, char **syminfo, - const struct nls_table *nls_codepage); + const struct nls_table *nls_codepage, int remap); extern int CIFSSMBQuerySymLink(const unsigned int xid, struct cifs_tcon *tcon, __u16 fid, char **symlinkinfo, const struct nls_table *nls_codepage); diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 84650a5..672ef35 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -625,9 +625,8 @@ CIFSSMBNegotiate(const unsigned int xid, struct cifs_ses *ses) server->negflavor = CIFS_NEGFLAVOR_UNENCAP; memcpy(ses->server->cryptkey, pSMBr->u.EncryptionKey, CIFS_CRYPTO_KEY_SIZE); - } else if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC || - server->capabilities & CAP_EXTENDED_SECURITY) && - (pSMBr->EncryptionKeyLength == 0)) { + } else if (pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC || + server->capabilities & CAP_EXTENDED_SECURITY) { server->negflavor = CIFS_NEGFLAVOR_EXTENDED; rc = decode_ext_sec_blob(ses, pSMBr); } else if (server->sec_mode & SECMODE_PW_ENCRYPT) { @@ -2784,7 +2783,7 @@ copyRetry: int CIFSUnixCreateSymLink(const unsigned int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, - const struct nls_table *nls_codepage) + const struct nls_table *nls_codepage, int remap) { TRANSACTION2_SPI_REQ *pSMB = NULL; TRANSACTION2_SPI_RSP *pSMBr = NULL; @@ -2804,9 +2803,9 @@ createSymLinkRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifs_strtoUTF16((__le16 *) pSMB->FileName, fromName, - /* find define for this maxpathcomponent */ - PATH_MAX, nls_codepage); + cifsConvertToUTF16((__le16 *) pSMB->FileName, fromName, + /* find define for this maxpathcomponent */ + PATH_MAX, nls_codepage, remap); name_len++; /* trailing null */ name_len *= 2; @@ -2828,9 +2827,9 @@ createSymLinkRetry: data_offset = (char *) (&pSMB->hdr.Protocol) + offset; if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len_target = - cifs_strtoUTF16((__le16 *) data_offset, toName, PATH_MAX - /* find define for this maxpathcomponent */ - , nls_codepage); + cifsConvertToUTF16((__le16 *) data_offset, toName, + /* find define for this maxpathcomponent */ + PATH_MAX, nls_codepage, remap); name_len_target++; /* trailing null */ name_len_target *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -3034,7 +3033,7 @@ winCreateHardLinkRetry: int CIFSSMBUnixQuerySymLink(const unsigned int xid, struct cifs_tcon *tcon, const unsigned char *searchName, char **symlinkinfo, - const struct nls_table *nls_codepage) + const struct nls_table *nls_codepage, int remap) { /* SMB_QUERY_FILE_UNIX_LINK */ TRANSACTION2_QPI_REQ *pSMB = NULL; @@ -3055,8 +3054,9 @@ querySymLinkRetry: if (pSMB->hdr.Flags2 & SMBFLG2_UNICODE) { name_len = - cifs_strtoUTF16((__le16 *) pSMB->FileName, searchName, - PATH_MAX, nls_codepage); + cifsConvertToUTF16((__le16 *) pSMB->FileName, + searchName, PATH_MAX, nls_codepage, + remap); name_len++; /* trailing null */ name_len *= 2; } else { /* BB improve the check for buffer overruns BB */ @@ -4917,7 +4917,7 @@ getDFSRetry: strncpy(pSMB->RequestFileName, search_name, name_len); } - if (ses->server && ses->server->sign) + if (ses->server->sign) pSMB->hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; pSMB->hdr.Uid = ses->Suid; diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index f3bfe08..773f4dc 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -280,6 +280,11 @@ static const match_table_t cifs_smb_version_tokens = { { Smb_21, SMB21_VERSION_STRING }, { Smb_30, SMB30_VERSION_STRING }, { Smb_302, SMB302_VERSION_STRING }, +#ifdef CONFIG_CIFS_SMB311 + { Smb_311, SMB311_VERSION_STRING }, + { Smb_311, ALT_SMB311_VERSION_STRING }, +#endif /* SMB311 */ + { Smb_version_err, NULL } }; static int ip_connect(struct TCP_Server_Info *server); @@ -386,6 +391,7 @@ cifs_reconnect(struct TCP_Server_Info *server) rc = generic_ip_connect(server); if (rc) { cifs_dbg(FYI, "reconnect error %d\n", rc); + mutex_unlock(&server->srv_mutex); msleep(3000); } else { atomic_inc(&tcpSesReconnectCount); @@ -393,8 +399,8 @@ cifs_reconnect(struct TCP_Server_Info *server) if (server->tcpStatus != CifsExiting) server->tcpStatus = CifsNeedNegotiate; spin_unlock(&GlobalMid_Lock); + mutex_unlock(&server->srv_mutex); } - mutex_unlock(&server->srv_mutex); } while (server->tcpStatus == CifsNeedReconnect); return rc; @@ -1132,6 +1138,12 @@ cifs_parse_smb_version(char *value, struct smb_vol *vol) vol->ops = &smb30_operations; /* currently identical with 3.0 */ vol->vals = &smb302_values; break; +#ifdef CONFIG_CIFS_SMB311 + case Smb_311: + vol->ops = &smb311_operations; + vol->vals = &smb311_values; + break; +#endif /* SMB311 */ #endif default: cifs_dbg(VFS, "Unknown vers= option specified: %s\n", value); @@ -3460,6 +3472,8 @@ try_mount_again: else if (ses) cifs_put_smb_ses(ses); + cifs_sb->mnt_cifs_flags &= ~CIFS_MOUNT_POSIX_PATHS; + free_xid(xid); } #endif diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index 338d569..c3eb998 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -620,8 +620,7 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, umode_t mode, } rc = CIFSSMBUnixSetPathInfo(xid, tcon, full_path, &args, cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + cifs_remap(cifs_sb)); if (rc) goto mknod_out; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index cafbf10..3f50cee 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -140,8 +140,7 @@ int cifs_posix_open(char *full_path, struct inode **pinode, posix_flags = cifs_posix_convert_flags(f_flags); rc = CIFSPOSIXCreate(xid, tcon, posix_flags, mode, pnetfid, presp_data, poplock, full_path, cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + cifs_remap(cifs_sb)); cifs_put_tlink(tlink); if (rc) @@ -1553,8 +1552,8 @@ cifs_setlk(struct file *file, struct file_lock *flock, __u32 type, rc = server->ops->mand_unlock_range(cfile, flock, xid); out: - if (flock->fl_flags & FL_POSIX) - posix_lock_file_wait(file, flock); + if (flock->fl_flags & FL_POSIX && !rc) + rc = posix_lock_file_wait(file, flock); return rc; } diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 55b5811..f621b44 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -373,8 +373,7 @@ int cifs_get_inode_info_unix(struct inode **pinode, /* could have done a find first instead but this returns more info */ rc = CIFSSMBUnixQPathInfo(xid, tcon, full_path, &find_data, - cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + cifs_sb->local_nls, cifs_remap(cifs_sb)); cifs_put_tlink(tlink); if (!rc) { @@ -402,9 +401,25 @@ int cifs_get_inode_info_unix(struct inode **pinode, rc = -ENOMEM; } else { /* we already have inode, update it */ + + /* if uniqueid is different, return error */ + if (unlikely(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM && + CIFS_I(*pinode)->uniqueid != fattr.cf_uniqueid)) { + rc = -ESTALE; + goto cgiiu_exit; + } + + /* if filetype is different, return error */ + if (unlikely(((*pinode)->i_mode & S_IFMT) != + (fattr.cf_mode & S_IFMT))) { + rc = -ESTALE; + goto cgiiu_exit; + } + cifs_fattr_to_inode(*pinode, &fattr); } +cgiiu_exit: return rc; } @@ -839,6 +854,15 @@ cifs_get_inode_info(struct inode **inode, const char *full_path, if (!*inode) rc = -ENOMEM; } else { + /* we already have inode, update it */ + + /* if filetype is different, return error */ + if (unlikely(((*inode)->i_mode & S_IFMT) != + (fattr.cf_mode & S_IFMT))) { + rc = -ESTALE; + goto cgii_exit; + } + cifs_fattr_to_inode(*inode, &fattr); } @@ -2215,8 +2239,7 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs) pTcon = tlink_tcon(tlink); rc = CIFSSMBUnixSetPathInfo(xid, pTcon, full_path, args, cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & - CIFS_MOUNT_MAP_SPECIAL_CHR); + cifs_remap(cifs_sb)); cifs_put_tlink(tlink); } diff --git a/fs/cifs/ioctl.c b/fs/cifs/ioctl.c index 8b7898b..49b8b6e 100644 --- a/fs/cifs/ioctl.c +++ b/fs/cifs/ioctl.c @@ -31,12 +31,15 @@ #include "cifsproto.h" #include "cifs_debug.h" #include "cifsfs.h" +#include <linux/btrfs.h> #define CIFS_IOCTL_MAGIC 0xCF #define CIFS_IOC_COPYCHUNK_FILE _IOW(CIFS_IOCTL_MAGIC, 3, int) +#define CIFS_IOC_SET_INTEGRITY _IO(CIFS_IOCTL_MAGIC, 4) static long cifs_ioctl_clone(unsigned int xid, struct file *dst_file, - unsigned long srcfd, u64 off, u64 len, u64 destoff) + unsigned long srcfd, u64 off, u64 len, u64 destoff, + bool dup_extents) { int rc; struct cifsFileInfo *smb_file_target = dst_file->private_data; @@ -109,9 +112,14 @@ static long cifs_ioctl_clone(unsigned int xid, struct file *dst_file, truncate_inode_pages_range(&target_inode->i_data, destoff, PAGE_CACHE_ALIGN(destoff + len)-1); - if (target_tcon->ses->server->ops->clone_range) + if (dup_extents && target_tcon->ses->server->ops->duplicate_extents) + rc = target_tcon->ses->server->ops->duplicate_extents(xid, + smb_file_src, smb_file_target, off, len, destoff); + else if (!dup_extents && target_tcon->ses->server->ops->clone_range) rc = target_tcon->ses->server->ops->clone_range(xid, smb_file_src, smb_file_target, off, len, destoff); + else + rc = -EOPNOTSUPP; /* force revalidate of size and timestamps of target file now that target is updated on the server */ @@ -205,7 +213,20 @@ long cifs_ioctl(struct file *filep, unsigned int command, unsigned long arg) } break; case CIFS_IOC_COPYCHUNK_FILE: - rc = cifs_ioctl_clone(xid, filep, arg, 0, 0, 0); + rc = cifs_ioctl_clone(xid, filep, arg, 0, 0, 0, false); + break; + case BTRFS_IOC_CLONE: + rc = cifs_ioctl_clone(xid, filep, arg, 0, 0, 0, true); + break; + case CIFS_IOC_SET_INTEGRITY: + if (pSMBFile == NULL) + break; + tcon = tlink_tcon(pSMBFile->tlink); + if (tcon->ses->server->ops->set_integrity) + rc = tcon->ses->server->ops->set_integrity(xid, + tcon, pSMBFile); + else + rc = -EOPNOTSUPP; break; default: cifs_dbg(FYI, "unsupported ioctl\n"); diff --git a/fs/cifs/link.c b/fs/cifs/link.c index 546f86a..e3548f7 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -715,7 +715,8 @@ cifs_symlink(struct inode *inode, struct dentry *direntry, const char *symname) rc = create_mf_symlink(xid, pTcon, cifs_sb, full_path, symname); else if (pTcon->unix_ext) rc = CIFSUnixCreateSymLink(xid, pTcon, full_path, symname, - cifs_sb->local_nls); + cifs_sb->local_nls, + cifs_remap(cifs_sb)); /* else rc = CIFSCreateReparseSymLink(xid, pTcon, fromName, toName, cifs_sb_target->local_nls); */ diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index b4a4723..b1eede3 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -90,6 +90,8 @@ cifs_prime_dcache(struct dentry *parent, struct qstr *name, if (dentry) { inode = d_inode(dentry); if (inode) { + if (d_mountpoint(dentry)) + goto out; /* * If we're generating inode numbers, then we don't * want to clobber the existing one with the one that diff --git a/fs/cifs/smb1ops.c b/fs/cifs/smb1ops.c index 7bfdd60..fc537c2 100644 --- a/fs/cifs/smb1ops.c +++ b/fs/cifs/smb1ops.c @@ -960,7 +960,8 @@ cifs_query_symlink(const unsigned int xid, struct cifs_tcon *tcon, /* Check for unix extensions */ if (cap_unix(tcon->ses)) { rc = CIFSSMBUnixQuerySymLink(xid, tcon, full_path, target_path, - cifs_sb->local_nls); + cifs_sb->local_nls, + cifs_remap(cifs_sb)); if (rc == -EREMOTE) rc = cifs_unix_dfs_readlink(xid, tcon, full_path, target_path, diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c index 54daee5..df91bcf 100644 --- a/fs/cifs/smb2ops.c +++ b/fs/cifs/smb2ops.c @@ -806,6 +806,53 @@ smb2_set_file_size(const unsigned int xid, struct cifs_tcon *tcon, cfile->fid.volatile_fid, cfile->pid, &eof, false); } +#ifdef CONFIG_CIFS_SMB311 +static int +smb2_duplicate_extents(const unsigned int xid, + struct cifsFileInfo *srcfile, + struct cifsFileInfo *trgtfile, u64 src_off, + u64 len, u64 dest_off) +{ + int rc; + unsigned int ret_data_len; + char *retbuf = NULL; + struct duplicate_extents_to_file dup_ext_buf; + struct cifs_tcon *tcon = tlink_tcon(trgtfile->tlink); + + /* server fileays advertise duplicate extent support with this flag */ + if ((le32_to_cpu(tcon->fsAttrInfo.Attributes) & + FILE_SUPPORTS_BLOCK_REFCOUNTING) == 0) + return -EOPNOTSUPP; + + dup_ext_buf.VolatileFileHandle = srcfile->fid.volatile_fid; + dup_ext_buf.PersistentFileHandle = srcfile->fid.persistent_fid; + dup_ext_buf.SourceFileOffset = cpu_to_le64(src_off); + dup_ext_buf.TargetFileOffset = cpu_to_le64(dest_off); + dup_ext_buf.ByteCount = cpu_to_le64(len); + cifs_dbg(FYI, "duplicate extents: src off %lld dst off %lld len %lld", + src_off, dest_off, len); + + rc = smb2_set_file_size(xid, tcon, trgtfile, dest_off + len, false); + if (rc) + goto duplicate_extents_out; + + rc = SMB2_ioctl(xid, tcon, trgtfile->fid.persistent_fid, + trgtfile->fid.volatile_fid, + FSCTL_DUPLICATE_EXTENTS_TO_FILE, + true /* is_fsctl */, (char *)&dup_ext_buf, + sizeof(struct duplicate_extents_to_file), + (char **)&retbuf, + &ret_data_len); + + if (ret_data_len > 0) + cifs_dbg(FYI, "non-zero response length in duplicate extents"); + +duplicate_extents_out: + return rc; +} +#endif /* CONFIG_CIFS_SMB311 */ + + static int smb2_set_compression(const unsigned int xid, struct cifs_tcon *tcon, struct cifsFileInfo *cfile) @@ -815,6 +862,28 @@ smb2_set_compression(const unsigned int xid, struct cifs_tcon *tcon, } static int +smb3_set_integrity(const unsigned int xid, struct cifs_tcon *tcon, + struct cifsFileInfo *cfile) +{ + struct fsctl_set_integrity_information_req integr_info; + char *retbuf = NULL; + unsigned int ret_data_len; + + integr_info.ChecksumAlgorithm = cpu_to_le16(CHECKSUM_TYPE_UNCHANGED); + integr_info.Flags = 0; + integr_info.Reserved = 0; + + return SMB2_ioctl(xid, tcon, cfile->fid.persistent_fid, + cfile->fid.volatile_fid, + FSCTL_SET_INTEGRITY_INFORMATION, + true /* is_fsctl */, (char *)&integr_info, + sizeof(struct fsctl_set_integrity_information_req), + (char **)&retbuf, + &ret_data_len); + +} + +static int smb2_query_dir_first(const unsigned int xid, struct cifs_tcon *tcon, const char *path, struct cifs_sb_info *cifs_sb, struct cifs_fid *fid, __u16 search_flags, @@ -1624,6 +1693,7 @@ struct smb_version_operations smb30_operations = { .new_lease_key = smb2_new_lease_key, .generate_signingkey = generate_smb3signingkey, .calc_signature = smb3_calc_signature, + .set_integrity = smb3_set_integrity, .is_read_op = smb21_is_read_op, .set_oplock_level = smb3_set_oplock_level, .create_lease_buf = smb3_create_lease_buf, @@ -1635,6 +1705,94 @@ struct smb_version_operations smb30_operations = { .fallocate = smb3_fallocate, }; +#ifdef CONFIG_CIFS_SMB311 +struct smb_version_operations smb311_operations = { + .compare_fids = smb2_compare_fids, + .setup_request = smb2_setup_request, + .setup_async_request = smb2_setup_async_request, + .check_receive = smb2_check_receive, + .add_credits = smb2_add_credits, + .set_credits = smb2_set_credits, + .get_credits_field = smb2_get_credits_field, + .get_credits = smb2_get_credits, + .wait_mtu_credits = smb2_wait_mtu_credits, + .get_next_mid = smb2_get_next_mid, + .read_data_offset = smb2_read_data_offset, + .read_data_length = smb2_read_data_length, + .map_error = map_smb2_to_linux_error, + .find_mid = smb2_find_mid, + .check_message = smb2_check_message, + .dump_detail = smb2_dump_detail, + .clear_stats = smb2_clear_stats, + .print_stats = smb2_print_stats, + .dump_share_caps = smb2_dump_share_caps, + .is_oplock_break = smb2_is_valid_oplock_break, + .downgrade_oplock = smb2_downgrade_oplock, + .need_neg = smb2_need_neg, + .negotiate = smb2_negotiate, + .negotiate_wsize = smb2_negotiate_wsize, + .negotiate_rsize = smb2_negotiate_rsize, + .sess_setup = SMB2_sess_setup, + .logoff = SMB2_logoff, + .tree_connect = SMB2_tcon, + .tree_disconnect = SMB2_tdis, + .qfs_tcon = smb3_qfs_tcon, + .is_path_accessible = smb2_is_path_accessible, + .can_echo = smb2_can_echo, + .echo = SMB2_echo, + .query_path_info = smb2_query_path_info, + .get_srv_inum = smb2_get_srv_inum, + .query_file_info = smb2_query_file_info, + .set_path_size = smb2_set_path_size, + .set_file_size = smb2_set_file_size, + .set_file_info = smb2_set_file_info, + .set_compression = smb2_set_compression, + .mkdir = smb2_mkdir, + .mkdir_setinfo = smb2_mkdir_setinfo, + .rmdir = smb2_rmdir, + .unlink = smb2_unlink, + .rename = smb2_rename_path, + .create_hardlink = smb2_create_hardlink, + .query_symlink = smb2_query_symlink, + .query_mf_symlink = smb3_query_mf_symlink, + .create_mf_symlink = smb3_create_mf_symlink, + .open = smb2_open_file, + .set_fid = smb2_set_fid, + .close = smb2_close_file, + .flush = smb2_flush_file, + .async_readv = smb2_async_readv, + .async_writev = smb2_async_writev, + .sync_read = smb2_sync_read, + .sync_write = smb2_sync_write, + .query_dir_first = smb2_query_dir_first, + .query_dir_next = smb2_query_dir_next, + .close_dir = smb2_close_dir, + .calc_smb_size = smb2_calc_size, + .is_status_pending = smb2_is_status_pending, + .oplock_response = smb2_oplock_response, + .queryfs = smb2_queryfs, + .mand_lock = smb2_mand_lock, + .mand_unlock_range = smb2_unlock_range, + .push_mand_locks = smb2_push_mandatory_locks, + .get_lease_key = smb2_get_lease_key, + .set_lease_key = smb2_set_lease_key, + .new_lease_key = smb2_new_lease_key, + .generate_signingkey = generate_smb3signingkey, + .calc_signature = smb3_calc_signature, + .set_integrity = smb3_set_integrity, + .is_read_op = smb21_is_read_op, + .set_oplock_level = smb3_set_oplock_level, + .create_lease_buf = smb3_create_lease_buf, + .parse_lease_buf = smb3_parse_lease_buf, + .clone_range = smb2_clone_range, + .duplicate_extents = smb2_duplicate_extents, +/* .validate_negotiate = smb3_validate_negotiate, */ /* not used in 3.11 */ + .wp_retry_size = smb2_wp_retry_size, + .dir_needs_close = smb2_dir_needs_close, + .fallocate = smb3_fallocate, +}; +#endif /* CIFS_SMB311 */ + struct smb_version_values smb20_values = { .version_string = SMB20_VERSION_STRING, .protocol_id = SMB20_PROT_ID, @@ -1714,3 +1872,25 @@ struct smb_version_values smb302_values = { .signing_required = SMB2_NEGOTIATE_SIGNING_REQUIRED, .create_lease_size = sizeof(struct create_lease_v2), }; + +#ifdef CONFIG_CIFS_SMB311 +struct smb_version_values smb311_values = { + .version_string = SMB311_VERSION_STRING, + .protocol_id = SMB311_PROT_ID, + .req_capabilities = SMB2_GLOBAL_CAP_DFS | SMB2_GLOBAL_CAP_LEASING | SMB2_GLOBAL_CAP_LARGE_MTU, + .large_lock_type = 0, + .exclusive_lock_type = SMB2_LOCKFLAG_EXCLUSIVE_LOCK, + .shared_lock_type = SMB2_LOCKFLAG_SHARED_LOCK, + .unlock_lock_type = SMB2_LOCKFLAG_UNLOCK, + .header_size = sizeof(struct smb2_hdr), + .max_header_size = MAX_SMB2_HDR_SIZE, + .read_rsp_size = sizeof(struct smb2_read_rsp) - 1, + .lock_cmd = SMB2_LOCK, + .cap_unix = 0, + .cap_nt_find = SMB2_NT_FIND, + .cap_large_files = SMB2_LARGE_FILES, + .signing_enabled = SMB2_NEGOTIATE_SIGNING_ENABLED | SMB2_NEGOTIATE_SIGNING_REQUIRED, + .signing_required = SMB2_NEGOTIATE_SIGNING_REQUIRED, + .create_lease_size = sizeof(struct create_lease_v2), +}; +#endif /* SMB311 */ diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c index 65cd7a8..b8b4f08 100644 --- a/fs/cifs/smb2pdu.c +++ b/fs/cifs/smb2pdu.c @@ -110,7 +110,7 @@ smb2_hdr_assemble(struct smb2_hdr *hdr, __le16 smb2_cmd /* command */ , /* GLOBAL_CAP_LARGE_MTU will only be set if dialect > SMB2.02 */ /* See sections 2.2.4 and 3.2.4.1.5 of MS-SMB2 */ - if ((tcon->ses) && + if ((tcon->ses) && (tcon->ses->server) && (tcon->ses->server->capabilities & SMB2_GLOBAL_CAP_LARGE_MTU)) hdr->CreditCharge = cpu_to_le16(1); /* else CreditCharge MBZ */ @@ -304,6 +304,59 @@ small_smb2_init(__le16 smb2_command, struct cifs_tcon *tcon, return rc; } +#ifdef CONFIG_CIFS_SMB311 +/* offset is sizeof smb2_negotiate_req - 4 but rounded up to 8 bytes */ +#define OFFSET_OF_NEG_CONTEXT 0x68 /* sizeof(struct smb2_negotiate_req) - 4 */ + + +#define SMB2_PREAUTH_INTEGRITY_CAPABILITIES cpu_to_le16(1) +#define SMB2_ENCRYPTION_CAPABILITIES cpu_to_le16(2) + +static void +build_preauth_ctxt(struct smb2_preauth_neg_context *pneg_ctxt) +{ + pneg_ctxt->ContextType = SMB2_PREAUTH_INTEGRITY_CAPABILITIES; + pneg_ctxt->DataLength = cpu_to_le16(38); + pneg_ctxt->HashAlgorithmCount = cpu_to_le16(1); + pneg_ctxt->SaltLength = cpu_to_le16(SMB311_SALT_SIZE); + get_random_bytes(pneg_ctxt->Salt, SMB311_SALT_SIZE); + pneg_ctxt->HashAlgorithms = SMB2_PREAUTH_INTEGRITY_SHA512; +} + +static void +build_encrypt_ctxt(struct smb2_encryption_neg_context *pneg_ctxt) +{ + pneg_ctxt->ContextType = SMB2_ENCRYPTION_CAPABILITIES; + pneg_ctxt->DataLength = cpu_to_le16(6); + pneg_ctxt->CipherCount = cpu_to_le16(2); + pneg_ctxt->Ciphers[0] = SMB2_ENCRYPTION_AES128_GCM; + pneg_ctxt->Ciphers[1] = SMB2_ENCRYPTION_AES128_CCM; +} + +static void +assemble_neg_contexts(struct smb2_negotiate_req *req) +{ + + /* +4 is to account for the RFC1001 len field */ + char *pneg_ctxt = (char *)req + OFFSET_OF_NEG_CONTEXT + 4; + + build_preauth_ctxt((struct smb2_preauth_neg_context *)pneg_ctxt); + /* Add 2 to size to round to 8 byte boundary */ + pneg_ctxt += 2 + sizeof(struct smb2_preauth_neg_context); + build_encrypt_ctxt((struct smb2_encryption_neg_context *)pneg_ctxt); + req->NegotiateContextOffset = cpu_to_le32(OFFSET_OF_NEG_CONTEXT); + req->NegotiateContextCount = cpu_to_le16(2); + inc_rfc1001_len(req, 4 + sizeof(struct smb2_preauth_neg_context) + 2 + + sizeof(struct smb2_encryption_neg_context)); /* calculate hash */ +} +#else +static void assemble_neg_contexts(struct smb2_negotiate_req *req) +{ + return; +} +#endif /* SMB311 */ + + /* * * SMB2 Worker functions follow: @@ -363,10 +416,12 @@ SMB2_negotiate(const unsigned int xid, struct cifs_ses *ses) /* ClientGUID must be zero for SMB2.02 dialect */ if (ses->server->vals->protocol_id == SMB20_PROT_ID) memset(req->ClientGUID, 0, SMB2_CLIENT_GUID_SIZE); - else + else { memcpy(req->ClientGUID, server->client_guid, SMB2_CLIENT_GUID_SIZE); - + if (ses->server->vals->protocol_id == SMB311_PROT_ID) + assemble_neg_contexts(req); + } iov[0].iov_base = (char *)req; /* 4 for rfc1002 length field */ iov[0].iov_len = get_rfc1002_length(req) + 4; @@ -393,8 +448,12 @@ SMB2_negotiate(const unsigned int xid, struct cifs_ses *ses) cifs_dbg(FYI, "negotiated smb3.0 dialect\n"); else if (rsp->DialectRevision == cpu_to_le16(SMB302_PROT_ID)) cifs_dbg(FYI, "negotiated smb3.02 dialect\n"); +#ifdef CONFIG_CIFS_SMB311 + else if (rsp->DialectRevision == cpu_to_le16(SMB311_PROT_ID)) + cifs_dbg(FYI, "negotiated smb3.1.1 dialect\n"); +#endif /* SMB311 */ else { - cifs_dbg(VFS, "Illegal dialect returned by server %d\n", + cifs_dbg(VFS, "Illegal dialect returned by server 0x%x\n", le16_to_cpu(rsp->DialectRevision)); rc = -EIO; goto neg_exit; @@ -572,7 +631,7 @@ ssetup_ntlmssp_authenticate: return rc; req->hdr.SessionId = 0; /* First session, not a reauthenticate */ - req->VcNumber = 0; /* MBZ */ + req->Flags = 0; /* MBZ */ /* to enable echos and oplocks */ req->hdr.CreditRequest = cpu_to_le16(3); diff --git a/fs/cifs/smb2pdu.h b/fs/cifs/smb2pdu.h index 70867d5..4511082 100644 --- a/fs/cifs/smb2pdu.h +++ b/fs/cifs/smb2pdu.h @@ -136,9 +136,6 @@ struct smb2_transform_hdr { __u64 SessionId; } __packed; -/* Encryption Algorithms */ -#define SMB2_ENCRYPTION_AES128_CCM cpu_to_le16(0x0001) - /* * SMB2 flag definitions */ @@ -191,7 +188,10 @@ struct smb2_negotiate_req { __le16 Reserved; /* MBZ */ __le32 Capabilities; __u8 ClientGUID[SMB2_CLIENT_GUID_SIZE]; - __le64 ClientStartTime; /* MBZ */ + /* In SMB3.02 and earlier next three were MBZ le64 ClientStartTime */ + __le32 NegotiateContextOffset; /* SMB3.1.1 only. MBZ earlier */ + __le16 NegotiateContextCount; /* SMB3.1.1 only. MBZ earlier */ + __le16 Reserved2; __le16 Dialects[1]; /* One dialect (vers=) at a time for now */ } __packed; @@ -200,6 +200,7 @@ struct smb2_negotiate_req { #define SMB21_PROT_ID 0x0210 #define SMB30_PROT_ID 0x0300 #define SMB302_PROT_ID 0x0302 +#define SMB311_PROT_ID 0x0311 #define BAD_PROT_ID 0xFFFF /* SecurityMode flags */ @@ -217,12 +218,38 @@ struct smb2_negotiate_req { #define SMB2_NT_FIND 0x00100000 #define SMB2_LARGE_FILES 0x00200000 +#define SMB311_SALT_SIZE 32 +/* Hash Algorithm Types */ +#define SMB2_PREAUTH_INTEGRITY_SHA512 cpu_to_le16(0x0001) + +struct smb2_preauth_neg_context { + __le16 ContextType; /* 1 */ + __le16 DataLength; + __le32 Reserved; + __le16 HashAlgorithmCount; /* 1 */ + __le16 SaltLength; + __le16 HashAlgorithms; /* HashAlgorithms[0] since only one defined */ + __u8 Salt[SMB311_SALT_SIZE]; +} __packed; + +/* Encryption Algorithms Ciphers */ +#define SMB2_ENCRYPTION_AES128_CCM cpu_to_le16(0x0001) +#define SMB2_ENCRYPTION_AES128_GCM cpu_to_le16(0x0002) + +struct smb2_encryption_neg_context { + __le16 ContextType; /* 2 */ + __le16 DataLength; + __le32 Reserved; + __le16 CipherCount; /* AES-128-GCM and AES-128-CCM */ + __le16 Ciphers[2]; /* Ciphers[0] since only one used now */ +} __packed; + struct smb2_negotiate_rsp { struct smb2_hdr hdr; __le16 StructureSize; /* Must be 65 */ __le16 SecurityMode; __le16 DialectRevision; - __le16 Reserved; /* MBZ */ + __le16 NegotiateContextCount; /* Prior to SMB3.1.1 was Reserved & MBZ */ __u8 ServerGUID[16]; __le32 Capabilities; __le32 MaxTransactSize; @@ -232,14 +259,18 @@ struct smb2_negotiate_rsp { __le64 ServerStartTime; __le16 SecurityBufferOffset; __le16 SecurityBufferLength; - __le32 Reserved2; /* may be any value, ignore */ + __le32 NegotiateContextOffset; /* Pre:SMB3.1.1 was reserved/ignored */ __u8 Buffer[1]; /* variable length GSS security buffer */ } __packed; +/* Flags */ +#define SMB2_SESSION_REQ_FLAG_BINDING 0x01 +#define SMB2_SESSION_REQ_FLAG_ENCRYPT_DATA 0x04 + struct smb2_sess_setup_req { struct smb2_hdr hdr; __le16 StructureSize; /* Must be 25 */ - __u8 VcNumber; + __u8 Flags; __u8 SecurityMode; __le32 Capabilities; __le32 Channel; @@ -274,10 +305,13 @@ struct smb2_logoff_rsp { __le16 Reserved; } __packed; +/* Flags/Reserved for SMB3.1.1 */ +#define SMB2_SHAREFLAG_CLUSTER_RECONNECT 0x0001 + struct smb2_tree_connect_req { struct smb2_hdr hdr; __le16 StructureSize; /* Must be 9 */ - __le16 Reserved; + __le16 Reserved; /* Flags in SMB3.1.1 */ __le16 PathOffset; __le16 PathLength; __u8 Buffer[1]; /* variable length */ @@ -587,6 +621,29 @@ struct copychunk_ioctl_rsp { __le32 TotalBytesWritten; } __packed; +struct fsctl_set_integrity_information_req { + __le16 ChecksumAlgorithm; + __le16 Reserved; + __le32 Flags; +} __packed; + +struct fsctl_get_integrity_information_rsp { + __le16 ChecksumAlgorithm; + __le16 Reserved; + __le32 Flags; + __le32 ChecksumChunkSizeInBytes; + __le32 ClusterSizeInBytes; +} __packed; + +/* Integrity ChecksumAlgorithm choices for above */ +#define CHECKSUM_TYPE_NONE 0x0000 +#define CHECKSUM_TYPE_CRC64 0x0002 +#define CHECKSUM_TYPE_UNCHANGED 0xFFFF /* set only */ + +/* Integrity flags for above */ +#define FSCTL_INTEGRITY_FLAG_CHECKSUM_ENFORCEMENT_OFF 0x00000001 + + struct validate_negotiate_info_req { __le32 Capabilities; __u8 Guid[SMB2_CLIENT_GUID_SIZE]; @@ -620,6 +677,14 @@ struct compress_ioctl { __le16 CompressionState; /* See cifspdu.h for possible flag values */ } __packed; +struct duplicate_extents_to_file { + __u64 PersistentFileHandle; /* source file handle, opaque endianness */ + __u64 VolatileFileHandle; + __le64 SourceFileOffset; + __le64 TargetFileOffset; + __le64 ByteCount; /* Bytes to be copied */ +} __packed; + struct smb2_ioctl_req { struct smb2_hdr hdr; __le16 StructureSize; /* Must be 57 */ diff --git a/fs/cifs/smbfsctl.h b/fs/cifs/smbfsctl.h index 83efa59..a639d0d 100644 --- a/fs/cifs/smbfsctl.h +++ b/fs/cifs/smbfsctl.h @@ -75,10 +75,13 @@ #define FSCTL_QUERY_SPARING_INFO 0x00090138 /* BB add struct */ #define FSCTL_SET_ZERO_ON_DEALLOC 0x00090194 /* BB add struct */ #define FSCTL_SET_SHORT_NAME_BEHAVIOR 0x000901B4 /* BB add struct */ +#define FSCTL_GET_INTEGRITY_INFORMATION 0x0009027C #define FSCTL_QUERY_ALLOCATED_RANGES 0x000940CF /* BB add struct */ #define FSCTL_SET_DEFECT_MANAGEMENT 0x00098134 /* BB add struct */ #define FSCTL_FILE_LEVEL_TRIM 0x00098208 /* BB add struct */ +#define FSCTL_DUPLICATE_EXTENTS_TO_FILE 0x00098344 #define FSCTL_SIS_LINK_FILES 0x0009C104 +#define FSCTL_SET_INTEGRITY_INFORMATION 0x0009C280 #define FSCTL_PIPE_PEEK 0x0011400C /* BB add struct */ #define FSCTL_PIPE_TRANSCEIVE 0x0011C017 /* BB add struct */ /* strange that the number for this op is not sequential with previous op */ diff --git a/fs/configfs/item.c b/fs/configfs/item.c index e65f9ff..4d6a30e 100644 --- a/fs/configfs/item.c +++ b/fs/configfs/item.c @@ -47,12 +47,11 @@ static void config_item_release(struct kref *kref); * config_item_init - initialize item. * @item: item in question. */ -void config_item_init(struct config_item *item) +static void config_item_init(struct config_item *item) { kref_init(&item->ci_kref); INIT_LIST_HEAD(&item->ci_entry); } -EXPORT_SYMBOL(config_item_init); /** * config_item_set_name - Set the name of an item diff --git a/fs/configfs/mount.c b/fs/configfs/mount.c index 5373567..a8f3b58 100644 --- a/fs/configfs/mount.c +++ b/fs/configfs/mount.c @@ -129,8 +129,6 @@ void configfs_release_fs(void) } -static struct kobject *config_kobj; - static int __init configfs_init(void) { int err = -ENOMEM; @@ -141,8 +139,8 @@ static int __init configfs_init(void) if (!configfs_dir_cachep) goto out; - config_kobj = kobject_create_and_add("config", kernel_kobj); - if (!config_kobj) + err = sysfs_create_mount_point(kernel_kobj, "config"); + if (err) goto out2; err = register_filesystem(&configfs_fs_type); @@ -152,7 +150,7 @@ static int __init configfs_init(void) return 0; out3: pr_err("Unable to register filesystem!\n"); - kobject_put(config_kobj); + sysfs_remove_mount_point(kernel_kobj, "config"); out2: kmem_cache_destroy(configfs_dir_cachep); configfs_dir_cachep = NULL; @@ -163,7 +161,7 @@ out: static void __exit configfs_exit(void) { unregister_filesystem(&configfs_fs_type); - kobject_put(config_kobj); + sysfs_remove_mount_point(kernel_kobj, "config"); kmem_cache_destroy(configfs_dir_cachep); configfs_dir_cachep = NULL; } diff --git a/fs/coredump.c b/fs/coredump.c index 5b771b3..c5ecde6 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -70,7 +70,8 @@ static int expand_corename(struct core_name *cn, int size) return 0; } -static int cn_vprintf(struct core_name *cn, const char *fmt, va_list arg) +static __printf(2, 0) int cn_vprintf(struct core_name *cn, const char *fmt, + va_list arg) { int free, need; va_list arg_copy; @@ -93,7 +94,7 @@ again: return -ENOMEM; } -static int cn_printf(struct core_name *cn, const char *fmt, ...) +static __printf(2, 3) int cn_printf(struct core_name *cn, const char *fmt, ...) { va_list arg; int ret; @@ -105,7 +106,8 @@ static int cn_printf(struct core_name *cn, const char *fmt, ...) return ret; } -static int cn_esc_printf(struct core_name *cn, const char *fmt, ...) +static __printf(2, 3) +int cn_esc_printf(struct core_name *cn, const char *fmt, ...) { int cur = cn->used; va_list arg; @@ -209,11 +211,15 @@ static int format_corename(struct core_name *cn, struct coredump_params *cprm) break; /* uid */ case 'u': - err = cn_printf(cn, "%d", cred->uid); + err = cn_printf(cn, "%u", + from_kuid(&init_user_ns, + cred->uid)); break; /* gid */ case 'g': - err = cn_printf(cn, "%d", cred->gid); + err = cn_printf(cn, "%u", + from_kgid(&init_user_ns, + cred->gid)); break; case 'd': err = cn_printf(cn, "%d", @@ -221,7 +227,8 @@ static int format_corename(struct core_name *cn, struct coredump_params *cprm) break; /* signal that caused the coredump */ case 's': - err = cn_printf(cn, "%ld", cprm->siginfo->si_signo); + err = cn_printf(cn, "%d", + cprm->siginfo->si_signo); break; /* UNIX time of coredump */ case 't': { @@ -311,14 +311,21 @@ static int dax_insert_mapping(struct inode *inode, struct buffer_head *bh, out: i_mmap_unlock_read(mapping); - if (bh->b_end_io) - bh->b_end_io(bh, 1); - return error; } -static int do_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf, - get_block_t get_block) +/** + * __dax_fault - handle a page fault on a DAX file + * @vma: The virtual memory area where the fault occurred + * @vmf: The description of the fault + * @get_block: The filesystem method used to translate file offsets to blocks + * + * When a page fault occurs, filesystems may call this helper in their + * fault handler for DAX files. __dax_fault() assumes the caller has done all + * the necessary locking for the page fault to proceed successfully. + */ +int __dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf, + get_block_t get_block, dax_iodone_t complete_unwritten) { struct file *file = vma->vm_file; struct address_space *mapping = file->f_mapping; @@ -419,7 +426,19 @@ static int do_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf, page_cache_release(page); } + /* + * If we successfully insert the new mapping over an unwritten extent, + * we need to ensure we convert the unwritten extent. If there is an + * error inserting the mapping, the filesystem needs to leave it as + * unwritten to prevent exposure of the stale underlying data to + * userspace, but we still need to call the completion function so + * the private resources on the mapping buffer can be released. We + * indicate what the callback should do via the uptodate variable, same + * as for normal BH based IO completions. + */ error = dax_insert_mapping(inode, &bh, vma, vmf); + if (buffer_unwritten(&bh)) + complete_unwritten(&bh, !error); out: if (error == -ENOMEM) @@ -436,6 +455,7 @@ static int do_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf, } goto out; } +EXPORT_SYMBOL(__dax_fault); /** * dax_fault - handle a page fault on a DAX file @@ -447,7 +467,7 @@ static int do_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf, * fault handler for DAX files. */ int dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf, - get_block_t get_block) + get_block_t get_block, dax_iodone_t complete_unwritten) { int result; struct super_block *sb = file_inode(vma->vm_file)->i_sb; @@ -456,7 +476,7 @@ int dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf, sb_start_pagefault(sb); file_update_time(vma->vm_file); } - result = do_dax_fault(vma, vmf, get_block); + result = __dax_fault(vma, vmf, get_block, complete_unwritten); if (vmf->flags & FAULT_FLAG_WRITE) sb_end_pagefault(sb); diff --git a/fs/dcache.c b/fs/dcache.c index c4ce351..7a3f3e5 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -322,17 +322,17 @@ static void dentry_free(struct dentry *dentry) } /** - * dentry_rcuwalk_barrier - invalidate in-progress rcu-walk lookups + * dentry_rcuwalk_invalidate - invalidate in-progress rcu-walk lookups * @dentry: the target dentry * After this call, in-progress rcu-walk path lookup will fail. This * should be called after unhashing, and after changing d_inode (if * the dentry has not already been unhashed). */ -static inline void dentry_rcuwalk_barrier(struct dentry *dentry) +static inline void dentry_rcuwalk_invalidate(struct dentry *dentry) { - assert_spin_locked(&dentry->d_lock); - /* Go through a barrier */ - write_seqcount_barrier(&dentry->d_seq); + lockdep_assert_held(&dentry->d_lock); + /* Go through am invalidation barrier */ + write_seqcount_invalidate(&dentry->d_seq); } /* @@ -372,7 +372,7 @@ static void dentry_unlink_inode(struct dentry * dentry) struct inode *inode = dentry->d_inode; __d_clear_type_and_inode(dentry); hlist_del_init(&dentry->d_u.d_alias); - dentry_rcuwalk_barrier(dentry); + dentry_rcuwalk_invalidate(dentry); spin_unlock(&dentry->d_lock); spin_unlock(&inode->i_lock); if (!inode->i_nlink) @@ -494,7 +494,7 @@ void __d_drop(struct dentry *dentry) __hlist_bl_del(&dentry->d_hash); dentry->d_hash.pprev = NULL; hlist_bl_unlock(b); - dentry_rcuwalk_barrier(dentry); + dentry_rcuwalk_invalidate(dentry); } } EXPORT_SYMBOL(__d_drop); @@ -1755,7 +1755,7 @@ static void __d_instantiate(struct dentry *dentry, struct inode *inode) if (inode) hlist_add_head(&dentry->d_u.d_alias, &inode->i_dentry); __d_set_inode_and_type(dentry, inode, add_flags); - dentry_rcuwalk_barrier(dentry); + dentry_rcuwalk_invalidate(dentry); spin_unlock(&dentry->d_lock); fsnotify_d_instantiate(dentry, inode); } @@ -2930,17 +2930,6 @@ restart: vfsmnt = &mnt->mnt; continue; } - /* - * Filesystems needing to implement special "root names" - * should do so with ->d_dname() - */ - if (IS_ROOT(dentry) && - (dentry->d_name.len != 1 || - dentry->d_name.name[0] != '/')) { - WARN(1, "Root dentry has weird name <%.*s>\n", - (int) dentry->d_name.len, - dentry->d_name.name); - } if (!error) error = is_mounted(vfsmnt) ? 1 : 2; break; diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index ef86ad6..c711be8 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -711,20 +711,17 @@ bool debugfs_initialized(void) } EXPORT_SYMBOL_GPL(debugfs_initialized); - -static struct kobject *debug_kobj; - static int __init debugfs_init(void) { int retval; - debug_kobj = kobject_create_and_add("debug", kernel_kobj); - if (!debug_kobj) - return -EINVAL; + retval = sysfs_create_mount_point(kernel_kobj, "debug"); + if (retval) + return retval; retval = register_filesystem(&debug_fs_type); if (retval) - kobject_put(debug_kobj); + sysfs_remove_mount_point(kernel_kobj, "debug"); else debugfs_registered = true; diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index add5663..c35ffdc 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -142,6 +142,8 @@ static inline struct super_block *pts_sb_from_inode(struct inode *inode) if (inode->i_sb->s_magic == DEVPTS_SUPER_MAGIC) return inode->i_sb; #endif + if (!devpts_mnt) + return NULL; return devpts_mnt->mnt_sb; } @@ -525,10 +527,14 @@ static struct file_system_type devpts_fs_type = { int devpts_new_index(struct inode *ptmx_inode) { struct super_block *sb = pts_sb_from_inode(ptmx_inode); - struct pts_fs_info *fsi = DEVPTS_SB(sb); + struct pts_fs_info *fsi; int index; int ida_ret; + if (!sb) + return -ENODEV; + + fsi = DEVPTS_SB(sb); retry: if (!ida_pre_get(&fsi->allocated_ptys, GFP_KERNEL)) return -ENOMEM; @@ -584,11 +590,18 @@ struct inode *devpts_pty_new(struct inode *ptmx_inode, dev_t device, int index, struct dentry *dentry; struct super_block *sb = pts_sb_from_inode(ptmx_inode); struct inode *inode; - struct dentry *root = sb->s_root; - struct pts_fs_info *fsi = DEVPTS_SB(sb); - struct pts_mount_opts *opts = &fsi->mount_opts; + struct dentry *root; + struct pts_fs_info *fsi; + struct pts_mount_opts *opts; char s[12]; + if (!sb) + return ERR_PTR(-ENODEV); + + root = sb->s_root; + fsi = DEVPTS_SB(sb); + opts = &fsi->mount_opts; + inode = new_inode(sb); if (!inode) return ERR_PTR(-ENOMEM); @@ -676,12 +689,16 @@ static int __init init_devpts_fs(void) struct ctl_table_header *table; if (!err) { + struct vfsmount *mnt; + table = register_sysctl_table(pty_root_table); - devpts_mnt = kern_mount(&devpts_fs_type); - if (IS_ERR(devpts_mnt)) { - err = PTR_ERR(devpts_mnt); + mnt = kern_mount(&devpts_fs_type); + if (IS_ERR(mnt)) { + err = PTR_ERR(mnt); unregister_filesystem(&devpts_fs_type); unregister_sysctl_table(table); + } else { + devpts_mnt = mnt; } } return err; diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index d08e079..754fd6c 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -921,8 +921,8 @@ static int tcp_accept_from_sock(struct connection *con) mutex_unlock(&connections_lock); memset(&peeraddr, 0, sizeof(peeraddr)); - result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM, - IPPROTO_TCP, &newsock); + result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family, + SOCK_STREAM, IPPROTO_TCP, &newsock); if (result < 0) return -ENOMEM; @@ -1173,8 +1173,8 @@ static void tcp_connect_to_sock(struct connection *con) goto out; /* Create a socket to communicate with */ - result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM, - IPPROTO_TCP, &sock); + result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family, + SOCK_STREAM, IPPROTO_TCP, &sock); if (result < 0) goto out_err; @@ -1258,8 +1258,8 @@ static struct socket *tcp_create_listen_sock(struct connection *con, addr_len = sizeof(struct sockaddr_in6); /* Create a socket to communicate with */ - result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_STREAM, - IPPROTO_TCP, &sock); + result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family, + SOCK_STREAM, IPPROTO_TCP, &sock); if (result < 0) { log_print("Can't create listening comms socket"); goto create_out; @@ -1365,8 +1365,8 @@ static int sctp_listen_for_all(void) log_print("Using SCTP for communications"); - result = sock_create_kern(dlm_local_addr[0]->ss_family, SOCK_SEQPACKET, - IPPROTO_SCTP, &sock); + result = sock_create_kern(&init_net, dlm_local_addr[0]->ss_family, + SOCK_SEQPACKET, IPPROTO_SCTP, &sock); if (result < 0) { log_print("Can't create comms socket, check SCTP is loaded"); goto out; diff --git a/fs/efs/super.c b/fs/efs/super.c index 7fca462..c8411a3 100644 --- a/fs/efs/super.c +++ b/fs/efs/super.c @@ -67,7 +67,7 @@ static struct kmem_cache * efs_inode_cachep; static struct inode *efs_alloc_inode(struct super_block *sb) { struct efs_inode_info *ei; - ei = (struct efs_inode_info *)kmem_cache_alloc(efs_inode_cachep, GFP_KERNEL); + ei = kmem_cache_alloc(efs_inode_cachep, GFP_KERNEL); if (!ei) return NULL; return &ei->vfs_inode; @@ -659,6 +659,9 @@ int setup_arg_pages(struct linux_binprm *bprm, if (stack_base > STACK_SIZE_MAX) stack_base = STACK_SIZE_MAX; + /* Add space for stack randomization. */ + stack_base += (STACK_RND_MASK << PAGE_SHIFT); + /* Make sure we didn't let the argument array grow too large. */ if (vma->vm_end - vma->vm_start > stack_base) return -ENOMEM; diff --git a/fs/ext2/file.c b/fs/ext2/file.c index 3a0a6c6..3b57c9f 100644 --- a/fs/ext2/file.c +++ b/fs/ext2/file.c @@ -28,12 +28,12 @@ #ifdef CONFIG_FS_DAX static int ext2_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf) { - return dax_fault(vma, vmf, ext2_get_block); + return dax_fault(vma, vmf, ext2_get_block, NULL); } static int ext2_dax_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { - return dax_mkwrite(vma, vmf, ext2_get_block); + return dax_mkwrite(vma, vmf, ext2_get_block, NULL); } static const struct vm_operations_struct ext2_dax_vm_ops = { diff --git a/fs/ext2/super.c b/fs/ext2/super.c index d0e746e..900e19c 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -882,6 +882,7 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) sb->s_flags = (sb->s_flags & ~MS_POSIXACL) | ((EXT2_SB(sb)->s_mount_opt & EXT2_MOUNT_POSIX_ACL) ? MS_POSIXACL : 0); + sb->s_iflags |= SB_I_CGROUPWB; if (le32_to_cpu(es->s_rev_level) == EXT2_GOOD_OLD_REV && (EXT2_HAS_COMPAT_FEATURE(sb, ~0U) || diff --git a/fs/ext3/super.c b/fs/ext3/super.c index a9312f0..5ed0044 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -1908,7 +1908,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) sbi->s_mount_state = le16_to_cpu(es->s_state); sbi->s_addr_per_block_bits = ilog2(EXT3_ADDR_PER_BLOCK(sb)); sbi->s_desc_per_block_bits = ilog2(EXT3_DESC_PER_BLOCK(sb)); - for (i=0; i < 4; i++) + for (i = 0; i < 4; i++) sbi->s_hash_seed[i] = le32_to_cpu(es->s_hash_seed[i]); sbi->s_def_hash_version = es->s_def_hash_version; i = le32_to_cpu(es->s_flags); diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig index 024f228..bf8bc8a 100644 --- a/fs/ext4/Kconfig +++ b/fs/ext4/Kconfig @@ -72,6 +72,7 @@ config EXT4_ENCRYPTION select CRYPTO_ECB select CRYPTO_XTS select CRYPTO_CTS + select CRYPTO_CTR select CRYPTO_SHA256 select KEYS select ENCRYPTED_KEYS diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 955bf49..cd6ea29 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -369,7 +369,7 @@ static void ext4_validate_block_bitmap(struct super_block *sb, struct ext4_group_info *grp = ext4_get_group_info(sb, block_group); struct ext4_sb_info *sbi = EXT4_SB(sb); - if (buffer_verified(bh)) + if (buffer_verified(bh) || EXT4_MB_GRP_BBITMAP_CORRUPT(grp)) return; ext4_lock_group(sb, block_group); @@ -446,7 +446,7 @@ ext4_read_block_bitmap_nowait(struct super_block *sb, ext4_group_t block_group) unlock_buffer(bh); if (err) ext4_error(sb, "Checksum bad for grp %u", block_group); - return bh; + goto verify; } ext4_unlock_group(sb, block_group); if (buffer_uptodate(bh)) { diff --git a/fs/ext4/crypto.c b/fs/ext4/crypto.c index 8ff1527..4573155 100644 --- a/fs/ext4/crypto.c +++ b/fs/ext4/crypto.c @@ -55,6 +55,9 @@ static mempool_t *ext4_bounce_page_pool; static LIST_HEAD(ext4_free_crypto_ctxs); static DEFINE_SPINLOCK(ext4_crypto_ctx_lock); +static struct kmem_cache *ext4_crypto_ctx_cachep; +struct kmem_cache *ext4_crypt_info_cachep; + /** * ext4_release_crypto_ctx() - Releases an encryption context * @ctx: The encryption context to release. @@ -68,18 +71,12 @@ void ext4_release_crypto_ctx(struct ext4_crypto_ctx *ctx) { unsigned long flags; - if (ctx->bounce_page) { - if (ctx->flags & EXT4_BOUNCE_PAGE_REQUIRES_FREE_ENCRYPT_FL) - __free_page(ctx->bounce_page); - else - mempool_free(ctx->bounce_page, ext4_bounce_page_pool); - ctx->bounce_page = NULL; - } - ctx->control_page = NULL; + if (ctx->flags & EXT4_WRITE_PATH_FL && ctx->w.bounce_page) + mempool_free(ctx->w.bounce_page, ext4_bounce_page_pool); + ctx->w.bounce_page = NULL; + ctx->w.control_page = NULL; if (ctx->flags & EXT4_CTX_REQUIRES_FREE_ENCRYPT_FL) { - if (ctx->tfm) - crypto_free_tfm(ctx->tfm); - kfree(ctx); + kmem_cache_free(ext4_crypto_ctx_cachep, ctx); } else { spin_lock_irqsave(&ext4_crypto_ctx_lock, flags); list_add(&ctx->free_list, &ext4_free_crypto_ctxs); @@ -88,23 +85,6 @@ void ext4_release_crypto_ctx(struct ext4_crypto_ctx *ctx) } /** - * ext4_alloc_and_init_crypto_ctx() - Allocates and inits an encryption context - * @mask: The allocation mask. - * - * Return: An allocated and initialized encryption context on success. An error - * value or NULL otherwise. - */ -static struct ext4_crypto_ctx *ext4_alloc_and_init_crypto_ctx(gfp_t mask) -{ - struct ext4_crypto_ctx *ctx = kzalloc(sizeof(struct ext4_crypto_ctx), - mask); - - if (!ctx) - return ERR_PTR(-ENOMEM); - return ctx; -} - -/** * ext4_get_crypto_ctx() - Gets an encryption context * @inode: The inode for which we are doing the crypto * @@ -118,10 +98,10 @@ struct ext4_crypto_ctx *ext4_get_crypto_ctx(struct inode *inode) struct ext4_crypto_ctx *ctx = NULL; int res = 0; unsigned long flags; - struct ext4_encryption_key *key = &EXT4_I(inode)->i_encryption_key; + struct ext4_crypt_info *ci = EXT4_I(inode)->i_crypt_info; - if (!ext4_read_workqueue) - ext4_init_crypto(); + if (ci == NULL) + return ERR_PTR(-ENOKEY); /* * We first try getting the ctx from a free list because in @@ -140,50 +120,16 @@ struct ext4_crypto_ctx *ext4_get_crypto_ctx(struct inode *inode) list_del(&ctx->free_list); spin_unlock_irqrestore(&ext4_crypto_ctx_lock, flags); if (!ctx) { - ctx = ext4_alloc_and_init_crypto_ctx(GFP_NOFS); - if (IS_ERR(ctx)) { - res = PTR_ERR(ctx); + ctx = kmem_cache_zalloc(ext4_crypto_ctx_cachep, GFP_NOFS); + if (!ctx) { + res = -ENOMEM; goto out; } ctx->flags |= EXT4_CTX_REQUIRES_FREE_ENCRYPT_FL; } else { ctx->flags &= ~EXT4_CTX_REQUIRES_FREE_ENCRYPT_FL; } - - /* Allocate a new Crypto API context if we don't already have - * one or if it isn't the right mode. */ - BUG_ON(key->mode == EXT4_ENCRYPTION_MODE_INVALID); - if (ctx->tfm && (ctx->mode != key->mode)) { - crypto_free_tfm(ctx->tfm); - ctx->tfm = NULL; - ctx->mode = EXT4_ENCRYPTION_MODE_INVALID; - } - if (!ctx->tfm) { - switch (key->mode) { - case EXT4_ENCRYPTION_MODE_AES_256_XTS: - ctx->tfm = crypto_ablkcipher_tfm( - crypto_alloc_ablkcipher("xts(aes)", 0, 0)); - break; - case EXT4_ENCRYPTION_MODE_AES_256_GCM: - /* TODO(mhalcrow): AEAD w/ gcm(aes); - * crypto_aead_setauthsize() */ - ctx->tfm = ERR_PTR(-ENOTSUPP); - break; - default: - BUG(); - } - if (IS_ERR_OR_NULL(ctx->tfm)) { - res = PTR_ERR(ctx->tfm); - ctx->tfm = NULL; - goto out; - } - ctx->mode = key->mode; - } - BUG_ON(key->size != ext4_encryption_key_size(key->mode)); - - /* There shouldn't be a bounce page attached to the crypto - * context at this point. */ - BUG_ON(ctx->bounce_page); + ctx->flags &= ~EXT4_WRITE_PATH_FL; out: if (res) { @@ -204,20 +150,8 @@ void ext4_exit_crypto(void) { struct ext4_crypto_ctx *pos, *n; - list_for_each_entry_safe(pos, n, &ext4_free_crypto_ctxs, free_list) { - if (pos->bounce_page) { - if (pos->flags & - EXT4_BOUNCE_PAGE_REQUIRES_FREE_ENCRYPT_FL) { - __free_page(pos->bounce_page); - } else { - mempool_free(pos->bounce_page, - ext4_bounce_page_pool); - } - } - if (pos->tfm) - crypto_free_tfm(pos->tfm); - kfree(pos); - } + list_for_each_entry_safe(pos, n, &ext4_free_crypto_ctxs, free_list) + kmem_cache_free(ext4_crypto_ctx_cachep, pos); INIT_LIST_HEAD(&ext4_free_crypto_ctxs); if (ext4_bounce_page_pool) mempool_destroy(ext4_bounce_page_pool); @@ -225,6 +159,12 @@ void ext4_exit_crypto(void) if (ext4_read_workqueue) destroy_workqueue(ext4_read_workqueue); ext4_read_workqueue = NULL; + if (ext4_crypto_ctx_cachep) + kmem_cache_destroy(ext4_crypto_ctx_cachep); + ext4_crypto_ctx_cachep = NULL; + if (ext4_crypt_info_cachep) + kmem_cache_destroy(ext4_crypt_info_cachep); + ext4_crypt_info_cachep = NULL; } /** @@ -237,23 +177,31 @@ void ext4_exit_crypto(void) */ int ext4_init_crypto(void) { - int i, res; + int i, res = -ENOMEM; mutex_lock(&crypto_init); if (ext4_read_workqueue) goto already_initialized; ext4_read_workqueue = alloc_workqueue("ext4_crypto", WQ_HIGHPRI, 0); - if (!ext4_read_workqueue) { - res = -ENOMEM; + if (!ext4_read_workqueue) + goto fail; + + ext4_crypto_ctx_cachep = KMEM_CACHE(ext4_crypto_ctx, + SLAB_RECLAIM_ACCOUNT); + if (!ext4_crypto_ctx_cachep) + goto fail; + + ext4_crypt_info_cachep = KMEM_CACHE(ext4_crypt_info, + SLAB_RECLAIM_ACCOUNT); + if (!ext4_crypt_info_cachep) goto fail; - } for (i = 0; i < num_prealloc_crypto_ctxs; i++) { struct ext4_crypto_ctx *ctx; - ctx = ext4_alloc_and_init_crypto_ctx(GFP_KERNEL); - if (IS_ERR(ctx)) { - res = PTR_ERR(ctx); + ctx = kmem_cache_zalloc(ext4_crypto_ctx_cachep, GFP_NOFS); + if (!ctx) { + res = -ENOMEM; goto fail; } list_add(&ctx->free_list, &ext4_free_crypto_ctxs); @@ -317,32 +265,11 @@ static int ext4_page_crypto(struct ext4_crypto_ctx *ctx, struct ablkcipher_request *req = NULL; DECLARE_EXT4_COMPLETION_RESULT(ecr); struct scatterlist dst, src; - struct ext4_inode_info *ei = EXT4_I(inode); - struct crypto_ablkcipher *atfm = __crypto_ablkcipher_cast(ctx->tfm); + struct ext4_crypt_info *ci = EXT4_I(inode)->i_crypt_info; + struct crypto_ablkcipher *tfm = ci->ci_ctfm; int res = 0; - BUG_ON(!ctx->tfm); - BUG_ON(ctx->mode != ei->i_encryption_key.mode); - - if (ctx->mode != EXT4_ENCRYPTION_MODE_AES_256_XTS) { - printk_ratelimited(KERN_ERR - "%s: unsupported crypto algorithm: %d\n", - __func__, ctx->mode); - return -ENOTSUPP; - } - - crypto_ablkcipher_clear_flags(atfm, ~0); - crypto_tfm_set_flags(ctx->tfm, CRYPTO_TFM_REQ_WEAK_KEY); - - res = crypto_ablkcipher_setkey(atfm, ei->i_encryption_key.raw, - ei->i_encryption_key.size); - if (res) { - printk_ratelimited(KERN_ERR - "%s: crypto_ablkcipher_setkey() failed\n", - __func__); - return res; - } - req = ablkcipher_request_alloc(atfm, GFP_NOFS); + req = ablkcipher_request_alloc(tfm, GFP_NOFS); if (!req) { printk_ratelimited(KERN_ERR "%s: crypto_request_alloc() failed\n", @@ -384,6 +311,15 @@ static int ext4_page_crypto(struct ext4_crypto_ctx *ctx, return 0; } +static struct page *alloc_bounce_page(struct ext4_crypto_ctx *ctx) +{ + ctx->w.bounce_page = mempool_alloc(ext4_bounce_page_pool, GFP_NOWAIT); + if (ctx->w.bounce_page == NULL) + return ERR_PTR(-ENOMEM); + ctx->flags |= EXT4_WRITE_PATH_FL; + return ctx->w.bounce_page; +} + /** * ext4_encrypt() - Encrypts a page * @inode: The inode for which the encryption should take place @@ -413,27 +349,17 @@ struct page *ext4_encrypt(struct inode *inode, return (struct page *) ctx; /* The encryption operation will require a bounce page. */ - ciphertext_page = alloc_page(GFP_NOFS); - if (!ciphertext_page) { - /* This is a potential bottleneck, but at least we'll have - * forward progress. */ - ciphertext_page = mempool_alloc(ext4_bounce_page_pool, - GFP_NOFS); - if (WARN_ON_ONCE(!ciphertext_page)) { - ciphertext_page = mempool_alloc(ext4_bounce_page_pool, - GFP_NOFS | __GFP_WAIT); - } - ctx->flags &= ~EXT4_BOUNCE_PAGE_REQUIRES_FREE_ENCRYPT_FL; - } else { - ctx->flags |= EXT4_BOUNCE_PAGE_REQUIRES_FREE_ENCRYPT_FL; - } - ctx->bounce_page = ciphertext_page; - ctx->control_page = plaintext_page; + ciphertext_page = alloc_bounce_page(ctx); + if (IS_ERR(ciphertext_page)) + goto errout; + ctx->w.control_page = plaintext_page; err = ext4_page_crypto(ctx, inode, EXT4_ENCRYPT, plaintext_page->index, plaintext_page, ciphertext_page); if (err) { + ciphertext_page = ERR_PTR(err); + errout: ext4_release_crypto_ctx(ctx); - return ERR_PTR(err); + return ciphertext_page; } SetPagePrivate(ciphertext_page); set_page_private(ciphertext_page, (unsigned long)ctx); @@ -470,8 +396,8 @@ int ext4_decrypt_one(struct inode *inode, struct page *page) struct ext4_crypto_ctx *ctx = ext4_get_crypto_ctx(inode); - if (!ctx) - return -ENOMEM; + if (IS_ERR(ctx)) + return PTR_ERR(ctx); ret = ext4_decrypt(ctx, page); ext4_release_crypto_ctx(ctx); return ret; @@ -493,21 +419,11 @@ int ext4_encrypted_zeroout(struct inode *inode, struct ext4_extent *ex) if (IS_ERR(ctx)) return PTR_ERR(ctx); - ciphertext_page = alloc_page(GFP_NOFS); - if (!ciphertext_page) { - /* This is a potential bottleneck, but at least we'll have - * forward progress. */ - ciphertext_page = mempool_alloc(ext4_bounce_page_pool, - GFP_NOFS); - if (WARN_ON_ONCE(!ciphertext_page)) { - ciphertext_page = mempool_alloc(ext4_bounce_page_pool, - GFP_NOFS | __GFP_WAIT); - } - ctx->flags &= ~EXT4_BOUNCE_PAGE_REQUIRES_FREE_ENCRYPT_FL; - } else { - ctx->flags |= EXT4_BOUNCE_PAGE_REQUIRES_FREE_ENCRYPT_FL; + ciphertext_page = alloc_bounce_page(ctx); + if (IS_ERR(ciphertext_page)) { + err = PTR_ERR(ciphertext_page); + goto errout; } - ctx->bounce_page = ciphertext_page; while (len--) { err = ext4_page_crypto(ctx, inode, EXT4_ENCRYPT, lblk, @@ -529,6 +445,7 @@ int ext4_encrypted_zeroout(struct inode *inode, struct ext4_extent *ex) goto errout; } err = submit_bio_wait(WRITE, bio); + bio_put(bio); if (err) goto errout; } diff --git a/fs/ext4/crypto_fname.c b/fs/ext4/crypto_fname.c index fded02f..7dc4eb5 100644 --- a/fs/ext4/crypto_fname.c +++ b/fs/ext4/crypto_fname.c @@ -48,6 +48,12 @@ bool ext4_valid_filenames_enc_mode(uint32_t mode) return (mode == EXT4_ENCRYPTION_MODE_AES_256_CTS); } +static unsigned max_name_len(struct inode *inode) +{ + return S_ISLNK(inode->i_mode) ? inode->i_sb->s_blocksize : + EXT4_NAME_LEN; +} + /** * ext4_fname_encrypt() - * @@ -55,43 +61,52 @@ bool ext4_valid_filenames_enc_mode(uint32_t mode) * ciphertext. Errors are returned as negative numbers. We trust the caller to * allocate sufficient memory to oname string. */ -static int ext4_fname_encrypt(struct ext4_fname_crypto_ctx *ctx, +static int ext4_fname_encrypt(struct inode *inode, const struct qstr *iname, struct ext4_str *oname) { u32 ciphertext_len; struct ablkcipher_request *req = NULL; DECLARE_EXT4_COMPLETION_RESULT(ecr); - struct crypto_ablkcipher *tfm = ctx->ctfm; + struct ext4_crypt_info *ci = EXT4_I(inode)->i_crypt_info; + struct crypto_ablkcipher *tfm = ci->ci_ctfm; int res = 0; char iv[EXT4_CRYPTO_BLOCK_SIZE]; - struct scatterlist sg[1]; - int padding = 4 << (ctx->flags & EXT4_POLICY_FLAGS_PAD_MASK); - char *workbuf; + struct scatterlist src_sg, dst_sg; + int padding = 4 << (ci->ci_flags & EXT4_POLICY_FLAGS_PAD_MASK); + char *workbuf, buf[32], *alloc_buf = NULL; + unsigned lim = max_name_len(inode); - if (iname->len <= 0 || iname->len > ctx->lim) + if (iname->len <= 0 || iname->len > lim) return -EIO; ciphertext_len = (iname->len < EXT4_CRYPTO_BLOCK_SIZE) ? EXT4_CRYPTO_BLOCK_SIZE : iname->len; ciphertext_len = ext4_fname_crypto_round_up(ciphertext_len, padding); - ciphertext_len = (ciphertext_len > ctx->lim) - ? ctx->lim : ciphertext_len; + ciphertext_len = (ciphertext_len > lim) + ? lim : ciphertext_len; + + if (ciphertext_len <= sizeof(buf)) { + workbuf = buf; + } else { + alloc_buf = kmalloc(ciphertext_len, GFP_NOFS); + if (!alloc_buf) + return -ENOMEM; + workbuf = alloc_buf; + } /* Allocate request */ req = ablkcipher_request_alloc(tfm, GFP_NOFS); if (!req) { printk_ratelimited( KERN_ERR "%s: crypto_request_alloc() failed\n", __func__); + kfree(alloc_buf); return -ENOMEM; } ablkcipher_request_set_callback(req, CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP, ext4_dir_crypt_complete, &ecr); - /* Map the workpage */ - workbuf = kmap(ctx->workpage); - /* Copy the input */ memcpy(workbuf, iname->name, iname->len); if (iname->len < ciphertext_len) @@ -101,21 +116,16 @@ static int ext4_fname_encrypt(struct ext4_fname_crypto_ctx *ctx, memset(iv, 0, EXT4_CRYPTO_BLOCK_SIZE); /* Create encryption request */ - sg_init_table(sg, 1); - sg_set_page(sg, ctx->workpage, PAGE_SIZE, 0); - ablkcipher_request_set_crypt(req, sg, sg, ciphertext_len, iv); + sg_init_one(&src_sg, workbuf, ciphertext_len); + sg_init_one(&dst_sg, oname->name, ciphertext_len); + ablkcipher_request_set_crypt(req, &src_sg, &dst_sg, ciphertext_len, iv); res = crypto_ablkcipher_encrypt(req); if (res == -EINPROGRESS || res == -EBUSY) { BUG_ON(req->base.data != &ecr); wait_for_completion(&ecr.completion); res = ecr.res; } - if (res >= 0) { - /* Copy the result to output */ - memcpy(oname->name, workbuf, ciphertext_len); - res = ciphertext_len; - } - kunmap(ctx->workpage); + kfree(alloc_buf); ablkcipher_request_free(req); if (res < 0) { printk_ratelimited( @@ -132,20 +142,21 @@ static int ext4_fname_encrypt(struct ext4_fname_crypto_ctx *ctx, * Errors are returned as negative numbers. * We trust the caller to allocate sufficient memory to oname string. */ -static int ext4_fname_decrypt(struct ext4_fname_crypto_ctx *ctx, +static int ext4_fname_decrypt(struct inode *inode, const struct ext4_str *iname, struct ext4_str *oname) { struct ext4_str tmp_in[2], tmp_out[1]; struct ablkcipher_request *req = NULL; DECLARE_EXT4_COMPLETION_RESULT(ecr); - struct scatterlist sg[1]; - struct crypto_ablkcipher *tfm = ctx->ctfm; + struct scatterlist src_sg, dst_sg; + struct ext4_crypt_info *ci = EXT4_I(inode)->i_crypt_info; + struct crypto_ablkcipher *tfm = ci->ci_ctfm; int res = 0; char iv[EXT4_CRYPTO_BLOCK_SIZE]; - char *workbuf; + unsigned lim = max_name_len(inode); - if (iname->len <= 0 || iname->len > ctx->lim) + if (iname->len <= 0 || iname->len > lim) return -EIO; tmp_in[0].name = iname->name; @@ -163,31 +174,19 @@ static int ext4_fname_decrypt(struct ext4_fname_crypto_ctx *ctx, CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP, ext4_dir_crypt_complete, &ecr); - /* Map the workpage */ - workbuf = kmap(ctx->workpage); - - /* Copy the input */ - memcpy(workbuf, iname->name, iname->len); - /* Initialize IV */ memset(iv, 0, EXT4_CRYPTO_BLOCK_SIZE); /* Create encryption request */ - sg_init_table(sg, 1); - sg_set_page(sg, ctx->workpage, PAGE_SIZE, 0); - ablkcipher_request_set_crypt(req, sg, sg, iname->len, iv); + sg_init_one(&src_sg, iname->name, iname->len); + sg_init_one(&dst_sg, oname->name, oname->len); + ablkcipher_request_set_crypt(req, &src_sg, &dst_sg, iname->len, iv); res = crypto_ablkcipher_decrypt(req); if (res == -EINPROGRESS || res == -EBUSY) { BUG_ON(req->base.data != &ecr); wait_for_completion(&ecr.completion); res = ecr.res; } - if (res >= 0) { - /* Copy the result to output */ - memcpy(oname->name, workbuf, iname->len); - res = iname->len; - } - kunmap(ctx->workpage); ablkcipher_request_free(req); if (res < 0) { printk_ratelimited( @@ -254,207 +253,6 @@ static int digest_decode(const char *src, int len, char *dst) } /** - * ext4_free_fname_crypto_ctx() - - * - * Frees up a crypto context. - */ -void ext4_free_fname_crypto_ctx(struct ext4_fname_crypto_ctx *ctx) -{ - if (ctx == NULL || IS_ERR(ctx)) - return; - - if (ctx->ctfm && !IS_ERR(ctx->ctfm)) - crypto_free_ablkcipher(ctx->ctfm); - if (ctx->htfm && !IS_ERR(ctx->htfm)) - crypto_free_hash(ctx->htfm); - if (ctx->workpage && !IS_ERR(ctx->workpage)) - __free_page(ctx->workpage); - kfree(ctx); -} - -/** - * ext4_put_fname_crypto_ctx() - - * - * Return: The crypto context onto free list. If the free list is above a - * threshold, completely frees up the context, and returns the memory. - * - * TODO: Currently we directly free the crypto context. Eventually we should - * add code it to return to free list. Such an approach will increase - * efficiency of directory lookup. - */ -void ext4_put_fname_crypto_ctx(struct ext4_fname_crypto_ctx **ctx) -{ - if (*ctx == NULL || IS_ERR(*ctx)) - return; - ext4_free_fname_crypto_ctx(*ctx); - *ctx = NULL; -} - -/** - * ext4_search_fname_crypto_ctx() - - */ -static struct ext4_fname_crypto_ctx *ext4_search_fname_crypto_ctx( - const struct ext4_encryption_key *key) -{ - return NULL; -} - -/** - * ext4_alloc_fname_crypto_ctx() - - */ -struct ext4_fname_crypto_ctx *ext4_alloc_fname_crypto_ctx( - const struct ext4_encryption_key *key) -{ - struct ext4_fname_crypto_ctx *ctx; - - ctx = kmalloc(sizeof(struct ext4_fname_crypto_ctx), GFP_NOFS); - if (ctx == NULL) - return ERR_PTR(-ENOMEM); - if (key->mode == EXT4_ENCRYPTION_MODE_INVALID) { - /* This will automatically set key mode to invalid - * As enum for ENCRYPTION_MODE_INVALID is zero */ - memset(&ctx->key, 0, sizeof(ctx->key)); - } else { - memcpy(&ctx->key, key, sizeof(struct ext4_encryption_key)); - } - ctx->has_valid_key = (EXT4_ENCRYPTION_MODE_INVALID == key->mode) - ? 0 : 1; - ctx->ctfm_key_is_ready = 0; - ctx->ctfm = NULL; - ctx->htfm = NULL; - ctx->workpage = NULL; - return ctx; -} - -/** - * ext4_get_fname_crypto_ctx() - - * - * Allocates a free crypto context and initializes it to hold - * the crypto material for the inode. - * - * Return: NULL if not encrypted. Error value on error. Valid pointer otherwise. - */ -struct ext4_fname_crypto_ctx *ext4_get_fname_crypto_ctx( - struct inode *inode, u32 max_ciphertext_len) -{ - struct ext4_fname_crypto_ctx *ctx; - struct ext4_inode_info *ei = EXT4_I(inode); - int res; - - /* Check if the crypto policy is set on the inode */ - res = ext4_encrypted_inode(inode); - if (res == 0) - return NULL; - - if (!ext4_has_encryption_key(inode)) - ext4_generate_encryption_key(inode); - - /* Get a crypto context based on the key. - * A new context is allocated if no context matches the requested key. - */ - ctx = ext4_search_fname_crypto_ctx(&(ei->i_encryption_key)); - if (ctx == NULL) - ctx = ext4_alloc_fname_crypto_ctx(&(ei->i_encryption_key)); - if (IS_ERR(ctx)) - return ctx; - - ctx->flags = ei->i_crypt_policy_flags; - if (ctx->has_valid_key) { - if (ctx->key.mode != EXT4_ENCRYPTION_MODE_AES_256_CTS) { - printk_once(KERN_WARNING - "ext4: unsupported key mode %d\n", - ctx->key.mode); - return ERR_PTR(-ENOKEY); - } - - /* As a first cut, we will allocate new tfm in every call. - * later, we will keep the tfm around, in case the key gets - * re-used */ - if (ctx->ctfm == NULL) { - ctx->ctfm = crypto_alloc_ablkcipher("cts(cbc(aes))", - 0, 0); - } - if (IS_ERR(ctx->ctfm)) { - res = PTR_ERR(ctx->ctfm); - printk( - KERN_DEBUG "%s: error (%d) allocating crypto tfm\n", - __func__, res); - ctx->ctfm = NULL; - ext4_put_fname_crypto_ctx(&ctx); - return ERR_PTR(res); - } - if (ctx->ctfm == NULL) { - printk( - KERN_DEBUG "%s: could not allocate crypto tfm\n", - __func__); - ext4_put_fname_crypto_ctx(&ctx); - return ERR_PTR(-ENOMEM); - } - if (ctx->workpage == NULL) - ctx->workpage = alloc_page(GFP_NOFS); - if (IS_ERR(ctx->workpage)) { - res = PTR_ERR(ctx->workpage); - printk( - KERN_DEBUG "%s: error (%d) allocating work page\n", - __func__, res); - ctx->workpage = NULL; - ext4_put_fname_crypto_ctx(&ctx); - return ERR_PTR(res); - } - if (ctx->workpage == NULL) { - printk( - KERN_DEBUG "%s: could not allocate work page\n", - __func__); - ext4_put_fname_crypto_ctx(&ctx); - return ERR_PTR(-ENOMEM); - } - ctx->lim = max_ciphertext_len; - crypto_ablkcipher_clear_flags(ctx->ctfm, ~0); - crypto_tfm_set_flags(crypto_ablkcipher_tfm(ctx->ctfm), - CRYPTO_TFM_REQ_WEAK_KEY); - - /* If we are lucky, we will get a context that is already - * set up with the right key. Else, we will have to - * set the key */ - if (!ctx->ctfm_key_is_ready) { - /* Since our crypto objectives for filename encryption - * are pretty weak, - * we directly use the inode master key */ - res = crypto_ablkcipher_setkey(ctx->ctfm, - ctx->key.raw, ctx->key.size); - if (res) { - ext4_put_fname_crypto_ctx(&ctx); - return ERR_PTR(-EIO); - } - ctx->ctfm_key_is_ready = 1; - } else { - /* In the current implementation, key should never be - * marked "ready" for a context that has just been - * allocated. So we should never reach here */ - BUG(); - } - } - if (ctx->htfm == NULL) - ctx->htfm = crypto_alloc_hash("sha256", 0, CRYPTO_ALG_ASYNC); - if (IS_ERR(ctx->htfm)) { - res = PTR_ERR(ctx->htfm); - printk(KERN_DEBUG "%s: error (%d) allocating hash tfm\n", - __func__, res); - ctx->htfm = NULL; - ext4_put_fname_crypto_ctx(&ctx); - return ERR_PTR(res); - } - if (ctx->htfm == NULL) { - printk(KERN_DEBUG "%s: could not allocate hash tfm\n", - __func__); - ext4_put_fname_crypto_ctx(&ctx); - return ERR_PTR(-ENOMEM); - } - - return ctx; -} - -/** * ext4_fname_crypto_round_up() - * * Return: The next multiple of block size @@ -464,44 +262,29 @@ u32 ext4_fname_crypto_round_up(u32 size, u32 blksize) return ((size+blksize-1)/blksize)*blksize; } -/** - * ext4_fname_crypto_namelen_on_disk() - - */ -int ext4_fname_crypto_namelen_on_disk(struct ext4_fname_crypto_ctx *ctx, - u32 namelen) +unsigned ext4_fname_encrypted_size(struct inode *inode, u32 ilen) { - u32 ciphertext_len; - int padding = 4 << (ctx->flags & EXT4_POLICY_FLAGS_PAD_MASK); - - if (ctx == NULL) - return -EIO; - if (!(ctx->has_valid_key)) - return -EACCES; - ciphertext_len = (namelen < EXT4_CRYPTO_BLOCK_SIZE) ? - EXT4_CRYPTO_BLOCK_SIZE : namelen; - ciphertext_len = ext4_fname_crypto_round_up(ciphertext_len, padding); - ciphertext_len = (ciphertext_len > ctx->lim) - ? ctx->lim : ciphertext_len; - return (int) ciphertext_len; + struct ext4_crypt_info *ci = EXT4_I(inode)->i_crypt_info; + int padding = 32; + + if (ci) + padding = 4 << (ci->ci_flags & EXT4_POLICY_FLAGS_PAD_MASK); + if (ilen < EXT4_CRYPTO_BLOCK_SIZE) + ilen = EXT4_CRYPTO_BLOCK_SIZE; + return ext4_fname_crypto_round_up(ilen, padding); } -/** - * ext4_fname_crypto_alloc_obuff() - +/* + * ext4_fname_crypto_alloc_buffer() - * * Allocates an output buffer that is sufficient for the crypto operation * specified by the context and the direction. */ -int ext4_fname_crypto_alloc_buffer(struct ext4_fname_crypto_ctx *ctx, +int ext4_fname_crypto_alloc_buffer(struct inode *inode, u32 ilen, struct ext4_str *crypto_str) { - unsigned int olen; - int padding = 4 << (ctx->flags & EXT4_POLICY_FLAGS_PAD_MASK); + unsigned int olen = ext4_fname_encrypted_size(inode, ilen); - if (!ctx) - return -EIO; - if (padding < EXT4_CRYPTO_BLOCK_SIZE) - padding = EXT4_CRYPTO_BLOCK_SIZE; - olen = ext4_fname_crypto_round_up(ilen, padding); crypto_str->len = olen; if (olen < EXT4_FNAME_CRYPTO_DIGEST_SIZE*2) olen = EXT4_FNAME_CRYPTO_DIGEST_SIZE*2; @@ -529,7 +312,7 @@ void ext4_fname_crypto_free_buffer(struct ext4_str *crypto_str) /** * ext4_fname_disk_to_usr() - converts a filename from disk space to user space */ -int _ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, +int _ext4_fname_disk_to_usr(struct inode *inode, struct dx_hash_info *hinfo, const struct ext4_str *iname, struct ext4_str *oname) @@ -537,8 +320,6 @@ int _ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, char buf[24]; int ret; - if (ctx == NULL) - return -EIO; if (iname->len < 3) { /*Check for . and .. */ if (iname->name[0] == '.' && iname->name[iname->len-1] == '.') { @@ -548,8 +329,8 @@ int _ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, return oname->len; } } - if (ctx->has_valid_key) - return ext4_fname_decrypt(ctx, iname, oname); + if (EXT4_I(inode)->i_crypt_info) + return ext4_fname_decrypt(inode, iname, oname); if (iname->len <= EXT4_FNAME_CRYPTO_DIGEST_SIZE) { ret = digest_encode(iname->name, iname->len, oname->name); @@ -568,7 +349,7 @@ int _ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, return ret + 1; } -int ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, +int ext4_fname_disk_to_usr(struct inode *inode, struct dx_hash_info *hinfo, const struct ext4_dir_entry_2 *de, struct ext4_str *oname) @@ -576,21 +357,20 @@ int ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, struct ext4_str iname = {.name = (unsigned char *) de->name, .len = de->name_len }; - return _ext4_fname_disk_to_usr(ctx, hinfo, &iname, oname); + return _ext4_fname_disk_to_usr(inode, hinfo, &iname, oname); } /** * ext4_fname_usr_to_disk() - converts a filename from user space to disk space */ -int ext4_fname_usr_to_disk(struct ext4_fname_crypto_ctx *ctx, +int ext4_fname_usr_to_disk(struct inode *inode, const struct qstr *iname, struct ext4_str *oname) { int res; + struct ext4_crypt_info *ci = EXT4_I(inode)->i_crypt_info; - if (ctx == NULL) - return -EIO; if (iname->len < 3) { /*Check for . and .. */ if (iname->name[0] == '.' && @@ -601,8 +381,8 @@ int ext4_fname_usr_to_disk(struct ext4_fname_crypto_ctx *ctx, return oname->len; } } - if (ctx->has_valid_key) { - res = ext4_fname_encrypt(ctx, iname, oname); + if (ci) { + res = ext4_fname_encrypt(inode, iname, oname); return res; } /* Without a proper key, a user is not allowed to modify the filenames @@ -611,109 +391,79 @@ int ext4_fname_usr_to_disk(struct ext4_fname_crypto_ctx *ctx, return -EACCES; } -/* - * Calculate the htree hash from a filename from user space - */ -int ext4_fname_usr_to_hash(struct ext4_fname_crypto_ctx *ctx, - const struct qstr *iname, - struct dx_hash_info *hinfo) +int ext4_fname_setup_filename(struct inode *dir, const struct qstr *iname, + int lookup, struct ext4_filename *fname) { - struct ext4_str tmp; - int ret = 0; - char buf[EXT4_FNAME_CRYPTO_DIGEST_SIZE+1]; + struct ext4_crypt_info *ci; + int ret = 0, bigname = 0; + + memset(fname, 0, sizeof(struct ext4_filename)); + fname->usr_fname = iname; - if (!ctx || + if (!ext4_encrypted_inode(dir) || ((iname->name[0] == '.') && ((iname->len == 1) || ((iname->name[1] == '.') && (iname->len == 2))))) { - ext4fs_dirhash(iname->name, iname->len, hinfo); + fname->disk_name.name = (unsigned char *) iname->name; + fname->disk_name.len = iname->len; return 0; } - - if (!ctx->has_valid_key && iname->name[0] == '_') { - if (iname->len != 33) - return -ENOENT; - ret = digest_decode(iname->name+1, iname->len, buf); - if (ret != 24) - return -ENOENT; - memcpy(&hinfo->hash, buf, 4); - memcpy(&hinfo->minor_hash, buf + 4, 4); + ret = ext4_get_encryption_info(dir); + if (ret) + return ret; + ci = EXT4_I(dir)->i_crypt_info; + if (ci) { + ret = ext4_fname_crypto_alloc_buffer(dir, iname->len, + &fname->crypto_buf); + if (ret < 0) + return ret; + ret = ext4_fname_encrypt(dir, iname, &fname->crypto_buf); + if (ret < 0) + goto errout; + fname->disk_name.name = fname->crypto_buf.name; + fname->disk_name.len = fname->crypto_buf.len; return 0; } + if (!lookup) + return -EACCES; - if (!ctx->has_valid_key && iname->name[0] != '_') { - if (iname->len > 43) - return -ENOENT; - ret = digest_decode(iname->name, iname->len, buf); - ext4fs_dirhash(buf, ret, hinfo); - return 0; + /* We don't have the key and we are doing a lookup; decode the + * user-supplied name + */ + if (iname->name[0] == '_') + bigname = 1; + if ((bigname && (iname->len != 33)) || + (!bigname && (iname->len > 43))) + return -ENOENT; + + fname->crypto_buf.name = kmalloc(32, GFP_KERNEL); + if (fname->crypto_buf.name == NULL) + return -ENOMEM; + ret = digest_decode(iname->name + bigname, iname->len - bigname, + fname->crypto_buf.name); + if (ret < 0) { + ret = -ENOENT; + goto errout; } - - /* First encrypt the plaintext name */ - ret = ext4_fname_crypto_alloc_buffer(ctx, iname->len, &tmp); - if (ret < 0) - return ret; - - ret = ext4_fname_encrypt(ctx, iname, &tmp); - if (ret >= 0) { - ext4fs_dirhash(tmp.name, tmp.len, hinfo); - ret = 0; + fname->crypto_buf.len = ret; + if (bigname) { + memcpy(&fname->hinfo.hash, fname->crypto_buf.name, 4); + memcpy(&fname->hinfo.minor_hash, fname->crypto_buf.name + 4, 4); + } else { + fname->disk_name.name = fname->crypto_buf.name; + fname->disk_name.len = fname->crypto_buf.len; } - - ext4_fname_crypto_free_buffer(&tmp); + return 0; +errout: + kfree(fname->crypto_buf.name); + fname->crypto_buf.name = NULL; return ret; } -int ext4_fname_match(struct ext4_fname_crypto_ctx *ctx, struct ext4_str *cstr, - int len, const char * const name, - struct ext4_dir_entry_2 *de) +void ext4_fname_free_filename(struct ext4_filename *fname) { - int ret = -ENOENT; - int bigname = (*name == '_'); - - if (ctx->has_valid_key) { - if (cstr->name == NULL) { - struct qstr istr; - - ret = ext4_fname_crypto_alloc_buffer(ctx, len, cstr); - if (ret < 0) - goto errout; - istr.name = name; - istr.len = len; - ret = ext4_fname_encrypt(ctx, &istr, cstr); - if (ret < 0) - goto errout; - } - } else { - if (cstr->name == NULL) { - cstr->name = kmalloc(32, GFP_KERNEL); - if (cstr->name == NULL) - return -ENOMEM; - if ((bigname && (len != 33)) || - (!bigname && (len > 43))) - goto errout; - ret = digest_decode(name+bigname, len-bigname, - cstr->name); - if (ret < 0) { - ret = -ENOENT; - goto errout; - } - cstr->len = ret; - } - if (bigname) { - if (de->name_len < 16) - return 0; - ret = memcmp(de->name + de->name_len - 16, - cstr->name + 8, 16); - return (ret == 0) ? 1 : 0; - } - } - if (de->name_len != cstr->len) - return 0; - ret = memcmp(de->name, cstr->name, cstr->len); - return (ret == 0) ? 1 : 0; -errout: - kfree(cstr->name); - cstr->name = NULL; - return ret; + kfree(fname->crypto_buf.name); + fname->crypto_buf.name = NULL; + fname->usr_fname = NULL; + fname->disk_name.name = NULL; } diff --git a/fs/ext4/crypto_key.c b/fs/ext4/crypto_key.c index 52170d0..442d24e 100644 --- a/fs/ext4/crypto_key.c +++ b/fs/ext4/crypto_key.c @@ -84,14 +84,38 @@ out: return res; } -/** - * ext4_generate_encryption_key() - generates an encryption key - * @inode: The inode to generate the encryption key for. - */ -int ext4_generate_encryption_key(struct inode *inode) +void ext4_free_crypt_info(struct ext4_crypt_info *ci) +{ + if (!ci) + return; + + if (ci->ci_keyring_key) + key_put(ci->ci_keyring_key); + crypto_free_ablkcipher(ci->ci_ctfm); + kmem_cache_free(ext4_crypt_info_cachep, ci); +} + +void ext4_free_encryption_info(struct inode *inode, + struct ext4_crypt_info *ci) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_crypt_info *prev; + + if (ci == NULL) + ci = ACCESS_ONCE(ei->i_crypt_info); + if (ci == NULL) + return; + prev = cmpxchg(&ei->i_crypt_info, ci, NULL); + if (prev != ci) + return; + + ext4_free_crypt_info(ci); +} + +int _ext4_get_encryption_info(struct inode *inode) { struct ext4_inode_info *ei = EXT4_I(inode); - struct ext4_encryption_key *crypt_key = &ei->i_encryption_key; + struct ext4_crypt_info *crypt_info; char full_key_descriptor[EXT4_KEY_DESC_PREFIX_SIZE + (EXT4_KEY_DESCRIPTOR_SIZE * 2) + 1]; struct key *keyring_key = NULL; @@ -99,31 +123,76 @@ int ext4_generate_encryption_key(struct inode *inode) struct ext4_encryption_context ctx; struct user_key_payload *ukp; struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); - int res = ext4_xattr_get(inode, EXT4_XATTR_INDEX_ENCRYPTION, - EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, - &ctx, sizeof(ctx)); + struct crypto_ablkcipher *ctfm; + const char *cipher_str; + char raw_key[EXT4_MAX_KEY_SIZE]; + char mode; + int res; - if (res != sizeof(ctx)) { - if (res > 0) - res = -EINVAL; - goto out; + if (!ext4_read_workqueue) { + res = ext4_init_crypto(); + if (res) + return res; + } + +retry: + crypt_info = ACCESS_ONCE(ei->i_crypt_info); + if (crypt_info) { + if (!crypt_info->ci_keyring_key || + key_validate(crypt_info->ci_keyring_key) == 0) + return 0; + ext4_free_encryption_info(inode, crypt_info); + goto retry; } + + res = ext4_xattr_get(inode, EXT4_XATTR_INDEX_ENCRYPTION, + EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, + &ctx, sizeof(ctx)); + if (res < 0) { + if (!DUMMY_ENCRYPTION_ENABLED(sbi)) + return res; + ctx.contents_encryption_mode = EXT4_ENCRYPTION_MODE_AES_256_XTS; + ctx.filenames_encryption_mode = + EXT4_ENCRYPTION_MODE_AES_256_CTS; + ctx.flags = 0; + } else if (res != sizeof(ctx)) + return -EINVAL; res = 0; - ei->i_crypt_policy_flags = ctx.flags; + crypt_info = kmem_cache_alloc(ext4_crypt_info_cachep, GFP_KERNEL); + if (!crypt_info) + return -ENOMEM; + + crypt_info->ci_flags = ctx.flags; + crypt_info->ci_data_mode = ctx.contents_encryption_mode; + crypt_info->ci_filename_mode = ctx.filenames_encryption_mode; + crypt_info->ci_ctfm = NULL; + crypt_info->ci_keyring_key = NULL; + memcpy(crypt_info->ci_master_key, ctx.master_key_descriptor, + sizeof(crypt_info->ci_master_key)); if (S_ISREG(inode->i_mode)) - crypt_key->mode = ctx.contents_encryption_mode; + mode = crypt_info->ci_data_mode; else if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) - crypt_key->mode = ctx.filenames_encryption_mode; - else { - printk(KERN_ERR "ext4 crypto: Unsupported inode type.\n"); + mode = crypt_info->ci_filename_mode; + else BUG(); + switch (mode) { + case EXT4_ENCRYPTION_MODE_AES_256_XTS: + cipher_str = "xts(aes)"; + break; + case EXT4_ENCRYPTION_MODE_AES_256_CTS: + cipher_str = "cts(cbc(aes))"; + break; + default: + printk_once(KERN_WARNING + "ext4: unsupported key mode %d (ino %u)\n", + mode, (unsigned) inode->i_ino); + res = -ENOKEY; + goto out; } - crypt_key->size = ext4_encryption_key_size(crypt_key->mode); - BUG_ON(!crypt_key->size); if (DUMMY_ENCRYPTION_ENABLED(sbi)) { - memset(crypt_key->raw, 0x42, EXT4_AES_256_XTS_KEY_SIZE); - goto out; + memset(raw_key, 0x42, EXT4_AES_256_XTS_KEY_SIZE); + goto got_key; } memcpy(full_key_descriptor, EXT4_KEY_DESC_PREFIX, EXT4_KEY_DESC_PREFIX_SIZE); @@ -138,6 +207,7 @@ int ext4_generate_encryption_key(struct inode *inode) keyring_key = NULL; goto out; } + crypt_info->ci_keyring_key = keyring_key; BUG_ON(keyring_key->type != &key_type_logon); ukp = ((struct user_key_payload *)keyring_key->payload.data); if (ukp->datalen != sizeof(struct ext4_encryption_key)) { @@ -148,19 +218,43 @@ int ext4_generate_encryption_key(struct inode *inode) BUILD_BUG_ON(EXT4_AES_128_ECB_KEY_SIZE != EXT4_KEY_DERIVATION_NONCE_SIZE); BUG_ON(master_key->size != EXT4_AES_256_XTS_KEY_SIZE); - res = ext4_derive_key_aes(ctx.nonce, master_key->raw, crypt_key->raw); + res = ext4_derive_key_aes(ctx.nonce, master_key->raw, + raw_key); +got_key: + ctfm = crypto_alloc_ablkcipher(cipher_str, 0, 0); + if (!ctfm || IS_ERR(ctfm)) { + res = ctfm ? PTR_ERR(ctfm) : -ENOMEM; + printk(KERN_DEBUG + "%s: error %d (inode %u) allocating crypto tfm\n", + __func__, res, (unsigned) inode->i_ino); + goto out; + } + crypt_info->ci_ctfm = ctfm; + crypto_ablkcipher_clear_flags(ctfm, ~0); + crypto_tfm_set_flags(crypto_ablkcipher_tfm(ctfm), + CRYPTO_TFM_REQ_WEAK_KEY); + res = crypto_ablkcipher_setkey(ctfm, raw_key, + ext4_encryption_key_size(mode)); + if (res) + goto out; + memzero_explicit(raw_key, sizeof(raw_key)); + if (cmpxchg(&ei->i_crypt_info, NULL, crypt_info) != NULL) { + ext4_free_crypt_info(crypt_info); + goto retry; + } + return 0; + out: - if (keyring_key) - key_put(keyring_key); - if (res < 0) - crypt_key->mode = EXT4_ENCRYPTION_MODE_INVALID; + if (res == -ENOKEY) + res = 0; + ext4_free_crypt_info(crypt_info); + memzero_explicit(raw_key, sizeof(raw_key)); return res; } int ext4_has_encryption_key(struct inode *inode) { struct ext4_inode_info *ei = EXT4_I(inode); - struct ext4_encryption_key *crypt_key = &ei->i_encryption_key; - return (crypt_key->mode != EXT4_ENCRYPTION_MODE_INVALID); + return (ei->i_crypt_info != NULL); } diff --git a/fs/ext4/crypto_policy.c b/fs/ext4/crypto_policy.c index a6d6291..02c4e5d 100644 --- a/fs/ext4/crypto_policy.c +++ b/fs/ext4/crypto_policy.c @@ -51,6 +51,10 @@ static int ext4_create_encryption_context_from_policy( struct ext4_encryption_context ctx; int res = 0; + res = ext4_convert_inline_data(inode); + if (res) + return res; + ctx.format = EXT4_ENCRYPTION_CONTEXT_FORMAT_V1; memcpy(ctx.master_key_descriptor, policy->master_key_descriptor, EXT4_KEY_DESCRIPTOR_SIZE); @@ -89,6 +93,8 @@ int ext4_process_policy(const struct ext4_encryption_policy *policy, return -EINVAL; if (!ext4_inode_has_encryption_context(inode)) { + if (!S_ISDIR(inode->i_mode)) + return -EINVAL; if (!ext4_empty_dir(inode)) return -ENOTEMPTY; return ext4_create_encryption_context_from_policy(inode, @@ -126,7 +132,7 @@ int ext4_get_policy(struct inode *inode, struct ext4_encryption_policy *policy) int ext4_is_child_context_consistent_with_parent(struct inode *parent, struct inode *child) { - struct ext4_encryption_context parent_ctx, child_ctx; + struct ext4_crypt_info *parent_ci, *child_ci; int res; if ((parent == NULL) || (child == NULL)) { @@ -136,26 +142,28 @@ int ext4_is_child_context_consistent_with_parent(struct inode *parent, /* no restrictions if the parent directory is not encrypted */ if (!ext4_encrypted_inode(parent)) return 1; - res = ext4_xattr_get(parent, EXT4_XATTR_INDEX_ENCRYPTION, - EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, - &parent_ctx, sizeof(parent_ctx)); - if (res != sizeof(parent_ctx)) - return 0; /* if the child directory is not encrypted, this is always a problem */ if (!ext4_encrypted_inode(child)) return 0; - res = ext4_xattr_get(child, EXT4_XATTR_INDEX_ENCRYPTION, - EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, - &child_ctx, sizeof(child_ctx)); - if (res != sizeof(child_ctx)) + res = ext4_get_encryption_info(parent); + if (res) return 0; - return (memcmp(parent_ctx.master_key_descriptor, - child_ctx.master_key_descriptor, + res = ext4_get_encryption_info(child); + if (res) + return 0; + parent_ci = EXT4_I(parent)->i_crypt_info; + child_ci = EXT4_I(child)->i_crypt_info; + if (!parent_ci && !child_ci) + return 1; + if (!parent_ci || !child_ci) + return 0; + + return (memcmp(parent_ci->ci_master_key, + child_ci->ci_master_key, EXT4_KEY_DESCRIPTOR_SIZE) == 0 && - (parent_ctx.contents_encryption_mode == - child_ctx.contents_encryption_mode) && - (parent_ctx.filenames_encryption_mode == - child_ctx.filenames_encryption_mode)); + (parent_ci->ci_data_mode == child_ci->ci_data_mode) && + (parent_ci->ci_filename_mode == child_ci->ci_filename_mode) && + (parent_ci->ci_flags == child_ci->ci_flags)); } /** @@ -168,31 +176,40 @@ int ext4_is_child_context_consistent_with_parent(struct inode *parent, int ext4_inherit_context(struct inode *parent, struct inode *child) { struct ext4_encryption_context ctx; - int res = ext4_xattr_get(parent, EXT4_XATTR_INDEX_ENCRYPTION, - EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, - &ctx, sizeof(ctx)); + struct ext4_crypt_info *ci; + int res; + + res = ext4_get_encryption_info(parent); + if (res < 0) + return res; + ci = EXT4_I(parent)->i_crypt_info; + if (ci == NULL) + return -ENOKEY; - if (res != sizeof(ctx)) { - if (DUMMY_ENCRYPTION_ENABLED(EXT4_SB(parent->i_sb))) { - ctx.format = EXT4_ENCRYPTION_CONTEXT_FORMAT_V1; - ctx.contents_encryption_mode = - EXT4_ENCRYPTION_MODE_AES_256_XTS; - ctx.filenames_encryption_mode = - EXT4_ENCRYPTION_MODE_AES_256_CTS; - ctx.flags = 0; - memset(ctx.master_key_descriptor, 0x42, - EXT4_KEY_DESCRIPTOR_SIZE); - res = 0; - } else { - goto out; - } + ctx.format = EXT4_ENCRYPTION_CONTEXT_FORMAT_V1; + if (DUMMY_ENCRYPTION_ENABLED(EXT4_SB(parent->i_sb))) { + ctx.contents_encryption_mode = EXT4_ENCRYPTION_MODE_AES_256_XTS; + ctx.filenames_encryption_mode = + EXT4_ENCRYPTION_MODE_AES_256_CTS; + ctx.flags = 0; + memset(ctx.master_key_descriptor, 0x42, + EXT4_KEY_DESCRIPTOR_SIZE); + res = 0; + } else { + ctx.contents_encryption_mode = ci->ci_data_mode; + ctx.filenames_encryption_mode = ci->ci_filename_mode; + ctx.flags = ci->ci_flags; + memcpy(ctx.master_key_descriptor, ci->ci_master_key, + EXT4_KEY_DESCRIPTOR_SIZE); } get_random_bytes(ctx.nonce, EXT4_KEY_DERIVATION_NONCE_SIZE); res = ext4_xattr_set(child, EXT4_XATTR_INDEX_ENCRYPTION, EXT4_XATTR_NAME_ENCRYPTION_CONTEXT, &ctx, sizeof(ctx), 0); -out: - if (!res) + if (!res) { ext4_set_inode_flag(child, EXT4_INODE_ENCRYPT); + ext4_clear_inode_state(child, EXT4_STATE_MAY_INLINE_DATA); + res = ext4_get_encryption_info(child); + } return res; } diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index 5665d82..f9e1491 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -110,7 +110,6 @@ static int ext4_readdir(struct file *file, struct dir_context *ctx) struct super_block *sb = inode->i_sb; struct buffer_head *bh = NULL; int dir_has_error = 0; - struct ext4_fname_crypto_ctx *enc_ctx = NULL; struct ext4_str fname_crypto_str = {.name = NULL, .len = 0}; if (is_dx_dir(inode)) { @@ -134,16 +133,11 @@ static int ext4_readdir(struct file *file, struct dir_context *ctx) return err; } - enc_ctx = ext4_get_fname_crypto_ctx(inode, EXT4_NAME_LEN); - if (IS_ERR(enc_ctx)) - return PTR_ERR(enc_ctx); - if (enc_ctx) { - err = ext4_fname_crypto_alloc_buffer(enc_ctx, EXT4_NAME_LEN, + if (ext4_encrypted_inode(inode)) { + err = ext4_fname_crypto_alloc_buffer(inode, EXT4_NAME_LEN, &fname_crypto_str); - if (err < 0) { - ext4_put_fname_crypto_ctx(&enc_ctx); + if (err < 0) return err; - } } offset = ctx->pos & (sb->s_blocksize - 1); @@ -239,17 +233,19 @@ static int ext4_readdir(struct file *file, struct dir_context *ctx) offset += ext4_rec_len_from_disk(de->rec_len, sb->s_blocksize); if (le32_to_cpu(de->inode)) { - if (enc_ctx == NULL) { - /* Directory is not encrypted */ + if (!ext4_encrypted_inode(inode)) { if (!dir_emit(ctx, de->name, de->name_len, le32_to_cpu(de->inode), get_dtype(sb, de->file_type))) goto done; } else { + int save_len = fname_crypto_str.len; + /* Directory is encrypted */ - err = ext4_fname_disk_to_usr(enc_ctx, + err = ext4_fname_disk_to_usr(inode, NULL, de, &fname_crypto_str); + fname_crypto_str.len = save_len; if (err < 0) goto errout; if (!dir_emit(ctx, @@ -272,7 +268,6 @@ done: err = 0; errout: #ifdef CONFIG_EXT4_FS_ENCRYPTION - ext4_put_fname_crypto_ctx(&enc_ctx); ext4_fname_crypto_free_buffer(&fname_crypto_str); #endif brelse(bh); @@ -598,6 +593,13 @@ finished: return 0; } +static int ext4_dir_open(struct inode * inode, struct file * filp) +{ + if (ext4_encrypted_inode(inode)) + return ext4_get_encryption_info(inode) ? -EACCES : 0; + return 0; +} + static int ext4_release_dir(struct inode *inode, struct file *filp) { if (filp->private_data) @@ -640,5 +642,6 @@ const struct file_operations ext4_dir_operations = { .compat_ioctl = ext4_compat_ioctl, #endif .fsync = ext4_sync_file, + .open = ext4_dir_open, .release = ext4_release_dir, }; diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index ad358f2..f5e9f04 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -69,15 +69,6 @@ #define ext_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) #endif -#define EXT4_ERROR_INODE(inode, fmt, a...) \ - ext4_error_inode((inode), __func__, __LINE__, 0, (fmt), ## a) - -#define EXT4_ERROR_INODE_BLOCK(inode, block, fmt, a...) \ - ext4_error_inode((inode), __func__, __LINE__, (block), (fmt), ## a) - -#define EXT4_ERROR_FILE(file, block, fmt, a...) \ - ext4_error_file((file), __func__, __LINE__, (block), (fmt), ## a) - /* data type for block offset of block group */ typedef int ext4_grpblk_t; @@ -90,6 +81,11 @@ typedef __u32 ext4_lblk_t; /* data type for block group number */ typedef unsigned int ext4_group_t; +enum SHIFT_DIRECTION { + SHIFT_LEFT = 0, + SHIFT_RIGHT, +}; + /* * Flags used in mballoc's allocation_context flags field. * @@ -911,7 +907,6 @@ struct ext4_inode_info { /* on-disk additional length */ __u16 i_extra_isize; - char i_crypt_policy_flags; /* Indicate the inline data space. */ u16 i_inline_off; @@ -955,7 +950,7 @@ struct ext4_inode_info { #ifdef CONFIG_EXT4_FS_ENCRYPTION /* Encryption params */ - struct ext4_encryption_key i_encryption_key; + struct ext4_crypt_info *i_crypt_info; #endif }; @@ -1374,12 +1369,6 @@ struct ext4_sb_info { struct ratelimit_state s_err_ratelimit_state; struct ratelimit_state s_warning_ratelimit_state; struct ratelimit_state s_msg_ratelimit_state; - -#ifdef CONFIG_EXT4_FS_ENCRYPTION - /* Encryption */ - uint32_t s_file_encryption_mode; - uint32_t s_dir_encryption_mode; -#endif }; static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb) @@ -1838,6 +1827,17 @@ struct dx_hash_info */ #define HASH_NB_ALWAYS 1 +struct ext4_filename { + const struct qstr *usr_fname; + struct ext4_str disk_name; + struct dx_hash_info hinfo; +#ifdef CONFIG_EXT4_FS_ENCRYPTION + struct ext4_str crypto_buf; +#endif +}; + +#define fname_name(p) ((p)->disk_name.name) +#define fname_len(p) ((p)->disk_name.len) /* * Describe an inode's exact location on disk and in memory @@ -2054,6 +2054,7 @@ int ext4_get_policy(struct inode *inode, struct ext4_encryption_policy *policy); /* crypto.c */ +extern struct kmem_cache *ext4_crypt_info_cachep; bool ext4_valid_contents_enc_mode(uint32_t mode); uint32_t ext4_validate_encryption_key_size(uint32_t mode, uint32_t size); extern struct workqueue_struct *ext4_read_workqueue; @@ -2085,57 +2086,84 @@ static inline int ext4_sb_has_crypto(struct super_block *sb) /* crypto_fname.c */ bool ext4_valid_filenames_enc_mode(uint32_t mode); u32 ext4_fname_crypto_round_up(u32 size, u32 blksize); -int ext4_fname_crypto_alloc_buffer(struct ext4_fname_crypto_ctx *ctx, +unsigned ext4_fname_encrypted_size(struct inode *inode, u32 ilen); +int ext4_fname_crypto_alloc_buffer(struct inode *inode, u32 ilen, struct ext4_str *crypto_str); -int _ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, +int _ext4_fname_disk_to_usr(struct inode *inode, struct dx_hash_info *hinfo, const struct ext4_str *iname, struct ext4_str *oname); -int ext4_fname_disk_to_usr(struct ext4_fname_crypto_ctx *ctx, +int ext4_fname_disk_to_usr(struct inode *inode, struct dx_hash_info *hinfo, const struct ext4_dir_entry_2 *de, struct ext4_str *oname); -int ext4_fname_usr_to_disk(struct ext4_fname_crypto_ctx *ctx, +int ext4_fname_usr_to_disk(struct inode *inode, const struct qstr *iname, struct ext4_str *oname); -int ext4_fname_usr_to_hash(struct ext4_fname_crypto_ctx *ctx, - const struct qstr *iname, - struct dx_hash_info *hinfo); -int ext4_fname_crypto_namelen_on_disk(struct ext4_fname_crypto_ctx *ctx, - u32 namelen); -int ext4_fname_match(struct ext4_fname_crypto_ctx *ctx, struct ext4_str *cstr, - int len, const char * const name, - struct ext4_dir_entry_2 *de); - - #ifdef CONFIG_EXT4_FS_ENCRYPTION -void ext4_put_fname_crypto_ctx(struct ext4_fname_crypto_ctx **ctx); -struct ext4_fname_crypto_ctx *ext4_get_fname_crypto_ctx(struct inode *inode, - u32 max_len); void ext4_fname_crypto_free_buffer(struct ext4_str *crypto_str); +int ext4_fname_setup_filename(struct inode *dir, const struct qstr *iname, + int lookup, struct ext4_filename *fname); +void ext4_fname_free_filename(struct ext4_filename *fname); #else static inline -void ext4_put_fname_crypto_ctx(struct ext4_fname_crypto_ctx **ctx) { } -static inline -struct ext4_fname_crypto_ctx *ext4_get_fname_crypto_ctx(struct inode *inode, - u32 max_len) +int ext4_setup_fname_crypto(struct inode *inode) { - return NULL; + return 0; } static inline void ext4_fname_crypto_free_buffer(struct ext4_str *p) { } +static inline int ext4_fname_setup_filename(struct inode *dir, + const struct qstr *iname, + int lookup, struct ext4_filename *fname) +{ + fname->usr_fname = iname; + fname->disk_name.name = (unsigned char *) iname->name; + fname->disk_name.len = iname->len; + return 0; +} +static inline void ext4_fname_free_filename(struct ext4_filename *fname) { } #endif /* crypto_key.c */ -int ext4_generate_encryption_key(struct inode *inode); +void ext4_free_crypt_info(struct ext4_crypt_info *ci); +void ext4_free_encryption_info(struct inode *inode, struct ext4_crypt_info *ci); +int _ext4_get_encryption_info(struct inode *inode); #ifdef CONFIG_EXT4_FS_ENCRYPTION int ext4_has_encryption_key(struct inode *inode); + +static inline int ext4_get_encryption_info(struct inode *inode) +{ + struct ext4_crypt_info *ci = EXT4_I(inode)->i_crypt_info; + + if (!ci || + (ci->ci_keyring_key && + (ci->ci_keyring_key->flags & ((1 << KEY_FLAG_INVALIDATED) | + (1 << KEY_FLAG_REVOKED) | + (1 << KEY_FLAG_DEAD))))) + return _ext4_get_encryption_info(inode); + return 0; +} + +static inline struct ext4_crypt_info *ext4_encryption_info(struct inode *inode) +{ + return EXT4_I(inode)->i_crypt_info; +} + #else static inline int ext4_has_encryption_key(struct inode *inode) { return 0; } +static inline int ext4_get_encryption_info(struct inode *inode) +{ + return 0; +} +static inline struct ext4_crypt_info *ext4_encryption_info(struct inode *inode) +{ + return NULL; +} #endif @@ -2156,14 +2184,13 @@ extern void ext4_htree_free_dir_info(struct dir_private_info *p); extern int ext4_find_dest_de(struct inode *dir, struct inode *inode, struct buffer_head *bh, void *buf, int buf_size, - const char *name, int namelen, + struct ext4_filename *fname, struct ext4_dir_entry_2 **dest_de); int ext4_insert_dentry(struct inode *dir, - struct inode *inode, - struct ext4_dir_entry_2 *de, - int buf_size, - const struct qstr *iname, - const char *name, int namelen); + struct inode *inode, + struct ext4_dir_entry_2 *de, + int buf_size, + struct ext4_filename *fname); static inline void ext4_update_dx_flag(struct inode *inode) { if (!EXT4_HAS_COMPAT_FEATURE(inode->i_sb, @@ -2317,13 +2344,14 @@ extern int ext4_orphan_add(handle_t *, struct inode *); extern int ext4_orphan_del(handle_t *, struct inode *); extern int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash, __u32 start_minor_hash, __u32 *next_hash); -extern int search_dir(struct buffer_head *bh, - char *search_buf, - int buf_size, - struct inode *dir, - const struct qstr *d_name, - unsigned int offset, - struct ext4_dir_entry_2 **res_dir); +extern int ext4_search_dir(struct buffer_head *bh, + char *search_buf, + int buf_size, + struct inode *dir, + struct ext4_filename *fname, + const struct qstr *d_name, + unsigned int offset, + struct ext4_dir_entry_2 **res_dir); extern int ext4_generic_delete_entry(handle_t *handle, struct inode *dir, struct ext4_dir_entry_2 *de_del, @@ -2368,6 +2396,9 @@ void __ext4_abort(struct super_block *, const char *, unsigned int, extern __printf(4, 5) void __ext4_warning(struct super_block *, const char *, unsigned int, const char *, ...); +extern __printf(4, 5) +void __ext4_warning_inode(const struct inode *inode, const char *function, + unsigned int line, const char *fmt, ...); extern __printf(3, 4) void __ext4_msg(struct super_block *, const char *, const char *, ...); extern void __dump_mmp_msg(struct super_block *, struct mmp_struct *mmp, @@ -2378,6 +2409,15 @@ void __ext4_grp_locked_error(const char *, unsigned int, unsigned long, ext4_fsblk_t, const char *, ...); +#define EXT4_ERROR_INODE(inode, fmt, a...) \ + ext4_error_inode((inode), __func__, __LINE__, 0, (fmt), ## a) + +#define EXT4_ERROR_INODE_BLOCK(inode, block, fmt, a...) \ + ext4_error_inode((inode), __func__, __LINE__, (block), (fmt), ## a) + +#define EXT4_ERROR_FILE(file, block, fmt, a...) \ + ext4_error_file((file), __func__, __LINE__, (block), (fmt), ## a) + #ifdef CONFIG_PRINTK #define ext4_error_inode(inode, func, line, block, fmt, ...) \ @@ -2390,6 +2430,8 @@ void __ext4_grp_locked_error(const char *, unsigned int, __ext4_abort(sb, __func__, __LINE__, fmt, ##__VA_ARGS__) #define ext4_warning(sb, fmt, ...) \ __ext4_warning(sb, __func__, __LINE__, fmt, ##__VA_ARGS__) +#define ext4_warning_inode(inode, fmt, ...) \ + __ext4_warning_inode(inode, __func__, __LINE__, fmt, ##__VA_ARGS__) #define ext4_msg(sb, level, fmt, ...) \ __ext4_msg(sb, level, fmt, ##__VA_ARGS__) #define dump_mmp_msg(sb, mmp, msg) \ @@ -2425,6 +2467,11 @@ do { \ no_printk(fmt, ##__VA_ARGS__); \ __ext4_warning(sb, "", 0, " "); \ } while (0) +#define ext4_warning_inode(inode, fmt, ...) \ +do { \ + no_printk(fmt, ##__VA_ARGS__); \ + __ext4_warning_inode(inode, "", 0, " "); \ +} while (0) #define ext4_msg(sb, level, fmt, ...) \ do { \ no_printk(fmt, ##__VA_ARGS__); \ @@ -2768,7 +2815,9 @@ extern int ext4_da_write_inline_data_begin(struct address_space *mapping, extern int ext4_da_write_inline_data_end(struct inode *inode, loff_t pos, unsigned len, unsigned copied, struct page *page); -extern int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, +extern int ext4_try_add_inline_entry(handle_t *handle, + struct ext4_filename *fname, + struct dentry *dentry, struct inode *inode); extern int ext4_try_create_inline_dir(handle_t *handle, struct inode *parent, @@ -2782,6 +2831,7 @@ extern int htree_inlinedir_to_tree(struct file *dir_file, __u32 start_hash, __u32 start_minor_hash, int *has_inline_data); extern struct buffer_head *ext4_find_inline_entry(struct inode *dir, + struct ext4_filename *fname, const struct qstr *d_name, struct ext4_dir_entry_2 **res_dir, int *has_inline_data); @@ -2890,7 +2940,6 @@ extern int ext4_map_blocks(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, int flags); extern int ext4_ext_calc_metadata_amount(struct inode *inode, ext4_lblk_t lblocks); -extern int ext4_extent_tree_init(handle_t *, struct inode *); extern int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int num, struct ext4_ext_path *path); @@ -2914,6 +2963,7 @@ extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); extern int ext4_ext_precache(struct inode *inode); extern int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len); +extern int ext4_insert_range(struct inode *inode, loff_t offset, loff_t len); extern int ext4_swap_extents(handle_t *handle, struct inode *inode1, struct inode *inode2, ext4_lblk_t lblk1, ext4_lblk_t lblk2, ext4_lblk_t count, diff --git a/fs/ext4/ext4_crypto.h b/fs/ext4/ext4_crypto.h index d75159c..ac7d4e8 100644 --- a/fs/ext4/ext4_crypto.h +++ b/fs/ext4/ext4_crypto.h @@ -66,24 +66,39 @@ struct ext4_encryption_context { #define EXT4_KEY_DESC_PREFIX "ext4:" #define EXT4_KEY_DESC_PREFIX_SIZE 5 +/* This is passed in from userspace into the kernel keyring */ struct ext4_encryption_key { - uint32_t mode; - char raw[EXT4_MAX_KEY_SIZE]; - uint32_t size; + __u32 mode; + char raw[EXT4_MAX_KEY_SIZE]; + __u32 size; +} __attribute__((__packed__)); + +struct ext4_crypt_info { + char ci_data_mode; + char ci_filename_mode; + char ci_flags; + struct crypto_ablkcipher *ci_ctfm; + struct key *ci_keyring_key; + char ci_master_key[EXT4_KEY_DESCRIPTOR_SIZE]; }; #define EXT4_CTX_REQUIRES_FREE_ENCRYPT_FL 0x00000001 -#define EXT4_BOUNCE_PAGE_REQUIRES_FREE_ENCRYPT_FL 0x00000002 +#define EXT4_WRITE_PATH_FL 0x00000002 struct ext4_crypto_ctx { - struct crypto_tfm *tfm; /* Crypto API context */ - struct page *bounce_page; /* Ciphertext page on write path */ - struct page *control_page; /* Original page on write path */ - struct bio *bio; /* The bio for this context */ - struct work_struct work; /* Work queue for read complete path */ - struct list_head free_list; /* Free list */ - int flags; /* Flags */ - int mode; /* Encryption mode for tfm */ + union { + struct { + struct page *bounce_page; /* Ciphertext page */ + struct page *control_page; /* Original page */ + } w; + struct { + struct bio *bio; + struct work_struct work; + } r; + struct list_head free_list; /* Free list */ + }; + char flags; /* Flags */ + char mode; /* Encryption mode for tfm */ }; struct ext4_completion_result { @@ -121,18 +136,6 @@ struct ext4_str { u32 len; }; -struct ext4_fname_crypto_ctx { - u32 lim; - char tmp_buf[EXT4_CRYPTO_BLOCK_SIZE]; - struct crypto_ablkcipher *ctfm; - struct crypto_hash *htfm; - struct page *workpage; - struct ext4_encryption_key key; - unsigned flags : 8; - unsigned has_valid_key : 1; - unsigned ctfm_key_is_ready : 1; -}; - /** * For encrypted symlinks, the ciphertext length is stored at the beginning * of the string in little-endian format. diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index 3445035..d418431 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -87,6 +87,12 @@ int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle) ext4_put_nojournal(handle); return 0; } + + if (!handle->h_transaction) { + err = jbd2_journal_stop(handle); + return handle->h_err ? handle->h_err : err; + } + sb = handle->h_transaction->t_journal->j_private; err = handle->h_err; rc = jbd2_journal_stop(handle); diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index d74e0802..aadb728 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -39,6 +39,7 @@ #include <linux/slab.h> #include <asm/uaccess.h> #include <linux/fiemap.h> +#include <linux/backing-dev.h> #include "ext4_jbd2.h" #include "ext4_extents.h" #include "xattr.h" @@ -377,7 +378,7 @@ static int ext4_valid_extent(struct inode *inode, struct ext4_extent *ext) ext4_lblk_t lblock = le32_to_cpu(ext->ee_block); ext4_lblk_t last = lblock + len - 1; - if (lblock > last) + if (len == 0 || lblock > last) return 0; return ext4_data_block_valid(EXT4_SB(inode->i_sb), block, len); } @@ -4456,6 +4457,8 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ar.flags |= EXT4_MB_HINT_NOPREALLOC; if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) ar.flags |= EXT4_MB_DELALLOC_RESERVED; + if (flags & EXT4_GET_BLOCKS_METADATA_NOFAIL) + ar.flags |= EXT4_MB_USE_RESERVED; newblock = ext4_mb_new_blocks(handle, &ar, &err); if (!newblock) goto out2; @@ -4663,6 +4666,7 @@ static int ext4_alloc_file_blocks(struct file *file, ext4_lblk_t offset, int ret = 0; int ret2 = 0; int retries = 0; + int depth = 0; struct ext4_map_blocks map; unsigned int credits; loff_t epos; @@ -4677,13 +4681,32 @@ static int ext4_alloc_file_blocks(struct file *file, ext4_lblk_t offset, if (len <= EXT_UNWRITTEN_MAX_LEN) flags |= EXT4_GET_BLOCKS_NO_NORMALIZE; + /* Wait all existing dio workers, newcomers will block on i_mutex */ + ext4_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); + /* * credits to insert 1 extent into extent tree */ credits = ext4_chunk_trans_blocks(inode, len); + /* + * We can only call ext_depth() on extent based inodes + */ + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + depth = ext_depth(inode); + else + depth = -1; retry: while (ret >= 0 && len) { + /* + * Recalculate credits when extent tree depth changes. + */ + if (depth >= 0 && depth != ext_depth(inode)) { + credits = ext4_chunk_trans_blocks(inode, len); + depth = ext_depth(inode); + } + handle = ext4_journal_start(inode, EXT4_HT_MAP_BLOCKS, credits); if (IS_ERR(handle)) { @@ -4725,6 +4748,8 @@ retry: goto retry; } + ext4_inode_resume_unlocked_dio(inode); + return ret > 0 ? ret2 : ret; } @@ -4912,12 +4937,14 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) * bug we should fix.... */ if (ext4_encrypted_inode(inode) && - (mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE))) + (mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_INSERT_RANGE | + FALLOC_FL_ZERO_RANGE))) return -EOPNOTSUPP; /* Return error if mode is not supported */ if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE | - FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE)) + FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE | + FALLOC_FL_INSERT_RANGE)) return -EOPNOTSUPP; if (mode & FALLOC_FL_PUNCH_HOLE) @@ -4930,6 +4957,9 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) if (mode & FALLOC_FL_COLLAPSE_RANGE) return ext4_collapse_range(inode, offset, len); + if (mode & FALLOC_FL_INSERT_RANGE) + return ext4_insert_range(inode, offset, len); + if (mode & FALLOC_FL_ZERO_RANGE) return ext4_zero_range(file, offset, len, mode); @@ -5224,13 +5254,13 @@ ext4_access_path(handle_t *handle, struct inode *inode, /* * ext4_ext_shift_path_extents: * Shift the extents of a path structure lying between path[depth].p_ext - * and EXT_LAST_EXTENT(path[depth].p_hdr) downwards, by subtracting shift - * from starting block for each extent. + * and EXT_LAST_EXTENT(path[depth].p_hdr), by @shift blocks. @SHIFT tells + * if it is right shift or left shift operation. */ static int ext4_ext_shift_path_extents(struct ext4_ext_path *path, ext4_lblk_t shift, struct inode *inode, handle_t *handle, - ext4_lblk_t *start) + enum SHIFT_DIRECTION SHIFT) { int depth, err = 0; struct ext4_extent *ex_start, *ex_last; @@ -5252,19 +5282,25 @@ ext4_ext_shift_path_extents(struct ext4_ext_path *path, ext4_lblk_t shift, if (ex_start == EXT_FIRST_EXTENT(path[depth].p_hdr)) update = 1; - *start = le32_to_cpu(ex_last->ee_block) + - ext4_ext_get_actual_len(ex_last); - while (ex_start <= ex_last) { - le32_add_cpu(&ex_start->ee_block, -shift); - /* Try to merge to the left. */ - if ((ex_start > - EXT_FIRST_EXTENT(path[depth].p_hdr)) && - ext4_ext_try_to_merge_right(inode, - path, ex_start - 1)) + if (SHIFT == SHIFT_LEFT) { + le32_add_cpu(&ex_start->ee_block, + -shift); + /* Try to merge to the left. */ + if ((ex_start > + EXT_FIRST_EXTENT(path[depth].p_hdr)) + && + ext4_ext_try_to_merge_right(inode, + path, ex_start - 1)) + ex_last--; + else + ex_start++; + } else { + le32_add_cpu(&ex_last->ee_block, shift); + ext4_ext_try_to_merge_right(inode, path, + ex_last); ex_last--; - else - ex_start++; + } } err = ext4_ext_dirty(handle, inode, path + depth); if (err) @@ -5279,7 +5315,10 @@ ext4_ext_shift_path_extents(struct ext4_ext_path *path, ext4_lblk_t shift, if (err) goto out; - le32_add_cpu(&path[depth].p_idx->ei_block, -shift); + if (SHIFT == SHIFT_LEFT) + le32_add_cpu(&path[depth].p_idx->ei_block, -shift); + else + le32_add_cpu(&path[depth].p_idx->ei_block, shift); err = ext4_ext_dirty(handle, inode, path + depth); if (err) goto out; @@ -5297,19 +5336,20 @@ out: /* * ext4_ext_shift_extents: - * All the extents which lies in the range from start to the last allocated - * block for the file are shifted downwards by shift blocks. + * All the extents which lies in the range from @start to the last allocated + * block for the @inode are shifted either towards left or right (depending + * upon @SHIFT) by @shift blocks. * On success, 0 is returned, error otherwise. */ static int ext4_ext_shift_extents(struct inode *inode, handle_t *handle, - ext4_lblk_t start, ext4_lblk_t shift) + ext4_lblk_t start, ext4_lblk_t shift, + enum SHIFT_DIRECTION SHIFT) { struct ext4_ext_path *path; int ret = 0, depth; struct ext4_extent *extent; - ext4_lblk_t stop_block; - ext4_lblk_t ex_start, ex_end; + ext4_lblk_t stop, *iterator, ex_start, ex_end; /* Let path point to the last extent */ path = ext4_find_extent(inode, EXT_MAX_BLOCKS - 1, NULL, 0); @@ -5321,58 +5361,84 @@ ext4_ext_shift_extents(struct inode *inode, handle_t *handle, if (!extent) goto out; - stop_block = le32_to_cpu(extent->ee_block) + + stop = le32_to_cpu(extent->ee_block) + ext4_ext_get_actual_len(extent); - /* Nothing to shift, if hole is at the end of file */ - if (start >= stop_block) - goto out; + /* + * In case of left shift, Don't start shifting extents until we make + * sure the hole is big enough to accommodate the shift. + */ + if (SHIFT == SHIFT_LEFT) { + path = ext4_find_extent(inode, start - 1, &path, 0); + if (IS_ERR(path)) + return PTR_ERR(path); + depth = path->p_depth; + extent = path[depth].p_ext; + if (extent) { + ex_start = le32_to_cpu(extent->ee_block); + ex_end = le32_to_cpu(extent->ee_block) + + ext4_ext_get_actual_len(extent); + } else { + ex_start = 0; + ex_end = 0; + } - /* - * Don't start shifting extents until we make sure the hole is big - * enough to accomodate the shift. - */ - path = ext4_find_extent(inode, start - 1, &path, 0); - if (IS_ERR(path)) - return PTR_ERR(path); - depth = path->p_depth; - extent = path[depth].p_ext; - if (extent) { - ex_start = le32_to_cpu(extent->ee_block); - ex_end = le32_to_cpu(extent->ee_block) + - ext4_ext_get_actual_len(extent); - } else { - ex_start = 0; - ex_end = 0; + if ((start == ex_start && shift > ex_start) || + (shift > start - ex_end)) { + ext4_ext_drop_refs(path); + kfree(path); + return -EINVAL; + } } - if ((start == ex_start && shift > ex_start) || - (shift > start - ex_end)) - return -EINVAL; + /* + * In case of left shift, iterator points to start and it is increased + * till we reach stop. In case of right shift, iterator points to stop + * and it is decreased till we reach start. + */ + if (SHIFT == SHIFT_LEFT) + iterator = &start; + else + iterator = &stop; /* Its safe to start updating extents */ - while (start < stop_block) { - path = ext4_find_extent(inode, start, &path, 0); + while (start < stop) { + path = ext4_find_extent(inode, *iterator, &path, 0); if (IS_ERR(path)) return PTR_ERR(path); depth = path->p_depth; extent = path[depth].p_ext; if (!extent) { EXT4_ERROR_INODE(inode, "unexpected hole at %lu", - (unsigned long) start); + (unsigned long) *iterator); return -EIO; } - if (start > le32_to_cpu(extent->ee_block)) { + if (SHIFT == SHIFT_LEFT && *iterator > + le32_to_cpu(extent->ee_block)) { /* Hole, move to the next extent */ if (extent < EXT_LAST_EXTENT(path[depth].p_hdr)) { path[depth].p_ext++; } else { - start = ext4_ext_next_allocated_block(path); + *iterator = ext4_ext_next_allocated_block(path); continue; } } + + if (SHIFT == SHIFT_LEFT) { + extent = EXT_LAST_EXTENT(path[depth].p_hdr); + *iterator = le32_to_cpu(extent->ee_block) + + ext4_ext_get_actual_len(extent); + } else { + extent = EXT_FIRST_EXTENT(path[depth].p_hdr); + *iterator = le32_to_cpu(extent->ee_block) > 0 ? + le32_to_cpu(extent->ee_block) - 1 : 0; + /* Update path extent in case we need to stop */ + while (le32_to_cpu(extent->ee_block) < start) + extent++; + path[depth].p_ext = extent; + } ret = ext4_ext_shift_path_extents(path, shift, inode, - handle, &start); + handle, SHIFT); if (ret) break; } @@ -5396,6 +5462,14 @@ int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) loff_t new_size, ioffset; int ret; + /* + * We need to test this early because xfstests assumes that a + * collapse range of (0, 1) will return EOPNOTSUPP if the file + * system does not support collapse range. + */ + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + return -EOPNOTSUPP; + /* Collapse range works only on fs block size aligned offsets. */ if (offset & (EXT4_CLUSTER_SIZE(sb) - 1) || len & (EXT4_CLUSTER_SIZE(sb) - 1)) @@ -5477,7 +5551,7 @@ int ext4_collapse_range(struct inode *inode, loff_t offset, loff_t len) ext4_discard_preallocations(inode); ret = ext4_ext_shift_extents(inode, handle, punch_stop, - punch_stop - punch_start); + punch_stop - punch_start, SHIFT_LEFT); if (ret) { up_write(&EXT4_I(inode)->i_data_sem); goto out_stop; @@ -5502,6 +5576,174 @@ out_mutex: return ret; } +/* + * ext4_insert_range: + * This function implements the FALLOC_FL_INSERT_RANGE flag of fallocate. + * The data blocks starting from @offset to the EOF are shifted by @len + * towards right to create a hole in the @inode. Inode size is increased + * by len bytes. + * Returns 0 on success, error otherwise. + */ +int ext4_insert_range(struct inode *inode, loff_t offset, loff_t len) +{ + struct super_block *sb = inode->i_sb; + handle_t *handle; + struct ext4_ext_path *path; + struct ext4_extent *extent; + ext4_lblk_t offset_lblk, len_lblk, ee_start_lblk = 0; + unsigned int credits, ee_len; + int ret = 0, depth, split_flag = 0; + loff_t ioffset; + + /* + * We need to test this early because xfstests assumes that an + * insert range of (0, 1) will return EOPNOTSUPP if the file + * system does not support insert range. + */ + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + return -EOPNOTSUPP; + + /* Insert range works only on fs block size aligned offsets. */ + if (offset & (EXT4_CLUSTER_SIZE(sb) - 1) || + len & (EXT4_CLUSTER_SIZE(sb) - 1)) + return -EINVAL; + + if (!S_ISREG(inode->i_mode)) + return -EOPNOTSUPP; + + trace_ext4_insert_range(inode, offset, len); + + offset_lblk = offset >> EXT4_BLOCK_SIZE_BITS(sb); + len_lblk = len >> EXT4_BLOCK_SIZE_BITS(sb); + + /* Call ext4_force_commit to flush all data in case of data=journal */ + if (ext4_should_journal_data(inode)) { + ret = ext4_force_commit(inode->i_sb); + if (ret) + return ret; + } + + /* + * Need to round down to align start offset to page size boundary + * for page size > block size. + */ + ioffset = round_down(offset, PAGE_SIZE); + + /* Write out all dirty pages */ + ret = filemap_write_and_wait_range(inode->i_mapping, ioffset, + LLONG_MAX); + if (ret) + return ret; + + /* Take mutex lock */ + mutex_lock(&inode->i_mutex); + + /* Currently just for extent based files */ + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + ret = -EOPNOTSUPP; + goto out_mutex; + } + + /* Check for wrap through zero */ + if (inode->i_size + len > inode->i_sb->s_maxbytes) { + ret = -EFBIG; + goto out_mutex; + } + + /* Offset should be less than i_size */ + if (offset >= i_size_read(inode)) { + ret = -EINVAL; + goto out_mutex; + } + + truncate_pagecache(inode, ioffset); + + /* Wait for existing dio to complete */ + ext4_inode_block_unlocked_dio(inode); + inode_dio_wait(inode); + + credits = ext4_writepage_trans_blocks(inode); + handle = ext4_journal_start(inode, EXT4_HT_TRUNCATE, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + goto out_dio; + } + + /* Expand file to avoid data loss if there is error while shifting */ + inode->i_size += len; + EXT4_I(inode)->i_disksize += len; + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + ret = ext4_mark_inode_dirty(handle, inode); + if (ret) + goto out_stop; + + down_write(&EXT4_I(inode)->i_data_sem); + ext4_discard_preallocations(inode); + + path = ext4_find_extent(inode, offset_lblk, NULL, 0); + if (IS_ERR(path)) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_stop; + } + + depth = ext_depth(inode); + extent = path[depth].p_ext; + if (extent) { + ee_start_lblk = le32_to_cpu(extent->ee_block); + ee_len = ext4_ext_get_actual_len(extent); + + /* + * If offset_lblk is not the starting block of extent, split + * the extent @offset_lblk + */ + if ((offset_lblk > ee_start_lblk) && + (offset_lblk < (ee_start_lblk + ee_len))) { + if (ext4_ext_is_unwritten(extent)) + split_flag = EXT4_EXT_MARK_UNWRIT1 | + EXT4_EXT_MARK_UNWRIT2; + ret = ext4_split_extent_at(handle, inode, &path, + offset_lblk, split_flag, + EXT4_EX_NOCACHE | + EXT4_GET_BLOCKS_PRE_IO | + EXT4_GET_BLOCKS_METADATA_NOFAIL); + } + + ext4_ext_drop_refs(path); + kfree(path); + if (ret < 0) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_stop; + } + } + + ret = ext4_es_remove_extent(inode, offset_lblk, + EXT_MAX_BLOCKS - offset_lblk); + if (ret) { + up_write(&EXT4_I(inode)->i_data_sem); + goto out_stop; + } + + /* + * if offset_lblk lies in a hole which is at start of file, use + * ee_start_lblk to shift extents + */ + ret = ext4_ext_shift_extents(inode, handle, + ee_start_lblk > offset_lblk ? ee_start_lblk : offset_lblk, + len_lblk, SHIFT_RIGHT); + + up_write(&EXT4_I(inode)->i_data_sem); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + +out_stop: + ext4_journal_stop(handle); +out_dio: + ext4_inode_resume_unlocked_dio(inode); +out_mutex: + mutex_unlock(&inode->i_mutex); + return ret; +} + /** * ext4_swap_extents - Swap extents between two inodes * @@ -5534,7 +5776,7 @@ ext4_swap_extents(handle_t *handle, struct inode *inode1, BUG_ON(!rwsem_is_locked(&EXT4_I(inode1)->i_data_sem)); BUG_ON(!rwsem_is_locked(&EXT4_I(inode2)->i_data_sem)); BUG_ON(!mutex_is_locked(&inode1->i_mutex)); - BUG_ON(!mutex_is_locked(&inode1->i_mutex)); + BUG_ON(!mutex_is_locked(&inode2->i_mutex)); *erp = ext4_es_remove_extent(inode1, lblk1, count); if (unlikely(*erp)) diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 0613c25..bc313ac 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -192,15 +192,27 @@ out: } #ifdef CONFIG_FS_DAX +static void ext4_end_io_unwritten(struct buffer_head *bh, int uptodate) +{ + struct inode *inode = bh->b_assoc_map->host; + /* XXX: breaks on 32-bit > 16GB. Is that even supported? */ + loff_t offset = (loff_t)(uintptr_t)bh->b_private << inode->i_blkbits; + int err; + if (!uptodate) + return; + WARN_ON(!buffer_unwritten(bh)); + err = ext4_convert_unwritten_extents(NULL, inode, offset, bh->b_size); +} + static int ext4_dax_fault(struct vm_area_struct *vma, struct vm_fault *vmf) { - return dax_fault(vma, vmf, ext4_get_block); + return dax_fault(vma, vmf, ext4_get_block, ext4_end_io_unwritten); /* Is this the right get_block? */ } static int ext4_dax_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) { - return dax_mkwrite(vma, vmf, ext4_get_block); + return dax_mkwrite(vma, vmf, ext4_get_block, ext4_end_io_unwritten); } static const struct vm_operations_struct ext4_dax_vm_ops = { @@ -223,9 +235,11 @@ static int ext4_file_mmap(struct file *file, struct vm_area_struct *vma) struct inode *inode = file->f_mapping->host; if (ext4_encrypted_inode(inode)) { - int err = ext4_generate_encryption_key(inode); + int err = ext4_get_encryption_info(inode); if (err) return 0; + if (ext4_encryption_info(inode) == NULL) + return -ENOKEY; } file_accessed(file); if (IS_DAX(file_inode(file))) { @@ -278,6 +292,13 @@ static int ext4_file_open(struct inode * inode, struct file * filp) ext4_journal_stop(handle); } } + if (ext4_encrypted_inode(inode)) { + ret = ext4_get_encryption_info(inode); + if (ret) + return -EACCES; + if (ext4_encryption_info(inode) == NULL) + return -ENOKEY; + } /* * Set up the jbd2_inode if we are opening the inode for * writing and the journal is present @@ -287,13 +308,7 @@ static int ext4_file_open(struct inode * inode, struct file * filp) if (ret < 0) return ret; } - ret = dquot_file_open(inode, filp); - if (!ret && ext4_encrypted_inode(inode)) { - ret = ext4_generate_encryption_key(inode); - if (ret) - ret = -EACCES; - } - return ret; + return dquot_file_open(inode, filp); } /* diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 1eaa6cb..173c1ae 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -726,11 +726,25 @@ struct inode *__ext4_new_inode(handle_t *handle, struct inode *dir, ext4_group_t i; ext4_group_t flex_group; struct ext4_group_info *grp; + int encrypt = 0; /* Cannot create files in a deleted directory */ if (!dir || !dir->i_nlink) return ERR_PTR(-EPERM); + if ((ext4_encrypted_inode(dir) || + DUMMY_ENCRYPTION_ENABLED(EXT4_SB(dir->i_sb))) && + (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode))) { + err = ext4_get_encryption_info(dir); + if (err) + return ERR_PTR(err); + if (ext4_encryption_info(dir) == NULL) + return ERR_PTR(-EPERM); + if (!handle) + nblocks += EXT4_DATA_TRANS_BLOCKS(dir->i_sb); + encrypt = 1; + } + sb = dir->i_sb; ngroups = ext4_get_groups_count(sb); trace_ext4_request_inode(dir, mode); @@ -996,12 +1010,6 @@ got: ei->i_block_group = group; ei->i_last_alloc_group = ~0; - /* If the directory encrypted, then we should encrypt the inode. */ - if ((S_ISDIR(mode) || S_ISREG(mode) || S_ISLNK(mode)) && - (ext4_encrypted_inode(dir) || - DUMMY_ENCRYPTION_ENABLED(sbi))) - ext4_set_inode_flag(inode, EXT4_INODE_ENCRYPT); - ext4_set_inode_flags(inode); if (IS_DIRSYNC(inode)) ext4_handle_sync(handle); @@ -1034,28 +1042,9 @@ got: ext4_set_inode_state(inode, EXT4_STATE_NEW); ei->i_extra_isize = EXT4_SB(sb)->s_want_extra_isize; -#ifdef CONFIG_EXT4_FS_ENCRYPTION - if ((sbi->s_file_encryption_mode == EXT4_ENCRYPTION_MODE_INVALID) && - (sbi->s_dir_encryption_mode == EXT4_ENCRYPTION_MODE_INVALID)) { - ei->i_inline_off = 0; - if (EXT4_HAS_INCOMPAT_FEATURE(sb, - EXT4_FEATURE_INCOMPAT_INLINE_DATA)) - ext4_set_inode_state(inode, - EXT4_STATE_MAY_INLINE_DATA); - } else { - /* Inline data and encryption are incompatible - * We turn off inline data since encryption is enabled */ - ei->i_inline_off = 1; - if (EXT4_HAS_INCOMPAT_FEATURE(sb, - EXT4_FEATURE_INCOMPAT_INLINE_DATA)) - ext4_clear_inode_state(inode, - EXT4_STATE_MAY_INLINE_DATA); - } -#else ei->i_inline_off = 0; if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_INLINE_DATA)) ext4_set_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); -#endif ret = inode; err = dquot_alloc_inode(inode); if (err) @@ -1082,6 +1071,12 @@ got: ei->i_datasync_tid = handle->h_transaction->t_tid; } + if (encrypt) { + err = ext4_inherit_context(dir, inode); + if (err) + goto fail_free_drop; + } + err = ext4_mark_inode_dirty(handle, inode); if (err) { ext4_std_error(sb, err); diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 9588240..4f6ac49 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -565,7 +565,7 @@ int ext4_ind_map_blocks(handle_t *handle, struct inode *inode, EXT4_FEATURE_RO_COMPAT_BIGALLOC)) { EXT4_ERROR_INODE(inode, "Can't allocate blocks for " "non-extent mapped inodes with bigalloc"); - return -ENOSPC; + return -EUCLEAN; } /* Set up for the direct block allocation */ @@ -576,6 +576,8 @@ int ext4_ind_map_blocks(handle_t *handle, struct inode *inode, ar.flags = EXT4_MB_HINT_DATA; if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) ar.flags |= EXT4_MB_DELALLOC_RESERVED; + if (flags & EXT4_GET_BLOCKS_METADATA_NOFAIL) + ar.flags |= EXT4_MB_USE_RESERVED; ar.goal = ext4_find_goal(inode, map->m_lblk, partial); diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c index 095c7a2..cd944a7 100644 --- a/fs/ext4/inline.c +++ b/fs/ext4/inline.c @@ -995,20 +995,18 @@ void ext4_show_inline_dir(struct inode *dir, struct buffer_head *bh, * and -EEXIST if directory entry already exists. */ static int ext4_add_dirent_to_inline(handle_t *handle, + struct ext4_filename *fname, struct dentry *dentry, struct inode *inode, struct ext4_iloc *iloc, void *inline_start, int inline_size) { struct inode *dir = d_inode(dentry->d_parent); - const char *name = dentry->d_name.name; - int namelen = dentry->d_name.len; int err; struct ext4_dir_entry_2 *de; - err = ext4_find_dest_de(dir, inode, iloc->bh, - inline_start, inline_size, - name, namelen, &de); + err = ext4_find_dest_de(dir, inode, iloc->bh, inline_start, + inline_size, fname, &de); if (err) return err; @@ -1016,8 +1014,7 @@ static int ext4_add_dirent_to_inline(handle_t *handle, err = ext4_journal_get_write_access(handle, iloc->bh); if (err) return err; - ext4_insert_dentry(dir, inode, de, inline_size, &dentry->d_name, - name, namelen); + ext4_insert_dentry(dir, inode, de, inline_size, fname); ext4_show_inline_dir(dir, iloc->bh, inline_start, inline_size); @@ -1248,8 +1245,8 @@ out: * If succeeds, return 0. If not, extended the inline dir and copied data to * the new created block. */ -int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, - struct inode *inode) +int ext4_try_add_inline_entry(handle_t *handle, struct ext4_filename *fname, + struct dentry *dentry, struct inode *inode) { int ret, inline_size; void *inline_start; @@ -1268,7 +1265,7 @@ int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, EXT4_INLINE_DOTDOT_SIZE; inline_size = EXT4_MIN_INLINE_DATA_SIZE - EXT4_INLINE_DOTDOT_SIZE; - ret = ext4_add_dirent_to_inline(handle, dentry, inode, &iloc, + ret = ext4_add_dirent_to_inline(handle, fname, dentry, inode, &iloc, inline_start, inline_size); if (ret != -ENOSPC) goto out; @@ -1289,8 +1286,9 @@ int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, if (inline_size) { inline_start = ext4_get_inline_xattr_pos(dir, &iloc); - ret = ext4_add_dirent_to_inline(handle, dentry, inode, &iloc, - inline_start, inline_size); + ret = ext4_add_dirent_to_inline(handle, fname, dentry, + inode, &iloc, inline_start, + inline_size); if (ret != -ENOSPC) goto out; @@ -1611,6 +1609,7 @@ out: } struct buffer_head *ext4_find_inline_entry(struct inode *dir, + struct ext4_filename *fname, const struct qstr *d_name, struct ext4_dir_entry_2 **res_dir, int *has_inline_data) @@ -1632,8 +1631,8 @@ struct buffer_head *ext4_find_inline_entry(struct inode *dir, inline_start = (void *)ext4_raw_inode(&iloc)->i_block + EXT4_INLINE_DOTDOT_SIZE; inline_size = EXT4_MIN_INLINE_DATA_SIZE - EXT4_INLINE_DOTDOT_SIZE; - ret = search_dir(iloc.bh, inline_start, inline_size, - dir, d_name, 0, res_dir); + ret = ext4_search_dir(iloc.bh, inline_start, inline_size, + dir, fname, d_name, 0, res_dir); if (ret == 1) goto out_find; if (ret < 0) @@ -1645,8 +1644,8 @@ struct buffer_head *ext4_find_inline_entry(struct inode *dir, inline_start = ext4_get_inline_xattr_pos(dir, &iloc); inline_size = ext4_get_inline_size(dir) - EXT4_MIN_INLINE_DATA_SIZE; - ret = search_dir(iloc.bh, inline_start, inline_size, - dir, d_name, 0, res_dir); + ret = ext4_search_dir(iloc.bh, inline_start, inline_size, + dir, fname, d_name, 0, res_dir); if (ret == 1) goto out_find; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 066fdd7..41f8e55 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -656,18 +656,6 @@ has_zeroout: return retval; } -static void ext4_end_io_unwritten(struct buffer_head *bh, int uptodate) -{ - struct inode *inode = bh->b_assoc_map->host; - /* XXX: breaks on 32-bit > 16GB. Is that even supported? */ - loff_t offset = (loff_t)(uintptr_t)bh->b_private << inode->i_blkbits; - int err; - if (!uptodate) - return; - WARN_ON(!buffer_unwritten(bh)); - err = ext4_convert_unwritten_extents(NULL, inode, offset, bh->b_size); -} - /* Maximum number of blocks we map for direct IO at once. */ #define DIO_MAX_BLOCKS 4096 @@ -705,10 +693,15 @@ static int _ext4_get_block(struct inode *inode, sector_t iblock, map_bh(bh, inode->i_sb, map.m_pblk); bh->b_state = (bh->b_state & ~EXT4_MAP_FLAGS) | map.m_flags; - if (IS_DAX(inode) && buffer_unwritten(bh) && !io_end) { + if (IS_DAX(inode) && buffer_unwritten(bh)) { + /* + * dgc: I suspect unwritten conversion on ext4+DAX is + * fundamentally broken here when there are concurrent + * read/write in progress on this inode. + */ + WARN_ON_ONCE(io_end); bh->b_assoc_map = inode->i_mapping; bh->b_private = (void *)(unsigned long)iblock; - bh->b_end_io = ext4_end_io_unwritten; } if (io_end && io_end->flag & EXT4_IO_END_UNWRITTEN) set_buffer_defer_completion(bh); @@ -731,18 +724,18 @@ int ext4_get_block(struct inode *inode, sector_t iblock, * `handle' can be NULL if create is zero */ struct buffer_head *ext4_getblk(handle_t *handle, struct inode *inode, - ext4_lblk_t block, int create) + ext4_lblk_t block, int map_flags) { struct ext4_map_blocks map; struct buffer_head *bh; + int create = map_flags & EXT4_GET_BLOCKS_CREATE; int err; J_ASSERT(handle != NULL || create == 0); map.m_lblk = block; map.m_len = 1; - err = ext4_map_blocks(handle, inode, &map, - create ? EXT4_GET_BLOCKS_CREATE : 0); + err = ext4_map_blocks(handle, inode, &map, map_flags); if (err == 0) return create ? ERR_PTR(-ENOSPC) : NULL; @@ -788,11 +781,11 @@ errout: } struct buffer_head *ext4_bread(handle_t *handle, struct inode *inode, - ext4_lblk_t block, int create) + ext4_lblk_t block, int map_flags) { struct buffer_head *bh; - bh = ext4_getblk(handle, inode, block, create); + bh = ext4_getblk(handle, inode, block, map_flags); if (IS_ERR(bh)) return bh; if (!bh || buffer_uptodate(bh)) @@ -1261,13 +1254,12 @@ static int ext4_journalled_write_end(struct file *file, } /* - * Reserve a single cluster located at lblock + * Reserve space for a single cluster */ -static int ext4_da_reserve_space(struct inode *inode, ext4_lblk_t lblock) +static int ext4_da_reserve_space(struct inode *inode) { struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); struct ext4_inode_info *ei = EXT4_I(inode); - unsigned int md_needed; int ret; /* @@ -1279,25 +1271,14 @@ static int ext4_da_reserve_space(struct inode *inode, ext4_lblk_t lblock) if (ret) return ret; - /* - * recalculate the amount of metadata blocks to reserve - * in order to allocate nrblocks - * worse case is one extent per block - */ spin_lock(&ei->i_block_reservation_lock); - /* - * ext4_calc_metadata_amount() has side effects, which we have - * to be prepared undo if we fail to claim space. - */ - md_needed = 0; - trace_ext4_da_reserve_space(inode, 0); - if (ext4_claim_free_clusters(sbi, 1, 0)) { spin_unlock(&ei->i_block_reservation_lock); dquot_release_reservation_block(inode, EXT4_C2B(sbi, 1)); return -ENOSPC; } ei->i_reserved_data_blocks++; + trace_ext4_da_reserve_space(inode); spin_unlock(&ei->i_block_reservation_lock); return 0; /* success */ @@ -1566,9 +1547,9 @@ add_delayed: * then we don't need to reserve it again. However we still need * to reserve metadata for every block we're going to write. */ - if (EXT4_SB(inode->i_sb)->s_cluster_ratio <= 1 || + if (EXT4_SB(inode->i_sb)->s_cluster_ratio == 1 || !ext4_find_delalloc_cluster(inode, map->m_lblk)) { - ret = ext4_da_reserve_space(inode, iblock); + ret = ext4_da_reserve_space(inode); if (ret) { /* not enough space to reserve */ retval = ret; @@ -1701,19 +1682,32 @@ static int __ext4_journalled_writepage(struct page *page, ext4_walk_page_buffers(handle, page_bufs, 0, len, NULL, bget_one); } - /* As soon as we unlock the page, it can go away, but we have - * references to buffers so we are safe */ + /* + * We need to release the page lock before we start the + * journal, so grab a reference so the page won't disappear + * out from under us. + */ + get_page(page); unlock_page(page); handle = ext4_journal_start(inode, EXT4_HT_WRITE_PAGE, ext4_writepage_trans_blocks(inode)); if (IS_ERR(handle)) { ret = PTR_ERR(handle); - goto out; + put_page(page); + goto out_no_pagelock; } - BUG_ON(!ext4_handle_valid(handle)); + lock_page(page); + put_page(page); + if (page->mapping != mapping) { + /* The page got truncated from under us */ + ext4_journal_stop(handle); + ret = 0; + goto out; + } + if (inline_data) { BUFFER_TRACE(inode_bh, "get write access"); ret = ext4_journal_get_write_access(handle, inode_bh); @@ -1739,6 +1733,8 @@ static int __ext4_journalled_writepage(struct page *page, NULL, bput_one); ext4_set_inode_state(inode, EXT4_STATE_JDATA); out: + unlock_page(page); +out_no_pagelock: brelse(inode_bh); return ret; } @@ -4348,7 +4344,7 @@ static void ext4_update_other_inodes_time(struct super_block *sb, int inode_size = EXT4_INODE_SIZE(sb); oi.orig_ino = orig_ino; - ino = orig_ino & ~(inodes_per_block - 1); + ino = (orig_ino & ~(inodes_per_block - 1)) + 1; for (i = 0; i < inodes_per_block; i++, ino++, buf += inode_size) { if (ino == orig_ino) continue; @@ -4681,8 +4677,10 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) ext4_journal_stop(handle); } - if (attr->ia_valid & ATTR_SIZE && attr->ia_size != inode->i_size) { + if (attr->ia_valid & ATTR_SIZE) { handle_t *handle; + loff_t oldsize = inode->i_size; + int shrink = (attr->ia_size <= inode->i_size); if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) { struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); @@ -4690,24 +4688,26 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) if (attr->ia_size > sbi->s_bitmap_maxbytes) return -EFBIG; } + if (!S_ISREG(inode->i_mode)) + return -EINVAL; if (IS_I_VERSION(inode) && attr->ia_size != inode->i_size) inode_inc_iversion(inode); - if (S_ISREG(inode->i_mode) && + if (ext4_should_order_data(inode) && (attr->ia_size < inode->i_size)) { - if (ext4_should_order_data(inode)) { - error = ext4_begin_ordered_truncate(inode, + error = ext4_begin_ordered_truncate(inode, attr->ia_size); - if (error) - goto err_out; - } + if (error) + goto err_out; + } + if (attr->ia_size != inode->i_size) { handle = ext4_journal_start(inode, EXT4_HT_INODE, 3); if (IS_ERR(handle)) { error = PTR_ERR(handle); goto err_out; } - if (ext4_handle_valid(handle)) { + if (ext4_handle_valid(handle) && shrink) { error = ext4_orphan_add(handle, inode); orphan = 1; } @@ -4726,15 +4726,13 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) up_write(&EXT4_I(inode)->i_data_sem); ext4_journal_stop(handle); if (error) { - ext4_orphan_del(NULL, inode); + if (orphan) + ext4_orphan_del(NULL, inode); goto err_out; } - } else { - loff_t oldsize = inode->i_size; - - i_size_write(inode, attr->ia_size); - pagecache_isize_extended(inode, oldsize, inode->i_size); } + if (!shrink) + pagecache_isize_extended(inode, oldsize, inode->i_size); /* * Blocks are going to be removed from the inode. Wait @@ -4754,13 +4752,9 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) * in data=journal mode to make pages freeable. */ truncate_pagecache(inode, inode->i_size); + if (shrink) + ext4_truncate(inode); } - /* - * We want to call ext4_truncate() even if attr->ia_size == - * inode->i_size for cases like truncation of fallocated space - */ - if (attr->ia_valid & ATTR_SIZE) - ext4_truncate(inode); if (!rc) { setattr_copy(inode, attr); diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 2cb9e17..cb84512 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -31,14 +31,11 @@ static void memswap(void *a, void *b, size_t len) { unsigned char *ap, *bp; - unsigned char tmp; ap = (unsigned char *)a; bp = (unsigned char *)b; while (len-- > 0) { - tmp = *ap; - *ap = *bp; - *bp = tmp; + swap(*ap, *bp); ap++; bp++; } @@ -675,8 +672,8 @@ encryption_policy_out: if (err) return err; } - if (copy_to_user((void *) arg, sbi->s_es->s_encrypt_pw_salt, - 16)) + if (copy_to_user((void __user *) arg, + sbi->s_es->s_encrypt_pw_salt, 16)) return -EFAULT; return 0; } @@ -690,7 +687,7 @@ encryption_policy_out: err = ext4_get_policy(inode, &policy); if (err) return err; - if (copy_to_user((void *)arg, &policy, sizeof(policy))) + if (copy_to_user((void __user *)arg, &policy, sizeof(policy))) return -EFAULT; return 0; #else diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 8d1e602..f6aedf8 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -26,6 +26,7 @@ #include <linux/log2.h> #include <linux/module.h> #include <linux/slab.h> +#include <linux/backing-dev.h> #include <trace/events/ext4.h> #ifdef CONFIG_EXT4_DEBUG @@ -882,10 +883,8 @@ static int ext4_mb_init_cache(struct page *page, char *incore) /* wait for I/O completion */ for (i = 0, group = first_group; i < groups_per_page; i++, group++) { - if (bh[i] && ext4_wait_block_bitmap(sb, group, bh[i])) { + if (bh[i] && ext4_wait_block_bitmap(sb, group, bh[i])) err = -EIO; - goto out; - } } first_block = page->index * blocks_per_page; @@ -898,6 +897,11 @@ static int ext4_mb_init_cache(struct page *page, char *incore) /* skip initialized uptodate buddy */ continue; + if (!buffer_verified(bh[group - first_group])) + /* Skip faulty bitmaps */ + continue; + err = 0; + /* * data carry information regarding this * particular group in the format specified @@ -2008,7 +2012,12 @@ void ext4_mb_scan_aligned(struct ext4_allocation_context *ac, } } -/* This is now called BEFORE we load the buddy bitmap. */ +/* + * This is now called BEFORE we load the buddy bitmap. + * Returns either 1 or 0 indicating that the group is either suitable + * for the allocation or not. In addition it can also return negative + * error code when something goes wrong. + */ static int ext4_mb_good_group(struct ext4_allocation_context *ac, ext4_group_t group, int cr) { @@ -2031,7 +2040,7 @@ static int ext4_mb_good_group(struct ext4_allocation_context *ac, if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) { int ret = ext4_mb_init_group(ac->ac_sb, group); if (ret) - return 0; + return ret; } fragments = grp->bb_fragments; @@ -2078,7 +2087,7 @@ ext4_mb_regular_allocator(struct ext4_allocation_context *ac) { ext4_group_t ngroups, group, i; int cr; - int err = 0; + int err = 0, first_err = 0; struct ext4_sb_info *sbi; struct super_block *sb; struct ext4_buddy e4b; @@ -2145,6 +2154,7 @@ repeat: group = ac->ac_g_ex.fe_group; for (i = 0; i < ngroups; group++, i++) { + int ret = 0; cond_resched(); /* * Artificially restricted ngroups for non-extent @@ -2154,8 +2164,12 @@ repeat: group = 0; /* This now checks without needing the buddy page */ - if (!ext4_mb_good_group(ac, group, cr)) + ret = ext4_mb_good_group(ac, group, cr); + if (ret <= 0) { + if (!first_err) + first_err = ret; continue; + } err = ext4_mb_load_buddy(sb, group, &e4b); if (err) @@ -2167,9 +2181,12 @@ repeat: * We need to check again after locking the * block group */ - if (!ext4_mb_good_group(ac, group, cr)) { + ret = ext4_mb_good_group(ac, group, cr); + if (ret <= 0) { ext4_unlock_group(sb, group); ext4_mb_unload_buddy(&e4b); + if (!first_err) + first_err = ret; continue; } @@ -2216,6 +2233,8 @@ repeat: } } out: + if (!err && ac->ac_status != AC_STATUS_FOUND && first_err) + err = first_err; return err; } @@ -2257,12 +2276,9 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v) group--; if (group == 0) - seq_printf(seq, "#%-5s: %-5s %-5s %-5s " - "[ %-5s %-5s %-5s %-5s %-5s %-5s %-5s " - "%-5s %-5s %-5s %-5s %-5s %-5s %-5s ]\n", - "group", "free", "frags", "first", - "2^0", "2^1", "2^2", "2^3", "2^4", "2^5", "2^6", - "2^7", "2^8", "2^9", "2^10", "2^11", "2^12", "2^13"); + seq_puts(seq, "#group: free frags first [" + " 2^0 2^1 2^2 2^3 2^4 2^5 2^6 " + " 2^7 2^8 2^9 2^10 2^11 2^12 2^13 ]"); i = (sb->s_blocksize_bits + 2) * sizeof(sg.info.bb_counters[0]) + sizeof(struct ext4_group_info); diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index 370420b..fb6f117 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -166,12 +166,9 @@ mext_page_double_lock(struct inode *inode1, struct inode *inode2, */ wait_on_page_writeback(page[0]); wait_on_page_writeback(page[1]); - if (inode1 > inode2) { - struct page *tmp; - tmp = page[0]; - page[0] = page[1]; - page[1] = tmp; - } + if (inode1 > inode2) + swap(page[0], page[1]); + return 0; } @@ -574,12 +571,16 @@ ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 orig_blk, orig_inode->i_ino, donor_inode->i_ino); return -EINVAL; } - /* TODO: This is non obvious task to swap blocks for inodes with full - jornaling enabled */ + + /* TODO: it's not obvious how to swap blocks for inodes with full + journaling enabled */ if (ext4_should_journal_data(orig_inode) || ext4_should_journal_data(donor_inode)) { - return -EINVAL; + ext4_msg(orig_inode->i_sb, KERN_ERR, + "Online defrag not supported with data journaling"); + return -EOPNOTSUPP; } + /* Protect orig and donor inodes against a truncate */ lock_two_nondirectories(orig_inode, donor_inode); diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 5fdb9f6a..011dcfb 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -61,7 +61,7 @@ static struct buffer_head *ext4_append(handle_t *handle, *block = inode->i_size >> inode->i_sb->s_blocksize_bits; - bh = ext4_bread(handle, inode, *block, 1); + bh = ext4_bread(handle, inode, *block, EXT4_GET_BLOCKS_CREATE); if (IS_ERR(bh)) return bh; inode->i_size += inode->i_sb->s_blocksize; @@ -84,12 +84,13 @@ typedef enum { } dirblock_type_t; #define ext4_read_dirblock(inode, block, type) \ - __ext4_read_dirblock((inode), (block), (type), __LINE__) + __ext4_read_dirblock((inode), (block), (type), __func__, __LINE__) static struct buffer_head *__ext4_read_dirblock(struct inode *inode, - ext4_lblk_t block, - dirblock_type_t type, - unsigned int line) + ext4_lblk_t block, + dirblock_type_t type, + const char *func, + unsigned int line) { struct buffer_head *bh; struct ext4_dir_entry *dirent; @@ -97,15 +98,17 @@ static struct buffer_head *__ext4_read_dirblock(struct inode *inode, bh = ext4_bread(NULL, inode, block, 0); if (IS_ERR(bh)) { - __ext4_warning(inode->i_sb, __func__, line, - "error %ld reading directory block " - "(ino %lu, block %lu)", PTR_ERR(bh), inode->i_ino, - (unsigned long) block); + __ext4_warning(inode->i_sb, func, line, + "inode #%lu: lblock %lu: comm %s: " + "error %ld reading directory block", + inode->i_ino, (unsigned long)block, + current->comm, PTR_ERR(bh)); return bh; } if (!bh) { - ext4_error_inode(inode, __func__, line, block, "Directory hole found"); + ext4_error_inode(inode, func, line, block, + "Directory hole found"); return ERR_PTR(-EIO); } dirent = (struct ext4_dir_entry *) bh->b_data; @@ -119,7 +122,7 @@ static struct buffer_head *__ext4_read_dirblock(struct inode *inode, is_dx_block = 1; } if (!is_dx_block && type == INDEX) { - ext4_error_inode(inode, __func__, line, block, + ext4_error_inode(inode, func, line, block, "directory leaf block found instead of index block"); return ERR_PTR(-EIO); } @@ -136,8 +139,8 @@ static struct buffer_head *__ext4_read_dirblock(struct inode *inode, if (ext4_dx_csum_verify(inode, dirent)) set_buffer_verified(bh); else { - ext4_error_inode(inode, __func__, line, block, - "Directory index failed checksum"); + ext4_error_inode(inode, func, line, block, + "Directory index failed checksum"); brelse(bh); return ERR_PTR(-EIO); } @@ -146,8 +149,8 @@ static struct buffer_head *__ext4_read_dirblock(struct inode *inode, if (ext4_dirent_csum_verify(inode, dirent)) set_buffer_verified(bh); else { - ext4_error_inode(inode, __func__, line, block, - "Directory block failed checksum"); + ext4_error_inode(inode, func, line, block, + "Directory block failed checksum"); brelse(bh); return ERR_PTR(-EIO); } @@ -248,7 +251,7 @@ static void dx_set_count(struct dx_entry *entries, unsigned value); static void dx_set_limit(struct dx_entry *entries, unsigned value); static unsigned dx_root_limit(struct inode *dir, unsigned infosize); static unsigned dx_node_limit(struct inode *dir); -static struct dx_frame *dx_probe(const struct qstr *d_name, +static struct dx_frame *dx_probe(struct ext4_filename *fname, struct inode *dir, struct dx_hash_info *hinfo, struct dx_frame *frame); @@ -267,10 +270,10 @@ static int ext4_htree_next_block(struct inode *dir, __u32 hash, struct dx_frame *frames, __u32 *start_hash); static struct buffer_head * ext4_dx_find_entry(struct inode *dir, - const struct qstr *d_name, + struct ext4_filename *fname, struct ext4_dir_entry_2 **res_dir); -static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, - struct inode *inode); +static int ext4_dx_add_entry(handle_t *handle, struct ext4_filename *fname, + struct dentry *dentry, struct inode *inode); /* checksumming functions */ void initialize_dirent_tail(struct ext4_dir_entry_tail *t, @@ -327,10 +330,14 @@ static __le32 ext4_dirent_csum(struct inode *inode, return cpu_to_le32(csum); } -static void warn_no_space_for_csum(struct inode *inode) +#define warn_no_space_for_csum(inode) \ + __warn_no_space_for_csum((inode), __func__, __LINE__) + +static void __warn_no_space_for_csum(struct inode *inode, const char *func, + unsigned int line) { - ext4_warning(inode->i_sb, "no space in directory inode %lu leaf for " - "checksum. Please run e2fsck -D.", inode->i_ino); + __ext4_warning_inode(inode, func, line, + "No space for directory leaf checksum. Please run e2fsck -D."); } int ext4_dirent_csum_verify(struct inode *inode, struct ext4_dir_entry *dirent) @@ -607,17 +614,15 @@ static struct stats dx_show_leaf(struct inode *dir, char *name; struct ext4_str fname_crypto_str = {.name = NULL, .len = 0}; - struct ext4_fname_crypto_ctx *ctx = NULL; - int res; + int res = 0; name = de->name; len = de->name_len; - ctx = ext4_get_fname_crypto_ctx(dir, - EXT4_NAME_LEN); - if (IS_ERR(ctx)) { - printk(KERN_WARNING "Error acquiring" - " crypto ctxt--skipping crypto\n"); - ctx = NULL; + if (ext4_encrypted_inode(inode)) + res = ext4_get_encryption_info(dir); + if (res) { + printk(KERN_WARNING "Error setting up" + " fname crypto: %d\n", res); } if (ctx == NULL) { /* Directory is not encrypted */ @@ -637,7 +642,6 @@ static struct stats dx_show_leaf(struct inode *dir, "allocating crypto " "buffer--skipping " "crypto\n"); - ext4_put_fname_crypto_ctx(&ctx); ctx = NULL; } res = ext4_fname_disk_to_usr(ctx, NULL, de, @@ -658,7 +662,6 @@ static struct stats dx_show_leaf(struct inode *dir, printk("%*.s:(E)%x.%u ", len, name, h.hash, (unsigned) ((char *) de - base)); - ext4_put_fname_crypto_ctx(&ctx); ext4_fname_crypto_free_buffer( &fname_crypto_str); } @@ -724,7 +727,7 @@ struct stats dx_show_entries(struct dx_hash_info *hinfo, struct inode *dir, * back to userspace. */ static struct dx_frame * -dx_probe(const struct qstr *d_name, struct inode *dir, +dx_probe(struct ext4_filename *fname, struct inode *dir, struct dx_hash_info *hinfo, struct dx_frame *frame_in) { unsigned count, indirect; @@ -742,56 +745,41 @@ dx_probe(const struct qstr *d_name, struct inode *dir, if (root->info.hash_version != DX_HASH_TEA && root->info.hash_version != DX_HASH_HALF_MD4 && root->info.hash_version != DX_HASH_LEGACY) { - ext4_warning(dir->i_sb, "Unrecognised inode hash code %d", - root->info.hash_version); + ext4_warning_inode(dir, "Unrecognised inode hash code %u", + root->info.hash_version); goto fail; } + if (fname) + hinfo = &fname->hinfo; hinfo->hash_version = root->info.hash_version; if (hinfo->hash_version <= DX_HASH_TEA) hinfo->hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned; hinfo->seed = EXT4_SB(dir->i_sb)->s_hash_seed; -#ifdef CONFIG_EXT4_FS_ENCRYPTION - if (d_name) { - struct ext4_fname_crypto_ctx *ctx = NULL; - int res; - - /* Check if the directory is encrypted */ - ctx = ext4_get_fname_crypto_ctx(dir, EXT4_NAME_LEN); - if (IS_ERR(ctx)) { - ret_err = ERR_PTR(PTR_ERR(ctx)); - goto fail; - } - res = ext4_fname_usr_to_hash(ctx, d_name, hinfo); - if (res < 0) { - ret_err = ERR_PTR(res); - goto fail; - } - ext4_put_fname_crypto_ctx(&ctx); - } -#else - if (d_name) - ext4fs_dirhash(d_name->name, d_name->len, hinfo); -#endif + if (fname && fname_name(fname)) + ext4fs_dirhash(fname_name(fname), fname_len(fname), hinfo); hash = hinfo->hash; if (root->info.unused_flags & 1) { - ext4_warning(dir->i_sb, "Unimplemented inode hash flags: %#06x", - root->info.unused_flags); + ext4_warning_inode(dir, "Unimplemented hash flags: %#06x", + root->info.unused_flags); goto fail; } - if ((indirect = root->info.indirect_levels) > 1) { - ext4_warning(dir->i_sb, "Unimplemented inode hash depth: %#06x", - root->info.indirect_levels); + indirect = root->info.indirect_levels; + if (indirect > 1) { + ext4_warning_inode(dir, "Unimplemented hash depth: %#06x", + root->info.indirect_levels); goto fail; } - entries = (struct dx_entry *) (((char *)&root->info) + - root->info.info_length); + entries = (struct dx_entry *)(((char *)&root->info) + + root->info.info_length); if (dx_get_limit(entries) != dx_root_limit(dir, root->info.info_length)) { - ext4_warning(dir->i_sb, "dx entry: limit != root limit"); + ext4_warning_inode(dir, "dx entry: limit %u != root limit %u", + dx_get_limit(entries), + dx_root_limit(dir, root->info.info_length)); goto fail; } @@ -799,15 +787,16 @@ dx_probe(const struct qstr *d_name, struct inode *dir, while (1) { count = dx_get_count(entries); if (!count || count > dx_get_limit(entries)) { - ext4_warning(dir->i_sb, - "dx entry: no count or count > limit"); + ext4_warning_inode(dir, + "dx entry: count %u beyond limit %u", + count, dx_get_limit(entries)); goto fail; } p = entries + 1; q = entries + count - 1; while (p <= q) { - m = p + (q - p)/2; + m = p + (q - p) / 2; dxtrace(printk(".")); if (dx_get_hash(m) > hash) q = m - 1; @@ -831,7 +820,8 @@ dx_probe(const struct qstr *d_name, struct inode *dir, } at = p - 1; - dxtrace(printk(" %x->%u\n", at == entries? 0: dx_get_hash(at), dx_get_block(at))); + dxtrace(printk(" %x->%u\n", at == entries ? 0 : dx_get_hash(at), + dx_get_block(at))); frame->entries = entries; frame->at = at; if (!indirect--) @@ -845,9 +835,10 @@ dx_probe(const struct qstr *d_name, struct inode *dir, } entries = ((struct dx_node *) frame->bh->b_data)->entries; - if (dx_get_limit(entries) != dx_node_limit (dir)) { - ext4_warning(dir->i_sb, - "dx entry: limit != node limit"); + if (dx_get_limit(entries) != dx_node_limit(dir)) { + ext4_warning_inode(dir, + "dx entry: limit %u != node limit %u", + dx_get_limit(entries), dx_node_limit(dir)); goto fail; } } @@ -858,18 +849,17 @@ fail: } if (ret_err == ERR_PTR(ERR_BAD_DX_DIR)) - ext4_warning(dir->i_sb, - "Corrupt dir inode %lu, running e2fsck is " - "recommended.", dir->i_ino); + ext4_warning_inode(dir, + "Corrupt directory, running e2fsck is recommended"); return ret_err; } -static void dx_release (struct dx_frame *frames) +static void dx_release(struct dx_frame *frames) { if (frames[0].bh == NULL) return; - if (((struct dx_root *) frames[0].bh->b_data)->info.indirect_levels) + if (((struct dx_root *)frames[0].bh->b_data)->info.indirect_levels) brelse(frames[1].bh); brelse(frames[0].bh); } @@ -962,7 +952,6 @@ static int htree_dirblock_to_tree(struct file *dir_file, struct buffer_head *bh; struct ext4_dir_entry_2 *de, *top; int err = 0, count = 0; - struct ext4_fname_crypto_ctx *ctx = NULL; struct ext4_str fname_crypto_str = {.name = NULL, .len = 0}, tmp_str; dxtrace(printk(KERN_INFO "In htree dirblock_to_tree: block %lu\n", @@ -977,17 +966,15 @@ static int htree_dirblock_to_tree(struct file *dir_file, EXT4_DIR_REC_LEN(0)); #ifdef CONFIG_EXT4_FS_ENCRYPTION /* Check if the directory is encrypted */ - ctx = ext4_get_fname_crypto_ctx(dir, EXT4_NAME_LEN); - if (IS_ERR(ctx)) { - err = PTR_ERR(ctx); - brelse(bh); - return err; - } - if (ctx != NULL) { - err = ext4_fname_crypto_alloc_buffer(ctx, EXT4_NAME_LEN, + if (ext4_encrypted_inode(dir)) { + err = ext4_get_encryption_info(dir); + if (err < 0) { + brelse(bh); + return err; + } + err = ext4_fname_crypto_alloc_buffer(dir, EXT4_NAME_LEN, &fname_crypto_str); if (err < 0) { - ext4_put_fname_crypto_ctx(&ctx); brelse(bh); return err; } @@ -1008,16 +995,17 @@ static int htree_dirblock_to_tree(struct file *dir_file, continue; if (de->inode == 0) continue; - if (ctx == NULL) { - /* Directory is not encrypted */ + if (!ext4_encrypted_inode(dir)) { tmp_str.name = de->name; tmp_str.len = de->name_len; err = ext4_htree_store_dirent(dir_file, hinfo->hash, hinfo->minor_hash, de, &tmp_str); } else { + int save_len = fname_crypto_str.len; + /* Directory is encrypted */ - err = ext4_fname_disk_to_usr(ctx, hinfo, de, + err = ext4_fname_disk_to_usr(dir, hinfo, de, &fname_crypto_str); if (err < 0) { count = err; @@ -1026,6 +1014,7 @@ static int htree_dirblock_to_tree(struct file *dir_file, err = ext4_htree_store_dirent(dir_file, hinfo->hash, hinfo->minor_hash, de, &fname_crypto_str); + fname_crypto_str.len = save_len; } if (err != 0) { count = err; @@ -1036,7 +1025,6 @@ static int htree_dirblock_to_tree(struct file *dir_file, errout: brelse(bh); #ifdef CONFIG_EXT4_FS_ENCRYPTION - ext4_put_fname_crypto_ctx(&ctx); ext4_fname_crypto_free_buffer(&fname_crypto_str); #endif return count; @@ -1155,12 +1143,13 @@ errout: static inline int search_dirblock(struct buffer_head *bh, struct inode *dir, + struct ext4_filename *fname, const struct qstr *d_name, unsigned int offset, struct ext4_dir_entry_2 **res_dir) { - return search_dir(bh, bh->b_data, dir->i_sb->s_blocksize, dir, - d_name, offset, res_dir); + return ext4_search_dir(bh, bh->b_data, dir->i_sb->s_blocksize, dir, + fname, d_name, offset, res_dir); } /* @@ -1242,54 +1231,54 @@ static void dx_insert_block(struct dx_frame *frame, u32 hash, ext4_lblk_t block) * `len <= EXT4_NAME_LEN' is guaranteed by caller. * `de != NULL' is guaranteed by caller. */ -static inline int ext4_match(struct ext4_fname_crypto_ctx *ctx, - struct ext4_str *fname_crypto_str, - int len, const char * const name, +static inline int ext4_match(struct ext4_filename *fname, struct ext4_dir_entry_2 *de) { - int res; + const void *name = fname_name(fname); + u32 len = fname_len(fname); if (!de->inode) return 0; #ifdef CONFIG_EXT4_FS_ENCRYPTION - if (ctx) - return ext4_fname_match(ctx, fname_crypto_str, len, name, de); + if (unlikely(!name)) { + if (fname->usr_fname->name[0] == '_') { + int ret; + if (de->name_len < 16) + return 0; + ret = memcmp(de->name + de->name_len - 16, + fname->crypto_buf.name + 8, 16); + return (ret == 0) ? 1 : 0; + } + name = fname->crypto_buf.name; + len = fname->crypto_buf.len; + } #endif - if (len != de->name_len) + if (de->name_len != len) return 0; - res = memcmp(name, de->name, len); - return (res == 0) ? 1 : 0; + return (memcmp(de->name, name, len) == 0) ? 1 : 0; } /* * Returns 0 if not found, -1 on failure, and 1 on success */ -int search_dir(struct buffer_head *bh, char *search_buf, int buf_size, - struct inode *dir, const struct qstr *d_name, - unsigned int offset, struct ext4_dir_entry_2 **res_dir) +int ext4_search_dir(struct buffer_head *bh, char *search_buf, int buf_size, + struct inode *dir, struct ext4_filename *fname, + const struct qstr *d_name, + unsigned int offset, struct ext4_dir_entry_2 **res_dir) { struct ext4_dir_entry_2 * de; char * dlimit; int de_len; - const char *name = d_name->name; - int namelen = d_name->len; - struct ext4_fname_crypto_ctx *ctx = NULL; - struct ext4_str fname_crypto_str = {.name = NULL, .len = 0}; int res; - ctx = ext4_get_fname_crypto_ctx(dir, EXT4_NAME_LEN); - if (IS_ERR(ctx)) - return -1; - de = (struct ext4_dir_entry_2 *)search_buf; dlimit = search_buf + buf_size; while ((char *) de < dlimit) { /* this code is executed quadratically often */ /* do minimal checking `by hand' */ if ((char *) de + de->name_len <= dlimit) { - res = ext4_match(ctx, &fname_crypto_str, namelen, - name, de); + res = ext4_match(fname, de); if (res < 0) { res = -1; goto return_result; @@ -1322,8 +1311,6 @@ int search_dir(struct buffer_head *bh, char *search_buf, int buf_size, res = 0; return_result: - ext4_put_fname_crypto_ctx(&ctx); - ext4_fname_crypto_free_buffer(&fname_crypto_str); return res; } @@ -1370,7 +1357,8 @@ static struct buffer_head * ext4_find_entry (struct inode *dir, buffer */ int num = 0; ext4_lblk_t nblocks; - int i, namelen; + int i, namelen, retval; + struct ext4_filename fname; *res_dir = NULL; sb = dir->i_sb; @@ -1378,14 +1366,18 @@ static struct buffer_head * ext4_find_entry (struct inode *dir, if (namelen > EXT4_NAME_LEN) return NULL; + retval = ext4_fname_setup_filename(dir, d_name, 1, &fname); + if (retval) + return ERR_PTR(retval); + if (ext4_has_inline_data(dir)) { int has_inline_data = 1; - ret = ext4_find_inline_entry(dir, d_name, res_dir, + ret = ext4_find_inline_entry(dir, &fname, d_name, res_dir, &has_inline_data); if (has_inline_data) { if (inlined) *inlined = 1; - return ret; + goto cleanup_and_exit; } } @@ -1400,14 +1392,14 @@ static struct buffer_head * ext4_find_entry (struct inode *dir, goto restart; } if (is_dx(dir)) { - bh = ext4_dx_find_entry(dir, d_name, res_dir); + ret = ext4_dx_find_entry(dir, &fname, res_dir); /* * On success, or if the error was file not found, * return. Otherwise, fall back to doing a search the * old fashioned way. */ - if (!IS_ERR(bh) || PTR_ERR(bh) != ERR_BAD_DX_DIR) - return bh; + if (!IS_ERR(ret) || PTR_ERR(ret) != ERR_BAD_DX_DIR) + goto cleanup_and_exit; dxtrace(printk(KERN_DEBUG "ext4_find_entry: dx failed, " "falling back\n")); } @@ -1438,8 +1430,10 @@ restart: num++; bh = ext4_getblk(NULL, dir, b++, 0); if (unlikely(IS_ERR(bh))) { - if (ra_max == 0) - return bh; + if (ra_max == 0) { + ret = bh; + goto cleanup_and_exit; + } break; } bh_use[ra_max] = bh; @@ -1469,7 +1463,7 @@ restart: goto next; } set_buffer_verified(bh); - i = search_dirblock(bh, dir, d_name, + i = search_dirblock(bh, dir, &fname, d_name, block << EXT4_BLOCK_SIZE_BITS(sb), res_dir); if (i == 1) { EXT4_I(dir)->i_dir_start_lookup = block; @@ -1500,15 +1494,17 @@ cleanup_and_exit: /* Clean up the read-ahead blocks */ for (; ra_ptr < ra_max; ra_ptr++) brelse(bh_use[ra_ptr]); + ext4_fname_free_filename(&fname); return ret; } -static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct qstr *d_name, - struct ext4_dir_entry_2 **res_dir) +static struct buffer_head * ext4_dx_find_entry(struct inode *dir, + struct ext4_filename *fname, + struct ext4_dir_entry_2 **res_dir) { struct super_block * sb = dir->i_sb; - struct dx_hash_info hinfo; struct dx_frame frames[2], *frame; + const struct qstr *d_name = fname->usr_fname; struct buffer_head *bh; ext4_lblk_t block; int retval; @@ -1516,7 +1512,7 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q #ifdef CONFIG_EXT4_FS_ENCRYPTION *res_dir = NULL; #endif - frame = dx_probe(d_name, dir, &hinfo, frames); + frame = dx_probe(fname, dir, NULL, frames); if (IS_ERR(frame)) return (struct buffer_head *) frame; do { @@ -1525,7 +1521,7 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q if (IS_ERR(bh)) goto errout; - retval = search_dirblock(bh, dir, d_name, + retval = search_dirblock(bh, dir, fname, d_name, block << EXT4_BLOCK_SIZE_BITS(sb), res_dir); if (retval == 1) @@ -1537,12 +1533,12 @@ static struct buffer_head * ext4_dx_find_entry(struct inode *dir, const struct q } /* Check to see if we should continue to search */ - retval = ext4_htree_next_block(dir, hinfo.hash, frame, + retval = ext4_htree_next_block(dir, fname->hinfo.hash, frame, frames, NULL); if (retval < 0) { - ext4_warning(sb, - "error %d reading index page in directory #%lu", - retval, dir->i_ino); + ext4_warning_inode(dir, + "error %d reading directory index block", + retval); bh = ERR_PTR(retval); goto errout; } @@ -1796,32 +1792,16 @@ journal_error: int ext4_find_dest_de(struct inode *dir, struct inode *inode, struct buffer_head *bh, void *buf, int buf_size, - const char *name, int namelen, + struct ext4_filename *fname, struct ext4_dir_entry_2 **dest_de) { struct ext4_dir_entry_2 *de; - unsigned short reclen = EXT4_DIR_REC_LEN(namelen); + unsigned short reclen = EXT4_DIR_REC_LEN(fname_len(fname)); int nlen, rlen; unsigned int offset = 0; char *top; - struct ext4_fname_crypto_ctx *ctx = NULL; - struct ext4_str fname_crypto_str = {.name = NULL, .len = 0}; int res; - ctx = ext4_get_fname_crypto_ctx(dir, EXT4_NAME_LEN); - if (IS_ERR(ctx)) - return -1; - - if (ctx != NULL) { - /* Calculate record length needed to store the entry */ - res = ext4_fname_crypto_namelen_on_disk(ctx, namelen); - if (res < 0) { - ext4_put_fname_crypto_ctx(&ctx); - return res; - } - reclen = EXT4_DIR_REC_LEN(res); - } - de = (struct ext4_dir_entry_2 *)buf; top = buf + buf_size - reclen; while ((char *) de <= top) { @@ -1831,7 +1811,7 @@ int ext4_find_dest_de(struct inode *dir, struct inode *inode, goto return_result; } /* Provide crypto context and crypto buffer to ext4 match */ - res = ext4_match(ctx, &fname_crypto_str, namelen, name, de); + res = ext4_match(fname, de); if (res < 0) goto return_result; if (res > 0) { @@ -1853,8 +1833,6 @@ int ext4_find_dest_de(struct inode *dir, struct inode *inode, res = 0; } return_result: - ext4_put_fname_crypto_ctx(&ctx); - ext4_fname_crypto_free_buffer(&fname_crypto_str); return res; } @@ -1862,39 +1840,10 @@ int ext4_insert_dentry(struct inode *dir, struct inode *inode, struct ext4_dir_entry_2 *de, int buf_size, - const struct qstr *iname, - const char *name, int namelen) + struct ext4_filename *fname) { int nlen, rlen; - struct ext4_fname_crypto_ctx *ctx = NULL; - struct ext4_str fname_crypto_str = {.name = NULL, .len = 0}; - struct ext4_str tmp_str; - int res; - - ctx = ext4_get_fname_crypto_ctx(dir, EXT4_NAME_LEN); - if (IS_ERR(ctx)) - return -EIO; - /* By default, the input name would be written to the disk */ - tmp_str.name = (unsigned char *)name; - tmp_str.len = namelen; - if (ctx != NULL) { - /* Directory is encrypted */ - res = ext4_fname_crypto_alloc_buffer(ctx, EXT4_NAME_LEN, - &fname_crypto_str); - if (res < 0) { - ext4_put_fname_crypto_ctx(&ctx); - return -ENOMEM; - } - res = ext4_fname_usr_to_disk(ctx, iname, &fname_crypto_str); - if (res < 0) { - ext4_put_fname_crypto_ctx(&ctx); - ext4_fname_crypto_free_buffer(&fname_crypto_str); - return res; - } - tmp_str.name = fname_crypto_str.name; - tmp_str.len = fname_crypto_str.len; - } nlen = EXT4_DIR_REC_LEN(de->name_len); rlen = ext4_rec_len_from_disk(de->rec_len, buf_size); @@ -1908,11 +1857,8 @@ int ext4_insert_dentry(struct inode *dir, de->file_type = EXT4_FT_UNKNOWN; de->inode = cpu_to_le32(inode->i_ino); ext4_set_de_type(inode->i_sb, de, inode->i_mode); - de->name_len = tmp_str.len; - - memcpy(de->name, tmp_str.name, tmp_str.len); - ext4_put_fname_crypto_ctx(&ctx); - ext4_fname_crypto_free_buffer(&fname_crypto_str); + de->name_len = fname_len(fname); + memcpy(de->name, fname_name(fname), fname_len(fname)); return 0; } @@ -1924,13 +1870,11 @@ int ext4_insert_dentry(struct inode *dir, * space. It will return -ENOSPC if no space is available, and -EIO * and -EEXIST if directory entry already exists. */ -static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, +static int add_dirent_to_buf(handle_t *handle, struct ext4_filename *fname, + struct inode *dir, struct inode *inode, struct ext4_dir_entry_2 *de, struct buffer_head *bh) { - struct inode *dir = d_inode(dentry->d_parent); - const char *name = dentry->d_name.name; - int namelen = dentry->d_name.len; unsigned int blocksize = dir->i_sb->s_blocksize; int csum_size = 0; int err; @@ -1939,9 +1883,8 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, csum_size = sizeof(struct ext4_dir_entry_tail); if (!de) { - err = ext4_find_dest_de(dir, inode, - bh, bh->b_data, blocksize - csum_size, - name, namelen, &de); + err = ext4_find_dest_de(dir, inode, bh, bh->b_data, + blocksize - csum_size, fname, &de); if (err) return err; } @@ -1954,8 +1897,7 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, /* By now the buffer is marked for journaling. Due to crypto operations, * the following function call may fail */ - err = ext4_insert_dentry(dir, inode, de, blocksize, &dentry->d_name, - name, namelen); + err = ext4_insert_dentry(dir, inode, de, blocksize, fname); if (err < 0) return err; @@ -1985,17 +1927,11 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, * This converts a one block unindexed directory to a 3 block indexed * directory, and adds the dentry to the indexed directory. */ -static int make_indexed_dir(handle_t *handle, struct dentry *dentry, +static int make_indexed_dir(handle_t *handle, struct ext4_filename *fname, + struct dentry *dentry, struct inode *inode, struct buffer_head *bh) { struct inode *dir = d_inode(dentry->d_parent); -#ifdef CONFIG_EXT4_FS_ENCRYPTION - struct ext4_fname_crypto_ctx *ctx = NULL; - int res; -#else - const char *name = dentry->d_name.name; - int namelen = dentry->d_name.len; -#endif struct buffer_head *bh2; struct dx_root *root; struct dx_frame frames[2], *frame; @@ -2006,17 +1942,10 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, unsigned len; int retval; unsigned blocksize; - struct dx_hash_info hinfo; ext4_lblk_t block; struct fake_dirent *fde; int csum_size = 0; -#ifdef CONFIG_EXT4_FS_ENCRYPTION - ctx = ext4_get_fname_crypto_ctx(dir, EXT4_NAME_LEN); - if (IS_ERR(ctx)) - return PTR_ERR(ctx); -#endif - if (ext4_has_metadata_csum(inode->i_sb)) csum_size = sizeof(struct ext4_dir_entry_tail); @@ -2078,22 +2007,12 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, dx_set_limit(entries, dx_root_limit(dir, sizeof(root->info))); /* Initialize as for dx_probe */ - hinfo.hash_version = root->info.hash_version; - if (hinfo.hash_version <= DX_HASH_TEA) - hinfo.hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned; - hinfo.seed = EXT4_SB(dir->i_sb)->s_hash_seed; -#ifdef CONFIG_EXT4_FS_ENCRYPTION - res = ext4_fname_usr_to_hash(ctx, &dentry->d_name, &hinfo); - if (res < 0) { - ext4_put_fname_crypto_ctx(&ctx); - ext4_mark_inode_dirty(handle, dir); - brelse(bh); - return res; - } - ext4_put_fname_crypto_ctx(&ctx); -#else - ext4fs_dirhash(name, namelen, &hinfo); -#endif + fname->hinfo.hash_version = root->info.hash_version; + if (fname->hinfo.hash_version <= DX_HASH_TEA) + fname->hinfo.hash_version += EXT4_SB(dir->i_sb)->s_hash_unsigned; + fname->hinfo.seed = EXT4_SB(dir->i_sb)->s_hash_seed; + ext4fs_dirhash(fname_name(fname), fname_len(fname), &fname->hinfo); + memset(frames, 0, sizeof(frames)); frame = frames; frame->entries = entries; @@ -2108,14 +2027,14 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, if (retval) goto out_frames; - de = do_split(handle,dir, &bh, frame, &hinfo); + de = do_split(handle,dir, &bh, frame, &fname->hinfo); if (IS_ERR(de)) { retval = PTR_ERR(de); goto out_frames; } dx_release(frames); - retval = add_dirent_to_buf(handle, dentry, inode, de, bh); + retval = add_dirent_to_buf(handle, fname, dir, inode, de, bh); brelse(bh); return retval; out_frames: @@ -2147,6 +2066,7 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, struct ext4_dir_entry_2 *de; struct ext4_dir_entry_tail *t; struct super_block *sb; + struct ext4_filename fname; int retval; int dx_fallback=0; unsigned blocksize; @@ -2161,10 +2081,15 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, if (!dentry->d_name.len) return -EINVAL; + retval = ext4_fname_setup_filename(dir, &dentry->d_name, 0, &fname); + if (retval) + return retval; + if (ext4_has_inline_data(dir)) { - retval = ext4_try_add_inline_entry(handle, dentry, inode); + retval = ext4_try_add_inline_entry(handle, &fname, + dentry, inode); if (retval < 0) - return retval; + goto out; if (retval == 1) { retval = 0; goto out; @@ -2172,7 +2097,7 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, } if (is_dx(dir)) { - retval = ext4_dx_add_entry(handle, dentry, inode); + retval = ext4_dx_add_entry(handle, &fname, dentry, inode); if (!retval || (retval != ERR_BAD_DX_DIR)) goto out; ext4_clear_inode_flag(dir, EXT4_INODE_INDEX); @@ -2182,24 +2107,31 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, blocks = dir->i_size >> sb->s_blocksize_bits; for (block = 0; block < blocks; block++) { bh = ext4_read_dirblock(dir, block, DIRENT); - if (IS_ERR(bh)) - return PTR_ERR(bh); - - retval = add_dirent_to_buf(handle, dentry, inode, NULL, bh); + if (IS_ERR(bh)) { + retval = PTR_ERR(bh); + bh = NULL; + goto out; + } + retval = add_dirent_to_buf(handle, &fname, dir, inode, + NULL, bh); if (retval != -ENOSPC) goto out; if (blocks == 1 && !dx_fallback && EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_DIR_INDEX)) { - retval = make_indexed_dir(handle, dentry, inode, bh); + retval = make_indexed_dir(handle, &fname, dentry, + inode, bh); bh = NULL; /* make_indexed_dir releases bh */ goto out; } brelse(bh); } bh = ext4_append(handle, dir, &block); - if (IS_ERR(bh)) - return PTR_ERR(bh); + if (IS_ERR(bh)) { + retval = PTR_ERR(bh); + bh = NULL; + goto out; + } de = (struct ext4_dir_entry_2 *) bh->b_data; de->inode = 0; de->rec_len = ext4_rec_len_to_disk(blocksize - csum_size, blocksize); @@ -2209,8 +2141,9 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, initialize_dirent_tail(t, blocksize); } - retval = add_dirent_to_buf(handle, dentry, inode, de, bh); + retval = add_dirent_to_buf(handle, &fname, dir, inode, de, bh); out: + ext4_fname_free_filename(&fname); brelse(bh); if (retval == 0) ext4_set_inode_state(inode, EXT4_STATE_NEWENTRY); @@ -2220,19 +2153,18 @@ out: /* * Returns 0 for success, or a negative error value */ -static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, - struct inode *inode) +static int ext4_dx_add_entry(handle_t *handle, struct ext4_filename *fname, + struct dentry *dentry, struct inode *inode) { struct dx_frame frames[2], *frame; struct dx_entry *entries, *at; - struct dx_hash_info hinfo; struct buffer_head *bh; struct inode *dir = d_inode(dentry->d_parent); struct super_block *sb = dir->i_sb; struct ext4_dir_entry_2 *de; int err; - frame = dx_probe(&dentry->d_name, dir, &hinfo, frames); + frame = dx_probe(fname, dir, NULL, frames); if (IS_ERR(frame)) return PTR_ERR(frame); entries = frame->entries; @@ -2249,7 +2181,7 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, if (err) goto journal_error; - err = add_dirent_to_buf(handle, dentry, inode, NULL, bh); + err = add_dirent_to_buf(handle, fname, dir, inode, NULL, bh); if (err != -ENOSPC) goto cleanup; @@ -2267,7 +2199,7 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, if (levels && (dx_get_count(frames->entries) == dx_get_limit(frames->entries))) { - ext4_warning(sb, "Directory index full!"); + ext4_warning_inode(dir, "Directory index full!"); err = -ENOSPC; goto cleanup; } @@ -2345,12 +2277,12 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, goto cleanup; } } - de = do_split(handle, dir, &bh, frame, &hinfo); + de = do_split(handle, dir, &bh, frame, &fname->hinfo); if (IS_ERR(de)) { err = PTR_ERR(de); goto cleanup; } - err = add_dirent_to_buf(handle, dentry, inode, de, bh); + err = add_dirent_to_buf(handle, fname, dir, inode, de, bh); goto cleanup; journal_error: @@ -2517,20 +2449,7 @@ retry: inode->i_op = &ext4_file_inode_operations; inode->i_fop = &ext4_file_operations; ext4_set_aops(inode); - err = 0; -#ifdef CONFIG_EXT4_FS_ENCRYPTION - if (!err && (ext4_encrypted_inode(dir) || - DUMMY_ENCRYPTION_ENABLED(EXT4_SB(dir->i_sb)))) { - err = ext4_inherit_context(dir, inode); - if (err) { - clear_nlink(inode); - unlock_new_inode(inode); - iput(inode); - } - } -#endif - if (!err) - err = ext4_add_nondir(handle, dentry, inode); + err = ext4_add_nondir(handle, dentry, inode); if (!err && IS_DIRSYNC(dir)) ext4_handle_sync(handle); } @@ -2711,14 +2630,6 @@ retry: err = ext4_init_new_dir(handle, dir, inode); if (err) goto out_clear_inode; -#ifdef CONFIG_EXT4_FS_ENCRYPTION - if (ext4_encrypted_inode(dir) || - DUMMY_ENCRYPTION_ENABLED(EXT4_SB(dir->i_sb))) { - err = ext4_inherit_context(dir, inode); - if (err) - goto out_clear_inode; - } -#endif err = ext4_mark_inode_dirty(handle, inode); if (!err) err = ext4_add_entry(handle, dentry, inode); @@ -2779,12 +2690,9 @@ int ext4_empty_dir(struct inode *inode) de = (struct ext4_dir_entry_2 *) bh->b_data; de1 = ext4_next_entry(de, sb->s_blocksize); if (le32_to_cpu(de->inode) != inode->i_ino || - !le32_to_cpu(de1->inode) || - strcmp(".", de->name) || - strcmp("..", de1->name)) { - ext4_warning(inode->i_sb, - "bad directory (dir #%lu) - no `.' or `..'", - inode->i_ino); + le32_to_cpu(de1->inode) == 0 || + strcmp(".", de->name) || strcmp("..", de1->name)) { + ext4_warning_inode(inode, "directory missing '.' and/or '..'"); brelse(bh); return 1; } @@ -3037,8 +2945,9 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) if (retval) goto end_rmdir; if (!EXT4_DIR_LINK_EMPTY(inode)) - ext4_warning(inode->i_sb, - "empty directory has too many links (%d)", + ext4_warning_inode(inode, + "empty directory '%.*s' has too many links (%u)", + dentry->d_name.len, dentry->d_name.name, inode->i_nlink); inode->i_version++; clear_nlink(inode); @@ -3098,10 +3007,9 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) if (IS_DIRSYNC(dir)) ext4_handle_sync(handle); - if (!inode->i_nlink) { - ext4_warning(inode->i_sb, - "Deleting nonexistent file (%lu), %d", - inode->i_ino, inode->i_nlink); + if (inode->i_nlink == 0) { + ext4_warning_inode(inode, "Deleting file '%.*s' with no links", + dentry->d_name.len, dentry->d_name.name); set_nlink(inode, 1); } retval = ext4_delete_entry(handle, dir, de, bh); @@ -3140,10 +3048,23 @@ static int ext4_symlink(struct inode *dir, encryption_required = (ext4_encrypted_inode(dir) || DUMMY_ENCRYPTION_ENABLED(EXT4_SB(dir->i_sb))); - if (encryption_required) - disk_link.len = encrypted_symlink_data_len(len) + 1; - if (disk_link.len > dir->i_sb->s_blocksize) - return -ENAMETOOLONG; + if (encryption_required) { + err = ext4_get_encryption_info(dir); + if (err) + return err; + if (ext4_encryption_info(dir) == NULL) + return -EPERM; + disk_link.len = (ext4_fname_encrypted_size(dir, len) + + sizeof(struct ext4_encrypted_symlink_data)); + sd = kzalloc(disk_link.len, GFP_KERNEL); + if (!sd) + return -ENOMEM; + } + + if (disk_link.len > dir->i_sb->s_blocksize) { + err = -ENAMETOOLONG; + goto err_free_sd; + } dquot_initialize(dir); @@ -3174,34 +3095,19 @@ static int ext4_symlink(struct inode *dir, if (IS_ERR(inode)) { if (handle) ext4_journal_stop(handle); - return PTR_ERR(inode); + err = PTR_ERR(inode); + goto err_free_sd; } if (encryption_required) { - struct ext4_fname_crypto_ctx *ctx = NULL; struct qstr istr; struct ext4_str ostr; - sd = kzalloc(disk_link.len, GFP_NOFS); - if (!sd) { - err = -ENOMEM; - goto err_drop_inode; - } - err = ext4_inherit_context(dir, inode); - if (err) - goto err_drop_inode; - ctx = ext4_get_fname_crypto_ctx(inode, - inode->i_sb->s_blocksize); - if (IS_ERR_OR_NULL(ctx)) { - /* We just set the policy, so ctx should not be NULL */ - err = (ctx == NULL) ? -EIO : PTR_ERR(ctx); - goto err_drop_inode; - } istr.name = (const unsigned char *) symname; istr.len = len; ostr.name = sd->encrypted_path; - err = ext4_fname_usr_to_disk(ctx, &istr, &ostr); - ext4_put_fname_crypto_ctx(&ctx); + ostr.len = disk_link.len; + err = ext4_fname_usr_to_disk(inode, &istr, &ostr); if (err < 0) goto err_drop_inode; sd->len = cpu_to_le16(ostr.len); @@ -3271,10 +3177,11 @@ static int ext4_symlink(struct inode *dir, err_drop_inode: if (handle) ext4_journal_stop(handle); - kfree(sd); clear_nlink(inode); unlock_new_inode(inode); iput(inode); +err_free_sd: + kfree(sd); return err; } @@ -3490,9 +3397,9 @@ static void ext4_rename_delete(handle_t *handle, struct ext4_renament *ent, } if (retval) { - ext4_warning(ent->dir->i_sb, - "Deleting old file (%lu), %d, error=%d", - ent->dir->i_ino, ent->dir->i_nlink, retval); + ext4_warning_inode(ent->dir, + "Deleting old file: nlink %d, error=%d", + ent->dir->i_nlink, retval); } } @@ -3762,6 +3669,15 @@ static int ext4_cross_rename(struct inode *old_dir, struct dentry *old_dentry, u8 new_file_type; int retval; + if ((ext4_encrypted_inode(old_dir) || + ext4_encrypted_inode(new_dir)) && + (old_dir != new_dir) && + (!ext4_is_child_context_consistent_with_parent(new_dir, + old.inode) || + !ext4_is_child_context_consistent_with_parent(old_dir, + new.inode))) + return -EPERM; + dquot_initialize(old.dir); dquot_initialize(new.dir); diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 5765f88..5602450 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -84,7 +84,7 @@ static void ext4_finish_bio(struct bio *bio) /* The bounce data pages are unmapped. */ data_page = page; ctx = (struct ext4_crypto_ctx *)page_private(data_page); - page = ctx->control_page; + page = ctx->w.control_page; } #endif @@ -359,7 +359,6 @@ void ext4_io_submit(struct ext4_io_submit *io) if (bio) { bio_get(io->io_bio); submit_bio(io->io_op, io->io_bio); - BUG_ON(bio_flagged(io->io_bio, BIO_EOPNOTSUPP)); bio_put(io->io_bio); } io->io_bio = NULL; diff --git a/fs/ext4/readpage.c b/fs/ext4/readpage.c index 171b9ac..ec3ef93 100644 --- a/fs/ext4/readpage.c +++ b/fs/ext4/readpage.c @@ -54,8 +54,8 @@ static void completion_pages(struct work_struct *work) { #ifdef CONFIG_EXT4_FS_ENCRYPTION struct ext4_crypto_ctx *ctx = - container_of(work, struct ext4_crypto_ctx, work); - struct bio *bio = ctx->bio; + container_of(work, struct ext4_crypto_ctx, r.work); + struct bio *bio = ctx->r.bio; struct bio_vec *bv; int i; @@ -109,9 +109,9 @@ static void mpage_end_io(struct bio *bio, int err) if (err) { ext4_release_crypto_ctx(ctx); } else { - INIT_WORK(&ctx->work, completion_pages); - ctx->bio = bio; - queue_work(ext4_read_workqueue, &ctx->work); + INIT_WORK(&ctx->r.work, completion_pages); + ctx->r.bio = bio; + queue_work(ext4_read_workqueue, &ctx->r.work); return; } } diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 0ae853d..58987b5 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -24,6 +24,7 @@ #include <linux/slab.h> #include <linux/init.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <linux/parser.h> #include <linux/buffer_head.h> #include <linux/exportfs.h> @@ -294,6 +295,8 @@ static void __save_error_info(struct super_block *sb, const char *func, struct ext4_super_block *es = EXT4_SB(sb)->s_es; EXT4_SB(sb)->s_mount_state |= EXT4_ERROR_FS; + if (bdev_read_only(sb->s_bdev)) + return; es->s_state |= cpu_to_le16(EXT4_ERROR_FS); es->s_last_error_time = cpu_to_le32(get_seconds()); strncpy(es->s_last_error_func, func, sizeof(es->s_last_error_func)); @@ -589,14 +592,17 @@ void __ext4_msg(struct super_block *sb, va_end(args); } +#define ext4_warning_ratelimit(sb) \ + ___ratelimit(&(EXT4_SB(sb)->s_warning_ratelimit_state), \ + "EXT4-fs warning") + void __ext4_warning(struct super_block *sb, const char *function, unsigned int line, const char *fmt, ...) { struct va_format vaf; va_list args; - if (!___ratelimit(&(EXT4_SB(sb)->s_warning_ratelimit_state), - "EXT4-fs warning")) + if (!ext4_warning_ratelimit(sb)) return; va_start(args, fmt); @@ -607,6 +613,24 @@ void __ext4_warning(struct super_block *sb, const char *function, va_end(args); } +void __ext4_warning_inode(const struct inode *inode, const char *function, + unsigned int line, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + if (!ext4_warning_ratelimit(inode->i_sb)) + return; + + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + printk(KERN_WARNING "EXT4-fs warning (device %s): %s:%d: " + "inode #%lu: comm %s: %pV\n", inode->i_sb->s_id, + function, line, inode->i_ino, current->comm, &vaf); + va_end(args); +} + void __ext4_grp_locked_error(const char *function, unsigned int line, struct super_block *sb, ext4_group_t grp, unsigned long ino, ext4_fsblk_t block, @@ -805,6 +829,7 @@ static void ext4_put_super(struct super_block *sb) dump_orphan_list(sb, sbi); J_ASSERT(list_empty(&sbi->s_orphan)); + sync_blockdev(sb->s_bdev); invalidate_bdev(sb->s_bdev); if (sbi->journal_bdev && sbi->journal_bdev != sb->s_bdev) { /* @@ -877,9 +902,8 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) atomic_set(&ei->i_unwritten, 0); INIT_WORK(&ei->i_rsv_conversion_work, ext4_end_io_rsv_work); #ifdef CONFIG_EXT4_FS_ENCRYPTION - ei->i_encryption_key.mode = EXT4_ENCRYPTION_MODE_INVALID; + ei->i_crypt_info = NULL; #endif - return &ei->vfs_inode; } @@ -956,6 +980,10 @@ void ext4_clear_inode(struct inode *inode) jbd2_free_inode(EXT4_I(inode)->jinode); EXT4_I(inode)->jinode = NULL; } +#ifdef CONFIG_EXT4_FS_ENCRYPTION + if (EXT4_I(inode)->i_crypt_info) + ext4_free_encryption_info(inode, EXT4_I(inode)->i_crypt_info); +#endif } static struct inode *ext4_nfs_get_inode(struct super_block *sb, @@ -3418,7 +3446,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) unsigned long journal_devnum = 0; unsigned long def_mount_opts; struct inode *root; - char *cp; const char *descr; int ret = -ENOMEM; int blocksize, clustersize; @@ -3447,15 +3474,9 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) if (sb->s_bdev->bd_part) sbi->s_sectors_written_start = part_stat_read(sb->s_bdev->bd_part, sectors[1]); -#ifdef CONFIG_EXT4_FS_ENCRYPTION - /* Modes of operations for file and directory encryption. */ - sbi->s_file_encryption_mode = EXT4_ENCRYPTION_MODE_AES_256_XTS; - sbi->s_dir_encryption_mode = EXT4_ENCRYPTION_MODE_INVALID; -#endif /* Cleanup superblock name */ - for (cp = sb->s_id; (cp = strchr(cp, '/'));) - *cp = '!'; + strreplace(sb->s_id, '/', '!'); /* -EINVAL is default */ ret = -EINVAL; @@ -4065,7 +4086,15 @@ no_journal: } } - if (unlikely(sbi->s_mount_flags & EXT4_MF_TEST_DUMMY_ENCRYPTION) && + if ((DUMMY_ENCRYPTION_ENABLED(sbi) || + EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_ENCRYPT)) && + (blocksize != PAGE_CACHE_SIZE)) { + ext4_msg(sb, KERN_ERR, + "Unsupported blocksize for fs encryption"); + goto failed_mount_wq; + } + + if (DUMMY_ENCRYPTION_ENABLED(sbi) && !(sb->s_flags & MS_RDONLY) && !EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_ENCRYPT)) { EXT4_SET_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_ENCRYPT); @@ -4941,6 +4970,9 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) set_task_ioprio(sbi->s_journal->j_task, journal_ioprio); } + if (*flags & MS_LAZYTIME) + sb->s_flags |= MS_LAZYTIME; + if ((*flags & MS_RDONLY) != (sb->s_flags & MS_RDONLY)) { if (sbi->s_mount_flags & EXT4_MF_FS_ABORTED) { err = -EROFS; @@ -5408,6 +5440,7 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type, struct inode *inode = sb_dqopt(sb)->files[type]; ext4_lblk_t blk = off >> EXT4_BLOCK_SIZE_BITS(sb); int err, offset = off & (sb->s_blocksize - 1); + int retries = 0; struct buffer_head *bh; handle_t *handle = journal_current_handle(); @@ -5428,7 +5461,12 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type, return -EIO; } - bh = ext4_bread(handle, inode, blk, 1); + do { + bh = ext4_bread(handle, inode, blk, + EXT4_GET_BLOCKS_CREATE | + EXT4_GET_BLOCKS_METADATA_NOFAIL); + } while (IS_ERR(bh) && (PTR_ERR(bh) == -ENOSPC) && + ext4_should_retry_alloc(inode->i_sb, &retries)); if (IS_ERR(bh)) return PTR_ERR(bh); if (!bh) @@ -5645,6 +5683,7 @@ out7: static void __exit ext4_exit_fs(void) { + ext4_exit_crypto(); ext4_destroy_lazyinit_thread(); unregister_as_ext2(); unregister_as_ext3(); diff --git a/fs/ext4/symlink.c b/fs/ext4/symlink.c index ba5bd18..c677f2c 100644 --- a/fs/ext4/symlink.c +++ b/fs/ext4/symlink.c @@ -23,31 +23,28 @@ #include "xattr.h" #ifdef CONFIG_EXT4_FS_ENCRYPTION -static const char *ext4_follow_link(struct dentry *dentry, void **cookie) +static const char *ext4_encrypted_follow_link(struct dentry *dentry, void **cookie) { struct page *cpage = NULL; char *caddr, *paddr = NULL; struct ext4_str cstr, pstr; struct inode *inode = d_inode(dentry); - struct ext4_fname_crypto_ctx *ctx = NULL; struct ext4_encrypted_symlink_data *sd; loff_t size = min_t(loff_t, i_size_read(inode), PAGE_SIZE - 1); int res; u32 plen, max_size = inode->i_sb->s_blocksize; - ctx = ext4_get_fname_crypto_ctx(inode, inode->i_sb->s_blocksize); - if (IS_ERR(ctx)) - return ERR_CAST(ctx); + res = ext4_get_encryption_info(inode); + if (res) + return ERR_PTR(res); if (ext4_inode_is_fast_symlink(inode)) { caddr = (char *) EXT4_I(inode)->i_data; max_size = sizeof(EXT4_I(inode)->i_data); } else { cpage = read_mapping_page(inode->i_mapping, 0, NULL); - if (IS_ERR(cpage)) { - ext4_put_fname_crypto_ctx(&ctx); + if (IS_ERR(cpage)) return ERR_CAST(cpage); - } caddr = kmap(cpage); caddr[size] = 0; } @@ -71,20 +68,19 @@ static const char *ext4_follow_link(struct dentry *dentry, void **cookie) goto errout; } pstr.name = paddr; - res = _ext4_fname_disk_to_usr(ctx, NULL, &cstr, &pstr); + pstr.len = plen; + res = _ext4_fname_disk_to_usr(inode, NULL, &cstr, &pstr); if (res < 0) goto errout; /* Null-terminate the name */ if (res <= plen) paddr[res] = '\0'; - ext4_put_fname_crypto_ctx(&ctx); if (cpage) { kunmap(cpage); page_cache_release(cpage); } return *cookie = paddr; errout: - ext4_put_fname_crypto_ctx(&ctx); if (cpage) { kunmap(cpage); page_cache_release(cpage); @@ -95,7 +91,7 @@ errout: const struct inode_operations ext4_encrypted_symlink_inode_operations = { .readlink = generic_readlink, - .follow_link = ext4_follow_link, + .follow_link = ext4_encrypted_follow_link, .put_link = kfree_put_link, .setattr = ext4_setattr, .setxattr = generic_setxattr, diff --git a/fs/f2fs/Kconfig b/fs/f2fs/Kconfig index 05f0f66..c629762 100644 --- a/fs/f2fs/Kconfig +++ b/fs/f2fs/Kconfig @@ -72,6 +72,25 @@ config F2FS_CHECK_FS If you want to improve the performance, say N. +config F2FS_FS_ENCRYPTION + bool "F2FS Encryption" + depends on F2FS_FS + depends on F2FS_FS_XATTR + select CRYPTO_AES + select CRYPTO_CBC + select CRYPTO_ECB + select CRYPTO_XTS + select CRYPTO_CTS + select CRYPTO_CTR + select CRYPTO_SHA256 + select KEYS + select ENCRYPTED_KEYS + help + Enable encryption of f2fs files and directories. This + feature is similar to ecryptfs, but it is more memory + efficient since it avoids caching the encrypted and + decrypted pages in the page cache. + config F2FS_IO_TRACE bool "F2FS IO tracer" depends on F2FS_FS diff --git a/fs/f2fs/Makefile b/fs/f2fs/Makefile index d923977..396be1a 100644 --- a/fs/f2fs/Makefile +++ b/fs/f2fs/Makefile @@ -6,3 +6,5 @@ f2fs-$(CONFIG_F2FS_STAT_FS) += debug.o f2fs-$(CONFIG_F2FS_FS_XATTR) += xattr.o f2fs-$(CONFIG_F2FS_FS_POSIX_ACL) += acl.o f2fs-$(CONFIG_F2FS_IO_TRACE) += trace.o +f2fs-$(CONFIG_F2FS_FS_ENCRYPTION) += crypto_policy.o crypto.o \ + crypto_key.o crypto_fname.o diff --git a/fs/f2fs/acl.c b/fs/f2fs/acl.c index 4320ffa..c8f25f7 100644 --- a/fs/f2fs/acl.c +++ b/fs/f2fs/acl.c @@ -334,51 +334,45 @@ static int f2fs_acl_create(struct inode *dir, umode_t *mode, struct page *dpage) { struct posix_acl *p; + struct posix_acl *clone; int ret; + *acl = NULL; + *default_acl = NULL; + if (S_ISLNK(*mode) || !IS_POSIXACL(dir)) - goto no_acl; + return 0; p = __f2fs_get_acl(dir, ACL_TYPE_DEFAULT, dpage); - if (IS_ERR(p)) { - if (p == ERR_PTR(-EOPNOTSUPP)) - goto apply_umask; - return PTR_ERR(p); + if (!p || p == ERR_PTR(-EOPNOTSUPP)) { + *mode &= ~current_umask(); + return 0; } + if (IS_ERR(p)) + return PTR_ERR(p); - if (!p) - goto apply_umask; - - *acl = f2fs_acl_clone(p, GFP_NOFS); - if (!*acl) + clone = f2fs_acl_clone(p, GFP_NOFS); + if (!clone) goto no_mem; - ret = f2fs_acl_create_masq(*acl, mode); + ret = f2fs_acl_create_masq(clone, mode); if (ret < 0) goto no_mem_clone; - if (ret == 0) { - posix_acl_release(*acl); - *acl = NULL; - } + if (ret == 0) + posix_acl_release(clone); + else + *acl = clone; - if (!S_ISDIR(*mode)) { + if (!S_ISDIR(*mode)) posix_acl_release(p); - *default_acl = NULL; - } else { + else *default_acl = p; - } - return 0; -apply_umask: - *mode &= ~current_umask(); -no_acl: - *default_acl = NULL; - *acl = NULL; return 0; no_mem_clone: - posix_acl_release(*acl); + posix_acl_release(clone); no_mem: posix_acl_release(p); return -ENOMEM; diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index a5e17a2..b70bbe1 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -52,9 +52,11 @@ struct page *get_meta_page(struct f2fs_sb_info *sbi, pgoff_t index) struct address_space *mapping = META_MAPPING(sbi); struct page *page; struct f2fs_io_info fio = { + .sbi = sbi, .type = META, .rw = READ_SYNC | REQ_META | REQ_PRIO, .blk_addr = index, + .encrypted_page = NULL, }; repeat: page = grab_cache_page(mapping, index); @@ -65,7 +67,9 @@ repeat: if (PageUptodate(page)) goto out; - if (f2fs_submit_page_bio(sbi, page, &fio)) + fio.page = page; + + if (f2fs_submit_page_bio(&fio)) goto repeat; lock_page(page); @@ -77,8 +81,7 @@ out: return page; } -static inline bool is_valid_blkaddr(struct f2fs_sb_info *sbi, - block_t blkaddr, int type) +bool is_valid_blkaddr(struct f2fs_sb_info *sbi, block_t blkaddr, int type) { switch (type) { case META_NAT: @@ -118,8 +121,10 @@ int ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages, int type struct page *page; block_t blkno = start; struct f2fs_io_info fio = { + .sbi = sbi, .type = META, - .rw = READ_SYNC | REQ_META | REQ_PRIO + .rw = READ_SYNC | REQ_META | REQ_PRIO, + .encrypted_page = NULL, }; for (; nrpages-- > 0; blkno++) { @@ -161,7 +166,8 @@ int ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages, int type continue; } - f2fs_submit_page_mbio(sbi, page, &fio); + fio.page = page; + f2fs_submit_page_mbio(&fio); f2fs_put_page(page, 0); } out: @@ -510,7 +516,12 @@ static void write_orphan_inodes(struct f2fs_sb_info *sbi, block_t start_blk) grab_meta_page(sbi, start_blk + index); index = 1; - spin_lock(&im->ino_lock); + + /* + * we don't need to do spin_lock(&im->ino_lock) here, since all the + * orphan inode operations are covered under f2fs_lock_op(). + * And, spin_lock should be avoided due to page operations below. + */ head = &im->ino_list; /* loop for each orphan inode entry and write them in Jornal block */ @@ -550,8 +561,6 @@ static void write_orphan_inodes(struct f2fs_sb_info *sbi, block_t start_blk) set_page_dirty(page); f2fs_put_page(page, 1); } - - spin_unlock(&im->ino_lock); } static struct page *validate_checkpoint(struct f2fs_sb_info *sbi, @@ -879,10 +888,8 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) unsigned long orphan_num = sbi->im[ORPHAN_INO].ino_num; nid_t last_nid = nm_i->next_scan_nid; block_t start_blk; - struct page *cp_page; unsigned int data_sum_blocks, orphan_blocks; __u32 crc32 = 0; - void *kaddr; int i; int cp_payload_blks = __cp_payload(sbi); @@ -979,19 +986,11 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) start_blk = __start_cp_addr(sbi); /* write out checkpoint buffer at block 0 */ - cp_page = grab_meta_page(sbi, start_blk++); - kaddr = page_address(cp_page); - memcpy(kaddr, ckpt, F2FS_BLKSIZE); - set_page_dirty(cp_page); - f2fs_put_page(cp_page, 1); - - for (i = 1; i < 1 + cp_payload_blks; i++) { - cp_page = grab_meta_page(sbi, start_blk++); - kaddr = page_address(cp_page); - memcpy(kaddr, (char *)ckpt + i * F2FS_BLKSIZE, F2FS_BLKSIZE); - set_page_dirty(cp_page); - f2fs_put_page(cp_page, 1); - } + update_meta_page(sbi, ckpt, start_blk++); + + for (i = 1; i < 1 + cp_payload_blks; i++) + update_meta_page(sbi, (char *)ckpt + i * F2FS_BLKSIZE, + start_blk++); if (orphan_num) { write_orphan_inodes(sbi, start_blk); @@ -1006,11 +1005,7 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) } /* writeout checkpoint block */ - cp_page = grab_meta_page(sbi, start_blk); - kaddr = page_address(cp_page); - memcpy(kaddr, ckpt, F2FS_BLKSIZE); - set_page_dirty(cp_page); - f2fs_put_page(cp_page, 1); + update_meta_page(sbi, ckpt, start_blk); /* wait for previous submitted node/meta pages writeback */ wait_on_all_pages_writeback(sbi); @@ -1036,7 +1031,7 @@ static void do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) if (unlikely(f2fs_cp_error(sbi))) return; - clear_prefree_segments(sbi); + clear_prefree_segments(sbi, cpc); clear_sbi_flag(sbi, SBI_IS_DIRTY); } @@ -1051,7 +1046,8 @@ void write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc) mutex_lock(&sbi->cp_mutex); if (!is_sbi_flag_set(sbi, SBI_IS_DIRTY) && - (cpc->reason == CP_FASTBOOT || cpc->reason == CP_SYNC)) + (cpc->reason == CP_FASTBOOT || cpc->reason == CP_SYNC || + (cpc->reason == CP_DISCARD && !sbi->discard_blks))) goto out; if (unlikely(f2fs_cp_error(sbi))) goto out; diff --git a/fs/f2fs/crypto.c b/fs/f2fs/crypto.c new file mode 100644 index 0000000..4a62ef1 --- /dev/null +++ b/fs/f2fs/crypto.c @@ -0,0 +1,491 @@ +/* + * linux/fs/f2fs/crypto.c + * + * Copied from linux/fs/ext4/crypto.c + * + * Copyright (C) 2015, Google, Inc. + * Copyright (C) 2015, Motorola Mobility + * + * This contains encryption functions for f2fs + * + * Written by Michael Halcrow, 2014. + * + * Filename encryption additions + * Uday Savagaonkar, 2014 + * Encryption policy handling additions + * Ildar Muslukhov, 2014 + * Remove ext4_encrypted_zeroout(), + * add f2fs_restore_and_release_control_page() + * Jaegeuk Kim, 2015. + * + * This has not yet undergone a rigorous security audit. + * + * The usage of AES-XTS should conform to recommendations in NIST + * Special Publication 800-38E and IEEE P1619/D16. + */ +#include <crypto/hash.h> +#include <crypto/sha.h> +#include <keys/user-type.h> +#include <keys/encrypted-type.h> +#include <linux/crypto.h> +#include <linux/ecryptfs.h> +#include <linux/gfp.h> +#include <linux/kernel.h> +#include <linux/key.h> +#include <linux/list.h> +#include <linux/mempool.h> +#include <linux/module.h> +#include <linux/mutex.h> +#include <linux/random.h> +#include <linux/scatterlist.h> +#include <linux/spinlock_types.h> +#include <linux/f2fs_fs.h> +#include <linux/ratelimit.h> +#include <linux/bio.h> + +#include "f2fs.h" +#include "xattr.h" + +/* Encryption added and removed here! (L: */ + +static unsigned int num_prealloc_crypto_pages = 32; +static unsigned int num_prealloc_crypto_ctxs = 128; + +module_param(num_prealloc_crypto_pages, uint, 0444); +MODULE_PARM_DESC(num_prealloc_crypto_pages, + "Number of crypto pages to preallocate"); +module_param(num_prealloc_crypto_ctxs, uint, 0444); +MODULE_PARM_DESC(num_prealloc_crypto_ctxs, + "Number of crypto contexts to preallocate"); + +static mempool_t *f2fs_bounce_page_pool; + +static LIST_HEAD(f2fs_free_crypto_ctxs); +static DEFINE_SPINLOCK(f2fs_crypto_ctx_lock); + +static struct workqueue_struct *f2fs_read_workqueue; +static DEFINE_MUTEX(crypto_init); + +static struct kmem_cache *f2fs_crypto_ctx_cachep; +struct kmem_cache *f2fs_crypt_info_cachep; + +/** + * f2fs_release_crypto_ctx() - Releases an encryption context + * @ctx: The encryption context to release. + * + * If the encryption context was allocated from the pre-allocated pool, returns + * it to that pool. Else, frees it. + * + * If there's a bounce page in the context, this frees that. + */ +void f2fs_release_crypto_ctx(struct f2fs_crypto_ctx *ctx) +{ + unsigned long flags; + + if (ctx->flags & F2FS_WRITE_PATH_FL && ctx->w.bounce_page) { + mempool_free(ctx->w.bounce_page, f2fs_bounce_page_pool); + ctx->w.bounce_page = NULL; + } + ctx->w.control_page = NULL; + if (ctx->flags & F2FS_CTX_REQUIRES_FREE_ENCRYPT_FL) { + kmem_cache_free(f2fs_crypto_ctx_cachep, ctx); + } else { + spin_lock_irqsave(&f2fs_crypto_ctx_lock, flags); + list_add(&ctx->free_list, &f2fs_free_crypto_ctxs); + spin_unlock_irqrestore(&f2fs_crypto_ctx_lock, flags); + } +} + +/** + * f2fs_get_crypto_ctx() - Gets an encryption context + * @inode: The inode for which we are doing the crypto + * + * Allocates and initializes an encryption context. + * + * Return: An allocated and initialized encryption context on success; error + * value or NULL otherwise. + */ +struct f2fs_crypto_ctx *f2fs_get_crypto_ctx(struct inode *inode) +{ + struct f2fs_crypto_ctx *ctx = NULL; + unsigned long flags; + struct f2fs_crypt_info *ci = F2FS_I(inode)->i_crypt_info; + + if (ci == NULL) + return ERR_PTR(-ENOKEY); + + /* + * We first try getting the ctx from a free list because in + * the common case the ctx will have an allocated and + * initialized crypto tfm, so it's probably a worthwhile + * optimization. For the bounce page, we first try getting it + * from the kernel allocator because that's just about as fast + * as getting it from a list and because a cache of free pages + * should generally be a "last resort" option for a filesystem + * to be able to do its job. + */ + spin_lock_irqsave(&f2fs_crypto_ctx_lock, flags); + ctx = list_first_entry_or_null(&f2fs_free_crypto_ctxs, + struct f2fs_crypto_ctx, free_list); + if (ctx) + list_del(&ctx->free_list); + spin_unlock_irqrestore(&f2fs_crypto_ctx_lock, flags); + if (!ctx) { + ctx = kmem_cache_zalloc(f2fs_crypto_ctx_cachep, GFP_NOFS); + if (!ctx) + return ERR_PTR(-ENOMEM); + ctx->flags |= F2FS_CTX_REQUIRES_FREE_ENCRYPT_FL; + } else { + ctx->flags &= ~F2FS_CTX_REQUIRES_FREE_ENCRYPT_FL; + } + ctx->flags &= ~F2FS_WRITE_PATH_FL; + return ctx; +} + +/* + * Call f2fs_decrypt on every single page, reusing the encryption + * context. + */ +static void completion_pages(struct work_struct *work) +{ + struct f2fs_crypto_ctx *ctx = + container_of(work, struct f2fs_crypto_ctx, r.work); + struct bio *bio = ctx->r.bio; + struct bio_vec *bv; + int i; + + bio_for_each_segment_all(bv, bio, i) { + struct page *page = bv->bv_page; + int ret = f2fs_decrypt(ctx, page); + + if (ret) { + WARN_ON_ONCE(1); + SetPageError(page); + } else + SetPageUptodate(page); + unlock_page(page); + } + f2fs_release_crypto_ctx(ctx); + bio_put(bio); +} + +void f2fs_end_io_crypto_work(struct f2fs_crypto_ctx *ctx, struct bio *bio) +{ + INIT_WORK(&ctx->r.work, completion_pages); + ctx->r.bio = bio; + queue_work(f2fs_read_workqueue, &ctx->r.work); +} + +static void f2fs_crypto_destroy(void) +{ + struct f2fs_crypto_ctx *pos, *n; + + list_for_each_entry_safe(pos, n, &f2fs_free_crypto_ctxs, free_list) + kmem_cache_free(f2fs_crypto_ctx_cachep, pos); + INIT_LIST_HEAD(&f2fs_free_crypto_ctxs); + if (f2fs_bounce_page_pool) + mempool_destroy(f2fs_bounce_page_pool); + f2fs_bounce_page_pool = NULL; +} + +/** + * f2fs_crypto_initialize() - Set up for f2fs encryption. + * + * We only call this when we start accessing encrypted files, since it + * results in memory getting allocated that wouldn't otherwise be used. + * + * Return: Zero on success, non-zero otherwise. + */ +int f2fs_crypto_initialize(void) +{ + int i, res = -ENOMEM; + + if (f2fs_bounce_page_pool) + return 0; + + mutex_lock(&crypto_init); + if (f2fs_bounce_page_pool) + goto already_initialized; + + for (i = 0; i < num_prealloc_crypto_ctxs; i++) { + struct f2fs_crypto_ctx *ctx; + + ctx = kmem_cache_zalloc(f2fs_crypto_ctx_cachep, GFP_KERNEL); + if (!ctx) + goto fail; + list_add(&ctx->free_list, &f2fs_free_crypto_ctxs); + } + + /* must be allocated at the last step to avoid race condition above */ + f2fs_bounce_page_pool = + mempool_create_page_pool(num_prealloc_crypto_pages, 0); + if (!f2fs_bounce_page_pool) + goto fail; + +already_initialized: + mutex_unlock(&crypto_init); + return 0; +fail: + f2fs_crypto_destroy(); + mutex_unlock(&crypto_init); + return res; +} + +/** + * f2fs_exit_crypto() - Shutdown the f2fs encryption system + */ +void f2fs_exit_crypto(void) +{ + f2fs_crypto_destroy(); + + if (f2fs_read_workqueue) + destroy_workqueue(f2fs_read_workqueue); + if (f2fs_crypto_ctx_cachep) + kmem_cache_destroy(f2fs_crypto_ctx_cachep); + if (f2fs_crypt_info_cachep) + kmem_cache_destroy(f2fs_crypt_info_cachep); +} + +int __init f2fs_init_crypto(void) +{ + int res = -ENOMEM; + + f2fs_read_workqueue = alloc_workqueue("f2fs_crypto", WQ_HIGHPRI, 0); + if (!f2fs_read_workqueue) + goto fail; + + f2fs_crypto_ctx_cachep = KMEM_CACHE(f2fs_crypto_ctx, + SLAB_RECLAIM_ACCOUNT); + if (!f2fs_crypto_ctx_cachep) + goto fail; + + f2fs_crypt_info_cachep = KMEM_CACHE(f2fs_crypt_info, + SLAB_RECLAIM_ACCOUNT); + if (!f2fs_crypt_info_cachep) + goto fail; + + return 0; +fail: + f2fs_exit_crypto(); + return res; +} + +void f2fs_restore_and_release_control_page(struct page **page) +{ + struct f2fs_crypto_ctx *ctx; + struct page *bounce_page; + + /* The bounce data pages are unmapped. */ + if ((*page)->mapping) + return; + + /* The bounce data page is unmapped. */ + bounce_page = *page; + ctx = (struct f2fs_crypto_ctx *)page_private(bounce_page); + + /* restore control page */ + *page = ctx->w.control_page; + + f2fs_restore_control_page(bounce_page); +} + +void f2fs_restore_control_page(struct page *data_page) +{ + struct f2fs_crypto_ctx *ctx = + (struct f2fs_crypto_ctx *)page_private(data_page); + + set_page_private(data_page, (unsigned long)NULL); + ClearPagePrivate(data_page); + unlock_page(data_page); + f2fs_release_crypto_ctx(ctx); +} + +/** + * f2fs_crypt_complete() - The completion callback for page encryption + * @req: The asynchronous encryption request context + * @res: The result of the encryption operation + */ +static void f2fs_crypt_complete(struct crypto_async_request *req, int res) +{ + struct f2fs_completion_result *ecr = req->data; + + if (res == -EINPROGRESS) + return; + ecr->res = res; + complete(&ecr->completion); +} + +typedef enum { + F2FS_DECRYPT = 0, + F2FS_ENCRYPT, +} f2fs_direction_t; + +static int f2fs_page_crypto(struct f2fs_crypto_ctx *ctx, + struct inode *inode, + f2fs_direction_t rw, + pgoff_t index, + struct page *src_page, + struct page *dest_page) +{ + u8 xts_tweak[F2FS_XTS_TWEAK_SIZE]; + struct ablkcipher_request *req = NULL; + DECLARE_F2FS_COMPLETION_RESULT(ecr); + struct scatterlist dst, src; + struct f2fs_crypt_info *ci = F2FS_I(inode)->i_crypt_info; + struct crypto_ablkcipher *tfm = ci->ci_ctfm; + int res = 0; + + req = ablkcipher_request_alloc(tfm, GFP_NOFS); + if (!req) { + printk_ratelimited(KERN_ERR + "%s: crypto_request_alloc() failed\n", + __func__); + return -ENOMEM; + } + ablkcipher_request_set_callback( + req, CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP, + f2fs_crypt_complete, &ecr); + + BUILD_BUG_ON(F2FS_XTS_TWEAK_SIZE < sizeof(index)); + memcpy(xts_tweak, &index, sizeof(index)); + memset(&xts_tweak[sizeof(index)], 0, + F2FS_XTS_TWEAK_SIZE - sizeof(index)); + + sg_init_table(&dst, 1); + sg_set_page(&dst, dest_page, PAGE_CACHE_SIZE, 0); + sg_init_table(&src, 1); + sg_set_page(&src, src_page, PAGE_CACHE_SIZE, 0); + ablkcipher_request_set_crypt(req, &src, &dst, PAGE_CACHE_SIZE, + xts_tweak); + if (rw == F2FS_DECRYPT) + res = crypto_ablkcipher_decrypt(req); + else + res = crypto_ablkcipher_encrypt(req); + if (res == -EINPROGRESS || res == -EBUSY) { + BUG_ON(req->base.data != &ecr); + wait_for_completion(&ecr.completion); + res = ecr.res; + } + ablkcipher_request_free(req); + if (res) { + printk_ratelimited(KERN_ERR + "%s: crypto_ablkcipher_encrypt() returned %d\n", + __func__, res); + return res; + } + return 0; +} + +static struct page *alloc_bounce_page(struct f2fs_crypto_ctx *ctx) +{ + ctx->w.bounce_page = mempool_alloc(f2fs_bounce_page_pool, GFP_NOWAIT); + if (ctx->w.bounce_page == NULL) + return ERR_PTR(-ENOMEM); + ctx->flags |= F2FS_WRITE_PATH_FL; + return ctx->w.bounce_page; +} + +/** + * f2fs_encrypt() - Encrypts a page + * @inode: The inode for which the encryption should take place + * @plaintext_page: The page to encrypt. Must be locked. + * + * Allocates a ciphertext page and encrypts plaintext_page into it using the ctx + * encryption context. + * + * Called on the page write path. The caller must call + * f2fs_restore_control_page() on the returned ciphertext page to + * release the bounce buffer and the encryption context. + * + * Return: An allocated page with the encrypted content on success. Else, an + * error value or NULL. + */ +struct page *f2fs_encrypt(struct inode *inode, + struct page *plaintext_page) +{ + struct f2fs_crypto_ctx *ctx; + struct page *ciphertext_page = NULL; + int err; + + BUG_ON(!PageLocked(plaintext_page)); + + ctx = f2fs_get_crypto_ctx(inode); + if (IS_ERR(ctx)) + return (struct page *)ctx; + + /* The encryption operation will require a bounce page. */ + ciphertext_page = alloc_bounce_page(ctx); + if (IS_ERR(ciphertext_page)) + goto err_out; + + ctx->w.control_page = plaintext_page; + err = f2fs_page_crypto(ctx, inode, F2FS_ENCRYPT, plaintext_page->index, + plaintext_page, ciphertext_page); + if (err) { + ciphertext_page = ERR_PTR(err); + goto err_out; + } + + SetPagePrivate(ciphertext_page); + set_page_private(ciphertext_page, (unsigned long)ctx); + lock_page(ciphertext_page); + return ciphertext_page; + +err_out: + f2fs_release_crypto_ctx(ctx); + return ciphertext_page; +} + +/** + * f2fs_decrypt() - Decrypts a page in-place + * @ctx: The encryption context. + * @page: The page to decrypt. Must be locked. + * + * Decrypts page in-place using the ctx encryption context. + * + * Called from the read completion callback. + * + * Return: Zero on success, non-zero otherwise. + */ +int f2fs_decrypt(struct f2fs_crypto_ctx *ctx, struct page *page) +{ + BUG_ON(!PageLocked(page)); + + return f2fs_page_crypto(ctx, page->mapping->host, + F2FS_DECRYPT, page->index, page, page); +} + +/* + * Convenience function which takes care of allocating and + * deallocating the encryption context + */ +int f2fs_decrypt_one(struct inode *inode, struct page *page) +{ + struct f2fs_crypto_ctx *ctx = f2fs_get_crypto_ctx(inode); + int ret; + + if (IS_ERR(ctx)) + return PTR_ERR(ctx); + ret = f2fs_decrypt(ctx, page); + f2fs_release_crypto_ctx(ctx); + return ret; +} + +bool f2fs_valid_contents_enc_mode(uint32_t mode) +{ + return (mode == F2FS_ENCRYPTION_MODE_AES_256_XTS); +} + +/** + * f2fs_validate_encryption_key_size() - Validate the encryption key size + * @mode: The key mode. + * @size: The key size to validate. + * + * Return: The validated key size for @mode. Zero if invalid. + */ +uint32_t f2fs_validate_encryption_key_size(uint32_t mode, uint32_t size) +{ + if (size == f2fs_encryption_key_size(mode)) + return size; + return 0; +} diff --git a/fs/f2fs/crypto_fname.c b/fs/f2fs/crypto_fname.c new file mode 100644 index 0000000..ab377d4 --- /dev/null +++ b/fs/f2fs/crypto_fname.c @@ -0,0 +1,440 @@ +/* + * linux/fs/f2fs/crypto_fname.c + * + * Copied from linux/fs/ext4/crypto.c + * + * Copyright (C) 2015, Google, Inc. + * Copyright (C) 2015, Motorola Mobility + * + * This contains functions for filename crypto management in f2fs + * + * Written by Uday Savagaonkar, 2014. + * + * Adjust f2fs dentry structure + * Jaegeuk Kim, 2015. + * + * This has not yet undergone a rigorous security audit. + */ +#include <crypto/hash.h> +#include <crypto/sha.h> +#include <keys/encrypted-type.h> +#include <keys/user-type.h> +#include <linux/crypto.h> +#include <linux/gfp.h> +#include <linux/kernel.h> +#include <linux/key.h> +#include <linux/list.h> +#include <linux/mempool.h> +#include <linux/random.h> +#include <linux/scatterlist.h> +#include <linux/spinlock_types.h> +#include <linux/f2fs_fs.h> +#include <linux/ratelimit.h> + +#include "f2fs.h" +#include "f2fs_crypto.h" +#include "xattr.h" + +/** + * f2fs_dir_crypt_complete() - + */ +static void f2fs_dir_crypt_complete(struct crypto_async_request *req, int res) +{ + struct f2fs_completion_result *ecr = req->data; + + if (res == -EINPROGRESS) + return; + ecr->res = res; + complete(&ecr->completion); +} + +bool f2fs_valid_filenames_enc_mode(uint32_t mode) +{ + return (mode == F2FS_ENCRYPTION_MODE_AES_256_CTS); +} + +static unsigned max_name_len(struct inode *inode) +{ + return S_ISLNK(inode->i_mode) ? inode->i_sb->s_blocksize : + F2FS_NAME_LEN; +} + +/** + * f2fs_fname_encrypt() - + * + * This function encrypts the input filename, and returns the length of the + * ciphertext. Errors are returned as negative numbers. We trust the caller to + * allocate sufficient memory to oname string. + */ +static int f2fs_fname_encrypt(struct inode *inode, + const struct qstr *iname, struct f2fs_str *oname) +{ + u32 ciphertext_len; + struct ablkcipher_request *req = NULL; + DECLARE_F2FS_COMPLETION_RESULT(ecr); + struct f2fs_crypt_info *ci = F2FS_I(inode)->i_crypt_info; + struct crypto_ablkcipher *tfm = ci->ci_ctfm; + int res = 0; + char iv[F2FS_CRYPTO_BLOCK_SIZE]; + struct scatterlist src_sg, dst_sg; + int padding = 4 << (ci->ci_flags & F2FS_POLICY_FLAGS_PAD_MASK); + char *workbuf, buf[32], *alloc_buf = NULL; + unsigned lim = max_name_len(inode); + + if (iname->len <= 0 || iname->len > lim) + return -EIO; + + ciphertext_len = (iname->len < F2FS_CRYPTO_BLOCK_SIZE) ? + F2FS_CRYPTO_BLOCK_SIZE : iname->len; + ciphertext_len = f2fs_fname_crypto_round_up(ciphertext_len, padding); + ciphertext_len = (ciphertext_len > lim) ? lim : ciphertext_len; + + if (ciphertext_len <= sizeof(buf)) { + workbuf = buf; + } else { + alloc_buf = kmalloc(ciphertext_len, GFP_NOFS); + if (!alloc_buf) + return -ENOMEM; + workbuf = alloc_buf; + } + + /* Allocate request */ + req = ablkcipher_request_alloc(tfm, GFP_NOFS); + if (!req) { + printk_ratelimited(KERN_ERR + "%s: crypto_request_alloc() failed\n", __func__); + kfree(alloc_buf); + return -ENOMEM; + } + ablkcipher_request_set_callback(req, + CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP, + f2fs_dir_crypt_complete, &ecr); + + /* Copy the input */ + memcpy(workbuf, iname->name, iname->len); + if (iname->len < ciphertext_len) + memset(workbuf + iname->len, 0, ciphertext_len - iname->len); + + /* Initialize IV */ + memset(iv, 0, F2FS_CRYPTO_BLOCK_SIZE); + + /* Create encryption request */ + sg_init_one(&src_sg, workbuf, ciphertext_len); + sg_init_one(&dst_sg, oname->name, ciphertext_len); + ablkcipher_request_set_crypt(req, &src_sg, &dst_sg, ciphertext_len, iv); + res = crypto_ablkcipher_encrypt(req); + if (res == -EINPROGRESS || res == -EBUSY) { + BUG_ON(req->base.data != &ecr); + wait_for_completion(&ecr.completion); + res = ecr.res; + } + kfree(alloc_buf); + ablkcipher_request_free(req); + if (res < 0) { + printk_ratelimited(KERN_ERR + "%s: Error (error code %d)\n", __func__, res); + } + oname->len = ciphertext_len; + return res; +} + +/* + * f2fs_fname_decrypt() + * This function decrypts the input filename, and returns + * the length of the plaintext. + * Errors are returned as negative numbers. + * We trust the caller to allocate sufficient memory to oname string. + */ +static int f2fs_fname_decrypt(struct inode *inode, + const struct f2fs_str *iname, struct f2fs_str *oname) +{ + struct ablkcipher_request *req = NULL; + DECLARE_F2FS_COMPLETION_RESULT(ecr); + struct scatterlist src_sg, dst_sg; + struct f2fs_crypt_info *ci = F2FS_I(inode)->i_crypt_info; + struct crypto_ablkcipher *tfm = ci->ci_ctfm; + int res = 0; + char iv[F2FS_CRYPTO_BLOCK_SIZE]; + unsigned lim = max_name_len(inode); + + if (iname->len <= 0 || iname->len > lim) + return -EIO; + + /* Allocate request */ + req = ablkcipher_request_alloc(tfm, GFP_NOFS); + if (!req) { + printk_ratelimited(KERN_ERR + "%s: crypto_request_alloc() failed\n", __func__); + return -ENOMEM; + } + ablkcipher_request_set_callback(req, + CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP, + f2fs_dir_crypt_complete, &ecr); + + /* Initialize IV */ + memset(iv, 0, F2FS_CRYPTO_BLOCK_SIZE); + + /* Create decryption request */ + sg_init_one(&src_sg, iname->name, iname->len); + sg_init_one(&dst_sg, oname->name, oname->len); + ablkcipher_request_set_crypt(req, &src_sg, &dst_sg, iname->len, iv); + res = crypto_ablkcipher_decrypt(req); + if (res == -EINPROGRESS || res == -EBUSY) { + BUG_ON(req->base.data != &ecr); + wait_for_completion(&ecr.completion); + res = ecr.res; + } + ablkcipher_request_free(req); + if (res < 0) { + printk_ratelimited(KERN_ERR + "%s: Error in f2fs_fname_decrypt (error code %d)\n", + __func__, res); + return res; + } + + oname->len = strnlen(oname->name, iname->len); + return oname->len; +} + +static const char *lookup_table = + "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+,"; + +/** + * f2fs_fname_encode_digest() - + * + * Encodes the input digest using characters from the set [a-zA-Z0-9_+]. + * The encoded string is roughly 4/3 times the size of the input string. + */ +static int digest_encode(const char *src, int len, char *dst) +{ + int i = 0, bits = 0, ac = 0; + char *cp = dst; + + while (i < len) { + ac += (((unsigned char) src[i]) << bits); + bits += 8; + do { + *cp++ = lookup_table[ac & 0x3f]; + ac >>= 6; + bits -= 6; + } while (bits >= 6); + i++; + } + if (bits) + *cp++ = lookup_table[ac & 0x3f]; + return cp - dst; +} + +static int digest_decode(const char *src, int len, char *dst) +{ + int i = 0, bits = 0, ac = 0; + const char *p; + char *cp = dst; + + while (i < len) { + p = strchr(lookup_table, src[i]); + if (p == NULL || src[i] == 0) + return -2; + ac += (p - lookup_table) << bits; + bits += 6; + if (bits >= 8) { + *cp++ = ac & 0xff; + ac >>= 8; + bits -= 8; + } + i++; + } + if (ac) + return -1; + return cp - dst; +} + +/** + * f2fs_fname_crypto_round_up() - + * + * Return: The next multiple of block size + */ +u32 f2fs_fname_crypto_round_up(u32 size, u32 blksize) +{ + return ((size + blksize - 1) / blksize) * blksize; +} + +/** + * f2fs_fname_crypto_alloc_obuff() - + * + * Allocates an output buffer that is sufficient for the crypto operation + * specified by the context and the direction. + */ +int f2fs_fname_crypto_alloc_buffer(struct inode *inode, + u32 ilen, struct f2fs_str *crypto_str) +{ + unsigned int olen; + int padding = 16; + struct f2fs_crypt_info *ci = F2FS_I(inode)->i_crypt_info; + + if (ci) + padding = 4 << (ci->ci_flags & F2FS_POLICY_FLAGS_PAD_MASK); + if (padding < F2FS_CRYPTO_BLOCK_SIZE) + padding = F2FS_CRYPTO_BLOCK_SIZE; + olen = f2fs_fname_crypto_round_up(ilen, padding); + crypto_str->len = olen; + if (olen < F2FS_FNAME_CRYPTO_DIGEST_SIZE * 2) + olen = F2FS_FNAME_CRYPTO_DIGEST_SIZE * 2; + /* Allocated buffer can hold one more character to null-terminate the + * string */ + crypto_str->name = kmalloc(olen + 1, GFP_NOFS); + if (!(crypto_str->name)) + return -ENOMEM; + return 0; +} + +/** + * f2fs_fname_crypto_free_buffer() - + * + * Frees the buffer allocated for crypto operation. + */ +void f2fs_fname_crypto_free_buffer(struct f2fs_str *crypto_str) +{ + if (!crypto_str) + return; + kfree(crypto_str->name); + crypto_str->name = NULL; +} + +/** + * f2fs_fname_disk_to_usr() - converts a filename from disk space to user space + */ +int f2fs_fname_disk_to_usr(struct inode *inode, + f2fs_hash_t *hash, + const struct f2fs_str *iname, + struct f2fs_str *oname) +{ + const struct qstr qname = FSTR_TO_QSTR(iname); + char buf[24]; + int ret; + + if (is_dot_dotdot(&qname)) { + oname->name[0] = '.'; + oname->name[iname->len - 1] = '.'; + oname->len = iname->len; + return oname->len; + } + + if (F2FS_I(inode)->i_crypt_info) + return f2fs_fname_decrypt(inode, iname, oname); + + if (iname->len <= F2FS_FNAME_CRYPTO_DIGEST_SIZE) { + ret = digest_encode(iname->name, iname->len, oname->name); + oname->len = ret; + return ret; + } + if (hash) { + memcpy(buf, hash, 4); + memset(buf + 4, 0, 4); + } else + memset(buf, 0, 8); + memcpy(buf + 8, iname->name + iname->len - 16, 16); + oname->name[0] = '_'; + ret = digest_encode(buf, 24, oname->name + 1); + oname->len = ret + 1; + return ret + 1; +} + +/** + * f2fs_fname_usr_to_disk() - converts a filename from user space to disk space + */ +int f2fs_fname_usr_to_disk(struct inode *inode, + const struct qstr *iname, + struct f2fs_str *oname) +{ + int res; + struct f2fs_crypt_info *ci = F2FS_I(inode)->i_crypt_info; + + if (is_dot_dotdot(iname)) { + oname->name[0] = '.'; + oname->name[iname->len - 1] = '.'; + oname->len = iname->len; + return oname->len; + } + + if (ci) { + res = f2fs_fname_encrypt(inode, iname, oname); + return res; + } + /* Without a proper key, a user is not allowed to modify the filenames + * in a directory. Consequently, a user space name cannot be mapped to + * a disk-space name */ + return -EACCES; +} + +int f2fs_fname_setup_filename(struct inode *dir, const struct qstr *iname, + int lookup, struct f2fs_filename *fname) +{ + struct f2fs_crypt_info *ci; + int ret = 0, bigname = 0; + + memset(fname, 0, sizeof(struct f2fs_filename)); + fname->usr_fname = iname; + + if (!f2fs_encrypted_inode(dir) || is_dot_dotdot(iname)) { + fname->disk_name.name = (unsigned char *)iname->name; + fname->disk_name.len = iname->len; + return 0; + } + ret = f2fs_get_encryption_info(dir); + if (ret) + return ret; + ci = F2FS_I(dir)->i_crypt_info; + if (ci) { + ret = f2fs_fname_crypto_alloc_buffer(dir, iname->len, + &fname->crypto_buf); + if (ret < 0) + return ret; + ret = f2fs_fname_encrypt(dir, iname, &fname->crypto_buf); + if (ret < 0) + goto errout; + fname->disk_name.name = fname->crypto_buf.name; + fname->disk_name.len = fname->crypto_buf.len; + return 0; + } + if (!lookup) + return -EACCES; + + /* We don't have the key and we are doing a lookup; decode the + * user-supplied name + */ + if (iname->name[0] == '_') + bigname = 1; + if ((bigname && (iname->len != 33)) || + (!bigname && (iname->len > 43))) + return -ENOENT; + + fname->crypto_buf.name = kmalloc(32, GFP_KERNEL); + if (fname->crypto_buf.name == NULL) + return -ENOMEM; + ret = digest_decode(iname->name + bigname, iname->len - bigname, + fname->crypto_buf.name); + if (ret < 0) { + ret = -ENOENT; + goto errout; + } + fname->crypto_buf.len = ret; + if (bigname) { + memcpy(&fname->hash, fname->crypto_buf.name, 4); + } else { + fname->disk_name.name = fname->crypto_buf.name; + fname->disk_name.len = fname->crypto_buf.len; + } + return 0; +errout: + f2fs_fname_crypto_free_buffer(&fname->crypto_buf); + return ret; +} + +void f2fs_fname_free_filename(struct f2fs_filename *fname) +{ + kfree(fname->crypto_buf.name); + fname->crypto_buf.name = NULL; + fname->usr_fname = NULL; + fname->disk_name.name = NULL; +} diff --git a/fs/f2fs/crypto_key.c b/fs/f2fs/crypto_key.c new file mode 100644 index 0000000..95b8f93 --- /dev/null +++ b/fs/f2fs/crypto_key.c @@ -0,0 +1,255 @@ +/* + * linux/fs/f2fs/crypto_key.c + * + * Copied from linux/fs/f2fs/crypto_key.c + * + * Copyright (C) 2015, Google, Inc. + * + * This contains encryption key functions for f2fs + * + * Written by Michael Halcrow, Ildar Muslukhov, and Uday Savagaonkar, 2015. + */ +#include <keys/encrypted-type.h> +#include <keys/user-type.h> +#include <linux/random.h> +#include <linux/scatterlist.h> +#include <uapi/linux/keyctl.h> +#include <crypto/hash.h> +#include <linux/f2fs_fs.h> + +#include "f2fs.h" +#include "xattr.h" + +static void derive_crypt_complete(struct crypto_async_request *req, int rc) +{ + struct f2fs_completion_result *ecr = req->data; + + if (rc == -EINPROGRESS) + return; + + ecr->res = rc; + complete(&ecr->completion); +} + +/** + * f2fs_derive_key_aes() - Derive a key using AES-128-ECB + * @deriving_key: Encryption key used for derivatio. + * @source_key: Source key to which to apply derivation. + * @derived_key: Derived key. + * + * Return: Zero on success; non-zero otherwise. + */ +static int f2fs_derive_key_aes(char deriving_key[F2FS_AES_128_ECB_KEY_SIZE], + char source_key[F2FS_AES_256_XTS_KEY_SIZE], + char derived_key[F2FS_AES_256_XTS_KEY_SIZE]) +{ + int res = 0; + struct ablkcipher_request *req = NULL; + DECLARE_F2FS_COMPLETION_RESULT(ecr); + struct scatterlist src_sg, dst_sg; + struct crypto_ablkcipher *tfm = crypto_alloc_ablkcipher("ecb(aes)", 0, + 0); + + if (IS_ERR(tfm)) { + res = PTR_ERR(tfm); + tfm = NULL; + goto out; + } + crypto_ablkcipher_set_flags(tfm, CRYPTO_TFM_REQ_WEAK_KEY); + req = ablkcipher_request_alloc(tfm, GFP_NOFS); + if (!req) { + res = -ENOMEM; + goto out; + } + ablkcipher_request_set_callback(req, + CRYPTO_TFM_REQ_MAY_BACKLOG | CRYPTO_TFM_REQ_MAY_SLEEP, + derive_crypt_complete, &ecr); + res = crypto_ablkcipher_setkey(tfm, deriving_key, + F2FS_AES_128_ECB_KEY_SIZE); + if (res < 0) + goto out; + + sg_init_one(&src_sg, source_key, F2FS_AES_256_XTS_KEY_SIZE); + sg_init_one(&dst_sg, derived_key, F2FS_AES_256_XTS_KEY_SIZE); + ablkcipher_request_set_crypt(req, &src_sg, &dst_sg, + F2FS_AES_256_XTS_KEY_SIZE, NULL); + res = crypto_ablkcipher_encrypt(req); + if (res == -EINPROGRESS || res == -EBUSY) { + BUG_ON(req->base.data != &ecr); + wait_for_completion(&ecr.completion); + res = ecr.res; + } +out: + if (req) + ablkcipher_request_free(req); + if (tfm) + crypto_free_ablkcipher(tfm); + return res; +} + +static void f2fs_free_crypt_info(struct f2fs_crypt_info *ci) +{ + if (!ci) + return; + + if (ci->ci_keyring_key) + key_put(ci->ci_keyring_key); + crypto_free_ablkcipher(ci->ci_ctfm); + kmem_cache_free(f2fs_crypt_info_cachep, ci); +} + +void f2fs_free_encryption_info(struct inode *inode, struct f2fs_crypt_info *ci) +{ + struct f2fs_inode_info *fi = F2FS_I(inode); + struct f2fs_crypt_info *prev; + + if (ci == NULL) + ci = ACCESS_ONCE(fi->i_crypt_info); + if (ci == NULL) + return; + prev = cmpxchg(&fi->i_crypt_info, ci, NULL); + if (prev != ci) + return; + + f2fs_free_crypt_info(ci); +} + +int _f2fs_get_encryption_info(struct inode *inode) +{ + struct f2fs_inode_info *fi = F2FS_I(inode); + struct f2fs_crypt_info *crypt_info; + char full_key_descriptor[F2FS_KEY_DESC_PREFIX_SIZE + + (F2FS_KEY_DESCRIPTOR_SIZE * 2) + 1]; + struct key *keyring_key = NULL; + struct f2fs_encryption_key *master_key; + struct f2fs_encryption_context ctx; + struct user_key_payload *ukp; + struct crypto_ablkcipher *ctfm; + const char *cipher_str; + char raw_key[F2FS_MAX_KEY_SIZE]; + char mode; + int res; + + res = f2fs_crypto_initialize(); + if (res) + return res; +retry: + crypt_info = ACCESS_ONCE(fi->i_crypt_info); + if (crypt_info) { + if (!crypt_info->ci_keyring_key || + key_validate(crypt_info->ci_keyring_key) == 0) + return 0; + f2fs_free_encryption_info(inode, crypt_info); + goto retry; + } + + res = f2fs_getxattr(inode, F2FS_XATTR_INDEX_ENCRYPTION, + F2FS_XATTR_NAME_ENCRYPTION_CONTEXT, + &ctx, sizeof(ctx), NULL); + if (res < 0) + return res; + else if (res != sizeof(ctx)) + return -EINVAL; + res = 0; + + crypt_info = kmem_cache_alloc(f2fs_crypt_info_cachep, GFP_NOFS); + if (!crypt_info) + return -ENOMEM; + + crypt_info->ci_flags = ctx.flags; + crypt_info->ci_data_mode = ctx.contents_encryption_mode; + crypt_info->ci_filename_mode = ctx.filenames_encryption_mode; + crypt_info->ci_ctfm = NULL; + crypt_info->ci_keyring_key = NULL; + memcpy(crypt_info->ci_master_key, ctx.master_key_descriptor, + sizeof(crypt_info->ci_master_key)); + if (S_ISREG(inode->i_mode)) + mode = crypt_info->ci_data_mode; + else if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) + mode = crypt_info->ci_filename_mode; + else + BUG(); + + switch (mode) { + case F2FS_ENCRYPTION_MODE_AES_256_XTS: + cipher_str = "xts(aes)"; + break; + case F2FS_ENCRYPTION_MODE_AES_256_CTS: + cipher_str = "cts(cbc(aes))"; + break; + default: + printk_once(KERN_WARNING + "f2fs: unsupported key mode %d (ino %u)\n", + mode, (unsigned) inode->i_ino); + res = -ENOKEY; + goto out; + } + + memcpy(full_key_descriptor, F2FS_KEY_DESC_PREFIX, + F2FS_KEY_DESC_PREFIX_SIZE); + sprintf(full_key_descriptor + F2FS_KEY_DESC_PREFIX_SIZE, + "%*phN", F2FS_KEY_DESCRIPTOR_SIZE, + ctx.master_key_descriptor); + full_key_descriptor[F2FS_KEY_DESC_PREFIX_SIZE + + (2 * F2FS_KEY_DESCRIPTOR_SIZE)] = '\0'; + keyring_key = request_key(&key_type_logon, full_key_descriptor, NULL); + if (IS_ERR(keyring_key)) { + res = PTR_ERR(keyring_key); + keyring_key = NULL; + goto out; + } + crypt_info->ci_keyring_key = keyring_key; + BUG_ON(keyring_key->type != &key_type_logon); + ukp = ((struct user_key_payload *)keyring_key->payload.data); + if (ukp->datalen != sizeof(struct f2fs_encryption_key)) { + res = -EINVAL; + goto out; + } + master_key = (struct f2fs_encryption_key *)ukp->data; + BUILD_BUG_ON(F2FS_AES_128_ECB_KEY_SIZE != + F2FS_KEY_DERIVATION_NONCE_SIZE); + BUG_ON(master_key->size != F2FS_AES_256_XTS_KEY_SIZE); + res = f2fs_derive_key_aes(ctx.nonce, master_key->raw, + raw_key); + if (res) + goto out; + + ctfm = crypto_alloc_ablkcipher(cipher_str, 0, 0); + if (!ctfm || IS_ERR(ctfm)) { + res = ctfm ? PTR_ERR(ctfm) : -ENOMEM; + printk(KERN_DEBUG + "%s: error %d (inode %u) allocating crypto tfm\n", + __func__, res, (unsigned) inode->i_ino); + goto out; + } + crypt_info->ci_ctfm = ctfm; + crypto_ablkcipher_clear_flags(ctfm, ~0); + crypto_tfm_set_flags(crypto_ablkcipher_tfm(ctfm), + CRYPTO_TFM_REQ_WEAK_KEY); + res = crypto_ablkcipher_setkey(ctfm, raw_key, + f2fs_encryption_key_size(mode)); + if (res) + goto out; + + memzero_explicit(raw_key, sizeof(raw_key)); + if (cmpxchg(&fi->i_crypt_info, NULL, crypt_info) != NULL) { + f2fs_free_crypt_info(crypt_info); + goto retry; + } + return 0; + +out: + if (res == -ENOKEY && !S_ISREG(inode->i_mode)) + res = 0; + + f2fs_free_crypt_info(crypt_info); + memzero_explicit(raw_key, sizeof(raw_key)); + return res; +} + +int f2fs_has_encryption_key(struct inode *inode) +{ + struct f2fs_inode_info *fi = F2FS_I(inode); + + return (fi->i_crypt_info != NULL); +} diff --git a/fs/f2fs/crypto_policy.c b/fs/f2fs/crypto_policy.c new file mode 100644 index 0000000..d4a96af --- /dev/null +++ b/fs/f2fs/crypto_policy.c @@ -0,0 +1,209 @@ +/* + * copied from linux/fs/ext4/crypto_policy.c + * + * Copyright (C) 2015, Google, Inc. + * Copyright (C) 2015, Motorola Mobility. + * + * This contains encryption policy functions for f2fs with some modifications + * to support f2fs-specific xattr APIs. + * + * Written by Michael Halcrow, 2015. + * Modified by Jaegeuk Kim, 2015. + */ +#include <linux/random.h> +#include <linux/string.h> +#include <linux/types.h> +#include <linux/f2fs_fs.h> + +#include "f2fs.h" +#include "xattr.h" + +static int f2fs_inode_has_encryption_context(struct inode *inode) +{ + int res = f2fs_getxattr(inode, F2FS_XATTR_INDEX_ENCRYPTION, + F2FS_XATTR_NAME_ENCRYPTION_CONTEXT, NULL, 0, NULL); + return (res > 0); +} + +/* + * check whether the policy is consistent with the encryption context + * for the inode + */ +static int f2fs_is_encryption_context_consistent_with_policy( + struct inode *inode, const struct f2fs_encryption_policy *policy) +{ + struct f2fs_encryption_context ctx; + int res = f2fs_getxattr(inode, F2FS_XATTR_INDEX_ENCRYPTION, + F2FS_XATTR_NAME_ENCRYPTION_CONTEXT, &ctx, + sizeof(ctx), NULL); + + if (res != sizeof(ctx)) + return 0; + + return (memcmp(ctx.master_key_descriptor, policy->master_key_descriptor, + F2FS_KEY_DESCRIPTOR_SIZE) == 0 && + (ctx.flags == policy->flags) && + (ctx.contents_encryption_mode == + policy->contents_encryption_mode) && + (ctx.filenames_encryption_mode == + policy->filenames_encryption_mode)); +} + +static int f2fs_create_encryption_context_from_policy( + struct inode *inode, const struct f2fs_encryption_policy *policy) +{ + struct f2fs_encryption_context ctx; + + ctx.format = F2FS_ENCRYPTION_CONTEXT_FORMAT_V1; + memcpy(ctx.master_key_descriptor, policy->master_key_descriptor, + F2FS_KEY_DESCRIPTOR_SIZE); + + if (!f2fs_valid_contents_enc_mode(policy->contents_encryption_mode)) { + printk(KERN_WARNING + "%s: Invalid contents encryption mode %d\n", __func__, + policy->contents_encryption_mode); + return -EINVAL; + } + + if (!f2fs_valid_filenames_enc_mode(policy->filenames_encryption_mode)) { + printk(KERN_WARNING + "%s: Invalid filenames encryption mode %d\n", __func__, + policy->filenames_encryption_mode); + return -EINVAL; + } + + if (policy->flags & ~F2FS_POLICY_FLAGS_VALID) + return -EINVAL; + + ctx.contents_encryption_mode = policy->contents_encryption_mode; + ctx.filenames_encryption_mode = policy->filenames_encryption_mode; + ctx.flags = policy->flags; + BUILD_BUG_ON(sizeof(ctx.nonce) != F2FS_KEY_DERIVATION_NONCE_SIZE); + get_random_bytes(ctx.nonce, F2FS_KEY_DERIVATION_NONCE_SIZE); + + return f2fs_setxattr(inode, F2FS_XATTR_INDEX_ENCRYPTION, + F2FS_XATTR_NAME_ENCRYPTION_CONTEXT, &ctx, + sizeof(ctx), NULL, XATTR_CREATE); +} + +int f2fs_process_policy(const struct f2fs_encryption_policy *policy, + struct inode *inode) +{ + if (policy->version != 0) + return -EINVAL; + + if (!S_ISDIR(inode->i_mode)) + return -EINVAL; + + if (!f2fs_inode_has_encryption_context(inode)) { + if (!f2fs_empty_dir(inode)) + return -ENOTEMPTY; + return f2fs_create_encryption_context_from_policy(inode, + policy); + } + + if (f2fs_is_encryption_context_consistent_with_policy(inode, policy)) + return 0; + + printk(KERN_WARNING "%s: Policy inconsistent with encryption context\n", + __func__); + return -EINVAL; +} + +int f2fs_get_policy(struct inode *inode, struct f2fs_encryption_policy *policy) +{ + struct f2fs_encryption_context ctx; + int res; + + if (!f2fs_encrypted_inode(inode)) + return -ENODATA; + + res = f2fs_getxattr(inode, F2FS_XATTR_INDEX_ENCRYPTION, + F2FS_XATTR_NAME_ENCRYPTION_CONTEXT, + &ctx, sizeof(ctx), NULL); + if (res != sizeof(ctx)) + return -ENODATA; + if (ctx.format != F2FS_ENCRYPTION_CONTEXT_FORMAT_V1) + return -EINVAL; + + policy->version = 0; + policy->contents_encryption_mode = ctx.contents_encryption_mode; + policy->filenames_encryption_mode = ctx.filenames_encryption_mode; + policy->flags = ctx.flags; + memcpy(&policy->master_key_descriptor, ctx.master_key_descriptor, + F2FS_KEY_DESCRIPTOR_SIZE); + return 0; +} + +int f2fs_is_child_context_consistent_with_parent(struct inode *parent, + struct inode *child) +{ + struct f2fs_crypt_info *parent_ci, *child_ci; + int res; + + if ((parent == NULL) || (child == NULL)) { + pr_err("parent %p child %p\n", parent, child); + BUG_ON(1); + } + + /* no restrictions if the parent directory is not encrypted */ + if (!f2fs_encrypted_inode(parent)) + return 1; + /* if the child directory is not encrypted, this is always a problem */ + if (!f2fs_encrypted_inode(child)) + return 0; + res = f2fs_get_encryption_info(parent); + if (res) + return 0; + res = f2fs_get_encryption_info(child); + if (res) + return 0; + parent_ci = F2FS_I(parent)->i_crypt_info; + child_ci = F2FS_I(child)->i_crypt_info; + if (!parent_ci && !child_ci) + return 1; + if (!parent_ci || !child_ci) + return 0; + + return (memcmp(parent_ci->ci_master_key, + child_ci->ci_master_key, + F2FS_KEY_DESCRIPTOR_SIZE) == 0 && + (parent_ci->ci_data_mode == child_ci->ci_data_mode) && + (parent_ci->ci_filename_mode == child_ci->ci_filename_mode) && + (parent_ci->ci_flags == child_ci->ci_flags)); +} + +/** + * f2fs_inherit_context() - Sets a child context from its parent + * @parent: Parent inode from which the context is inherited. + * @child: Child inode that inherits the context from @parent. + * + * Return: Zero on success, non-zero otherwise + */ +int f2fs_inherit_context(struct inode *parent, struct inode *child, + struct page *ipage) +{ + struct f2fs_encryption_context ctx; + struct f2fs_crypt_info *ci; + int res; + + res = f2fs_get_encryption_info(parent); + if (res < 0) + return res; + + ci = F2FS_I(parent)->i_crypt_info; + BUG_ON(ci == NULL); + + ctx.format = F2FS_ENCRYPTION_CONTEXT_FORMAT_V1; + + ctx.contents_encryption_mode = ci->ci_data_mode; + ctx.filenames_encryption_mode = ci->ci_filename_mode; + ctx.flags = ci->ci_flags; + memcpy(ctx.master_key_descriptor, ci->ci_master_key, + F2FS_KEY_DESCRIPTOR_SIZE); + + get_random_bytes(ctx.nonce, F2FS_KEY_DERIVATION_NONCE_SIZE); + return f2fs_setxattr(child, F2FS_XATTR_INDEX_ENCRYPTION, + F2FS_XATTR_NAME_ENCRYPTION_CONTEXT, &ctx, + sizeof(ctx), ipage, XATTR_CREATE); +} diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 1e1aae6..9bedfa8 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -18,6 +18,7 @@ #include <linux/bio.h> #include <linux/prefetch.h> #include <linux/uio.h> +#include <linux/cleancache.h> #include "f2fs.h" #include "node.h" @@ -33,6 +34,15 @@ static void f2fs_read_end_io(struct bio *bio, int err) struct bio_vec *bvec; int i; + if (f2fs_bio_encrypted(bio)) { + if (err) { + f2fs_release_crypto_ctx(bio->bi_private); + } else { + f2fs_end_io_crypto_work(bio->bi_private, bio); + return; + } + } + bio_for_each_segment_all(bvec, bio, i) { struct page *page = bvec->bv_page; @@ -56,6 +66,8 @@ static void f2fs_write_end_io(struct bio *bio, int err) bio_for_each_segment_all(bvec, bio, i) { struct page *page = bvec->bv_page; + f2fs_restore_and_release_control_page(&page); + if (unlikely(err)) { set_page_dirty(page); set_bit(AS_EIO, &page->mapping->flags); @@ -86,7 +98,7 @@ static struct bio *__bio_alloc(struct f2fs_sb_info *sbi, block_t blk_addr, bio->bi_bdev = sbi->sb->s_bdev; bio->bi_iter.bi_sector = SECTOR_FROM_BLOCK(blk_addr); bio->bi_end_io = is_read ? f2fs_read_end_io : f2fs_write_end_io; - bio->bi_private = sbi; + bio->bi_private = is_read ? NULL : sbi; return bio; } @@ -133,16 +145,16 @@ void f2fs_submit_merged_bio(struct f2fs_sb_info *sbi, * Fill the locked page with data located in the block address. * Return unlocked page. */ -int f2fs_submit_page_bio(struct f2fs_sb_info *sbi, struct page *page, - struct f2fs_io_info *fio) +int f2fs_submit_page_bio(struct f2fs_io_info *fio) { struct bio *bio; + struct page *page = fio->encrypted_page ? fio->encrypted_page : fio->page; trace_f2fs_submit_page_bio(page, fio); - f2fs_trace_ios(page, fio, 0); + f2fs_trace_ios(fio, 0); /* Allocate a new bio */ - bio = __bio_alloc(sbi, fio->blk_addr, 1, is_read_io(fio->rw)); + bio = __bio_alloc(fio->sbi, fio->blk_addr, 1, is_read_io(fio->rw)); if (bio_add_page(bio, page, PAGE_CACHE_SIZE, 0) < PAGE_CACHE_SIZE) { bio_put(bio); @@ -154,12 +166,13 @@ int f2fs_submit_page_bio(struct f2fs_sb_info *sbi, struct page *page, return 0; } -void f2fs_submit_page_mbio(struct f2fs_sb_info *sbi, struct page *page, - struct f2fs_io_info *fio) +void f2fs_submit_page_mbio(struct f2fs_io_info *fio) { + struct f2fs_sb_info *sbi = fio->sbi; enum page_type btype = PAGE_TYPE_OF_BIO(fio->type); struct f2fs_bio_info *io; bool is_read = is_read_io(fio->rw); + struct page *bio_page; io = is_read ? &sbi->read_io : &sbi->write_io[btype]; @@ -181,17 +194,19 @@ alloc_new: io->fio = *fio; } - if (bio_add_page(io->bio, page, PAGE_CACHE_SIZE, 0) < + bio_page = fio->encrypted_page ? fio->encrypted_page : fio->page; + + if (bio_add_page(io->bio, bio_page, PAGE_CACHE_SIZE, 0) < PAGE_CACHE_SIZE) { __submit_merged_bio(io); goto alloc_new; } io->last_block_in_bio = fio->blk_addr; - f2fs_trace_ios(page, fio, 0); + f2fs_trace_ios(fio, 0); up_write(&io->io_rwsem); - trace_f2fs_submit_page_mbio(page, fio); + trace_f2fs_submit_page_mbio(fio->page, fio); } /* @@ -251,19 +266,6 @@ int f2fs_reserve_block(struct dnode_of_data *dn, pgoff_t index) return err; } -static void f2fs_map_bh(struct super_block *sb, pgoff_t pgofs, - struct extent_info *ei, struct buffer_head *bh_result) -{ - unsigned int blkbits = sb->s_blocksize_bits; - size_t max_size = bh_result->b_size; - size_t mapped_size; - - clear_buffer_new(bh_result); - map_bh(bh_result, sb, ei->blk + pgofs - ei->fofs); - mapped_size = (ei->fofs + ei->len - pgofs) << blkbits; - bh_result->b_size = min(max_size, mapped_size); -} - static bool lookup_extent_info(struct inode *inode, pgoff_t pgofs, struct extent_info *ei) { @@ -905,7 +907,7 @@ void f2fs_update_extent_cache(struct dnode_of_data *dn) sync_inode_page(dn); } -struct page *find_data_page(struct inode *inode, pgoff_t index, bool sync) +struct page *get_read_data_page(struct inode *inode, pgoff_t index, int rw) { struct address_space *mapping = inode->i_mapping; struct dnode_of_data dn; @@ -913,83 +915,15 @@ struct page *find_data_page(struct inode *inode, pgoff_t index, bool sync) struct extent_info ei; int err; struct f2fs_io_info fio = { + .sbi = F2FS_I_SB(inode), .type = DATA, - .rw = sync ? READ_SYNC : READA, + .rw = rw, + .encrypted_page = NULL, }; - /* - * If sync is false, it needs to check its block allocation. - * This is need and triggered by two flows: - * gc and truncate_partial_data_page. - */ - if (!sync) - goto search; - - page = find_get_page(mapping, index); - if (page && PageUptodate(page)) - return page; - f2fs_put_page(page, 0); -search: - if (f2fs_lookup_extent_cache(inode, index, &ei)) { - dn.data_blkaddr = ei.blk + index - ei.fofs; - goto got_it; - } - - set_new_dnode(&dn, inode, NULL, NULL, 0); - err = get_dnode_of_data(&dn, index, LOOKUP_NODE); - if (err) - return ERR_PTR(err); - f2fs_put_dnode(&dn); + if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) + return read_mapping_page(mapping, index, NULL); - if (dn.data_blkaddr == NULL_ADDR) - return ERR_PTR(-ENOENT); - - /* By fallocate(), there is no cached page, but with NEW_ADDR */ - if (unlikely(dn.data_blkaddr == NEW_ADDR)) - return ERR_PTR(-EINVAL); - -got_it: - page = grab_cache_page(mapping, index); - if (!page) - return ERR_PTR(-ENOMEM); - - if (PageUptodate(page)) { - unlock_page(page); - return page; - } - - fio.blk_addr = dn.data_blkaddr; - err = f2fs_submit_page_bio(F2FS_I_SB(inode), page, &fio); - if (err) - return ERR_PTR(err); - - if (sync) { - wait_on_page_locked(page); - if (unlikely(!PageUptodate(page))) { - f2fs_put_page(page, 0); - return ERR_PTR(-EIO); - } - } - return page; -} - -/* - * If it tries to access a hole, return an error. - * Because, the callers, functions in dir.c and GC, should be able to know - * whether this page exists or not. - */ -struct page *get_lock_data_page(struct inode *inode, pgoff_t index) -{ - struct address_space *mapping = inode->i_mapping; - struct dnode_of_data dn; - struct page *page; - struct extent_info ei; - int err; - struct f2fs_io_info fio = { - .type = DATA, - .rw = READ_SYNC, - }; -repeat: page = grab_cache_page(mapping, index); if (!page) return ERR_PTR(-ENOMEM); @@ -1011,10 +945,11 @@ repeat: f2fs_put_page(page, 1); return ERR_PTR(-ENOENT); } - got_it: - if (PageUptodate(page)) + if (PageUptodate(page)) { + unlock_page(page); return page; + } /* * A new dentry page is allocated but not able to be written, since its @@ -1025,14 +960,58 @@ got_it: if (dn.data_blkaddr == NEW_ADDR) { zero_user_segment(page, 0, PAGE_CACHE_SIZE); SetPageUptodate(page); + unlock_page(page); return page; } fio.blk_addr = dn.data_blkaddr; - err = f2fs_submit_page_bio(F2FS_I_SB(inode), page, &fio); + fio.page = page; + err = f2fs_submit_page_bio(&fio); if (err) return ERR_PTR(err); + return page; +} + +struct page *find_data_page(struct inode *inode, pgoff_t index) +{ + struct address_space *mapping = inode->i_mapping; + struct page *page; + + page = find_get_page(mapping, index); + if (page && PageUptodate(page)) + return page; + f2fs_put_page(page, 0); + page = get_read_data_page(inode, index, READ_SYNC); + if (IS_ERR(page)) + return page; + + if (PageUptodate(page)) + return page; + + wait_on_page_locked(page); + if (unlikely(!PageUptodate(page))) { + f2fs_put_page(page, 0); + return ERR_PTR(-EIO); + } + return page; +} + +/* + * If it tries to access a hole, return an error. + * Because, the callers, functions in dir.c and GC, should be able to know + * whether this page exists or not. + */ +struct page *get_lock_data_page(struct inode *inode, pgoff_t index) +{ + struct address_space *mapping = inode->i_mapping; + struct page *page; +repeat: + page = get_read_data_page(inode, index, READ_SYNC); + if (IS_ERR(page)) + return page; + + /* wait for read completion */ lock_page(page); if (unlikely(!PageUptodate(page))) { f2fs_put_page(page, 1); @@ -1060,46 +1039,37 @@ struct page *get_new_data_page(struct inode *inode, struct page *page; struct dnode_of_data dn; int err; +repeat: + page = grab_cache_page(mapping, index); + if (!page) + return ERR_PTR(-ENOMEM); set_new_dnode(&dn, inode, ipage, NULL, 0); err = f2fs_reserve_block(&dn, index); - if (err) + if (err) { + f2fs_put_page(page, 1); return ERR_PTR(err); -repeat: - page = grab_cache_page(mapping, index); - if (!page) { - err = -ENOMEM; - goto put_err; } + if (!ipage) + f2fs_put_dnode(&dn); if (PageUptodate(page)) - return page; + goto got_it; if (dn.data_blkaddr == NEW_ADDR) { zero_user_segment(page, 0, PAGE_CACHE_SIZE); SetPageUptodate(page); } else { - struct f2fs_io_info fio = { - .type = DATA, - .rw = READ_SYNC, - .blk_addr = dn.data_blkaddr, - }; - err = f2fs_submit_page_bio(F2FS_I_SB(inode), page, &fio); - if (err) - goto put_err; + f2fs_put_page(page, 1); - lock_page(page); - if (unlikely(!PageUptodate(page))) { - f2fs_put_page(page, 1); - err = -EIO; - goto put_err; - } - if (unlikely(page->mapping != mapping)) { - f2fs_put_page(page, 1); + page = get_read_data_page(inode, index, READ_SYNC); + if (IS_ERR(page)) goto repeat; - } - } + /* wait for read completion */ + lock_page(page); + } +got_it: if (new_i_size && i_size_read(inode) < ((index + 1) << PAGE_CACHE_SHIFT)) { i_size_write(inode, ((index + 1) << PAGE_CACHE_SHIFT)); @@ -1107,10 +1077,6 @@ repeat: set_inode_flag(F2FS_I(inode), FI_UPDATE_DIR); } return page; - -put_err: - f2fs_put_dnode(&dn); - return ERR_PTR(err); } static int __allocate_data_block(struct dnode_of_data *dn) @@ -1208,18 +1174,18 @@ out: } /* - * get_data_block() now supported readahead/bmap/rw direct_IO with mapped bh. + * f2fs_map_blocks() now supported readahead/bmap/rw direct_IO with + * f2fs_map_blocks structure. * If original data blocks are allocated, then give them to blockdev. * Otherwise, * a. preallocate requested block addresses * b. do not use extent cache for better performance * c. give the block addresses to blockdev */ -static int __get_data_block(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create, bool fiemap) +static int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, + int create, bool fiemap) { - unsigned int blkbits = inode->i_sb->s_blocksize_bits; - unsigned maxblocks = bh_result->b_size >> blkbits; + unsigned int maxblocks = map->m_len; struct dnode_of_data dn; int mode = create ? ALLOC_NODE : LOOKUP_NODE_RA; pgoff_t pgofs, end_offset; @@ -1227,11 +1193,16 @@ static int __get_data_block(struct inode *inode, sector_t iblock, struct extent_info ei; bool allocated = false; - /* Get the page offset from the block offset(iblock) */ - pgofs = (pgoff_t)(iblock >> (PAGE_CACHE_SHIFT - blkbits)); + map->m_len = 0; + map->m_flags = 0; + + /* it only supports block size == page size */ + pgofs = (pgoff_t)map->m_lblk; if (f2fs_lookup_extent_cache(inode, pgofs, &ei)) { - f2fs_map_bh(inode->i_sb, pgofs, &ei, bh_result); + map->m_pblk = ei.blk + pgofs - ei.fofs; + map->m_len = min((pgoff_t)maxblocks, ei.fofs + ei.len - pgofs); + map->m_flags = F2FS_MAP_MAPPED; goto out; } @@ -1250,21 +1221,23 @@ static int __get_data_block(struct inode *inode, sector_t iblock, goto put_out; if (dn.data_blkaddr != NULL_ADDR) { - clear_buffer_new(bh_result); - map_bh(bh_result, inode->i_sb, dn.data_blkaddr); + map->m_flags = F2FS_MAP_MAPPED; + map->m_pblk = dn.data_blkaddr; + if (dn.data_blkaddr == NEW_ADDR) + map->m_flags |= F2FS_MAP_UNWRITTEN; } else if (create) { err = __allocate_data_block(&dn); if (err) goto put_out; allocated = true; - set_buffer_new(bh_result); - map_bh(bh_result, inode->i_sb, dn.data_blkaddr); + map->m_flags = F2FS_MAP_NEW | F2FS_MAP_MAPPED; + map->m_pblk = dn.data_blkaddr; } else { goto put_out; } end_offset = ADDRS_PER_PAGE(dn.node_page, F2FS_I(inode)); - bh_result->b_size = (((size_t)1) << blkbits); + map->m_len = 1; dn.ofs_in_node++; pgofs++; @@ -1288,22 +1261,25 @@ get_next: end_offset = ADDRS_PER_PAGE(dn.node_page, F2FS_I(inode)); } - if (maxblocks > (bh_result->b_size >> blkbits)) { + if (maxblocks > map->m_len) { block_t blkaddr = datablock_addr(dn.node_page, dn.ofs_in_node); if (blkaddr == NULL_ADDR && create) { err = __allocate_data_block(&dn); if (err) goto sync_out; allocated = true; - set_buffer_new(bh_result); + map->m_flags |= F2FS_MAP_NEW; blkaddr = dn.data_blkaddr; } /* Give more consecutive addresses for the readahead */ - if (blkaddr == (bh_result->b_blocknr + ofs)) { + if ((map->m_pblk != NEW_ADDR && + blkaddr == (map->m_pblk + ofs)) || + (map->m_pblk == NEW_ADDR && + blkaddr == NEW_ADDR)) { ofs++; dn.ofs_in_node++; pgofs++; - bh_result->b_size += (((size_t)1) << blkbits); + map->m_len++; goto get_next; } } @@ -1316,10 +1292,28 @@ unlock_out: if (create) f2fs_unlock_op(F2FS_I_SB(inode)); out: - trace_f2fs_get_data_block(inode, iblock, bh_result, err); + trace_f2fs_map_blocks(inode, map, err); return err; } +static int __get_data_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh, int create, bool fiemap) +{ + struct f2fs_map_blocks map; + int ret; + + map.m_lblk = iblock; + map.m_len = bh->b_size >> inode->i_blkbits; + + ret = f2fs_map_blocks(inode, &map, create, fiemap); + if (!ret) { + map_bh(bh, inode->i_sb, map.m_pblk); + bh->b_state = (bh->b_state & ~F2FS_MAP_FLAGS) | map.m_flags; + bh->b_size = map.m_len << inode->i_blkbits; + } + return ret; +} + static int get_data_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) { @@ -1332,11 +1326,268 @@ static int get_data_block_fiemap(struct inode *inode, sector_t iblock, return __get_data_block(inode, iblock, bh_result, create, true); } +static inline sector_t logical_to_blk(struct inode *inode, loff_t offset) +{ + return (offset >> inode->i_blkbits); +} + +static inline loff_t blk_to_logical(struct inode *inode, sector_t blk) +{ + return (blk << inode->i_blkbits); +} + int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, u64 start, u64 len) { - return generic_block_fiemap(inode, fieinfo, - start, len, get_data_block_fiemap); + struct buffer_head map_bh; + sector_t start_blk, last_blk; + loff_t isize = i_size_read(inode); + u64 logical = 0, phys = 0, size = 0; + u32 flags = 0; + bool past_eof = false, whole_file = false; + int ret = 0; + + ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC); + if (ret) + return ret; + + mutex_lock(&inode->i_mutex); + + if (len >= isize) { + whole_file = true; + len = isize; + } + + if (logical_to_blk(inode, len) == 0) + len = blk_to_logical(inode, 1); + + start_blk = logical_to_blk(inode, start); + last_blk = logical_to_blk(inode, start + len - 1); +next: + memset(&map_bh, 0, sizeof(struct buffer_head)); + map_bh.b_size = len; + + ret = get_data_block_fiemap(inode, start_blk, &map_bh, 0); + if (ret) + goto out; + + /* HOLE */ + if (!buffer_mapped(&map_bh)) { + start_blk++; + + if (!past_eof && blk_to_logical(inode, start_blk) >= isize) + past_eof = 1; + + if (past_eof && size) { + flags |= FIEMAP_EXTENT_LAST; + ret = fiemap_fill_next_extent(fieinfo, logical, + phys, size, flags); + } else if (size) { + ret = fiemap_fill_next_extent(fieinfo, logical, + phys, size, flags); + size = 0; + } + + /* if we have holes up to/past EOF then we're done */ + if (start_blk > last_blk || past_eof || ret) + goto out; + } else { + if (start_blk > last_blk && !whole_file) { + ret = fiemap_fill_next_extent(fieinfo, logical, + phys, size, flags); + goto out; + } + + /* + * if size != 0 then we know we already have an extent + * to add, so add it. + */ + if (size) { + ret = fiemap_fill_next_extent(fieinfo, logical, + phys, size, flags); + if (ret) + goto out; + } + + logical = blk_to_logical(inode, start_blk); + phys = blk_to_logical(inode, map_bh.b_blocknr); + size = map_bh.b_size; + flags = 0; + if (buffer_unwritten(&map_bh)) + flags = FIEMAP_EXTENT_UNWRITTEN; + + start_blk += logical_to_blk(inode, size); + + /* + * If we are past the EOF, then we need to make sure as + * soon as we find a hole that the last extent we found + * is marked with FIEMAP_EXTENT_LAST + */ + if (!past_eof && logical + size >= isize) + past_eof = true; + } + cond_resched(); + if (fatal_signal_pending(current)) + ret = -EINTR; + else + goto next; +out: + if (ret == 1) + ret = 0; + + mutex_unlock(&inode->i_mutex); + return ret; +} + +/* + * This function was originally taken from fs/mpage.c, and customized for f2fs. + * Major change was from block_size == page_size in f2fs by default. + */ +static int f2fs_mpage_readpages(struct address_space *mapping, + struct list_head *pages, struct page *page, + unsigned nr_pages) +{ + struct bio *bio = NULL; + unsigned page_idx; + sector_t last_block_in_bio = 0; + struct inode *inode = mapping->host; + const unsigned blkbits = inode->i_blkbits; + const unsigned blocksize = 1 << blkbits; + sector_t block_in_file; + sector_t last_block; + sector_t last_block_in_file; + sector_t block_nr; + struct block_device *bdev = inode->i_sb->s_bdev; + struct f2fs_map_blocks map; + + map.m_pblk = 0; + map.m_lblk = 0; + map.m_len = 0; + map.m_flags = 0; + + for (page_idx = 0; nr_pages; page_idx++, nr_pages--) { + + prefetchw(&page->flags); + if (pages) { + page = list_entry(pages->prev, struct page, lru); + list_del(&page->lru); + if (add_to_page_cache_lru(page, mapping, + page->index, GFP_KERNEL)) + goto next_page; + } + + block_in_file = (sector_t)page->index; + last_block = block_in_file + nr_pages; + last_block_in_file = (i_size_read(inode) + blocksize - 1) >> + blkbits; + if (last_block > last_block_in_file) + last_block = last_block_in_file; + + /* + * Map blocks using the previous result first. + */ + if ((map.m_flags & F2FS_MAP_MAPPED) && + block_in_file > map.m_lblk && + block_in_file < (map.m_lblk + map.m_len)) + goto got_it; + + /* + * Then do more f2fs_map_blocks() calls until we are + * done with this page. + */ + map.m_flags = 0; + + if (block_in_file < last_block) { + map.m_lblk = block_in_file; + map.m_len = last_block - block_in_file; + + if (f2fs_map_blocks(inode, &map, 0, false)) + goto set_error_page; + } +got_it: + if ((map.m_flags & F2FS_MAP_MAPPED)) { + block_nr = map.m_pblk + block_in_file - map.m_lblk; + SetPageMappedToDisk(page); + + if (!PageUptodate(page) && !cleancache_get_page(page)) { + SetPageUptodate(page); + goto confused; + } + } else { + zero_user_segment(page, 0, PAGE_CACHE_SIZE); + SetPageUptodate(page); + unlock_page(page); + goto next_page; + } + + /* + * This page will go to BIO. Do we need to send this + * BIO off first? + */ + if (bio && (last_block_in_bio != block_nr - 1)) { +submit_and_realloc: + submit_bio(READ, bio); + bio = NULL; + } + if (bio == NULL) { + struct f2fs_crypto_ctx *ctx = NULL; + + if (f2fs_encrypted_inode(inode) && + S_ISREG(inode->i_mode)) { + struct page *cpage; + + ctx = f2fs_get_crypto_ctx(inode); + if (IS_ERR(ctx)) + goto set_error_page; + + /* wait the page to be moved by cleaning */ + cpage = find_lock_page( + META_MAPPING(F2FS_I_SB(inode)), + block_nr); + if (cpage) { + f2fs_wait_on_page_writeback(cpage, + DATA); + f2fs_put_page(cpage, 1); + } + } + + bio = bio_alloc(GFP_KERNEL, + min_t(int, nr_pages, bio_get_nr_vecs(bdev))); + if (!bio) { + if (ctx) + f2fs_release_crypto_ctx(ctx); + goto set_error_page; + } + bio->bi_bdev = bdev; + bio->bi_iter.bi_sector = SECTOR_FROM_BLOCK(block_nr); + bio->bi_end_io = f2fs_read_end_io; + bio->bi_private = ctx; + } + + if (bio_add_page(bio, page, blocksize, 0) < blocksize) + goto submit_and_realloc; + + last_block_in_bio = block_nr; + goto next_page; +set_error_page: + SetPageError(page); + zero_user_segment(page, 0, PAGE_CACHE_SIZE); + unlock_page(page); + goto next_page; +confused: + if (bio) { + submit_bio(READ, bio); + bio = NULL; + } + unlock_page(page); +next_page: + if (pages) + page_cache_release(page); + } + BUG_ON(pages && !list_empty(pages)); + if (bio) + submit_bio(READ, bio); + return 0; } static int f2fs_read_data_page(struct file *file, struct page *page) @@ -1350,8 +1601,7 @@ static int f2fs_read_data_page(struct file *file, struct page *page) if (f2fs_has_inline_data(inode)) ret = f2fs_read_inline_data(inode, page); if (ret == -EAGAIN) - ret = mpage_readpage(page, get_data_block); - + ret = f2fs_mpage_readpages(page->mapping, NULL, page, 1); return ret; } @@ -1365,11 +1615,12 @@ static int f2fs_read_data_pages(struct file *file, if (f2fs_has_inline_data(inode)) return 0; - return mpage_readpages(mapping, pages, nr_pages, get_data_block); + return f2fs_mpage_readpages(mapping, pages, NULL, nr_pages); } -int do_write_data_page(struct page *page, struct f2fs_io_info *fio) +int do_write_data_page(struct f2fs_io_info *fio) { + struct page *page = fio->page; struct inode *inode = page->mapping->host; struct dnode_of_data dn; int err = 0; @@ -1387,6 +1638,14 @@ int do_write_data_page(struct page *page, struct f2fs_io_info *fio) goto out_writepage; } + if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) { + fio->encrypted_page = f2fs_encrypt(inode, fio->page); + if (IS_ERR(fio->encrypted_page)) { + err = PTR_ERR(fio->encrypted_page); + goto out_writepage; + } + } + set_page_writeback(page); /* @@ -1396,11 +1655,11 @@ int do_write_data_page(struct page *page, struct f2fs_io_info *fio) if (unlikely(fio->blk_addr != NEW_ADDR && !is_cold_data(page) && need_inplace_update(inode))) { - rewrite_data_page(page, fio); + rewrite_data_page(fio); set_inode_flag(F2FS_I(inode), FI_UPDATE_WRITE); trace_f2fs_do_write_data_page(page, IPU); } else { - write_data_page(page, &dn, fio); + write_data_page(&dn, fio); set_data_blkaddr(&dn); f2fs_update_extent_cache(&dn); trace_f2fs_do_write_data_page(page, OPU); @@ -1425,8 +1684,11 @@ static int f2fs_write_data_page(struct page *page, bool need_balance_fs = false; int err = 0; struct f2fs_io_info fio = { + .sbi = sbi, .type = DATA, .rw = (wbc->sync_mode == WB_SYNC_ALL) ? WRITE_SYNC : WRITE, + .page = page, + .encrypted_page = NULL, }; trace_f2fs_writepage(page, DATA); @@ -1456,7 +1718,7 @@ write: if (S_ISDIR(inode->i_mode)) { if (unlikely(f2fs_cp_error(sbi))) goto redirty_out; - err = do_write_data_page(page, &fio); + err = do_write_data_page(&fio); goto done; } @@ -1476,7 +1738,7 @@ write: if (f2fs_has_inline_data(inode)) err = f2fs_write_inline_data(inode, page); if (err == -EAGAIN) - err = do_write_data_page(page, &fio); + err = do_write_data_page(&fio); f2fs_unlock_op(sbi); done: if (err && err != -ENOENT) @@ -1645,11 +1907,14 @@ put_next: zero_user_segment(page, 0, PAGE_CACHE_SIZE); } else { struct f2fs_io_info fio = { + .sbi = sbi, .type = DATA, .rw = READ_SYNC, .blk_addr = dn.data_blkaddr, + .page = page, + .encrypted_page = NULL, }; - err = f2fs_submit_page_bio(sbi, page, &fio); + err = f2fs_submit_page_bio(&fio); if (err) goto fail; @@ -1663,6 +1928,15 @@ put_next: f2fs_put_page(page, 1); goto repeat; } + + /* avoid symlink page */ + if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) { + err = f2fs_decrypt_one(inode, page); + if (err) { + f2fs_put_page(page, 1); + goto fail; + } + } } out: SetPageUptodate(page); @@ -1733,6 +2007,9 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter, return err; } + if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) + return 0; + if (check_direct_IO(inode, iter, offset)) return 0; diff --git a/fs/f2fs/debug.c b/fs/f2fs/debug.c index f5388f3..75176e0 100644 --- a/fs/f2fs/debug.c +++ b/fs/f2fs/debug.c @@ -94,7 +94,8 @@ static void update_general_status(struct f2fs_sb_info *sbi) static void update_sit_info(struct f2fs_sb_info *sbi) { struct f2fs_stat_info *si = F2FS_STAT(sbi); - unsigned int blks_per_sec, hblks_per_sec, total_vblocks, bimodal, dist; + unsigned long long blks_per_sec, hblks_per_sec, total_vblocks; + unsigned long long bimodal, dist; unsigned int segno, vblocks; int ndirty = 0; @@ -112,10 +113,10 @@ static void update_sit_info(struct f2fs_sb_info *sbi) ndirty++; } } - dist = MAIN_SECS(sbi) * hblks_per_sec * hblks_per_sec / 100; - si->bimodal = bimodal / dist; + dist = div_u64(MAIN_SECS(sbi) * hblks_per_sec * hblks_per_sec, 100); + si->bimodal = div_u64(bimodal, dist); if (si->dirty_count) - si->avg_vblocks = total_vblocks / ndirty; + si->avg_vblocks = div_u64(total_vblocks, ndirty); else si->avg_vblocks = 0; } @@ -143,7 +144,7 @@ static void update_mem_info(struct f2fs_sb_info *sbi) si->base_mem += sizeof(struct sit_info); si->base_mem += MAIN_SEGS(sbi) * sizeof(struct seg_entry); si->base_mem += f2fs_bitmap_size(MAIN_SEGS(sbi)); - si->base_mem += 2 * SIT_VBLOCK_MAP_SIZE * MAIN_SEGS(sbi); + si->base_mem += 3 * SIT_VBLOCK_MAP_SIZE * MAIN_SEGS(sbi); si->base_mem += SIT_VBLOCK_MAP_SIZE; if (sbi->segs_per_sec > 1) si->base_mem += MAIN_SECS(sbi) * sizeof(struct sec_entry); diff --git a/fs/f2fs/dir.c b/fs/f2fs/dir.c index 3a3302a..a34ebd8 100644 --- a/fs/f2fs/dir.c +++ b/fs/f2fs/dir.c @@ -76,20 +76,10 @@ static unsigned long dir_block_index(unsigned int level, return bidx; } -static bool early_match_name(size_t namelen, f2fs_hash_t namehash, - struct f2fs_dir_entry *de) -{ - if (le16_to_cpu(de->name_len) != namelen) - return false; - - if (de->hash_code != namehash) - return false; - - return true; -} - static struct f2fs_dir_entry *find_in_block(struct page *dentry_page, - struct qstr *name, int *max_slots, + struct f2fs_filename *fname, + f2fs_hash_t namehash, + int *max_slots, struct page **res_page) { struct f2fs_dentry_block *dentry_blk; @@ -98,9 +88,8 @@ static struct f2fs_dir_entry *find_in_block(struct page *dentry_page, dentry_blk = (struct f2fs_dentry_block *)kmap(dentry_page); - make_dentry_ptr(&d, (void *)dentry_blk, 1); - de = find_target_dentry(name, max_slots, &d); - + make_dentry_ptr(NULL, &d, (void *)dentry_blk, 1); + de = find_target_dentry(fname, namehash, max_slots, &d); if (de) *res_page = dentry_page; else @@ -114,13 +103,15 @@ static struct f2fs_dir_entry *find_in_block(struct page *dentry_page, return de; } -struct f2fs_dir_entry *find_target_dentry(struct qstr *name, int *max_slots, - struct f2fs_dentry_ptr *d) +struct f2fs_dir_entry *find_target_dentry(struct f2fs_filename *fname, + f2fs_hash_t namehash, int *max_slots, + struct f2fs_dentry_ptr *d) { struct f2fs_dir_entry *de; unsigned long bit_pos = 0; - f2fs_hash_t namehash = f2fs_dentry_hash(name); int max_len = 0; + struct f2fs_str de_name = FSTR_INIT(NULL, 0); + struct f2fs_str *name = &fname->disk_name; if (max_slots) *max_slots = 0; @@ -132,8 +123,18 @@ struct f2fs_dir_entry *find_target_dentry(struct qstr *name, int *max_slots, } de = &d->dentry[bit_pos]; - if (early_match_name(name->len, namehash, de) && - !memcmp(d->filename[bit_pos], name->name, name->len)) + + /* encrypted case */ + de_name.name = d->filename[bit_pos]; + de_name.len = le16_to_cpu(de->name_len); + + /* show encrypted name */ + if (fname->hash) { + if (de->hash_code == fname->hash) + goto found; + } else if (de_name.len == name->len && + de->hash_code == namehash && + !memcmp(de_name.name, name->name, name->len)) goto found; if (max_slots && max_len > *max_slots) @@ -155,16 +156,21 @@ found: } static struct f2fs_dir_entry *find_in_level(struct inode *dir, - unsigned int level, struct qstr *name, - f2fs_hash_t namehash, struct page **res_page) + unsigned int level, + struct f2fs_filename *fname, + struct page **res_page) { - int s = GET_DENTRY_SLOTS(name->len); + struct qstr name = FSTR_TO_QSTR(&fname->disk_name); + int s = GET_DENTRY_SLOTS(name.len); unsigned int nbucket, nblock; unsigned int bidx, end_block; struct page *dentry_page; struct f2fs_dir_entry *de = NULL; bool room = false; int max_slots; + f2fs_hash_t namehash; + + namehash = f2fs_dentry_hash(&name); f2fs_bug_on(F2FS_I_SB(dir), level > MAX_DIR_HASH_DEPTH); @@ -177,13 +183,14 @@ static struct f2fs_dir_entry *find_in_level(struct inode *dir, for (; bidx < end_block; bidx++) { /* no need to allocate new dentry pages to all the indices */ - dentry_page = find_data_page(dir, bidx, true); + dentry_page = find_data_page(dir, bidx); if (IS_ERR(dentry_page)) { room = true; continue; } - de = find_in_block(dentry_page, name, &max_slots, res_page); + de = find_in_block(dentry_page, fname, namehash, &max_slots, + res_page); if (de) break; @@ -211,30 +218,34 @@ struct f2fs_dir_entry *f2fs_find_entry(struct inode *dir, { unsigned long npages = dir_blocks(dir); struct f2fs_dir_entry *de = NULL; - f2fs_hash_t name_hash; unsigned int max_depth; unsigned int level; + struct f2fs_filename fname; + int err; *res_page = NULL; - if (f2fs_has_inline_dentry(dir)) - return find_in_inline_dir(dir, child, res_page); + err = f2fs_fname_setup_filename(dir, child, 1, &fname); + if (err) + return NULL; + + if (f2fs_has_inline_dentry(dir)) { + de = find_in_inline_dir(dir, &fname, res_page); + goto out; + } if (npages == 0) - return NULL; + goto out; - name_hash = f2fs_dentry_hash(child); max_depth = F2FS_I(dir)->i_current_depth; for (level = 0; level < max_depth; level++) { - de = find_in_level(dir, level, child, name_hash, res_page); + de = find_in_level(dir, level, &fname, res_page); if (de) break; } - if (!de && F2FS_I(dir)->chash != name_hash) { - F2FS_I(dir)->chash = name_hash; - F2FS_I(dir)->clevel = level - 1; - } +out: + f2fs_fname_free_filename(&fname); return de; } @@ -303,10 +314,14 @@ static void init_dent_inode(const struct qstr *name, struct page *ipage) set_page_dirty(ipage); } -int update_dent_inode(struct inode *inode, const struct qstr *name) +int update_dent_inode(struct inode *inode, struct inode *to, + const struct qstr *name) { struct page *page; + if (file_enc_name(to)) + return 0; + page = get_node_page(F2FS_I_SB(inode), inode->i_ino); if (IS_ERR(page)) return PTR_ERR(page); @@ -356,7 +371,7 @@ static int make_empty_dir(struct inode *inode, dentry_blk = kmap_atomic(dentry_page); - make_dentry_ptr(&d, (void *)dentry_blk, 1); + make_dentry_ptr(NULL, &d, (void *)dentry_blk, 1); do_make_empty_dir(inode, parent, &d); kunmap_atomic(dentry_blk); @@ -390,6 +405,12 @@ struct page *init_inode_metadata(struct inode *inode, struct inode *dir, err = f2fs_init_security(inode, dir, name, page); if (err) goto put_error; + + if (f2fs_encrypted_inode(dir) && f2fs_may_encrypt(inode)) { + err = f2fs_inherit_context(dir, inode, page); + if (err) + goto put_error; + } } else { page = get_node_page(F2FS_I_SB(dir), inode->i_ino); if (IS_ERR(page)) @@ -501,24 +522,33 @@ int __f2fs_add_link(struct inode *dir, const struct qstr *name, unsigned long bidx, block; f2fs_hash_t dentry_hash; unsigned int nbucket, nblock; - size_t namelen = name->len; struct page *dentry_page = NULL; struct f2fs_dentry_block *dentry_blk = NULL; struct f2fs_dentry_ptr d; - int slots = GET_DENTRY_SLOTS(namelen); struct page *page = NULL; - int err = 0; + struct f2fs_filename fname; + struct qstr new_name; + int slots, err; + + err = f2fs_fname_setup_filename(dir, name, 0, &fname); + if (err) + return err; + + new_name.name = fname_name(&fname); + new_name.len = fname_len(&fname); if (f2fs_has_inline_dentry(dir)) { - err = f2fs_add_inline_entry(dir, name, inode, ino, mode); + err = f2fs_add_inline_entry(dir, &new_name, inode, ino, mode); if (!err || err != -EAGAIN) - return err; + goto out; else err = 0; } - dentry_hash = f2fs_dentry_hash(name); level = 0; + slots = GET_DENTRY_SLOTS(new_name.len); + dentry_hash = f2fs_dentry_hash(&new_name); + current_depth = F2FS_I(dir)->i_current_depth; if (F2FS_I(dir)->chash == dentry_hash) { level = F2FS_I(dir)->clevel; @@ -526,8 +556,10 @@ int __f2fs_add_link(struct inode *dir, const struct qstr *name, } start: - if (unlikely(current_depth == MAX_DIR_HASH_DEPTH)) - return -ENOSPC; + if (unlikely(current_depth == MAX_DIR_HASH_DEPTH)) { + err = -ENOSPC; + goto out; + } /* Increase the depth, if required */ if (level == current_depth) @@ -541,8 +573,10 @@ start: for (block = bidx; block <= (bidx + nblock - 1); block++) { dentry_page = get_new_data_page(dir, NULL, block, true); - if (IS_ERR(dentry_page)) - return PTR_ERR(dentry_page); + if (IS_ERR(dentry_page)) { + err = PTR_ERR(dentry_page); + goto out; + } dentry_blk = kmap(dentry_page); bit_pos = room_for_filename(&dentry_blk->dentry_bitmap, @@ -562,15 +596,17 @@ add_dentry: if (inode) { down_write(&F2FS_I(inode)->i_sem); - page = init_inode_metadata(inode, dir, name, NULL); + page = init_inode_metadata(inode, dir, &new_name, NULL); if (IS_ERR(page)) { err = PTR_ERR(page); goto fail; } + if (f2fs_encrypted_inode(dir)) + file_set_enc_name(inode); } - make_dentry_ptr(&d, (void *)dentry_blk, 1); - f2fs_update_dentry(ino, mode, &d, name, dentry_hash, bit_pos); + make_dentry_ptr(NULL, &d, (void *)dentry_blk, 1); + f2fs_update_dentry(ino, mode, &d, &new_name, dentry_hash, bit_pos); set_page_dirty(dentry_page); @@ -592,6 +628,8 @@ fail: } kunmap(dentry_page); f2fs_put_page(dentry_page, 1); +out: + f2fs_fname_free_filename(&fname); return err; } @@ -729,11 +767,12 @@ bool f2fs_empty_dir(struct inode *dir) } bool f2fs_fill_dentries(struct dir_context *ctx, struct f2fs_dentry_ptr *d, - unsigned int start_pos) + unsigned int start_pos, struct f2fs_str *fstr) { unsigned char d_type = DT_UNKNOWN; unsigned int bit_pos; struct f2fs_dir_entry *de = NULL; + struct f2fs_str de_name = FSTR_INIT(NULL, 0); bit_pos = ((unsigned long)ctx->pos % d->max); @@ -747,8 +786,24 @@ bool f2fs_fill_dentries(struct dir_context *ctx, struct f2fs_dentry_ptr *d, d_type = f2fs_filetype_table[de->file_type]; else d_type = DT_UNKNOWN; - if (!dir_emit(ctx, d->filename[bit_pos], - le16_to_cpu(de->name_len), + + /* encrypted case */ + de_name.name = d->filename[bit_pos]; + de_name.len = le16_to_cpu(de->name_len); + + if (f2fs_encrypted_inode(d->inode)) { + int save_len = fstr->len; + int ret; + + ret = f2fs_fname_disk_to_usr(d->inode, &de->hash_code, + &de_name, fstr); + de_name = *fstr; + fstr->len = save_len; + if (ret < 0) + return true; + } + + if (!dir_emit(ctx, de_name.name, de_name.len, le32_to_cpu(de->ino), d_type)) return true; @@ -767,9 +822,24 @@ static int f2fs_readdir(struct file *file, struct dir_context *ctx) struct file_ra_state *ra = &file->f_ra; unsigned int n = ((unsigned long)ctx->pos / NR_DENTRY_IN_BLOCK); struct f2fs_dentry_ptr d; + struct f2fs_str fstr = FSTR_INIT(NULL, 0); + int err = 0; - if (f2fs_has_inline_dentry(inode)) - return f2fs_read_inline_dir(file, ctx); + if (f2fs_encrypted_inode(inode)) { + err = f2fs_get_encryption_info(inode); + if (err) + return err; + + err = f2fs_fname_crypto_alloc_buffer(inode, F2FS_NAME_LEN, + &fstr); + if (err < 0) + return err; + } + + if (f2fs_has_inline_dentry(inode)) { + err = f2fs_read_inline_dir(file, ctx, &fstr); + goto out; + } /* readahead for multi pages of dir */ if (npages - n > 1 && !ra_has_index(ra, n)) @@ -783,9 +853,9 @@ static int f2fs_readdir(struct file *file, struct dir_context *ctx) dentry_blk = kmap(dentry_page); - make_dentry_ptr(&d, (void *)dentry_blk, 1); + make_dentry_ptr(inode, &d, (void *)dentry_blk, 1); - if (f2fs_fill_dentries(ctx, &d, n * NR_DENTRY_IN_BLOCK)) + if (f2fs_fill_dentries(ctx, &d, n * NR_DENTRY_IN_BLOCK, &fstr)) goto stop; ctx->pos = (n + 1) * NR_DENTRY_IN_BLOCK; @@ -798,8 +868,9 @@ stop: kunmap(dentry_page); f2fs_put_page(dentry_page, 1); } - - return 0; +out: + f2fs_fname_crypto_free_buffer(&fstr); + return err; } const struct file_operations f2fs_dir_operations = { @@ -808,4 +879,7 @@ const struct file_operations f2fs_dir_operations = { .iterate = f2fs_readdir, .fsync = f2fs_sync_file, .unlocked_ioctl = f2fs_ioctl, +#ifdef CONFIG_COMPAT + .compat_ioctl = f2fs_compat_ioctl, +#endif }; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 8de34ab..a8327ed 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -70,6 +70,15 @@ struct f2fs_mount_info { unsigned int opt; }; +#define F2FS_FEATURE_ENCRYPT 0x0001 + +#define F2FS_HAS_FEATURE(sb, mask) \ + ((F2FS_SB(sb)->raw_super->feature & cpu_to_le32(mask)) != 0) +#define F2FS_SET_FEATURE(sb, mask) \ + F2FS_SB(sb)->raw_super->feature |= cpu_to_le32(mask) +#define F2FS_CLEAR_FEATURE(sb, mask) \ + F2FS_SB(sb)->raw_super->feature &= ~cpu_to_le32(mask) + #define CRCPOLY_LE 0xedb88320 static inline __u32 f2fs_crc32(void *buf, size_t len) @@ -110,6 +119,8 @@ enum { #define DEF_BATCHED_TRIM_SECTIONS 32 #define BATCHED_TRIM_SEGMENTS(sbi) \ (SM_I(sbi)->trim_sections * (sbi)->segs_per_sec) +#define BATCHED_TRIM_BLOCKS(sbi) \ + (BATCHED_TRIM_SEGMENTS(sbi) << (sbi)->log_blocks_per_seg) struct cp_control { int reason; @@ -218,6 +229,13 @@ static inline bool __has_cursum_space(struct f2fs_summary_block *sum, int size, #define F2FS_IOC_RELEASE_VOLATILE_WRITE _IO(F2FS_IOCTL_MAGIC, 4) #define F2FS_IOC_ABORT_VOLATILE_WRITE _IO(F2FS_IOCTL_MAGIC, 5) +#define F2FS_IOC_SET_ENCRYPTION_POLICY \ + _IOR('f', 19, struct f2fs_encryption_policy) +#define F2FS_IOC_GET_ENCRYPTION_PWSALT \ + _IOW('f', 20, __u8[16]) +#define F2FS_IOC_GET_ENCRYPTION_POLICY \ + _IOW('f', 21, struct f2fs_encryption_policy) + /* * should be same as XFS_IOC_GOINGDOWN. * Flags for going down operation used by FS_IOC_GOINGDOWN @@ -239,16 +257,38 @@ static inline bool __has_cursum_space(struct f2fs_summary_block *sum, int size, * For INODE and NODE manager */ /* for directory operations */ +struct f2fs_str { + unsigned char *name; + u32 len; +}; + +struct f2fs_filename { + const struct qstr *usr_fname; + struct f2fs_str disk_name; + f2fs_hash_t hash; +#ifdef CONFIG_F2FS_FS_ENCRYPTION + struct f2fs_str crypto_buf; +#endif +}; + +#define FSTR_INIT(n, l) { .name = n, .len = l } +#define FSTR_TO_QSTR(f) QSTR_INIT((f)->name, (f)->len) +#define fname_name(p) ((p)->disk_name.name) +#define fname_len(p) ((p)->disk_name.len) + struct f2fs_dentry_ptr { + struct inode *inode; const void *bitmap; struct f2fs_dir_entry *dentry; __u8 (*filename)[F2FS_SLOT_LEN]; int max; }; -static inline void make_dentry_ptr(struct f2fs_dentry_ptr *d, - void *src, int type) +static inline void make_dentry_ptr(struct inode *inode, + struct f2fs_dentry_ptr *d, void *src, int type) { + d->inode = inode; + if (type == 1) { struct f2fs_dentry_block *t = (struct f2fs_dentry_block *)src; d->max = NR_DENTRY_IN_BLOCK; @@ -315,10 +355,51 @@ struct extent_tree { }; /* + * This structure is taken from ext4_map_blocks. + * + * Note that, however, f2fs uses NEW and MAPPED flags for f2fs_map_blocks(). + */ +#define F2FS_MAP_NEW (1 << BH_New) +#define F2FS_MAP_MAPPED (1 << BH_Mapped) +#define F2FS_MAP_UNWRITTEN (1 << BH_Unwritten) +#define F2FS_MAP_FLAGS (F2FS_MAP_NEW | F2FS_MAP_MAPPED |\ + F2FS_MAP_UNWRITTEN) + +struct f2fs_map_blocks { + block_t m_pblk; + block_t m_lblk; + unsigned int m_len; + unsigned int m_flags; +}; + +/* * i_advise uses FADVISE_XXX_BIT. We can add additional hints later. */ #define FADVISE_COLD_BIT 0x01 #define FADVISE_LOST_PINO_BIT 0x02 +#define FADVISE_ENCRYPT_BIT 0x04 +#define FADVISE_ENC_NAME_BIT 0x08 + +#define file_is_cold(inode) is_file(inode, FADVISE_COLD_BIT) +#define file_wrong_pino(inode) is_file(inode, FADVISE_LOST_PINO_BIT) +#define file_set_cold(inode) set_file(inode, FADVISE_COLD_BIT) +#define file_lost_pino(inode) set_file(inode, FADVISE_LOST_PINO_BIT) +#define file_clear_cold(inode) clear_file(inode, FADVISE_COLD_BIT) +#define file_got_pino(inode) clear_file(inode, FADVISE_LOST_PINO_BIT) +#define file_is_encrypt(inode) is_file(inode, FADVISE_ENCRYPT_BIT) +#define file_set_encrypt(inode) set_file(inode, FADVISE_ENCRYPT_BIT) +#define file_clear_encrypt(inode) clear_file(inode, FADVISE_ENCRYPT_BIT) +#define file_enc_name(inode) is_file(inode, FADVISE_ENC_NAME_BIT) +#define file_set_enc_name(inode) set_file(inode, FADVISE_ENC_NAME_BIT) + +/* Encryption algorithms */ +#define F2FS_ENCRYPTION_MODE_INVALID 0 +#define F2FS_ENCRYPTION_MODE_AES_256_XTS 1 +#define F2FS_ENCRYPTION_MODE_AES_256_GCM 2 +#define F2FS_ENCRYPTION_MODE_AES_256_CBC 3 +#define F2FS_ENCRYPTION_MODE_AES_256_CTS 4 + +#include "f2fs_crypto.h" #define DEF_DIR_LEVEL 0 @@ -346,6 +427,11 @@ struct f2fs_inode_info { struct radix_tree_root inmem_root; /* radix tree for inmem pages */ struct list_head inmem_pages; /* inmemory pages managed by f2fs */ struct mutex inmem_lock; /* lock for inmemory pages */ + +#ifdef CONFIG_F2FS_FS_ENCRYPTION + /* Encryption params */ + struct f2fs_crypt_info *i_crypt_info; +#endif }; static inline void get_extent_info(struct extent_info *ext, @@ -571,9 +657,12 @@ enum page_type { }; struct f2fs_io_info { + struct f2fs_sb_info *sbi; /* f2fs_sb_info pointer */ enum page_type type; /* contains DATA/NODE/META/META_FLUSH */ int rw; /* contains R/RS/W/WS with REQ_META/REQ_PRIO */ block_t blk_addr; /* block address to be written */ + struct page *page; /* page to be written */ + struct page *encrypted_page; /* encrypted page */ }; #define is_read_io(rw) (((rw) & 1) == READ) @@ -666,6 +755,7 @@ struct f2fs_sb_info { block_t user_block_count; /* # of user blocks */ block_t total_valid_block_count; /* # of valid blocks */ block_t alloc_valid_block_count; /* # of allocated blocks */ + block_t discard_blks; /* discard command candidats */ block_t last_valid_block_count; /* for recovery */ u32 s_next_generation; /* for NFS support */ atomic_t nr_pages[NR_COUNT_TYPE]; /* # of pages, see count_type */ @@ -1193,6 +1283,24 @@ static inline int f2fs_test_bit(unsigned int nr, char *addr) return mask & *addr; } +static inline void f2fs_set_bit(unsigned int nr, char *addr) +{ + int mask; + + addr += (nr >> 3); + mask = 1 << (7 - (nr & 0x07)); + *addr |= mask; +} + +static inline void f2fs_clear_bit(unsigned int nr, char *addr) +{ + int mask; + + addr += (nr >> 3); + mask = 1 << (7 - (nr & 0x07)); + *addr &= ~mask; +} + static inline int f2fs_test_and_set_bit(unsigned int nr, char *addr) { int mask; @@ -1391,6 +1499,21 @@ static inline void f2fs_dentry_kunmap(struct inode *dir, struct page *page) kunmap(page); } +static inline int is_file(struct inode *inode, int type) +{ + return F2FS_I(inode)->i_advise & type; +} + +static inline void set_file(struct inode *inode, int type) +{ + F2FS_I(inode)->i_advise |= type; +} + +static inline void clear_file(struct inode *inode, int type) +{ + F2FS_I(inode)->i_advise &= ~type; +} + static inline int f2fs_readonly(struct super_block *sb) { return sb->s_flags & MS_RDONLY; @@ -1407,6 +1530,17 @@ static inline void f2fs_stop_checkpoint(struct f2fs_sb_info *sbi) sbi->sb->s_flags |= MS_RDONLY; } +static inline bool is_dot_dotdot(const struct qstr *str) +{ + if (str->len == 1 && str->name[0] == '.') + return true; + + if (str->len == 2 && str->name[0] == '.' && str->name[1] == '.') + return true; + + return false; +} + #define get_inode_mode(i) \ ((is_inode_flag_set(F2FS_I(i), FI_ACL_MODE)) ? \ (F2FS_I(i)->i_acl_mode) : ((i)->i_mode)) @@ -1453,10 +1587,11 @@ struct dentry *f2fs_get_parent(struct dentry *child); */ extern unsigned char f2fs_filetype_table[F2FS_FT_MAX]; void set_de_type(struct f2fs_dir_entry *, umode_t); -struct f2fs_dir_entry *find_target_dentry(struct qstr *, int *, - struct f2fs_dentry_ptr *); + +struct f2fs_dir_entry *find_target_dentry(struct f2fs_filename *, + f2fs_hash_t, int *, struct f2fs_dentry_ptr *); bool f2fs_fill_dentries(struct dir_context *, struct f2fs_dentry_ptr *, - unsigned int); + unsigned int, struct f2fs_str *); void do_make_empty_dir(struct inode *, struct inode *, struct f2fs_dentry_ptr *); struct page *init_inode_metadata(struct inode *, struct inode *, @@ -1470,7 +1605,7 @@ struct f2fs_dir_entry *f2fs_parent_dir(struct inode *, struct page **); ino_t f2fs_inode_by_name(struct inode *, struct qstr *); void f2fs_set_link(struct inode *, struct f2fs_dir_entry *, struct page *, struct inode *); -int update_dent_inode(struct inode *, const struct qstr *); +int update_dent_inode(struct inode *, struct inode *, const struct qstr *); void f2fs_update_dentry(nid_t ino, umode_t mode, struct f2fs_dentry_ptr *, const struct qstr *, f2fs_hash_t , unsigned int); int __f2fs_add_link(struct inode *, const struct qstr *, struct inode *, nid_t, @@ -1478,7 +1613,6 @@ int __f2fs_add_link(struct inode *, const struct qstr *, struct inode *, nid_t, void f2fs_delete_entry(struct f2fs_dir_entry *, struct page *, struct inode *, struct inode *); int f2fs_do_tmpfile(struct inode *, struct inode *); -int f2fs_make_empty(struct inode *, struct inode *); bool f2fs_empty_dir(struct inode *); static inline int f2fs_add_link(struct dentry *dentry, struct inode *inode) @@ -1490,6 +1624,7 @@ static inline int f2fs_add_link(struct dentry *dentry, struct inode *inode) /* * super.c */ +int f2fs_commit_super(struct f2fs_sb_info *, bool); int f2fs_sync_fs(struct super_block *, int); extern __printf(3, 4) void f2fs_msg(struct super_block *, const char *, const char *, ...); @@ -1506,8 +1641,8 @@ struct dnode_of_data; struct node_info; bool available_free_memory(struct f2fs_sb_info *, int); +int need_dentry_mark(struct f2fs_sb_info *, nid_t); bool is_checkpointed_node(struct f2fs_sb_info *, nid_t); -bool has_fsynced_inode(struct f2fs_sb_info *, nid_t); bool need_inode_block_update(struct f2fs_sb_info *, nid_t); void get_node_info(struct f2fs_sb_info *, nid_t, struct node_info *); int get_dnode_of_data(struct dnode_of_data *, pgoff_t, int); @@ -1548,21 +1683,20 @@ int create_flush_cmd_control(struct f2fs_sb_info *); void destroy_flush_cmd_control(struct f2fs_sb_info *); void invalidate_blocks(struct f2fs_sb_info *, block_t); void refresh_sit_entry(struct f2fs_sb_info *, block_t, block_t); -void clear_prefree_segments(struct f2fs_sb_info *); +void clear_prefree_segments(struct f2fs_sb_info *, struct cp_control *); void release_discard_addrs(struct f2fs_sb_info *); void discard_next_dnode(struct f2fs_sb_info *, block_t); int npages_for_summary_flush(struct f2fs_sb_info *, bool); void allocate_new_segments(struct f2fs_sb_info *); int f2fs_trim_fs(struct f2fs_sb_info *, struct fstrim_range *); struct page *get_sum_page(struct f2fs_sb_info *, unsigned int); +void update_meta_page(struct f2fs_sb_info *, void *, block_t); void write_meta_page(struct f2fs_sb_info *, struct page *); -void write_node_page(struct f2fs_sb_info *, struct page *, - unsigned int, struct f2fs_io_info *); -void write_data_page(struct page *, struct dnode_of_data *, - struct f2fs_io_info *); -void rewrite_data_page(struct page *, struct f2fs_io_info *); -void recover_data_page(struct f2fs_sb_info *, struct page *, - struct f2fs_summary *, block_t, block_t); +void write_node_page(unsigned int, struct f2fs_io_info *); +void write_data_page(struct dnode_of_data *, struct f2fs_io_info *); +void rewrite_data_page(struct f2fs_io_info *); +void f2fs_replace_block(struct f2fs_sb_info *, struct dnode_of_data *, + block_t, block_t, unsigned char, bool); void allocate_data_block(struct f2fs_sb_info *, struct page *, block_t, block_t *, struct f2fs_summary *, int); void f2fs_wait_on_page_writeback(struct page *, enum page_type); @@ -1581,6 +1715,7 @@ void destroy_segment_manager_caches(void); */ struct page *grab_meta_page(struct f2fs_sb_info *, pgoff_t); struct page *get_meta_page(struct f2fs_sb_info *, pgoff_t); +bool is_valid_blkaddr(struct f2fs_sb_info *, block_t, int); int ra_meta_pages(struct f2fs_sb_info *, block_t, int, int); void ra_meta_pages_cond(struct f2fs_sb_info *, pgoff_t); long sync_meta_pages(struct f2fs_sb_info *, enum page_type, long); @@ -1607,10 +1742,8 @@ void destroy_checkpoint_caches(void); * data.c */ void f2fs_submit_merged_bio(struct f2fs_sb_info *, enum page_type, int); -int f2fs_submit_page_bio(struct f2fs_sb_info *, struct page *, - struct f2fs_io_info *); -void f2fs_submit_page_mbio(struct f2fs_sb_info *, struct page *, - struct f2fs_io_info *); +int f2fs_submit_page_bio(struct f2fs_io_info *); +void f2fs_submit_page_mbio(struct f2fs_io_info *); void set_data_blkaddr(struct dnode_of_data *); int reserve_new_block(struct dnode_of_data *); int f2fs_reserve_block(struct dnode_of_data *, pgoff_t); @@ -1619,10 +1752,11 @@ void f2fs_destroy_extent_tree(struct inode *); void f2fs_init_extent_cache(struct inode *, struct f2fs_extent *); void f2fs_update_extent_cache(struct dnode_of_data *); void f2fs_preserve_extent_tree(struct inode *); -struct page *find_data_page(struct inode *, pgoff_t, bool); +struct page *get_read_data_page(struct inode *, pgoff_t, int); +struct page *find_data_page(struct inode *, pgoff_t); struct page *get_lock_data_page(struct inode *, pgoff_t); struct page *get_new_data_page(struct inode *, struct page *, pgoff_t, bool); -int do_write_data_page(struct page *, struct f2fs_io_info *); +int do_write_data_page(struct f2fs_io_info *); int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *, u64, u64); void init_extent_cache_info(struct f2fs_sb_info *); int __init create_extent_cache(void); @@ -1787,13 +1921,15 @@ extern const struct address_space_operations f2fs_node_aops; extern const struct address_space_operations f2fs_meta_aops; extern const struct inode_operations f2fs_dir_inode_operations; extern const struct inode_operations f2fs_symlink_inode_operations; +extern const struct inode_operations f2fs_encrypted_symlink_inode_operations; extern const struct inode_operations f2fs_special_inode_operations; extern struct kmem_cache *inode_entry_slab; /* * inline.c */ -bool f2fs_may_inline(struct inode *); +bool f2fs_may_inline_data(struct inode *); +bool f2fs_may_inline_dentry(struct inode *); void read_inline_data(struct page *, struct page *); bool truncate_inline_inode(struct page *, u64); int f2fs_read_inline_data(struct inode *, struct page *); @@ -1801,8 +1937,8 @@ int f2fs_convert_inline_page(struct dnode_of_data *, struct page *); int f2fs_convert_inline_inode(struct inode *); int f2fs_write_inline_data(struct inode *, struct page *); bool recover_inline_data(struct inode *, struct page *); -struct f2fs_dir_entry *find_in_inline_dir(struct inode *, struct qstr *, - struct page **); +struct f2fs_dir_entry *find_in_inline_dir(struct inode *, + struct f2fs_filename *, struct page **); struct f2fs_dir_entry *f2fs_parent_inline_dir(struct inode *, struct page **); int make_empty_inline_dir(struct inode *inode, struct inode *, struct page *); int f2fs_add_inline_entry(struct inode *, const struct qstr *, struct inode *, @@ -1810,5 +1946,137 @@ int f2fs_add_inline_entry(struct inode *, const struct qstr *, struct inode *, void f2fs_delete_inline_entry(struct f2fs_dir_entry *, struct page *, struct inode *, struct inode *); bool f2fs_empty_inline_dir(struct inode *); -int f2fs_read_inline_dir(struct file *, struct dir_context *); +int f2fs_read_inline_dir(struct file *, struct dir_context *, + struct f2fs_str *); + +/* + * crypto support + */ +static inline int f2fs_encrypted_inode(struct inode *inode) +{ +#ifdef CONFIG_F2FS_FS_ENCRYPTION + return file_is_encrypt(inode); +#else + return 0; +#endif +} + +static inline void f2fs_set_encrypted_inode(struct inode *inode) +{ +#ifdef CONFIG_F2FS_FS_ENCRYPTION + file_set_encrypt(inode); +#endif +} + +static inline bool f2fs_bio_encrypted(struct bio *bio) +{ +#ifdef CONFIG_F2FS_FS_ENCRYPTION + return unlikely(bio->bi_private != NULL); +#else + return false; +#endif +} + +static inline int f2fs_sb_has_crypto(struct super_block *sb) +{ +#ifdef CONFIG_F2FS_FS_ENCRYPTION + return F2FS_HAS_FEATURE(sb, F2FS_FEATURE_ENCRYPT); +#else + return 0; +#endif +} + +static inline bool f2fs_may_encrypt(struct inode *inode) +{ +#ifdef CONFIG_F2FS_FS_ENCRYPTION + mode_t mode = inode->i_mode; + + return (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode)); +#else + return 0; +#endif +} + +/* crypto_policy.c */ +int f2fs_is_child_context_consistent_with_parent(struct inode *, + struct inode *); +int f2fs_inherit_context(struct inode *, struct inode *, struct page *); +int f2fs_process_policy(const struct f2fs_encryption_policy *, struct inode *); +int f2fs_get_policy(struct inode *, struct f2fs_encryption_policy *); + +/* crypt.c */ +extern struct kmem_cache *f2fs_crypt_info_cachep; +bool f2fs_valid_contents_enc_mode(uint32_t); +uint32_t f2fs_validate_encryption_key_size(uint32_t, uint32_t); +struct f2fs_crypto_ctx *f2fs_get_crypto_ctx(struct inode *); +void f2fs_release_crypto_ctx(struct f2fs_crypto_ctx *); +struct page *f2fs_encrypt(struct inode *, struct page *); +int f2fs_decrypt(struct f2fs_crypto_ctx *, struct page *); +int f2fs_decrypt_one(struct inode *, struct page *); +void f2fs_end_io_crypto_work(struct f2fs_crypto_ctx *, struct bio *); + +/* crypto_key.c */ +void f2fs_free_encryption_info(struct inode *, struct f2fs_crypt_info *); +int _f2fs_get_encryption_info(struct inode *inode); + +/* crypto_fname.c */ +bool f2fs_valid_filenames_enc_mode(uint32_t); +u32 f2fs_fname_crypto_round_up(u32, u32); +int f2fs_fname_crypto_alloc_buffer(struct inode *, u32, struct f2fs_str *); +int f2fs_fname_disk_to_usr(struct inode *, f2fs_hash_t *, + const struct f2fs_str *, struct f2fs_str *); +int f2fs_fname_usr_to_disk(struct inode *, const struct qstr *, + struct f2fs_str *); + +#ifdef CONFIG_F2FS_FS_ENCRYPTION +void f2fs_restore_and_release_control_page(struct page **); +void f2fs_restore_control_page(struct page *); + +int __init f2fs_init_crypto(void); +int f2fs_crypto_initialize(void); +void f2fs_exit_crypto(void); + +int f2fs_has_encryption_key(struct inode *); + +static inline int f2fs_get_encryption_info(struct inode *inode) +{ + struct f2fs_crypt_info *ci = F2FS_I(inode)->i_crypt_info; + + if (!ci || + (ci->ci_keyring_key && + (ci->ci_keyring_key->flags & ((1 << KEY_FLAG_INVALIDATED) | + (1 << KEY_FLAG_REVOKED) | + (1 << KEY_FLAG_DEAD))))) + return _f2fs_get_encryption_info(inode); + return 0; +} + +void f2fs_fname_crypto_free_buffer(struct f2fs_str *); +int f2fs_fname_setup_filename(struct inode *, const struct qstr *, + int lookup, struct f2fs_filename *); +void f2fs_fname_free_filename(struct f2fs_filename *); +#else +static inline void f2fs_restore_and_release_control_page(struct page **p) { } +static inline void f2fs_restore_control_page(struct page *p) { } + +static inline int __init f2fs_init_crypto(void) { return 0; } +static inline void f2fs_exit_crypto(void) { } + +static inline int f2fs_has_encryption_key(struct inode *i) { return 0; } +static inline int f2fs_get_encryption_info(struct inode *i) { return 0; } +static inline void f2fs_fname_crypto_free_buffer(struct f2fs_str *p) { } + +static inline int f2fs_fname_setup_filename(struct inode *dir, + const struct qstr *iname, + int lookup, struct f2fs_filename *fname) +{ + memset(fname, 0, sizeof(struct f2fs_filename)); + fname->usr_fname = iname; + fname->disk_name.name = (unsigned char *)iname->name; + fname->disk_name.len = iname->len; + return 0; +} + +static inline void f2fs_fname_free_filename(struct f2fs_filename *fname) { } +#endif #endif diff --git a/fs/f2fs/f2fs_crypto.h b/fs/f2fs/f2fs_crypto.h new file mode 100644 index 0000000..c2c1c2b --- /dev/null +++ b/fs/f2fs/f2fs_crypto.h @@ -0,0 +1,151 @@ +/* + * linux/fs/f2fs/f2fs_crypto.h + * + * Copied from linux/fs/ext4/ext4_crypto.h + * + * Copyright (C) 2015, Google, Inc. + * + * This contains encryption header content for f2fs + * + * Written by Michael Halcrow, 2015. + * Modified by Jaegeuk Kim, 2015. + */ +#ifndef _F2FS_CRYPTO_H +#define _F2FS_CRYPTO_H + +#include <linux/fs.h> + +#define F2FS_KEY_DESCRIPTOR_SIZE 8 + +/* Policy provided via an ioctl on the topmost directory */ +struct f2fs_encryption_policy { + char version; + char contents_encryption_mode; + char filenames_encryption_mode; + char flags; + char master_key_descriptor[F2FS_KEY_DESCRIPTOR_SIZE]; +} __attribute__((__packed__)); + +#define F2FS_ENCRYPTION_CONTEXT_FORMAT_V1 1 +#define F2FS_KEY_DERIVATION_NONCE_SIZE 16 + +#define F2FS_POLICY_FLAGS_PAD_4 0x00 +#define F2FS_POLICY_FLAGS_PAD_8 0x01 +#define F2FS_POLICY_FLAGS_PAD_16 0x02 +#define F2FS_POLICY_FLAGS_PAD_32 0x03 +#define F2FS_POLICY_FLAGS_PAD_MASK 0x03 +#define F2FS_POLICY_FLAGS_VALID 0x03 + +/** + * Encryption context for inode + * + * Protector format: + * 1 byte: Protector format (1 = this version) + * 1 byte: File contents encryption mode + * 1 byte: File names encryption mode + * 1 byte: Flags + * 8 bytes: Master Key descriptor + * 16 bytes: Encryption Key derivation nonce + */ +struct f2fs_encryption_context { + char format; + char contents_encryption_mode; + char filenames_encryption_mode; + char flags; + char master_key_descriptor[F2FS_KEY_DESCRIPTOR_SIZE]; + char nonce[F2FS_KEY_DERIVATION_NONCE_SIZE]; +} __attribute__((__packed__)); + +/* Encryption parameters */ +#define F2FS_XTS_TWEAK_SIZE 16 +#define F2FS_AES_128_ECB_KEY_SIZE 16 +#define F2FS_AES_256_GCM_KEY_SIZE 32 +#define F2FS_AES_256_CBC_KEY_SIZE 32 +#define F2FS_AES_256_CTS_KEY_SIZE 32 +#define F2FS_AES_256_XTS_KEY_SIZE 64 +#define F2FS_MAX_KEY_SIZE 64 + +#define F2FS_KEY_DESC_PREFIX "f2fs:" +#define F2FS_KEY_DESC_PREFIX_SIZE 5 + +struct f2fs_encryption_key { + __u32 mode; + char raw[F2FS_MAX_KEY_SIZE]; + __u32 size; +} __attribute__((__packed__)); + +struct f2fs_crypt_info { + char ci_data_mode; + char ci_filename_mode; + char ci_flags; + struct crypto_ablkcipher *ci_ctfm; + struct key *ci_keyring_key; + char ci_master_key[F2FS_KEY_DESCRIPTOR_SIZE]; +}; + +#define F2FS_CTX_REQUIRES_FREE_ENCRYPT_FL 0x00000001 +#define F2FS_WRITE_PATH_FL 0x00000002 + +struct f2fs_crypto_ctx { + union { + struct { + struct page *bounce_page; /* Ciphertext page */ + struct page *control_page; /* Original page */ + } w; + struct { + struct bio *bio; + struct work_struct work; + } r; + struct list_head free_list; /* Free list */ + }; + char flags; /* Flags */ +}; + +struct f2fs_completion_result { + struct completion completion; + int res; +}; + +#define DECLARE_F2FS_COMPLETION_RESULT(ecr) \ + struct f2fs_completion_result ecr = { \ + COMPLETION_INITIALIZER((ecr).completion), 0 } + +static inline int f2fs_encryption_key_size(int mode) +{ + switch (mode) { + case F2FS_ENCRYPTION_MODE_AES_256_XTS: + return F2FS_AES_256_XTS_KEY_SIZE; + case F2FS_ENCRYPTION_MODE_AES_256_GCM: + return F2FS_AES_256_GCM_KEY_SIZE; + case F2FS_ENCRYPTION_MODE_AES_256_CBC: + return F2FS_AES_256_CBC_KEY_SIZE; + case F2FS_ENCRYPTION_MODE_AES_256_CTS: + return F2FS_AES_256_CTS_KEY_SIZE; + default: + BUG(); + } + return 0; +} + +#define F2FS_FNAME_NUM_SCATTER_ENTRIES 4 +#define F2FS_CRYPTO_BLOCK_SIZE 16 +#define F2FS_FNAME_CRYPTO_DIGEST_SIZE 32 + +/** + * For encrypted symlinks, the ciphertext length is stored at the beginning + * of the string in little-endian format. + */ +struct f2fs_encrypted_symlink_data { + __le16 len; + char encrypted_path[1]; +} __attribute__((__packed__)); + +/** + * This function is used to calculate the disk space required to + * store a filename of length l in encrypted symlink format. + */ +static inline u32 encrypted_symlink_data_len(u32 l) +{ + return (l + sizeof(struct f2fs_encrypted_symlink_data) - 1); +} +#endif /* _F2FS_CRYPTO_H */ diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 2b52e48..ada2a3d 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -20,6 +20,7 @@ #include <linux/uaccess.h> #include <linux/mount.h> #include <linux/pagevec.h> +#include <linux/random.h> #include "f2fs.h" #include "node.h" @@ -105,7 +106,7 @@ static int get_parent_ino(struct inode *inode, nid_t *pino) if (!dentry) return 0; - if (update_dent_inode(inode, &dentry->d_name)) { + if (update_dent_inode(inode, inode, &dentry->d_name)) { dput(dentry); return 0; } @@ -122,6 +123,8 @@ static inline bool need_do_checkpoint(struct inode *inode) if (!S_ISREG(inode->i_mode) || inode->i_nlink != 1) need_cp = true; + else if (file_enc_name(inode) && need_dentry_mark(sbi, inode->i_ino)) + need_cp = true; else if (file_wrong_pino(inode)) need_cp = true; else if (!space_for_roll_forward(sbi)) @@ -271,7 +274,7 @@ flush_out: ret = f2fs_issue_flush(sbi); out: trace_f2fs_sync_file_exit(inode, need_cp, datasync, ret); - f2fs_trace_ios(NULL, NULL, 1); + f2fs_trace_ios(NULL, 1); return ret; } @@ -407,6 +410,12 @@ static int f2fs_file_mmap(struct file *file, struct vm_area_struct *vma) { struct inode *inode = file_inode(file); + if (f2fs_encrypted_inode(inode)) { + int err = f2fs_get_encryption_info(inode); + if (err) + return 0; + } + /* we don't need to use inline_data strictly */ if (f2fs_has_inline_data(inode)) { int err = f2fs_convert_inline_inode(inode); @@ -419,6 +428,18 @@ static int f2fs_file_mmap(struct file *file, struct vm_area_struct *vma) return 0; } +static int f2fs_file_open(struct inode *inode, struct file *filp) +{ + int ret = generic_file_open(inode, filp); + + if (!ret && f2fs_encrypted_inode(inode)) { + ret = f2fs_get_encryption_info(inode); + if (ret) + ret = -EACCES; + } + return ret; +} + int truncate_data_blocks_range(struct dnode_of_data *dn, int count) { int nr_free = 0, ofs = dn->ofs_in_node; @@ -461,28 +482,32 @@ void truncate_data_blocks(struct dnode_of_data *dn) } static int truncate_partial_data_page(struct inode *inode, u64 from, - bool force) + bool cache_only) { unsigned offset = from & (PAGE_CACHE_SIZE - 1); + pgoff_t index = from >> PAGE_CACHE_SHIFT; + struct address_space *mapping = inode->i_mapping; struct page *page; - if (!offset && !force) + if (!offset && !cache_only) return 0; - page = find_data_page(inode, from >> PAGE_CACHE_SHIFT, force); - if (IS_ERR(page)) + if (cache_only) { + page = grab_cache_page(mapping, index); + if (page && PageUptodate(page)) + goto truncate_out; + f2fs_put_page(page, 1); return 0; + } - lock_page(page); - if (unlikely(!PageUptodate(page) || - page->mapping != inode->i_mapping)) - goto out; - + page = get_lock_data_page(inode, index); + if (IS_ERR(page)) + return 0; +truncate_out: f2fs_wait_on_page_writeback(page, DATA); zero_user(page, offset, PAGE_CACHE_SIZE - offset); - if (!force) + if (!cache_only || !f2fs_encrypted_inode(inode) || !S_ISREG(inode->i_mode)) set_page_dirty(page); -out: f2fs_put_page(page, 1); return 0; } @@ -560,7 +585,7 @@ void f2fs_truncate(struct inode *inode) trace_f2fs_truncate(inode); /* we should check inline_data size */ - if (f2fs_has_inline_data(inode) && !f2fs_may_inline(inode)) { + if (f2fs_has_inline_data(inode) && !f2fs_may_inline_data(inode)) { if (f2fs_convert_inline_inode(inode)) return; } @@ -622,16 +647,20 @@ int f2fs_setattr(struct dentry *dentry, struct iattr *attr) return err; if (attr->ia_valid & ATTR_SIZE) { - if (attr->ia_size != i_size_read(inode)) { + if (f2fs_encrypted_inode(inode) && + f2fs_get_encryption_info(inode)) + return -EACCES; + + if (attr->ia_size <= i_size_read(inode)) { truncate_setsize(inode, attr->ia_size); f2fs_truncate(inode); f2fs_balance_fs(F2FS_I_SB(inode)); } else { /* - * giving a chance to truncate blocks past EOF which - * are fallocated with FALLOC_FL_KEEP_SIZE. + * do not trim all blocks after i_size if target size is + * larger than i_size. */ - f2fs_truncate(inode); + truncate_setsize(inode, attr->ia_size); } } @@ -718,10 +747,6 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len) if (!S_ISREG(inode->i_mode)) return -EOPNOTSUPP; - /* skip punching hole beyond i_size */ - if (offset >= inode->i_size) - return ret; - if (f2fs_has_inline_data(inode)) { ret = f2fs_convert_inline_inode(inode); if (ret) @@ -765,6 +790,320 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len) return ret; } +static int f2fs_do_collapse(struct inode *inode, pgoff_t start, pgoff_t end) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct dnode_of_data dn; + pgoff_t nrpages = (i_size_read(inode) + PAGE_SIZE - 1) / PAGE_SIZE; + int ret = 0; + + f2fs_lock_op(sbi); + + for (; end < nrpages; start++, end++) { + block_t new_addr, old_addr; + + set_new_dnode(&dn, inode, NULL, NULL, 0); + ret = get_dnode_of_data(&dn, end, LOOKUP_NODE_RA); + if (ret && ret != -ENOENT) { + goto out; + } else if (ret == -ENOENT) { + new_addr = NULL_ADDR; + } else { + new_addr = dn.data_blkaddr; + truncate_data_blocks_range(&dn, 1); + f2fs_put_dnode(&dn); + } + + if (new_addr == NULL_ADDR) { + set_new_dnode(&dn, inode, NULL, NULL, 0); + ret = get_dnode_of_data(&dn, start, LOOKUP_NODE_RA); + if (ret && ret != -ENOENT) + goto out; + else if (ret == -ENOENT) + continue; + + if (dn.data_blkaddr == NULL_ADDR) { + f2fs_put_dnode(&dn); + continue; + } else { + truncate_data_blocks_range(&dn, 1); + } + + f2fs_put_dnode(&dn); + } else { + struct page *ipage; + + ipage = get_node_page(sbi, inode->i_ino); + if (IS_ERR(ipage)) { + ret = PTR_ERR(ipage); + goto out; + } + + set_new_dnode(&dn, inode, ipage, NULL, 0); + ret = f2fs_reserve_block(&dn, start); + if (ret) + goto out; + + old_addr = dn.data_blkaddr; + if (old_addr != NEW_ADDR && new_addr == NEW_ADDR) { + dn.data_blkaddr = NULL_ADDR; + f2fs_update_extent_cache(&dn); + invalidate_blocks(sbi, old_addr); + + dn.data_blkaddr = new_addr; + set_data_blkaddr(&dn); + } else if (new_addr != NEW_ADDR) { + struct node_info ni; + + get_node_info(sbi, dn.nid, &ni); + f2fs_replace_block(sbi, &dn, old_addr, new_addr, + ni.version, true); + } + + f2fs_put_dnode(&dn); + } + } + ret = 0; +out: + f2fs_unlock_op(sbi); + return ret; +} + +static int f2fs_collapse_range(struct inode *inode, loff_t offset, loff_t len) +{ + pgoff_t pg_start, pg_end; + loff_t new_size; + int ret; + + if (!S_ISREG(inode->i_mode)) + return -EINVAL; + + if (offset + len >= i_size_read(inode)) + return -EINVAL; + + /* collapse range should be aligned to block size of f2fs. */ + if (offset & (F2FS_BLKSIZE - 1) || len & (F2FS_BLKSIZE - 1)) + return -EINVAL; + + pg_start = offset >> PAGE_CACHE_SHIFT; + pg_end = (offset + len) >> PAGE_CACHE_SHIFT; + + /* write out all dirty pages from offset */ + ret = filemap_write_and_wait_range(inode->i_mapping, offset, LLONG_MAX); + if (ret) + return ret; + + truncate_pagecache(inode, offset); + + ret = f2fs_do_collapse(inode, pg_start, pg_end); + if (ret) + return ret; + + new_size = i_size_read(inode) - len; + + ret = truncate_blocks(inode, new_size, true); + if (!ret) + i_size_write(inode, new_size); + + return ret; +} + +static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len, + int mode) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct address_space *mapping = inode->i_mapping; + pgoff_t index, pg_start, pg_end; + loff_t new_size = i_size_read(inode); + loff_t off_start, off_end; + int ret = 0; + + if (!S_ISREG(inode->i_mode)) + return -EINVAL; + + ret = inode_newsize_ok(inode, (len + offset)); + if (ret) + return ret; + + f2fs_balance_fs(sbi); + + if (f2fs_has_inline_data(inode)) { + ret = f2fs_convert_inline_inode(inode); + if (ret) + return ret; + } + + ret = filemap_write_and_wait_range(mapping, offset, offset + len - 1); + if (ret) + return ret; + + truncate_pagecache_range(inode, offset, offset + len - 1); + + pg_start = ((unsigned long long) offset) >> PAGE_CACHE_SHIFT; + pg_end = ((unsigned long long) offset + len) >> PAGE_CACHE_SHIFT; + + off_start = offset & (PAGE_CACHE_SIZE - 1); + off_end = (offset + len) & (PAGE_CACHE_SIZE - 1); + + if (pg_start == pg_end) { + fill_zero(inode, pg_start, off_start, off_end - off_start); + if (offset + len > new_size) + new_size = offset + len; + new_size = max_t(loff_t, new_size, offset + len); + } else { + if (off_start) { + fill_zero(inode, pg_start++, off_start, + PAGE_CACHE_SIZE - off_start); + new_size = max_t(loff_t, new_size, + pg_start << PAGE_CACHE_SHIFT); + } + + for (index = pg_start; index < pg_end; index++) { + struct dnode_of_data dn; + struct page *ipage; + + f2fs_lock_op(sbi); + + ipage = get_node_page(sbi, inode->i_ino); + if (IS_ERR(ipage)) { + ret = PTR_ERR(ipage); + f2fs_unlock_op(sbi); + goto out; + } + + set_new_dnode(&dn, inode, ipage, NULL, 0); + ret = f2fs_reserve_block(&dn, index); + if (ret) { + f2fs_unlock_op(sbi); + goto out; + } + + if (dn.data_blkaddr != NEW_ADDR) { + invalidate_blocks(sbi, dn.data_blkaddr); + + dn.data_blkaddr = NEW_ADDR; + set_data_blkaddr(&dn); + + dn.data_blkaddr = NULL_ADDR; + f2fs_update_extent_cache(&dn); + } + f2fs_put_dnode(&dn); + f2fs_unlock_op(sbi); + + new_size = max_t(loff_t, new_size, + (index + 1) << PAGE_CACHE_SHIFT); + } + + if (off_end) { + fill_zero(inode, pg_end, 0, off_end); + new_size = max_t(loff_t, new_size, offset + len); + } + } + +out: + if (!(mode & FALLOC_FL_KEEP_SIZE) && i_size_read(inode) < new_size) { + i_size_write(inode, new_size); + mark_inode_dirty(inode); + update_inode_page(inode); + } + + return ret; +} + +static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + pgoff_t pg_start, pg_end, delta, nrpages, idx; + loff_t new_size; + int ret; + + if (!S_ISREG(inode->i_mode)) + return -EINVAL; + + new_size = i_size_read(inode) + len; + if (new_size > inode->i_sb->s_maxbytes) + return -EFBIG; + + if (offset >= i_size_read(inode)) + return -EINVAL; + + /* insert range should be aligned to block size of f2fs. */ + if (offset & (F2FS_BLKSIZE - 1) || len & (F2FS_BLKSIZE - 1)) + return -EINVAL; + + f2fs_balance_fs(sbi); + + ret = truncate_blocks(inode, i_size_read(inode), true); + if (ret) + return ret; + + /* write out all dirty pages from offset */ + ret = filemap_write_and_wait_range(inode->i_mapping, offset, LLONG_MAX); + if (ret) + return ret; + + truncate_pagecache(inode, offset); + + pg_start = offset >> PAGE_CACHE_SHIFT; + pg_end = (offset + len) >> PAGE_CACHE_SHIFT; + delta = pg_end - pg_start; + nrpages = (i_size_read(inode) + PAGE_SIZE - 1) / PAGE_SIZE; + + for (idx = nrpages - 1; idx >= pg_start && idx != -1; idx--) { + struct dnode_of_data dn; + struct page *ipage; + block_t new_addr, old_addr; + + f2fs_lock_op(sbi); + + set_new_dnode(&dn, inode, NULL, NULL, 0); + ret = get_dnode_of_data(&dn, idx, LOOKUP_NODE_RA); + if (ret && ret != -ENOENT) { + goto out; + } else if (ret == -ENOENT) { + goto next; + } else if (dn.data_blkaddr == NULL_ADDR) { + f2fs_put_dnode(&dn); + goto next; + } else { + new_addr = dn.data_blkaddr; + truncate_data_blocks_range(&dn, 1); + f2fs_put_dnode(&dn); + } + + ipage = get_node_page(sbi, inode->i_ino); + if (IS_ERR(ipage)) { + ret = PTR_ERR(ipage); + goto out; + } + + set_new_dnode(&dn, inode, ipage, NULL, 0); + ret = f2fs_reserve_block(&dn, idx + delta); + if (ret) + goto out; + + old_addr = dn.data_blkaddr; + f2fs_bug_on(sbi, old_addr != NEW_ADDR); + + if (new_addr != NEW_ADDR) { + struct node_info ni; + + get_node_info(sbi, dn.nid, &ni); + f2fs_replace_block(sbi, &dn, old_addr, new_addr, + ni.version, true); + } + f2fs_put_dnode(&dn); +next: + f2fs_unlock_op(sbi); + } + + i_size_write(inode, new_size); + return 0; +out: + f2fs_unlock_op(sbi); + return ret; +} + static int expand_inode_data(struct inode *inode, loff_t offset, loff_t len, int mode) { @@ -830,23 +1169,40 @@ static long f2fs_fallocate(struct file *file, int mode, loff_t offset, loff_t len) { struct inode *inode = file_inode(file); - long ret; + long ret = 0; + + if (f2fs_encrypted_inode(inode) && + (mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_INSERT_RANGE))) + return -EOPNOTSUPP; - if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) + if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE | + FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE | + FALLOC_FL_INSERT_RANGE)) return -EOPNOTSUPP; mutex_lock(&inode->i_mutex); - if (mode & FALLOC_FL_PUNCH_HOLE) + if (mode & FALLOC_FL_PUNCH_HOLE) { + if (offset >= inode->i_size) + goto out; + ret = punch_hole(inode, offset, len); - else + } else if (mode & FALLOC_FL_COLLAPSE_RANGE) { + ret = f2fs_collapse_range(inode, offset, len); + } else if (mode & FALLOC_FL_ZERO_RANGE) { + ret = f2fs_zero_range(inode, offset, len, mode); + } else if (mode & FALLOC_FL_INSERT_RANGE) { + ret = f2fs_insert_range(inode, offset, len); + } else { ret = expand_inode_data(inode, offset, len, mode); + } if (!ret) { inode->i_mtime = inode->i_ctime = CURRENT_TIME; mark_inode_dirty(inode); } +out: mutex_unlock(&inode->i_mutex); trace_f2fs_fallocate(inode, mode, offset, len, ret); @@ -1035,11 +1391,9 @@ static int f2fs_ioc_abort_volatile_write(struct file *filp) clear_inode_flag(F2FS_I(inode), FI_ATOMIC_FILE); } - if (f2fs_is_volatile_file(inode)) { + if (f2fs_is_volatile_file(inode)) clear_inode_flag(F2FS_I(inode), FI_VOLATILE_FILE); - filemap_fdatawrite(inode->i_mapping); - set_inode_flag(F2FS_I(inode), FI_VOLATILE_FILE); - } + mnt_drop_write_file(filp); return ret; } @@ -1109,6 +1463,86 @@ static int f2fs_ioc_fitrim(struct file *filp, unsigned long arg) return 0; } +static bool uuid_is_nonzero(__u8 u[16]) +{ + int i; + + for (i = 0; i < 16; i++) + if (u[i]) + return true; + return false; +} + +static int f2fs_ioc_set_encryption_policy(struct file *filp, unsigned long arg) +{ +#ifdef CONFIG_F2FS_FS_ENCRYPTION + struct f2fs_encryption_policy policy; + struct inode *inode = file_inode(filp); + + if (copy_from_user(&policy, (struct f2fs_encryption_policy __user *)arg, + sizeof(policy))) + return -EFAULT; + + return f2fs_process_policy(&policy, inode); +#else + return -EOPNOTSUPP; +#endif +} + +static int f2fs_ioc_get_encryption_policy(struct file *filp, unsigned long arg) +{ +#ifdef CONFIG_F2FS_FS_ENCRYPTION + struct f2fs_encryption_policy policy; + struct inode *inode = file_inode(filp); + int err; + + err = f2fs_get_policy(inode, &policy); + if (err) + return err; + + if (copy_to_user((struct f2fs_encryption_policy __user *)arg, &policy, + sizeof(policy))) + return -EFAULT; + return 0; +#else + return -EOPNOTSUPP; +#endif +} + +static int f2fs_ioc_get_encryption_pwsalt(struct file *filp, unsigned long arg) +{ + struct inode *inode = file_inode(filp); + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + int err; + + if (!f2fs_sb_has_crypto(inode->i_sb)) + return -EOPNOTSUPP; + + if (uuid_is_nonzero(sbi->raw_super->encrypt_pw_salt)) + goto got_it; + + err = mnt_want_write_file(filp); + if (err) + return err; + + /* update superblock with uuid */ + generate_random_uuid(sbi->raw_super->encrypt_pw_salt); + + err = f2fs_commit_super(sbi, false); + + mnt_drop_write_file(filp); + if (err) { + /* undo new data */ + memset(sbi->raw_super->encrypt_pw_salt, 0, 16); + return err; + } +got_it: + if (copy_to_user((__u8 __user *)arg, sbi->raw_super->encrypt_pw_salt, + 16)) + return -EFAULT; + return 0; +} + long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { switch (cmd) { @@ -1132,11 +1566,29 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return f2fs_ioc_shutdown(filp, arg); case FITRIM: return f2fs_ioc_fitrim(filp, arg); + case F2FS_IOC_SET_ENCRYPTION_POLICY: + return f2fs_ioc_set_encryption_policy(filp, arg); + case F2FS_IOC_GET_ENCRYPTION_POLICY: + return f2fs_ioc_get_encryption_policy(filp, arg); + case F2FS_IOC_GET_ENCRYPTION_PWSALT: + return f2fs_ioc_get_encryption_pwsalt(filp, arg); default: return -ENOTTY; } } +static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) +{ + struct inode *inode = file_inode(iocb->ki_filp); + + if (f2fs_encrypted_inode(inode) && + !f2fs_has_encryption_key(inode) && + f2fs_get_encryption_info(inode)) + return -EACCES; + + return generic_file_write_iter(iocb, from); +} + #ifdef CONFIG_COMPAT long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -1157,8 +1609,8 @@ long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) const struct file_operations f2fs_file_operations = { .llseek = f2fs_llseek, .read_iter = generic_file_read_iter, - .write_iter = generic_file_write_iter, - .open = generic_file_open, + .write_iter = f2fs_file_write_iter, + .open = f2fs_file_open, .release = f2fs_release_file, .mmap = f2fs_file_mmap, .fsync = f2fs_sync_file, diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index ed58211..e1e7361 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -518,12 +518,79 @@ static int check_dnode(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, return 1; } -static void move_data_page(struct inode *inode, struct page *page, int gc_type) +static void move_encrypted_block(struct inode *inode, block_t bidx) { struct f2fs_io_info fio = { + .sbi = F2FS_I_SB(inode), .type = DATA, - .rw = WRITE_SYNC, + .rw = READ_SYNC, + .encrypted_page = NULL, }; + struct dnode_of_data dn; + struct f2fs_summary sum; + struct node_info ni; + struct page *page; + int err; + + /* do not read out */ + page = grab_cache_page(inode->i_mapping, bidx); + if (!page) + return; + + set_new_dnode(&dn, inode, NULL, NULL, 0); + err = get_dnode_of_data(&dn, bidx, LOOKUP_NODE); + if (err) + goto out; + + if (unlikely(dn.data_blkaddr == NULL_ADDR)) + goto put_out; + + get_node_info(fio.sbi, dn.nid, &ni); + set_summary(&sum, dn.nid, dn.ofs_in_node, ni.version); + + /* read page */ + fio.page = page; + fio.blk_addr = dn.data_blkaddr; + + fio.encrypted_page = grab_cache_page(META_MAPPING(fio.sbi), fio.blk_addr); + if (!fio.encrypted_page) + goto put_out; + + f2fs_submit_page_bio(&fio); + + /* allocate block address */ + f2fs_wait_on_page_writeback(dn.node_page, NODE); + + allocate_data_block(fio.sbi, NULL, fio.blk_addr, + &fio.blk_addr, &sum, CURSEG_COLD_DATA); + dn.data_blkaddr = fio.blk_addr; + + /* write page */ + lock_page(fio.encrypted_page); + set_page_writeback(fio.encrypted_page); + fio.rw = WRITE_SYNC; + f2fs_submit_page_mbio(&fio); + + set_data_blkaddr(&dn); + f2fs_update_extent_cache(&dn); + set_inode_flag(F2FS_I(inode), FI_APPEND_WRITE); + if (page->index == 0) + set_inode_flag(F2FS_I(inode), FI_FIRST_BLOCK_WRITTEN); + + f2fs_put_page(fio.encrypted_page, 1); +put_out: + f2fs_put_dnode(&dn); +out: + f2fs_put_page(page, 1); +} + +static void move_data_page(struct inode *inode, block_t bidx, int gc_type) +{ + struct page *page; + + page = get_lock_data_page(inode, bidx); + if (IS_ERR(page)) + return; if (gc_type == BG_GC) { if (PageWriteback(page)) @@ -531,12 +598,19 @@ static void move_data_page(struct inode *inode, struct page *page, int gc_type) set_page_dirty(page); set_cold_data(page); } else { + struct f2fs_io_info fio = { + .sbi = F2FS_I_SB(inode), + .type = DATA, + .rw = WRITE_SYNC, + .page = page, + .encrypted_page = NULL, + }; f2fs_wait_on_page_writeback(page, DATA); if (clear_page_dirty_for_io(page)) inode_dec_dirty_pages(inode); set_cold_data(page); - do_write_data_page(page, &fio); + do_write_data_page(&fio); clear_cold_data(page); } out: @@ -599,10 +673,16 @@ next_step: if (IS_ERR(inode) || is_bad_inode(inode)) continue; - start_bidx = start_bidx_of_node(nofs, F2FS_I(inode)); + /* if encrypted inode, let's go phase 3 */ + if (f2fs_encrypted_inode(inode) && + S_ISREG(inode->i_mode)) { + add_gc_inode(gc_list, inode); + continue; + } - data_page = find_data_page(inode, - start_bidx + ofs_in_node, false); + start_bidx = start_bidx_of_node(nofs, F2FS_I(inode)); + data_page = get_read_data_page(inode, + start_bidx + ofs_in_node, READA); if (IS_ERR(data_page)) { iput(inode); continue; @@ -616,12 +696,12 @@ next_step: /* phase 3 */ inode = find_gc_inode(gc_list, dni.ino); if (inode) { - start_bidx = start_bidx_of_node(nofs, F2FS_I(inode)); - data_page = get_lock_data_page(inode, - start_bidx + ofs_in_node); - if (IS_ERR(data_page)) - continue; - move_data_page(inode, data_page, gc_type); + start_bidx = start_bidx_of_node(nofs, F2FS_I(inode)) + + ofs_in_node; + if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) + move_encrypted_block(inode, start_bidx); + else + move_data_page(inode, start_bidx, gc_type); stat_inc_data_blk_count(sbi, 1, gc_type); } } @@ -670,6 +750,15 @@ static void do_garbage_collect(struct f2fs_sb_info *sbi, unsigned int segno, sum = page_address(sum_page); + /* + * this is to avoid deadlock: + * - lock_page(sum_page) - f2fs_replace_block + * - check_valid_map() - mutex_lock(sentry_lock) + * - mutex_lock(sentry_lock) - change_curseg() + * - lock_page(sum_page) + */ + unlock_page(sum_page); + switch (GET_SUM_TYPE((&sum->footer))) { case SUM_TYPE_NODE: gc_node_segment(sbi, sum->entries, segno, gc_type); @@ -683,7 +772,7 @@ static void do_garbage_collect(struct f2fs_sb_info *sbi, unsigned int segno, stat_inc_seg_count(sbi, GET_SUM_TYPE((&sum->footer)), gc_type); stat_inc_call_count(sbi->stat_info); - f2fs_put_page(sum_page, 1); + f2fs_put_page(sum_page, 0); } int f2fs_gc(struct f2fs_sb_info *sbi) diff --git a/fs/f2fs/hash.c b/fs/f2fs/hash.c index a844fcf..71b7206 100644 --- a/fs/f2fs/hash.c +++ b/fs/f2fs/hash.c @@ -79,8 +79,7 @@ f2fs_hash_t f2fs_dentry_hash(const struct qstr *name_info) const unsigned char *name = name_info->name; size_t len = name_info->len; - if ((len <= 2) && (name[0] == '.') && - (name[1] == '.' || name[1] == '\0')) + if (is_dot_dotdot(name_info)) return 0; /* Initialize the default seed for the hash checksum functions */ diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c index 8140e4f..38e75fb 100644 --- a/fs/f2fs/inline.c +++ b/fs/f2fs/inline.c @@ -13,7 +13,7 @@ #include "f2fs.h" -bool f2fs_may_inline(struct inode *inode) +bool f2fs_may_inline_data(struct inode *inode) { if (!test_opt(F2FS_I_SB(inode), INLINE_DATA)) return false; @@ -27,6 +27,20 @@ bool f2fs_may_inline(struct inode *inode) if (i_size_read(inode) > MAX_INLINE_DATA) return false; + if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) + return false; + + return true; +} + +bool f2fs_may_inline_dentry(struct inode *inode) +{ + if (!test_opt(F2FS_I_SB(inode), INLINE_DENTRY)) + return false; + + if (!S_ISDIR(inode->i_mode)) + return false; + return true; } @@ -95,8 +109,11 @@ int f2fs_convert_inline_page(struct dnode_of_data *dn, struct page *page) { void *src_addr, *dst_addr; struct f2fs_io_info fio = { + .sbi = F2FS_I_SB(dn->inode), .type = DATA, .rw = WRITE_SYNC | REQ_PRIO, + .page = page, + .encrypted_page = NULL, }; int dirty, err; @@ -130,7 +147,7 @@ no_update: /* write data page to try to make data consistent */ set_page_writeback(page); fio.blk_addr = dn->data_blkaddr; - write_data_page(page, dn, &fio); + write_data_page(dn, &fio); set_data_blkaddr(dn); f2fs_update_extent_cache(dn); f2fs_wait_on_page_writeback(page, DATA); @@ -267,23 +284,26 @@ process_inline: } struct f2fs_dir_entry *find_in_inline_dir(struct inode *dir, - struct qstr *name, struct page **res_page) + struct f2fs_filename *fname, struct page **res_page) { struct f2fs_sb_info *sbi = F2FS_SB(dir->i_sb); struct f2fs_inline_dentry *inline_dentry; + struct qstr name = FSTR_TO_QSTR(&fname->disk_name); struct f2fs_dir_entry *de; struct f2fs_dentry_ptr d; struct page *ipage; + f2fs_hash_t namehash; ipage = get_node_page(sbi, dir->i_ino); if (IS_ERR(ipage)) return NULL; - inline_dentry = inline_data_addr(ipage); + namehash = f2fs_dentry_hash(&name); - make_dentry_ptr(&d, (void *)inline_dentry, 2); - de = find_target_dentry(name, NULL, &d); + inline_dentry = inline_data_addr(ipage); + make_dentry_ptr(NULL, &d, (void *)inline_dentry, 2); + de = find_target_dentry(fname, namehash, NULL, &d); unlock_page(ipage); if (de) *res_page = ipage; @@ -325,7 +345,7 @@ int make_empty_inline_dir(struct inode *inode, struct inode *parent, dentry_blk = inline_data_addr(ipage); - make_dentry_ptr(&d, (void *)dentry_blk, 2); + make_dentry_ptr(NULL, &d, (void *)dentry_blk, 2); do_make_empty_dir(inode, parent, &d); set_page_dirty(ipage); @@ -429,7 +449,7 @@ int f2fs_add_inline_entry(struct inode *dir, const struct qstr *name, f2fs_wait_on_page_writeback(ipage, NODE); name_hash = f2fs_dentry_hash(name); - make_dentry_ptr(&d, (void *)dentry_blk, 2); + make_dentry_ptr(NULL, &d, (void *)dentry_blk, 2); f2fs_update_dentry(ino, mode, &d, name, name_hash, bit_pos); set_page_dirty(ipage); @@ -506,7 +526,8 @@ bool f2fs_empty_inline_dir(struct inode *dir) return true; } -int f2fs_read_inline_dir(struct file *file, struct dir_context *ctx) +int f2fs_read_inline_dir(struct file *file, struct dir_context *ctx, + struct f2fs_str *fstr) { struct inode *inode = file_inode(file); struct f2fs_inline_dentry *inline_dentry = NULL; @@ -522,9 +543,9 @@ int f2fs_read_inline_dir(struct file *file, struct dir_context *ctx) inline_dentry = inline_data_addr(ipage); - make_dentry_ptr(&d, (void *)inline_dentry, 2); + make_dentry_ptr(inode, &d, (void *)inline_dentry, 2); - if (!f2fs_fill_dentries(ctx, &d, 0)) + if (!f2fs_fill_dentries(ctx, &d, 0, fstr)) ctx->pos = NR_INLINE_DENTRY; f2fs_put_page(ipage, 1); diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index e622ec9..2550868 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -198,7 +198,10 @@ make_now: inode->i_mapping->a_ops = &f2fs_dblock_aops; mapping_set_gfp_mask(inode->i_mapping, GFP_F2FS_HIGH_ZERO); } else if (S_ISLNK(inode->i_mode)) { - inode->i_op = &f2fs_symlink_inode_operations; + if (f2fs_encrypted_inode(inode)) + inode->i_op = &f2fs_encrypted_symlink_inode_operations; + else + inode->i_op = &f2fs_symlink_inode_operations; inode->i_mapping->a_ops = &f2fs_dblock_aops; } else if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode) || S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) { @@ -359,6 +362,10 @@ no_delete: if (is_inode_flag_set(F2FS_I(inode), FI_UPDATE_WRITE)) add_dirty_inode(sbi, inode->i_ino, UPDATE_INO); out_clear: +#ifdef CONFIG_F2FS_FS_ENCRYPTION + if (F2FS_I(inode)->i_crypt_info) + f2fs_free_encryption_info(inode, F2FS_I(inode)->i_crypt_info); +#endif clear_inode(inode); } diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index 71765d0..fdbae21 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -56,11 +56,18 @@ static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode) goto out; } - if (f2fs_may_inline(inode)) + /* If the directory encrypted, then we should encrypt the inode. */ + if (f2fs_encrypted_inode(dir) && f2fs_may_encrypt(inode)) + f2fs_set_encrypted_inode(inode); + + if (f2fs_may_inline_data(inode)) set_inode_flag(F2FS_I(inode), FI_INLINE_DATA); - if (test_opt(sbi, INLINE_DENTRY) && S_ISDIR(inode->i_mode)) + if (f2fs_may_inline_dentry(inode)) set_inode_flag(F2FS_I(inode), FI_INLINE_DENTRY); + stat_inc_inline_inode(inode); + stat_inc_inline_dir(inode); + trace_f2fs_new_inode(inode, 0); mark_inode_dirty(inode); return inode; @@ -136,7 +143,6 @@ static int f2fs_create(struct inode *dir, struct dentry *dentry, umode_t mode, alloc_nid_done(sbi, ino); - stat_inc_inline_inode(inode); d_instantiate(dentry, inode); unlock_new_inode(inode); @@ -155,6 +161,10 @@ static int f2fs_link(struct dentry *old_dentry, struct inode *dir, struct f2fs_sb_info *sbi = F2FS_I_SB(dir); int err; + if (f2fs_encrypted_inode(dir) && + !f2fs_is_child_context_consistent_with_parent(dir, inode)) + return -EPERM; + f2fs_balance_fs(sbi); inode->i_ctime = CURRENT_TIME; @@ -232,32 +242,34 @@ static struct dentry *f2fs_lookup(struct inode *dir, struct dentry *dentry, struct inode *inode = NULL; struct f2fs_dir_entry *de; struct page *page; + nid_t ino; + int err = 0; if (dentry->d_name.len > F2FS_NAME_LEN) return ERR_PTR(-ENAMETOOLONG); de = f2fs_find_entry(dir, &dentry->d_name, &page); - if (de) { - nid_t ino = le32_to_cpu(de->ino); - f2fs_dentry_kunmap(dir, page); - f2fs_put_page(page, 0); + if (!de) + return d_splice_alias(inode, dentry); - inode = f2fs_iget(dir->i_sb, ino); - if (IS_ERR(inode)) - return ERR_CAST(inode); + ino = le32_to_cpu(de->ino); + f2fs_dentry_kunmap(dir, page); + f2fs_put_page(page, 0); - if (f2fs_has_inline_dots(inode)) { - int err; + inode = f2fs_iget(dir->i_sb, ino); + if (IS_ERR(inode)) + return ERR_CAST(inode); - err = __recover_dot_dentries(inode, dir->i_ino); - if (err) { - iget_failed(inode); - return ERR_PTR(err); - } - } + if (f2fs_has_inline_dots(inode)) { + err = __recover_dot_dentries(inode, dir->i_ino); + if (err) + goto err_out; } - return d_splice_alias(inode, dentry); + +err_out: + iget_failed(inode); + return ERR_PTR(err); } static int f2fs_unlink(struct inode *dir, struct dentry *dentry) @@ -312,16 +324,26 @@ static int f2fs_symlink(struct inode *dir, struct dentry *dentry, { struct f2fs_sb_info *sbi = F2FS_I_SB(dir); struct inode *inode; - size_t symlen = strlen(symname) + 1; + size_t len = strlen(symname); + size_t p_len; + char *p_str; + struct f2fs_str disk_link = FSTR_INIT(NULL, 0); + struct f2fs_encrypted_symlink_data *sd = NULL; int err; + if (len > dir->i_sb->s_blocksize) + return -ENAMETOOLONG; + f2fs_balance_fs(sbi); inode = f2fs_new_inode(dir, S_IFLNK | S_IRWXUGO); if (IS_ERR(inode)) return PTR_ERR(inode); - inode->i_op = &f2fs_symlink_inode_operations; + if (f2fs_encrypted_inode(inode)) + inode->i_op = &f2fs_encrypted_symlink_inode_operations; + else + inode->i_op = &f2fs_symlink_inode_operations; inode->i_mapping->a_ops = &f2fs_dblock_aops; f2fs_lock_op(sbi); @@ -329,10 +351,46 @@ static int f2fs_symlink(struct inode *dir, struct dentry *dentry, if (err) goto out; f2fs_unlock_op(sbi); - - err = page_symlink(inode, symname, symlen); alloc_nid_done(sbi, inode->i_ino); + if (f2fs_encrypted_inode(dir)) { + struct qstr istr = QSTR_INIT(symname, len); + + err = f2fs_get_encryption_info(inode); + if (err) + goto err_out; + + err = f2fs_fname_crypto_alloc_buffer(inode, len, &disk_link); + if (err) + goto err_out; + + err = f2fs_fname_usr_to_disk(inode, &istr, &disk_link); + if (err < 0) + goto err_out; + + p_len = encrypted_symlink_data_len(disk_link.len) + 1; + + if (p_len > dir->i_sb->s_blocksize) { + err = -ENAMETOOLONG; + goto err_out; + } + + sd = kzalloc(p_len, GFP_NOFS); + if (!sd) { + err = -ENOMEM; + goto err_out; + } + memcpy(sd->encrypted_path, disk_link.name, disk_link.len); + sd->len = cpu_to_le16(disk_link.len); + p_str = (char *)sd; + } else { + p_len = len + 1; + p_str = (char *)symname; + } + + err = page_symlink(inode, p_str, p_len); + +err_out: d_instantiate(dentry, inode); unlock_new_inode(inode); @@ -345,10 +403,14 @@ static int f2fs_symlink(struct inode *dir, struct dentry *dentry, * If the symlink path is stored into inline_data, there is no * performance regression. */ - filemap_write_and_wait_range(inode->i_mapping, 0, symlen - 1); + if (!err) + filemap_write_and_wait_range(inode->i_mapping, 0, p_len - 1); if (IS_DIRSYNC(dir)) f2fs_sync_fs(sbi->sb, 1); + + kfree(sd); + f2fs_fname_crypto_free_buffer(&disk_link); return err; out: handle_failed_inode(inode); @@ -379,7 +441,6 @@ static int f2fs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) goto out_fail; f2fs_unlock_op(sbi); - stat_inc_inline_dir(inode); alloc_nid_done(sbi, inode->i_ino); d_instantiate(dentry, inode); @@ -441,19 +502,101 @@ out: return err; } +static int __f2fs_tmpfile(struct inode *dir, struct dentry *dentry, + umode_t mode, struct inode **whiteout) +{ + struct f2fs_sb_info *sbi = F2FS_I_SB(dir); + struct inode *inode; + int err; + + if (!whiteout) + f2fs_balance_fs(sbi); + + inode = f2fs_new_inode(dir, mode); + if (IS_ERR(inode)) + return PTR_ERR(inode); + + if (whiteout) { + init_special_inode(inode, inode->i_mode, WHITEOUT_DEV); + inode->i_op = &f2fs_special_inode_operations; + } else { + inode->i_op = &f2fs_file_inode_operations; + inode->i_fop = &f2fs_file_operations; + inode->i_mapping->a_ops = &f2fs_dblock_aops; + } + + f2fs_lock_op(sbi); + err = acquire_orphan_inode(sbi); + if (err) + goto out; + + err = f2fs_do_tmpfile(inode, dir); + if (err) + goto release_out; + + /* + * add this non-linked tmpfile to orphan list, in this way we could + * remove all unused data of tmpfile after abnormal power-off. + */ + add_orphan_inode(sbi, inode->i_ino); + f2fs_unlock_op(sbi); + + alloc_nid_done(sbi, inode->i_ino); + + if (whiteout) { + inode_dec_link_count(inode); + *whiteout = inode; + } else { + d_tmpfile(dentry, inode); + } + unlock_new_inode(inode); + return 0; + +release_out: + release_orphan_inode(sbi); +out: + handle_failed_inode(inode); + return err; +} + +static int f2fs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + if (f2fs_encrypted_inode(dir)) { + int err = f2fs_get_encryption_info(dir); + if (err) + return err; + } + + return __f2fs_tmpfile(dir, dentry, mode, NULL); +} + +static int f2fs_create_whiteout(struct inode *dir, struct inode **whiteout) +{ + return __f2fs_tmpfile(dir, NULL, S_IFCHR | WHITEOUT_MODE, whiteout); +} + static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, - struct inode *new_dir, struct dentry *new_dentry) + struct inode *new_dir, struct dentry *new_dentry, + unsigned int flags) { struct f2fs_sb_info *sbi = F2FS_I_SB(old_dir); struct inode *old_inode = d_inode(old_dentry); struct inode *new_inode = d_inode(new_dentry); + struct inode *whiteout = NULL; struct page *old_dir_page; - struct page *old_page, *new_page; + struct page *old_page, *new_page = NULL; struct f2fs_dir_entry *old_dir_entry = NULL; struct f2fs_dir_entry *old_entry; struct f2fs_dir_entry *new_entry; int err = -ENOENT; + if ((old_dir != new_dir) && f2fs_encrypted_inode(new_dir) && + !f2fs_is_child_context_consistent_with_parent(new_dir, + old_inode)) { + err = -EPERM; + goto out; + } + f2fs_balance_fs(sbi); old_entry = f2fs_find_entry(old_dir, &old_dentry->d_name, &old_page); @@ -467,17 +610,23 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, goto out_old; } + if (flags & RENAME_WHITEOUT) { + err = f2fs_create_whiteout(old_dir, &whiteout); + if (err) + goto out_dir; + } + if (new_inode) { err = -ENOTEMPTY; if (old_dir_entry && !f2fs_empty_dir(new_inode)) - goto out_dir; + goto out_whiteout; err = -ENOENT; new_entry = f2fs_find_entry(new_dir, &new_dentry->d_name, &new_page); if (!new_entry) - goto out_dir; + goto out_whiteout; f2fs_lock_op(sbi); @@ -485,7 +634,8 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, if (err) goto put_out_dir; - if (update_dent_inode(old_inode, &new_dentry->d_name)) { + if (update_dent_inode(old_inode, new_inode, + &new_dentry->d_name)) { release_orphan_inode(sbi); goto put_out_dir; } @@ -514,7 +664,7 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, err = f2fs_add_link(new_dentry, old_inode); if (err) { f2fs_unlock_op(sbi); - goto out_dir; + goto out_whiteout; } if (old_dir_entry) { @@ -525,6 +675,8 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, down_write(&F2FS_I(old_inode)->i_sem); file_lost_pino(old_inode); + if (new_inode && file_enc_name(new_inode)) + file_set_enc_name(old_inode); up_write(&F2FS_I(old_inode)->i_sem); old_inode->i_ctime = CURRENT_TIME; @@ -532,8 +684,18 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, f2fs_delete_entry(old_entry, old_page, old_dir, NULL); + if (whiteout) { + whiteout->i_state |= I_LINKABLE; + set_inode_flag(F2FS_I(whiteout), FI_INC_LINK); + err = f2fs_add_link(old_dentry, whiteout); + if (err) + goto put_out_dir; + whiteout->i_state &= ~I_LINKABLE; + iput(whiteout); + } + if (old_dir_entry) { - if (old_dir != new_dir) { + if (old_dir != new_dir && !whiteout) { f2fs_set_link(old_inode, old_dir_entry, old_dir_page, new_dir); update_inode_page(old_inode); @@ -554,8 +716,13 @@ static int f2fs_rename(struct inode *old_dir, struct dentry *old_dentry, put_out_dir: f2fs_unlock_op(sbi); - f2fs_dentry_kunmap(new_dir, new_page); - f2fs_put_page(new_page, 0); + if (new_page) { + f2fs_dentry_kunmap(new_dir, new_page); + f2fs_put_page(new_page, 0); + } +out_whiteout: + if (whiteout) + iput(whiteout); out_dir: if (old_dir_entry) { f2fs_dentry_kunmap(old_inode, old_dir_page); @@ -581,6 +748,14 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry, int old_nlink = 0, new_nlink = 0; int err = -ENOENT; + if ((f2fs_encrypted_inode(old_dir) || f2fs_encrypted_inode(new_dir)) && + (old_dir != new_dir) && + (!f2fs_is_child_context_consistent_with_parent(new_dir, + old_inode) || + !f2fs_is_child_context_consistent_with_parent(old_dir, + new_inode))) + return -EPERM; + f2fs_balance_fs(sbi); old_entry = f2fs_find_entry(old_dir, &old_dentry->d_name, &old_page); @@ -627,13 +802,17 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry, f2fs_lock_op(sbi); - err = update_dent_inode(old_inode, &new_dentry->d_name); + err = update_dent_inode(old_inode, new_inode, &new_dentry->d_name); if (err) goto out_unlock; + if (file_enc_name(new_inode)) + file_set_enc_name(old_inode); - err = update_dent_inode(new_inode, &old_dentry->d_name); + err = update_dent_inode(new_inode, old_inode, &old_dentry->d_name); if (err) goto out_undo; + if (file_enc_name(old_inode)) + file_set_enc_name(new_inode); /* update ".." directory entry info of old dentry */ if (old_dir_entry) @@ -691,8 +870,11 @@ static int f2fs_cross_rename(struct inode *old_dir, struct dentry *old_dentry, f2fs_sync_fs(sbi->sb, 1); return 0; out_undo: - /* Still we may fail to recover name info of f2fs_inode here */ - update_dent_inode(old_inode, &old_dentry->d_name); + /* + * Still we may fail to recover name info of f2fs_inode here + * Drop it, once its name is set as encrypted + */ + update_dent_inode(old_inode, old_inode, &old_dentry->d_name); out_unlock: f2fs_unlock_op(sbi); out_new_dir: @@ -719,7 +901,7 @@ static int f2fs_rename2(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry, unsigned int flags) { - if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE)) + if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT)) return -EINVAL; if (flags & RENAME_EXCHANGE) { @@ -730,53 +912,85 @@ static int f2fs_rename2(struct inode *old_dir, struct dentry *old_dentry, * VFS has already handled the new dentry existence case, * here, we just deal with "RENAME_NOREPLACE" as regular rename. */ - return f2fs_rename(old_dir, old_dentry, new_dir, new_dentry); + return f2fs_rename(old_dir, old_dentry, new_dir, new_dentry, flags); } -static int f2fs_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode) +#ifdef CONFIG_F2FS_FS_ENCRYPTION +static const char *f2fs_encrypted_follow_link(struct dentry *dentry, void **cookie) { - struct f2fs_sb_info *sbi = F2FS_I_SB(dir); - struct inode *inode; - int err; - - inode = f2fs_new_inode(dir, mode); - if (IS_ERR(inode)) - return PTR_ERR(inode); - - inode->i_op = &f2fs_file_inode_operations; - inode->i_fop = &f2fs_file_operations; - inode->i_mapping->a_ops = &f2fs_dblock_aops; - - f2fs_lock_op(sbi); - err = acquire_orphan_inode(sbi); - if (err) - goto out; - - err = f2fs_do_tmpfile(inode, dir); - if (err) - goto release_out; - - /* - * add this non-linked tmpfile to orphan list, in this way we could - * remove all unused data of tmpfile after abnormal power-off. - */ - add_orphan_inode(sbi, inode->i_ino); - f2fs_unlock_op(sbi); - - alloc_nid_done(sbi, inode->i_ino); - - stat_inc_inline_inode(inode); - d_tmpfile(dentry, inode); - unlock_new_inode(inode); - return 0; + struct page *cpage = NULL; + char *caddr, *paddr = NULL; + struct f2fs_str cstr; + struct f2fs_str pstr = FSTR_INIT(NULL, 0); + struct inode *inode = d_inode(dentry); + struct f2fs_encrypted_symlink_data *sd; + loff_t size = min_t(loff_t, i_size_read(inode), PAGE_SIZE - 1); + u32 max_size = inode->i_sb->s_blocksize; + int res; + + res = f2fs_get_encryption_info(inode); + if (res) + return ERR_PTR(res); + + cpage = read_mapping_page(inode->i_mapping, 0, NULL); + if (IS_ERR(cpage)) + return ERR_CAST(cpage); + caddr = kmap(cpage); + caddr[size] = 0; + + /* Symlink is encrypted */ + sd = (struct f2fs_encrypted_symlink_data *)caddr; + cstr.name = sd->encrypted_path; + cstr.len = le16_to_cpu(sd->len); + + /* this is broken symlink case */ + if (cstr.name[0] == 0 && cstr.len == 0) { + res = -ENOENT; + goto errout; + } -release_out: - release_orphan_inode(sbi); -out: - handle_failed_inode(inode); - return err; + if ((cstr.len + sizeof(struct f2fs_encrypted_symlink_data) - 1) > + max_size) { + /* Symlink data on the disk is corrupted */ + res = -EIO; + goto errout; + } + res = f2fs_fname_crypto_alloc_buffer(inode, cstr.len, &pstr); + if (res) + goto errout; + + res = f2fs_fname_disk_to_usr(inode, NULL, &cstr, &pstr); + if (res < 0) + goto errout; + + paddr = pstr.name; + + /* Null-terminate the name */ + paddr[res] = '\0'; + + kunmap(cpage); + page_cache_release(cpage); + return *cookie = paddr; +errout: + f2fs_fname_crypto_free_buffer(&pstr); + kunmap(cpage); + page_cache_release(cpage); + return ERR_PTR(res); } +const struct inode_operations f2fs_encrypted_symlink_inode_operations = { + .readlink = generic_readlink, + .follow_link = f2fs_encrypted_follow_link, + .put_link = kfree_put_link, + .getattr = f2fs_getattr, + .setattr = f2fs_setattr, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .listxattr = f2fs_listxattr, + .removexattr = generic_removexattr, +}; +#endif + const struct inode_operations f2fs_dir_inode_operations = { .create = f2fs_create, .lookup = f2fs_lookup, diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index 8ab0cf1..7dd63b7 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -53,7 +53,7 @@ bool available_free_memory(struct f2fs_sb_info *sbi, int type) PAGE_CACHE_SHIFT; res = mem_size < ((avail_ram * nm_i->ram_thresh / 100) >> 2); } else if (type == DIRTY_DENTS) { - if (sbi->sb->s_bdi->dirty_exceeded) + if (sbi->sb->s_bdi->wb.dirty_exceeded) return false; mem_size = get_pages(sbi, F2FS_DIRTY_DENTS); res = mem_size < ((avail_ram * nm_i->ram_thresh / 100) >> 1); @@ -70,7 +70,7 @@ bool available_free_memory(struct f2fs_sb_info *sbi, int type) sizeof(struct extent_node)) >> PAGE_CACHE_SHIFT; res = mem_size < ((avail_ram * nm_i->ram_thresh / 100) >> 1); } else { - if (sbi->sb->s_bdi->dirty_exceeded) + if (sbi->sb->s_bdi->wb.dirty_exceeded) return false; } return res; @@ -195,32 +195,35 @@ static unsigned int __gang_lookup_nat_set(struct f2fs_nm_info *nm_i, start, nr); } -bool is_checkpointed_node(struct f2fs_sb_info *sbi, nid_t nid) +int need_dentry_mark(struct f2fs_sb_info *sbi, nid_t nid) { struct f2fs_nm_info *nm_i = NM_I(sbi); struct nat_entry *e; - bool is_cp = true; + bool need = false; down_read(&nm_i->nat_tree_lock); e = __lookup_nat_cache(nm_i, nid); - if (e && !get_nat_flag(e, IS_CHECKPOINTED)) - is_cp = false; + if (e) { + if (!get_nat_flag(e, IS_CHECKPOINTED) && + !get_nat_flag(e, HAS_FSYNCED_INODE)) + need = true; + } up_read(&nm_i->nat_tree_lock); - return is_cp; + return need; } -bool has_fsynced_inode(struct f2fs_sb_info *sbi, nid_t ino) +bool is_checkpointed_node(struct f2fs_sb_info *sbi, nid_t nid) { struct f2fs_nm_info *nm_i = NM_I(sbi); struct nat_entry *e; - bool fsynced = false; + bool is_cp = true; down_read(&nm_i->nat_tree_lock); - e = __lookup_nat_cache(nm_i, ino); - if (e && get_nat_flag(e, HAS_FSYNCED_INODE)) - fsynced = true; + e = __lookup_nat_cache(nm_i, nid); + if (e && !get_nat_flag(e, IS_CHECKPOINTED)) + is_cp = false; up_read(&nm_i->nat_tree_lock); - return fsynced; + return is_cp; } bool need_inode_block_update(struct f2fs_sb_info *sbi, nid_t ino) @@ -312,7 +315,8 @@ static void set_node_addr(struct f2fs_sb_info *sbi, struct node_info *ni, __set_nat_cache_dirty(nm_i, e); /* update fsync_mark if its inode nat entry is still alive */ - e = __lookup_nat_cache(nm_i, ni->ino); + if (ni->nid != ni->ino) + e = __lookup_nat_cache(nm_i, ni->ino); if (e) { if (fsync_done && ni->nid == ni->ino) set_nat_flag(e, HAS_FSYNCED_INODE, true); @@ -995,8 +999,11 @@ static int read_node_page(struct page *page, int rw) struct f2fs_sb_info *sbi = F2FS_P_SB(page); struct node_info ni; struct f2fs_io_info fio = { + .sbi = sbi, .type = NODE, .rw = rw, + .page = page, + .encrypted_page = NULL, }; get_node_info(sbi, page->index, &ni); @@ -1011,7 +1018,7 @@ static int read_node_page(struct page *page, int rw) return LOCKED_PAGE; fio.blk_addr = ni.blk_addr; - return f2fs_submit_page_bio(sbi, page, &fio); + return f2fs_submit_page_bio(&fio); } /* @@ -1204,13 +1211,9 @@ continue_unlock: /* called by fsync() */ if (ino && IS_DNODE(page)) { set_fsync_mark(page, 1); - if (IS_INODE(page)) { - if (!is_checkpointed_node(sbi, ino) && - !has_fsynced_inode(sbi, ino)) - set_dentry_mark(page, 1); - else - set_dentry_mark(page, 0); - } + if (IS_INODE(page)) + set_dentry_mark(page, + need_dentry_mark(sbi, ino)); nwritten++; } else { set_fsync_mark(page, 0); @@ -1293,8 +1296,11 @@ static int f2fs_write_node_page(struct page *page, nid_t nid; struct node_info ni; struct f2fs_io_info fio = { + .sbi = sbi, .type = NODE, .rw = (wbc->sync_mode == WB_SYNC_ALL) ? WRITE_SYNC : WRITE, + .page = page, + .encrypted_page = NULL, }; trace_f2fs_writepage(page, NODE); @@ -1329,7 +1335,7 @@ static int f2fs_write_node_page(struct page *page, set_page_writeback(page); fio.blk_addr = ni.blk_addr; - write_node_page(sbi, page, nid, &fio); + write_node_page(nid, &fio); set_node_addr(sbi, &ni, fio.blk_addr, is_fsync_dnode(page)); dec_page_count(sbi, F2FS_DIRTY_NODES); up_read(&sbi->node_write); diff --git a/fs/f2fs/node.h b/fs/f2fs/node.h index c56026f..7427e95 100644 --- a/fs/f2fs/node.h +++ b/fs/f2fs/node.h @@ -343,28 +343,6 @@ static inline nid_t get_nid(struct page *p, int off, bool i) * - Mark cold node blocks in their node footer * - Mark cold data pages in page cache */ -static inline int is_file(struct inode *inode, int type) -{ - return F2FS_I(inode)->i_advise & type; -} - -static inline void set_file(struct inode *inode, int type) -{ - F2FS_I(inode)->i_advise |= type; -} - -static inline void clear_file(struct inode *inode, int type) -{ - F2FS_I(inode)->i_advise &= ~type; -} - -#define file_is_cold(inode) is_file(inode, FADVISE_COLD_BIT) -#define file_wrong_pino(inode) is_file(inode, FADVISE_LOST_PINO_BIT) -#define file_set_cold(inode) set_file(inode, FADVISE_COLD_BIT) -#define file_lost_pino(inode) set_file(inode, FADVISE_LOST_PINO_BIT) -#define file_clear_cold(inode) clear_file(inode, FADVISE_COLD_BIT) -#define file_got_pino(inode) clear_file(inode, FADVISE_LOST_PINO_BIT) - static inline int is_cold_data(struct page *page) { return PageChecked(page); diff --git a/fs/f2fs/recovery.c b/fs/f2fs/recovery.c index 8d8ea99..24a8c1d 100644 --- a/fs/f2fs/recovery.c +++ b/fs/f2fs/recovery.c @@ -83,6 +83,11 @@ static int recover_dentry(struct inode *inode, struct page *ipage) goto out; } + if (file_enc_name(inode)) { + iput(dir); + return 0; + } + name.len = le32_to_cpu(raw_inode->i_namelen); name.name = raw_inode->i_name; @@ -143,6 +148,7 @@ out: static void recover_inode(struct inode *inode, struct page *page) { struct f2fs_inode *raw = F2FS_INODE(page); + char *name; inode->i_mode = le16_to_cpu(raw->i_mode); i_size_write(inode, le64_to_cpu(raw->i_size)); @@ -153,8 +159,13 @@ static void recover_inode(struct inode *inode, struct page *page) inode->i_ctime.tv_nsec = le32_to_cpu(raw->i_ctime_nsec); inode->i_mtime.tv_nsec = le32_to_cpu(raw->i_mtime_nsec); + if (file_enc_name(inode)) + name = "<encrypted>"; + else + name = F2FS_INODE(page)->i_name; + f2fs_msg(inode->i_sb, KERN_NOTICE, "recover_inode: ino = %x, name = %s", - ino_of_node(page), F2FS_INODE(page)->i_name); + ino_of_node(page), name); } static int find_fsync_dnodes(struct f2fs_sb_info *sbi, struct list_head *head) @@ -174,7 +185,7 @@ static int find_fsync_dnodes(struct f2fs_sb_info *sbi, struct list_head *head) while (1) { struct fsync_inode_entry *entry; - if (blkaddr < MAIN_BLKADDR(sbi) || blkaddr >= MAX_BLKADDR(sbi)) + if (!is_valid_blkaddr(sbi, blkaddr, META_POR)) return 0; page = get_meta_page(sbi, blkaddr); @@ -349,7 +360,6 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode, struct f2fs_inode_info *fi = F2FS_I(inode); unsigned int start, end; struct dnode_of_data dn; - struct f2fs_summary sum; struct node_info ni; int err = 0, recovered = 0; @@ -396,7 +406,7 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode, dest = datablock_addr(page, dn.ofs_in_node); if (src != dest && dest != NEW_ADDR && dest != NULL_ADDR && - dest >= MAIN_BLKADDR(sbi) && dest < MAX_BLKADDR(sbi)) { + is_valid_blkaddr(sbi, dest, META_POR)) { if (src == NULL_ADDR) { err = reserve_new_block(&dn); @@ -409,13 +419,9 @@ static int do_recover_data(struct f2fs_sb_info *sbi, struct inode *inode, if (err) goto err; - set_summary(&sum, dn.nid, dn.ofs_in_node, ni.version); - /* write dummy data page */ - recover_data_page(sbi, NULL, &sum, src, dest); - dn.data_blkaddr = dest; - set_data_blkaddr(&dn); - f2fs_update_extent_cache(&dn); + f2fs_replace_block(sbi, &dn, src, dest, + ni.version, false); recovered++; } dn.ofs_in_node++; @@ -454,7 +460,7 @@ static int recover_data(struct f2fs_sb_info *sbi, while (1) { struct fsync_inode_entry *entry; - if (blkaddr < MAIN_BLKADDR(sbi) || blkaddr >= MAX_BLKADDR(sbi)) + if (!is_valid_blkaddr(sbi, blkaddr, META_POR)) break; ra_meta_pages_cond(sbi, blkaddr); diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index f939660..1eb3437 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -75,6 +75,14 @@ static inline unsigned long __reverse_ffs(unsigned long word) static unsigned long __find_rev_next_bit(const unsigned long *addr, unsigned long size, unsigned long offset) { + while (!f2fs_test_bit(offset, (unsigned char *)addr)) + offset++; + + if (offset > size) + offset = size; + + return offset; +#if 0 const unsigned long *p = addr + BIT_WORD(offset); unsigned long result = offset & ~(BITS_PER_LONG - 1); unsigned long tmp; @@ -121,11 +129,20 @@ found_first: return result + size; /* Nope. */ found_middle: return result + __reverse_ffs(tmp); +#endif } static unsigned long __find_rev_next_zero_bit(const unsigned long *addr, unsigned long size, unsigned long offset) { + while (f2fs_test_bit(offset, (unsigned char *)addr)) + offset++; + + if (offset > size) + offset = size; + + return offset; +#if 0 const unsigned long *p = addr + BIT_WORD(offset); unsigned long result = offset & ~(BITS_PER_LONG - 1); unsigned long tmp; @@ -173,6 +190,7 @@ found_first: return result + size; /* Nope. */ found_middle: return result + __reverse_ffz(tmp); +#endif } void register_inmem_page(struct inode *inode, struct page *page) @@ -216,8 +234,10 @@ void commit_inmem_pages(struct inode *inode, bool abort) struct inmem_pages *cur, *tmp; bool submit_bio = false; struct f2fs_io_info fio = { + .sbi = sbi, .type = DATA, .rw = WRITE_SYNC | REQ_PRIO, + .encrypted_page = NULL, }; /* @@ -241,7 +261,8 @@ void commit_inmem_pages(struct inode *inode, bool abort) if (clear_page_dirty_for_io(cur->page)) inode_dec_dirty_pages(inode); trace_f2fs_commit_inmem_page(cur->page, INMEM); - do_write_data_page(cur->page, &fio); + fio.page = cur->page; + do_write_data_page(&fio); submit_bio = true; } f2fs_put_page(cur->page, 1); @@ -466,22 +487,43 @@ static int f2fs_issue_discard(struct f2fs_sb_info *sbi, { sector_t start = SECTOR_FROM_BLOCK(blkstart); sector_t len = SECTOR_FROM_BLOCK(blklen); + struct seg_entry *se; + unsigned int offset; + block_t i; + + for (i = blkstart; i < blkstart + blklen; i++) { + se = get_seg_entry(sbi, GET_SEGNO(sbi, i)); + offset = GET_BLKOFF_FROM_SEG0(sbi, i); + + if (!f2fs_test_and_set_bit(offset, se->discard_map)) + sbi->discard_blks--; + } trace_f2fs_issue_discard(sbi->sb, blkstart, blklen); return blkdev_issue_discard(sbi->sb->s_bdev, start, len, GFP_NOFS, 0); } void discard_next_dnode(struct f2fs_sb_info *sbi, block_t blkaddr) { - if (f2fs_issue_discard(sbi, blkaddr, 1)) { - struct page *page = grab_meta_page(sbi, blkaddr); - /* zero-filled page */ - set_page_dirty(page); - f2fs_put_page(page, 1); + int err = -ENOTSUPP; + + if (test_opt(sbi, DISCARD)) { + struct seg_entry *se = get_seg_entry(sbi, + GET_SEGNO(sbi, blkaddr)); + unsigned int offset = GET_BLKOFF_FROM_SEG0(sbi, blkaddr); + + if (f2fs_test_bit(offset, se->discard_map)) + return; + + err = f2fs_issue_discard(sbi, blkaddr, 1); } + + if (err) + update_meta_page(sbi, NULL, blkaddr); } static void __add_discard_entry(struct f2fs_sb_info *sbi, - struct cp_control *cpc, unsigned int start, unsigned int end) + struct cp_control *cpc, struct seg_entry *se, + unsigned int start, unsigned int end) { struct list_head *head = &SM_I(sbi)->discard_list; struct discard_entry *new, *last; @@ -502,7 +544,6 @@ static void __add_discard_entry(struct f2fs_sb_info *sbi, list_add_tail(&new->list, head); done: SM_I(sbi)->nr_discards += end - start; - cpc->trimmed += end - start; } static void add_discard_addrs(struct f2fs_sb_info *sbi, struct cp_control *cpc) @@ -512,41 +553,24 @@ static void add_discard_addrs(struct f2fs_sb_info *sbi, struct cp_control *cpc) struct seg_entry *se = get_seg_entry(sbi, cpc->trim_start); unsigned long *cur_map = (unsigned long *)se->cur_valid_map; unsigned long *ckpt_map = (unsigned long *)se->ckpt_valid_map; + unsigned long *discard_map = (unsigned long *)se->discard_map; unsigned long *dmap = SIT_I(sbi)->tmp_map; unsigned int start = 0, end = -1; bool force = (cpc->reason == CP_DISCARD); int i; - if (!force && (!test_opt(sbi, DISCARD) || - SM_I(sbi)->nr_discards >= SM_I(sbi)->max_discards)) + if (se->valid_blocks == max_blocks) return; - if (force && !se->valid_blocks) { - struct dirty_seglist_info *dirty_i = DIRTY_I(sbi); - /* - * if this segment is registered in the prefree list, then - * we should skip adding a discard candidate, and let the - * checkpoint do that later. - */ - mutex_lock(&dirty_i->seglist_lock); - if (test_bit(cpc->trim_start, dirty_i->dirty_segmap[PRE])) { - mutex_unlock(&dirty_i->seglist_lock); - cpc->trimmed += sbi->blocks_per_seg; + if (!force) { + if (!test_opt(sbi, DISCARD) || !se->valid_blocks || + SM_I(sbi)->nr_discards >= SM_I(sbi)->max_discards) return; - } - mutex_unlock(&dirty_i->seglist_lock); - - __add_discard_entry(sbi, cpc, 0, sbi->blocks_per_seg); - return; } - /* zero block will be discarded through the prefree list */ - if (!se->valid_blocks || se->valid_blocks == max_blocks) - return; - /* SIT_VBLOCK_MAP_SIZE should be multiple of sizeof(unsigned long) */ for (i = 0; i < entries; i++) - dmap[i] = force ? ~ckpt_map[i] : + dmap[i] = force ? ~ckpt_map[i] & ~discard_map[i] : (cur_map[i] ^ ckpt_map[i]) & ckpt_map[i]; while (force || SM_I(sbi)->nr_discards <= SM_I(sbi)->max_discards) { @@ -555,11 +579,7 @@ static void add_discard_addrs(struct f2fs_sb_info *sbi, struct cp_control *cpc) break; end = __find_rev_next_zero_bit(dmap, max_blocks, start + 1); - - if (force && end - start < cpc->trim_minlen) - continue; - - __add_discard_entry(sbi, cpc, start, end); + __add_discard_entry(sbi, cpc, se, start, end); } } @@ -589,7 +609,7 @@ static void set_prefree_as_free_segments(struct f2fs_sb_info *sbi) mutex_unlock(&dirty_i->seglist_lock); } -void clear_prefree_segments(struct f2fs_sb_info *sbi) +void clear_prefree_segments(struct f2fs_sb_info *sbi, struct cp_control *cpc) { struct list_head *head = &(SM_I(sbi)->discard_list); struct discard_entry *entry, *this; @@ -622,7 +642,11 @@ void clear_prefree_segments(struct f2fs_sb_info *sbi) /* send small discards */ list_for_each_entry_safe(entry, this, head, list) { + if (cpc->reason == CP_DISCARD && entry->len < cpc->trim_minlen) + goto skip; f2fs_issue_discard(sbi, entry->blkaddr, entry->len); + cpc->trimmed += entry->len; +skip: list_del(&entry->list); SM_I(sbi)->nr_discards -= entry->len; kmem_cache_free(discard_entry_slab, entry); @@ -673,9 +697,13 @@ static void update_sit_entry(struct f2fs_sb_info *sbi, block_t blkaddr, int del) if (del > 0) { if (f2fs_test_and_set_bit(offset, se->cur_valid_map)) f2fs_bug_on(sbi, 1); + if (!f2fs_test_and_set_bit(offset, se->discard_map)) + sbi->discard_blks--; } else { if (!f2fs_test_and_clear_bit(offset, se->cur_valid_map)) f2fs_bug_on(sbi, 1); + if (f2fs_test_and_clear_bit(offset, se->discard_map)) + sbi->discard_blks++; } if (!f2fs_test_bit(offset, se->ckpt_valid_map)) se->ckpt_valid_blocks += del; @@ -769,16 +797,25 @@ struct page *get_sum_page(struct f2fs_sb_info *sbi, unsigned int segno) return get_meta_page(sbi, GET_SUM_BLOCK(sbi, segno)); } -static void write_sum_page(struct f2fs_sb_info *sbi, - struct f2fs_summary_block *sum_blk, block_t blk_addr) +void update_meta_page(struct f2fs_sb_info *sbi, void *src, block_t blk_addr) { struct page *page = grab_meta_page(sbi, blk_addr); - void *kaddr = page_address(page); - memcpy(kaddr, sum_blk, PAGE_CACHE_SIZE); + void *dst = page_address(page); + + if (src) + memcpy(dst, src, PAGE_CACHE_SIZE); + else + memset(dst, 0, PAGE_CACHE_SIZE); set_page_dirty(page); f2fs_put_page(page, 1); } +static void write_sum_page(struct f2fs_sb_info *sbi, + struct f2fs_summary_block *sum_blk, block_t blk_addr) +{ + update_meta_page(sbi, (void *)sum_blk, blk_addr); +} + static int is_next_segment_free(struct f2fs_sb_info *sbi, int type) { struct curseg_info *curseg = CURSEG_I(sbi, type); @@ -1060,8 +1097,7 @@ int f2fs_trim_fs(struct f2fs_sb_info *sbi, struct fstrim_range *range) unsigned int start_segno, end_segno; struct cp_control cpc; - if (range->minlen > SEGMENT_SIZE(sbi) || start >= MAX_BLKADDR(sbi) || - range->len < sbi->blocksize) + if (start >= MAX_BLKADDR(sbi) || range->len < sbi->blocksize) return -EINVAL; cpc.trimmed = 0; @@ -1073,12 +1109,19 @@ int f2fs_trim_fs(struct f2fs_sb_info *sbi, struct fstrim_range *range) end_segno = (end >= MAX_BLKADDR(sbi)) ? MAIN_SEGS(sbi) - 1 : GET_SEGNO(sbi, end); cpc.reason = CP_DISCARD; - cpc.trim_minlen = F2FS_BYTES_TO_BLK(range->minlen); + cpc.trim_minlen = max_t(__u64, 1, F2FS_BYTES_TO_BLK(range->minlen)); /* do checkpoint to issue discard commands safely */ for (; start_segno <= end_segno; start_segno = cpc.trim_end + 1) { cpc.trim_start = start_segno; - cpc.trim_end = min_t(unsigned int, rounddown(start_segno + + + if (sbi->discard_blks == 0) + break; + else if (sbi->discard_blks < BATCHED_TRIM_BLOCKS(sbi)) + cpc.trim_end = end_segno; + else + cpc.trim_end = min_t(unsigned int, + rounddown(start_segno + BATCHED_TRIM_SEGMENTS(sbi), sbi->segs_per_sec) - 1, end_segno); @@ -1206,84 +1249,95 @@ void allocate_data_block(struct f2fs_sb_info *sbi, struct page *page, mutex_unlock(&curseg->curseg_mutex); } -static void do_write_page(struct f2fs_sb_info *sbi, struct page *page, - struct f2fs_summary *sum, - struct f2fs_io_info *fio) +static void do_write_page(struct f2fs_summary *sum, struct f2fs_io_info *fio) { - int type = __get_segment_type(page, fio->type); + int type = __get_segment_type(fio->page, fio->type); - allocate_data_block(sbi, page, fio->blk_addr, &fio->blk_addr, sum, type); + allocate_data_block(fio->sbi, fio->page, fio->blk_addr, + &fio->blk_addr, sum, type); /* writeout dirty page into bdev */ - f2fs_submit_page_mbio(sbi, page, fio); + f2fs_submit_page_mbio(fio); } void write_meta_page(struct f2fs_sb_info *sbi, struct page *page) { struct f2fs_io_info fio = { + .sbi = sbi, .type = META, .rw = WRITE_SYNC | REQ_META | REQ_PRIO, .blk_addr = page->index, + .page = page, + .encrypted_page = NULL, }; set_page_writeback(page); - f2fs_submit_page_mbio(sbi, page, &fio); + f2fs_submit_page_mbio(&fio); } -void write_node_page(struct f2fs_sb_info *sbi, struct page *page, - unsigned int nid, struct f2fs_io_info *fio) +void write_node_page(unsigned int nid, struct f2fs_io_info *fio) { struct f2fs_summary sum; + set_summary(&sum, nid, 0, 0); - do_write_page(sbi, page, &sum, fio); + do_write_page(&sum, fio); } -void write_data_page(struct page *page, struct dnode_of_data *dn, - struct f2fs_io_info *fio) +void write_data_page(struct dnode_of_data *dn, struct f2fs_io_info *fio) { - struct f2fs_sb_info *sbi = F2FS_I_SB(dn->inode); + struct f2fs_sb_info *sbi = fio->sbi; struct f2fs_summary sum; struct node_info ni; f2fs_bug_on(sbi, dn->data_blkaddr == NULL_ADDR); get_node_info(sbi, dn->nid, &ni); set_summary(&sum, dn->nid, dn->ofs_in_node, ni.version); - do_write_page(sbi, page, &sum, fio); + do_write_page(&sum, fio); dn->data_blkaddr = fio->blk_addr; } -void rewrite_data_page(struct page *page, struct f2fs_io_info *fio) +void rewrite_data_page(struct f2fs_io_info *fio) { - stat_inc_inplace_blocks(F2FS_P_SB(page)); - f2fs_submit_page_mbio(F2FS_P_SB(page), page, fio); + stat_inc_inplace_blocks(fio->sbi); + f2fs_submit_page_mbio(fio); } -void recover_data_page(struct f2fs_sb_info *sbi, - struct page *page, struct f2fs_summary *sum, - block_t old_blkaddr, block_t new_blkaddr) +static void __f2fs_replace_block(struct f2fs_sb_info *sbi, + struct f2fs_summary *sum, + block_t old_blkaddr, block_t new_blkaddr, + bool recover_curseg) { struct sit_info *sit_i = SIT_I(sbi); struct curseg_info *curseg; unsigned int segno, old_cursegno; struct seg_entry *se; int type; + unsigned short old_blkoff; segno = GET_SEGNO(sbi, new_blkaddr); se = get_seg_entry(sbi, segno); type = se->type; - if (se->valid_blocks == 0 && !IS_CURSEG(sbi, segno)) { - if (old_blkaddr == NULL_ADDR) - type = CURSEG_COLD_DATA; - else + if (!recover_curseg) { + /* for recovery flow */ + if (se->valid_blocks == 0 && !IS_CURSEG(sbi, segno)) { + if (old_blkaddr == NULL_ADDR) + type = CURSEG_COLD_DATA; + else + type = CURSEG_WARM_DATA; + } + } else { + if (!IS_CURSEG(sbi, segno)) type = CURSEG_WARM_DATA; } + curseg = CURSEG_I(sbi, type); mutex_lock(&curseg->curseg_mutex); mutex_lock(&sit_i->sentry_lock); old_cursegno = curseg->segno; + old_blkoff = curseg->next_blkoff; /* change the current segment */ if (segno != curseg->segno) { @@ -1297,30 +1351,67 @@ void recover_data_page(struct f2fs_sb_info *sbi, refresh_sit_entry(sbi, old_blkaddr, new_blkaddr); locate_dirty_segment(sbi, old_cursegno); + if (recover_curseg) { + if (old_cursegno != curseg->segno) { + curseg->next_segno = old_cursegno; + change_curseg(sbi, type, true); + } + curseg->next_blkoff = old_blkoff; + } + mutex_unlock(&sit_i->sentry_lock); mutex_unlock(&curseg->curseg_mutex); } +void f2fs_replace_block(struct f2fs_sb_info *sbi, struct dnode_of_data *dn, + block_t old_addr, block_t new_addr, + unsigned char version, bool recover_curseg) +{ + struct f2fs_summary sum; + + set_summary(&sum, dn->nid, dn->ofs_in_node, version); + + __f2fs_replace_block(sbi, &sum, old_addr, new_addr, recover_curseg); + + dn->data_blkaddr = new_addr; + set_data_blkaddr(dn); + f2fs_update_extent_cache(dn); +} + static inline bool is_merged_page(struct f2fs_sb_info *sbi, struct page *page, enum page_type type) { enum page_type btype = PAGE_TYPE_OF_BIO(type); struct f2fs_bio_info *io = &sbi->write_io[btype]; struct bio_vec *bvec; + struct page *target; int i; down_read(&io->io_rwsem); - if (!io->bio) - goto out; + if (!io->bio) { + up_read(&io->io_rwsem); + return false; + } bio_for_each_segment_all(bvec, io->bio, i) { - if (page == bvec->bv_page) { + + if (bvec->bv_page->mapping) { + target = bvec->bv_page; + } else { + struct f2fs_crypto_ctx *ctx; + + /* encrypted page */ + ctx = (struct f2fs_crypto_ctx *)page_private( + bvec->bv_page); + target = ctx->w.control_page; + } + + if (page == target) { up_read(&io->io_rwsem); return true; } } -out: up_read(&io->io_rwsem); return false; } @@ -1857,8 +1948,11 @@ static int build_sit_info(struct f2fs_sb_info *sbi) = kzalloc(SIT_VBLOCK_MAP_SIZE, GFP_KERNEL); sit_i->sentries[start].ckpt_valid_map = kzalloc(SIT_VBLOCK_MAP_SIZE, GFP_KERNEL); - if (!sit_i->sentries[start].cur_valid_map - || !sit_i->sentries[start].ckpt_valid_map) + sit_i->sentries[start].discard_map + = kzalloc(SIT_VBLOCK_MAP_SIZE, GFP_KERNEL); + if (!sit_i->sentries[start].cur_valid_map || + !sit_i->sentries[start].ckpt_valid_map || + !sit_i->sentries[start].discard_map) return -ENOMEM; } @@ -1996,6 +2090,11 @@ static void build_sit_entries(struct f2fs_sb_info *sbi) got_it: check_block_count(sbi, start, &sit); seg_info_from_raw_sit(se, &sit); + + /* build discard map only one time */ + memcpy(se->discard_map, se->cur_valid_map, SIT_VBLOCK_MAP_SIZE); + sbi->discard_blks += sbi->blocks_per_seg - se->valid_blocks; + if (sbi->segs_per_sec > 1) { struct sec_entry *e = get_sec_entry(sbi, start); e->valid_blocks += se->valid_blocks; @@ -2245,6 +2344,7 @@ static void destroy_sit_info(struct f2fs_sb_info *sbi) for (start = 0; start < MAIN_SEGS(sbi); start++) { kfree(sit_i->sentries[start].cur_valid_map); kfree(sit_i->sentries[start].ckpt_valid_map); + kfree(sit_i->sentries[start].discard_map); } } kfree(sit_i->tmp_map); diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index 85d7fa7..79e7b87 100644 --- a/fs/f2fs/segment.h +++ b/fs/f2fs/segment.h @@ -9,6 +9,7 @@ * published by the Free Software Foundation. */ #include <linux/blkdev.h> +#include <linux/backing-dev.h> /* constant macro */ #define NULL_SEGNO ((unsigned int)(~0)) @@ -163,6 +164,7 @@ struct seg_entry { */ unsigned short ckpt_valid_blocks; unsigned char *ckpt_valid_map; + unsigned char *discard_map; unsigned char type; /* segment type like CURSEG_XXX_TYPE */ unsigned long long mtime; /* modification time of the segment */ }; @@ -713,7 +715,7 @@ static inline unsigned int max_hw_blocks(struct f2fs_sb_info *sbi) */ static inline int nr_pages_to_skip(struct f2fs_sb_info *sbi, int type) { - if (sbi->sb->s_bdi->dirty_exceeded) + if (sbi->sb->s_bdi->wb.dirty_exceeded) return 0; if (type == DATA) diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index b2dd1b0..a06b0b4 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -258,6 +258,7 @@ static void init_once(void *foo) static int parse_options(struct super_block *sb, char *options) { struct f2fs_sb_info *sbi = F2FS_SB(sb); + struct request_queue *q; substring_t args[MAX_OPT_ARGS]; char *p, *name; int arg = 0; @@ -302,7 +303,14 @@ static int parse_options(struct super_block *sb, char *options) return -EINVAL; break; case Opt_discard: - set_opt(sbi, DISCARD); + q = bdev_get_queue(sb->s_bdev); + if (blk_queue_discard(q)) { + set_opt(sbi, DISCARD); + } else { + f2fs_msg(sb, KERN_WARNING, + "mounting with \"discard\" option, but " + "the device does not support discard"); + } break; case Opt_noheap: set_opt(sbi, NOHEAP); @@ -416,6 +424,9 @@ static struct inode *f2fs_alloc_inode(struct super_block *sb) /* Will be used by directory only */ fi->i_dir_level = F2FS_SB(sb)->dir_level; +#ifdef CONFIG_F2FS_FS_ENCRYPTION + fi->i_crypt_info = NULL; +#endif return &fi->vfs_inode; } @@ -428,8 +439,31 @@ static int f2fs_drop_inode(struct inode *inode) * - f2fs_gc -> iput -> evict * - inode_wait_for_writeback(inode) */ - if (!inode_unhashed(inode) && inode->i_state & I_SYNC) + if (!inode_unhashed(inode) && inode->i_state & I_SYNC) { + if (!inode->i_nlink && !is_bad_inode(inode)) { + spin_unlock(&inode->i_lock); + + /* some remained atomic pages should discarded */ + if (f2fs_is_atomic_file(inode)) + commit_inmem_pages(inode, true); + + sb_start_intwrite(inode->i_sb); + i_size_write(inode, 0); + + if (F2FS_HAS_BLOCKS(inode)) + f2fs_truncate(inode); + + sb_end_intwrite(inode->i_sb); + +#ifdef CONFIG_F2FS_FS_ENCRYPTION + if (F2FS_I(inode)->i_crypt_info) + f2fs_free_encryption_info(inode, + F2FS_I(inode)->i_crypt_info); +#endif + spin_lock(&inode->i_lock); + } return 0; + } return generic_drop_inode(inode); } @@ -520,7 +554,7 @@ int f2fs_sync_fs(struct super_block *sb, int sync) } else { f2fs_balance_fs(sbi); } - f2fs_trace_ios(NULL, NULL, 1); + f2fs_trace_ios(NULL, 1); return 0; } @@ -658,6 +692,22 @@ static const struct file_operations f2fs_seq_segment_info_fops = { .release = single_release, }; +static void default_options(struct f2fs_sb_info *sbi) +{ + /* init some FS parameters */ + sbi->active_logs = NR_CURSEG_TYPE; + + set_opt(sbi, BG_GC); + set_opt(sbi, INLINE_DATA); + +#ifdef CONFIG_F2FS_FS_XATTR + set_opt(sbi, XATTR_USER); +#endif +#ifdef CONFIG_F2FS_FS_POSIX_ACL + set_opt(sbi, POSIX_ACL); +#endif +} + static int f2fs_remount(struct super_block *sb, int *flags, char *data) { struct f2fs_sb_info *sbi = F2FS_SB(sb); @@ -676,7 +726,7 @@ static int f2fs_remount(struct super_block *sb, int *flags, char *data) active_logs = sbi->active_logs; sbi->mount_opt.opt = 0; - sbi->active_logs = NR_CURSEG_TYPE; + default_options(sbi); /* parse mount options */ err = parse_options(sb, data); @@ -929,29 +979,36 @@ static void init_sb_info(struct f2fs_sb_info *sbi) */ static int read_raw_super_block(struct super_block *sb, struct f2fs_super_block **raw_super, - struct buffer_head **raw_super_buf) + struct buffer_head **raw_super_buf, + int *recovery) { int block = 0; + struct buffer_head *buffer; + struct f2fs_super_block *super; + int err = 0; retry: - *raw_super_buf = sb_bread(sb, block); - if (!*raw_super_buf) { + buffer = sb_bread(sb, block); + if (!buffer) { + *recovery = 1; f2fs_msg(sb, KERN_ERR, "Unable to read %dth superblock", block + 1); if (block == 0) { block++; goto retry; } else { - return -EIO; + err = -EIO; + goto out; } } - *raw_super = (struct f2fs_super_block *) - ((char *)(*raw_super_buf)->b_data + F2FS_SUPER_OFFSET); + super = (struct f2fs_super_block *) + ((char *)(buffer)->b_data + F2FS_SUPER_OFFSET); /* sanity checking of raw super */ - if (sanity_check_raw_super(sb, *raw_super)) { - brelse(*raw_super_buf); + if (sanity_check_raw_super(sb, super)) { + brelse(buffer); + *recovery = 1; f2fs_msg(sb, KERN_ERR, "Can't find valid F2FS filesystem in %dth superblock", block + 1); @@ -959,25 +1016,76 @@ retry: block++; goto retry; } else { - return -EINVAL; + err = -EINVAL; + goto out; } } + if (!*raw_super) { + *raw_super_buf = buffer; + *raw_super = super; + } else { + /* already have a valid superblock */ + brelse(buffer); + } + + /* check the validity of the second superblock */ + if (block == 0) { + block++; + goto retry; + } + +out: + /* No valid superblock */ + if (!*raw_super) + return err; + return 0; } +int f2fs_commit_super(struct f2fs_sb_info *sbi, bool recover) +{ + struct buffer_head *sbh = sbi->raw_super_buf; + sector_t block = sbh->b_blocknr; + int err; + + /* write back-up superblock first */ + sbh->b_blocknr = block ? 0 : 1; + mark_buffer_dirty(sbh); + err = sync_dirty_buffer(sbh); + + sbh->b_blocknr = block; + + /* if we are in recovery path, skip writing valid superblock */ + if (recover || err) + goto out; + + /* write current valid superblock */ + mark_buffer_dirty(sbh); + err = sync_dirty_buffer(sbh); +out: + clear_buffer_write_io_error(sbh); + set_buffer_uptodate(sbh); + return err; +} + static int f2fs_fill_super(struct super_block *sb, void *data, int silent) { struct f2fs_sb_info *sbi; - struct f2fs_super_block *raw_super = NULL; + struct f2fs_super_block *raw_super; struct buffer_head *raw_super_buf; struct inode *root; - long err = -EINVAL; + long err; bool retry = true, need_fsck = false; char *options = NULL; - int i; + int recovery, i; try_onemore: + err = -EINVAL; + raw_super = NULL; + raw_super_buf = NULL; + recovery = 0; + /* allocate memory for f2fs-specific super block info */ sbi = kzalloc(sizeof(struct f2fs_sb_info), GFP_KERNEL); if (!sbi) @@ -989,23 +1097,12 @@ try_onemore: goto free_sbi; } - err = read_raw_super_block(sb, &raw_super, &raw_super_buf); + err = read_raw_super_block(sb, &raw_super, &raw_super_buf, &recovery); if (err) goto free_sbi; sb->s_fs_info = sbi; - /* init some FS parameters */ - sbi->active_logs = NR_CURSEG_TYPE; - - set_opt(sbi, BG_GC); - set_opt(sbi, INLINE_DATA); - -#ifdef CONFIG_F2FS_FS_XATTR - set_opt(sbi, XATTR_USER); -#endif -#ifdef CONFIG_F2FS_FS_POSIX_ACL - set_opt(sbi, POSIX_ACL); -#endif + default_options(sbi); /* parse mount options */ options = kstrdup((const char *)data, GFP_KERNEL); if (data && !options) { @@ -1148,14 +1245,6 @@ try_onemore: proc_create_data("segment_info", S_IRUGO, sbi->s_proc, &f2fs_seq_segment_info_fops, sb); - if (test_opt(sbi, DISCARD)) { - struct request_queue *q = bdev_get_queue(sb->s_bdev); - if (!blk_queue_discard(q)) - f2fs_msg(sb, KERN_WARNING, - "mounting with \"discard\" option, but " - "the device does not support discard"); - } - sbi->s_kobj.kset = f2fs_kset; init_completion(&sbi->s_kobj_unregister); err = kobject_init_and_add(&sbi->s_kobj, &f2fs_ktype, NULL, @@ -1198,6 +1287,13 @@ try_onemore: goto free_kobj; } kfree(options); + + /* recover broken superblock */ + if (recovery && !f2fs_readonly(sb) && !bdev_read_only(sb->s_bdev)) { + f2fs_msg(sb, KERN_INFO, "Recover invalid superblock"); + f2fs_commit_super(sbi, true); + } + return 0; free_kobj: @@ -1305,13 +1401,18 @@ static int __init init_f2fs_fs(void) err = -ENOMEM; goto free_extent_cache; } - err = register_filesystem(&f2fs_fs_type); + err = f2fs_init_crypto(); if (err) goto free_kset; + err = register_filesystem(&f2fs_fs_type); + if (err) + goto free_crypto; f2fs_create_root_stats(); f2fs_proc_root = proc_mkdir("fs/f2fs", NULL); return 0; +free_crypto: + f2fs_exit_crypto(); free_kset: kset_unregister(f2fs_kset); free_extent_cache: @@ -1333,6 +1434,7 @@ static void __exit exit_f2fs_fs(void) remove_proc_entry("fs/f2fs", NULL); f2fs_destroy_root_stats(); unregister_filesystem(&f2fs_fs_type); + f2fs_exit_crypto(); destroy_extent_cache(); destroy_checkpoint_caches(); destroy_segment_manager_caches(); diff --git a/fs/f2fs/trace.c b/fs/f2fs/trace.c index 875aa81..145fb65 100644 --- a/fs/f2fs/trace.c +++ b/fs/f2fs/trace.c @@ -80,7 +80,7 @@ out: radix_tree_preload_end(); } -void f2fs_trace_ios(struct page *page, struct f2fs_io_info *fio, int flush) +void f2fs_trace_ios(struct f2fs_io_info *fio, int flush) { struct inode *inode; pid_t pid; @@ -91,8 +91,8 @@ void f2fs_trace_ios(struct page *page, struct f2fs_io_info *fio, int flush) return; } - inode = page->mapping->host; - pid = page_private(page); + inode = fio->page->mapping->host; + pid = page_private(fio->page); major = MAJOR(inode->i_sb->s_dev); minor = MINOR(inode->i_sb->s_dev); diff --git a/fs/f2fs/trace.h b/fs/f2fs/trace.h index 1041dbe..67db24a 100644 --- a/fs/f2fs/trace.h +++ b/fs/f2fs/trace.h @@ -33,12 +33,12 @@ struct last_io_info { }; extern void f2fs_trace_pid(struct page *); -extern void f2fs_trace_ios(struct page *, struct f2fs_io_info *, int); +extern void f2fs_trace_ios(struct f2fs_io_info *, int); extern void f2fs_build_trace_ios(void); extern void f2fs_destroy_trace_ios(void); #else #define f2fs_trace_pid(p) -#define f2fs_trace_ios(p, i, n) +#define f2fs_trace_ios(i, n) #define f2fs_build_trace_ios() #define f2fs_destroy_trace_ios() diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c index 9757f65..07449b98 100644 --- a/fs/f2fs/xattr.c +++ b/fs/f2fs/xattr.c @@ -584,6 +584,9 @@ static int __f2fs_setxattr(struct inode *inode, int index, inode->i_ctime = CURRENT_TIME; clear_inode_flag(fi, FI_ACL_MODE); } + if (index == F2FS_XATTR_INDEX_ENCRYPTION && + !strcmp(name, F2FS_XATTR_NAME_ENCRYPTION_CONTEXT)) + f2fs_set_encrypted_inode(inode); if (ipage) update_inode(inode, ipage); diff --git a/fs/f2fs/xattr.h b/fs/f2fs/xattr.h index 969d792..71a7100 100644 --- a/fs/f2fs/xattr.h +++ b/fs/f2fs/xattr.h @@ -35,6 +35,10 @@ #define F2FS_XATTR_INDEX_LUSTRE 5 #define F2FS_XATTR_INDEX_SECURITY 6 #define F2FS_XATTR_INDEX_ADVISE 7 +/* Should be same as EXT4_XATTR_INDEX_ENCRYPTION */ +#define F2FS_XATTR_INDEX_ENCRYPTION 9 + +#define F2FS_XATTR_NAME_ENCRYPTION_CONTEXT "c" struct f2fs_xattr_header { __le32 h_magic; /* magic number for identification */ diff --git a/fs/fat/file.c b/fs/fat/file.c index 442d50a..a08f103 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -11,6 +11,7 @@ #include <linux/compat.h> #include <linux/mount.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <linux/fsnotify.h> #include <linux/security.h> #include "fat.h" diff --git a/fs/fat/inode.c b/fs/fat/inode.c index c067746..509411d 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -18,6 +18,7 @@ #include <linux/parser.h> #include <linux/uio.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <asm/unaligned.h> #include "fat.h" diff --git a/fs/fhandle.c b/fs/fhandle.c index 999ff5c..d59712d 100644 --- a/fs/fhandle.c +++ b/fs/fhandle.c @@ -195,8 +195,9 @@ static int handle_to_path(int mountdirfd, struct file_handle __user *ufh, goto out_err; } /* copy the full handle */ - if (copy_from_user(handle, ufh, - sizeof(struct file_handle) + + *handle = f_handle; + if (copy_from_user(&handle->f_handle, + &ufh->f_handle, f_handle.handle_bytes)) { retval = -EFAULT; goto out_handle; diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 32a8bbd..f0520bc 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -27,6 +27,7 @@ #include <linux/backing-dev.h> #include <linux/tracepoint.h> #include <linux/device.h> +#include <linux/memcontrol.h> #include "internal.h" /* @@ -34,6 +35,10 @@ */ #define MIN_WRITEBACK_PAGES (4096UL >> (PAGE_CACHE_SHIFT - 10)) +struct wb_completion { + atomic_t cnt; +}; + /* * Passed into wb_writeback(), essentially a subset of writeback_control */ @@ -47,13 +52,29 @@ struct wb_writeback_work { unsigned int range_cyclic:1; unsigned int for_background:1; unsigned int for_sync:1; /* sync(2) WB_SYNC_ALL writeback */ + unsigned int auto_free:1; /* free on completion */ + unsigned int single_wait:1; + unsigned int single_done:1; enum wb_reason reason; /* why was writeback initiated? */ struct list_head list; /* pending work list */ - struct completion *done; /* set if the caller waits */ + struct wb_completion *done; /* set if the caller waits */ }; /* + * If one wants to wait for one or more wb_writeback_works, each work's + * ->done should be set to a wb_completion defined using the following + * macro. Once all work items are issued with wb_queue_work(), the caller + * can wait for the completion of all using wb_wait_for_completion(). Work + * items which are waited upon aren't freed automatically on completion. + */ +#define DEFINE_WB_COMPLETION_ONSTACK(cmpl) \ + struct wb_completion cmpl = { \ + .cnt = ATOMIC_INIT(1), \ + } + + +/* * If an inode is constantly having its pages dirtied, but then the * updates stop dirtytime_expire_interval seconds in the past, it's * possible for the worst case time between when an inode has its @@ -65,35 +86,6 @@ struct wb_writeback_work { */ unsigned int dirtytime_expire_interval = 12 * 60 * 60; -/** - * writeback_in_progress - determine whether there is writeback in progress - * @bdi: the device's backing_dev_info structure. - * - * Determine whether there is writeback waiting to be handled against a - * backing device. - */ -int writeback_in_progress(struct backing_dev_info *bdi) -{ - return test_bit(BDI_writeback_running, &bdi->state); -} -EXPORT_SYMBOL(writeback_in_progress); - -struct backing_dev_info *inode_to_bdi(struct inode *inode) -{ - struct super_block *sb; - - if (!inode) - return &noop_backing_dev_info; - - sb = inode->i_sb; -#ifdef CONFIG_BLOCK - if (sb_is_blkdev_sb(sb)) - return blk_get_backing_dev_info(I_BDEV(inode)); -#endif - return sb->s_bdi; -} -EXPORT_SYMBOL_GPL(inode_to_bdi); - static inline struct inode *wb_inode(struct list_head *head) { return list_entry(head, struct inode, i_wb_list); @@ -109,45 +101,830 @@ static inline struct inode *wb_inode(struct list_head *head) EXPORT_TRACEPOINT_SYMBOL_GPL(wbc_writepage); -static void bdi_wakeup_thread(struct backing_dev_info *bdi) +static bool wb_io_lists_populated(struct bdi_writeback *wb) +{ + if (wb_has_dirty_io(wb)) { + return false; + } else { + set_bit(WB_has_dirty_io, &wb->state); + WARN_ON_ONCE(!wb->avg_write_bandwidth); + atomic_long_add(wb->avg_write_bandwidth, + &wb->bdi->tot_write_bandwidth); + return true; + } +} + +static void wb_io_lists_depopulated(struct bdi_writeback *wb) { - spin_lock_bh(&bdi->wb_lock); - if (test_bit(BDI_registered, &bdi->state)) - mod_delayed_work(bdi_wq, &bdi->wb.dwork, 0); - spin_unlock_bh(&bdi->wb_lock); + if (wb_has_dirty_io(wb) && list_empty(&wb->b_dirty) && + list_empty(&wb->b_io) && list_empty(&wb->b_more_io)) { + clear_bit(WB_has_dirty_io, &wb->state); + WARN_ON_ONCE(atomic_long_sub_return(wb->avg_write_bandwidth, + &wb->bdi->tot_write_bandwidth) < 0); + } } -static void bdi_queue_work(struct backing_dev_info *bdi, - struct wb_writeback_work *work) +/** + * inode_wb_list_move_locked - move an inode onto a bdi_writeback IO list + * @inode: inode to be moved + * @wb: target bdi_writeback + * @head: one of @wb->b_{dirty|io|more_io} + * + * Move @inode->i_wb_list to @list of @wb and set %WB_has_dirty_io. + * Returns %true if @inode is the first occupant of the !dirty_time IO + * lists; otherwise, %false. + */ +static bool inode_wb_list_move_locked(struct inode *inode, + struct bdi_writeback *wb, + struct list_head *head) { - trace_writeback_queue(bdi, work); + assert_spin_locked(&wb->list_lock); + + list_move(&inode->i_wb_list, head); - spin_lock_bh(&bdi->wb_lock); - if (!test_bit(BDI_registered, &bdi->state)) { - if (work->done) - complete(work->done); + /* dirty_time doesn't count as dirty_io until expiration */ + if (head != &wb->b_dirty_time) + return wb_io_lists_populated(wb); + + wb_io_lists_depopulated(wb); + return false; +} + +/** + * inode_wb_list_del_locked - remove an inode from its bdi_writeback IO list + * @inode: inode to be removed + * @wb: bdi_writeback @inode is being removed from + * + * Remove @inode which may be on one of @wb->b_{dirty|io|more_io} lists and + * clear %WB_has_dirty_io if all are empty afterwards. + */ +static void inode_wb_list_del_locked(struct inode *inode, + struct bdi_writeback *wb) +{ + assert_spin_locked(&wb->list_lock); + + list_del_init(&inode->i_wb_list); + wb_io_lists_depopulated(wb); +} + +static void wb_wakeup(struct bdi_writeback *wb) +{ + spin_lock_bh(&wb->work_lock); + if (test_bit(WB_registered, &wb->state)) + mod_delayed_work(bdi_wq, &wb->dwork, 0); + spin_unlock_bh(&wb->work_lock); +} + +static void wb_queue_work(struct bdi_writeback *wb, + struct wb_writeback_work *work) +{ + trace_writeback_queue(wb->bdi, work); + + spin_lock_bh(&wb->work_lock); + if (!test_bit(WB_registered, &wb->state)) { + if (work->single_wait) + work->single_done = 1; goto out_unlock; } - list_add_tail(&work->list, &bdi->work_list); - mod_delayed_work(bdi_wq, &bdi->wb.dwork, 0); + if (work->done) + atomic_inc(&work->done->cnt); + list_add_tail(&work->list, &wb->work_list); + mod_delayed_work(bdi_wq, &wb->dwork, 0); out_unlock: - spin_unlock_bh(&bdi->wb_lock); + spin_unlock_bh(&wb->work_lock); +} + +/** + * wb_wait_for_completion - wait for completion of bdi_writeback_works + * @bdi: bdi work items were issued to + * @done: target wb_completion + * + * Wait for one or more work items issued to @bdi with their ->done field + * set to @done, which should have been defined with + * DEFINE_WB_COMPLETION_ONSTACK(). This function returns after all such + * work items are completed. Work items which are waited upon aren't freed + * automatically on completion. + */ +static void wb_wait_for_completion(struct backing_dev_info *bdi, + struct wb_completion *done) +{ + atomic_dec(&done->cnt); /* put down the initial count */ + wait_event(bdi->wb_waitq, !atomic_read(&done->cnt)); +} + +#ifdef CONFIG_CGROUP_WRITEBACK + +/* parameters for foreign inode detection, see wb_detach_inode() */ +#define WB_FRN_TIME_SHIFT 13 /* 1s = 2^13, upto 8 secs w/ 16bit */ +#define WB_FRN_TIME_AVG_SHIFT 3 /* avg = avg * 7/8 + new * 1/8 */ +#define WB_FRN_TIME_CUT_DIV 2 /* ignore rounds < avg / 2 */ +#define WB_FRN_TIME_PERIOD (2 * (1 << WB_FRN_TIME_SHIFT)) /* 2s */ + +#define WB_FRN_HIST_SLOTS 16 /* inode->i_wb_frn_history is 16bit */ +#define WB_FRN_HIST_UNIT (WB_FRN_TIME_PERIOD / WB_FRN_HIST_SLOTS) + /* each slot's duration is 2s / 16 */ +#define WB_FRN_HIST_THR_SLOTS (WB_FRN_HIST_SLOTS / 2) + /* if foreign slots >= 8, switch */ +#define WB_FRN_HIST_MAX_SLOTS (WB_FRN_HIST_THR_SLOTS / 2 + 1) + /* one round can affect upto 5 slots */ + +void __inode_attach_wb(struct inode *inode, struct page *page) +{ + struct backing_dev_info *bdi = inode_to_bdi(inode); + struct bdi_writeback *wb = NULL; + + if (inode_cgwb_enabled(inode)) { + struct cgroup_subsys_state *memcg_css; + + if (page) { + memcg_css = mem_cgroup_css_from_page(page); + wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC); + } else { + /* must pin memcg_css, see wb_get_create() */ + memcg_css = task_get_css(current, memory_cgrp_id); + wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC); + css_put(memcg_css); + } + } + + if (!wb) + wb = &bdi->wb; + + /* + * There may be multiple instances of this function racing to + * update the same inode. Use cmpxchg() to tell the winner. + */ + if (unlikely(cmpxchg(&inode->i_wb, NULL, wb))) + wb_put(wb); +} + +/** + * locked_inode_to_wb_and_lock_list - determine a locked inode's wb and lock it + * @inode: inode of interest with i_lock held + * + * Returns @inode's wb with its list_lock held. @inode->i_lock must be + * held on entry and is released on return. The returned wb is guaranteed + * to stay @inode's associated wb until its list_lock is released. + */ +static struct bdi_writeback * +locked_inode_to_wb_and_lock_list(struct inode *inode) + __releases(&inode->i_lock) + __acquires(&wb->list_lock) +{ + while (true) { + struct bdi_writeback *wb = inode_to_wb(inode); + + /* + * inode_to_wb() association is protected by both + * @inode->i_lock and @wb->list_lock but list_lock nests + * outside i_lock. Drop i_lock and verify that the + * association hasn't changed after acquiring list_lock. + */ + wb_get(wb); + spin_unlock(&inode->i_lock); + spin_lock(&wb->list_lock); + wb_put(wb); /* not gonna deref it anymore */ + + /* i_wb may have changed inbetween, can't use inode_to_wb() */ + if (likely(wb == inode->i_wb)) + return wb; /* @inode already has ref */ + + spin_unlock(&wb->list_lock); + cpu_relax(); + spin_lock(&inode->i_lock); + } +} + +/** + * inode_to_wb_and_lock_list - determine an inode's wb and lock it + * @inode: inode of interest + * + * Same as locked_inode_to_wb_and_lock_list() but @inode->i_lock isn't held + * on entry. + */ +static struct bdi_writeback *inode_to_wb_and_lock_list(struct inode *inode) + __acquires(&wb->list_lock) +{ + spin_lock(&inode->i_lock); + return locked_inode_to_wb_and_lock_list(inode); +} + +struct inode_switch_wbs_context { + struct inode *inode; + struct bdi_writeback *new_wb; + + struct rcu_head rcu_head; + struct work_struct work; +}; + +static void inode_switch_wbs_work_fn(struct work_struct *work) +{ + struct inode_switch_wbs_context *isw = + container_of(work, struct inode_switch_wbs_context, work); + struct inode *inode = isw->inode; + struct address_space *mapping = inode->i_mapping; + struct bdi_writeback *old_wb = inode->i_wb; + struct bdi_writeback *new_wb = isw->new_wb; + struct radix_tree_iter iter; + bool switched = false; + void **slot; + + /* + * By the time control reaches here, RCU grace period has passed + * since I_WB_SWITCH assertion and all wb stat update transactions + * between unlocked_inode_to_wb_begin/end() are guaranteed to be + * synchronizing against mapping->tree_lock. + * + * Grabbing old_wb->list_lock, inode->i_lock and mapping->tree_lock + * gives us exclusion against all wb related operations on @inode + * including IO list manipulations and stat updates. + */ + if (old_wb < new_wb) { + spin_lock(&old_wb->list_lock); + spin_lock_nested(&new_wb->list_lock, SINGLE_DEPTH_NESTING); + } else { + spin_lock(&new_wb->list_lock); + spin_lock_nested(&old_wb->list_lock, SINGLE_DEPTH_NESTING); + } + spin_lock(&inode->i_lock); + spin_lock_irq(&mapping->tree_lock); + + /* + * Once I_FREEING is visible under i_lock, the eviction path owns + * the inode and we shouldn't modify ->i_wb_list. + */ + if (unlikely(inode->i_state & I_FREEING)) + goto skip_switch; + + /* + * Count and transfer stats. Note that PAGECACHE_TAG_DIRTY points + * to possibly dirty pages while PAGECACHE_TAG_WRITEBACK points to + * pages actually under underwriteback. + */ + radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, 0, + PAGECACHE_TAG_DIRTY) { + struct page *page = radix_tree_deref_slot_protected(slot, + &mapping->tree_lock); + if (likely(page) && PageDirty(page)) { + __dec_wb_stat(old_wb, WB_RECLAIMABLE); + __inc_wb_stat(new_wb, WB_RECLAIMABLE); + } + } + + radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter, 0, + PAGECACHE_TAG_WRITEBACK) { + struct page *page = radix_tree_deref_slot_protected(slot, + &mapping->tree_lock); + if (likely(page)) { + WARN_ON_ONCE(!PageWriteback(page)); + __dec_wb_stat(old_wb, WB_WRITEBACK); + __inc_wb_stat(new_wb, WB_WRITEBACK); + } + } + + wb_get(new_wb); + + /* + * Transfer to @new_wb's IO list if necessary. The specific list + * @inode was on is ignored and the inode is put on ->b_dirty which + * is always correct including from ->b_dirty_time. The transfer + * preserves @inode->dirtied_when ordering. + */ + if (!list_empty(&inode->i_wb_list)) { + struct inode *pos; + + inode_wb_list_del_locked(inode, old_wb); + inode->i_wb = new_wb; + list_for_each_entry(pos, &new_wb->b_dirty, i_wb_list) + if (time_after_eq(inode->dirtied_when, + pos->dirtied_when)) + break; + inode_wb_list_move_locked(inode, new_wb, pos->i_wb_list.prev); + } else { + inode->i_wb = new_wb; + } + + /* ->i_wb_frn updates may race wbc_detach_inode() but doesn't matter */ + inode->i_wb_frn_winner = 0; + inode->i_wb_frn_avg_time = 0; + inode->i_wb_frn_history = 0; + switched = true; +skip_switch: + /* + * Paired with load_acquire in unlocked_inode_to_wb_begin() and + * ensures that the new wb is visible if they see !I_WB_SWITCH. + */ + smp_store_release(&inode->i_state, inode->i_state & ~I_WB_SWITCH); + + spin_unlock_irq(&mapping->tree_lock); + spin_unlock(&inode->i_lock); + spin_unlock(&new_wb->list_lock); + spin_unlock(&old_wb->list_lock); + + if (switched) { + wb_wakeup(new_wb); + wb_put(old_wb); + } + wb_put(new_wb); + + iput(inode); + kfree(isw); +} + +static void inode_switch_wbs_rcu_fn(struct rcu_head *rcu_head) +{ + struct inode_switch_wbs_context *isw = container_of(rcu_head, + struct inode_switch_wbs_context, rcu_head); + + /* needs to grab bh-unsafe locks, bounce to work item */ + INIT_WORK(&isw->work, inode_switch_wbs_work_fn); + schedule_work(&isw->work); +} + +/** + * inode_switch_wbs - change the wb association of an inode + * @inode: target inode + * @new_wb_id: ID of the new wb + * + * Switch @inode's wb association to the wb identified by @new_wb_id. The + * switching is performed asynchronously and may fail silently. + */ +static void inode_switch_wbs(struct inode *inode, int new_wb_id) +{ + struct backing_dev_info *bdi = inode_to_bdi(inode); + struct cgroup_subsys_state *memcg_css; + struct inode_switch_wbs_context *isw; + + /* noop if seems to be already in progress */ + if (inode->i_state & I_WB_SWITCH) + return; + + isw = kzalloc(sizeof(*isw), GFP_ATOMIC); + if (!isw) + return; + + /* find and pin the new wb */ + rcu_read_lock(); + memcg_css = css_from_id(new_wb_id, &memory_cgrp_subsys); + if (memcg_css) + isw->new_wb = wb_get_create(bdi, memcg_css, GFP_ATOMIC); + rcu_read_unlock(); + if (!isw->new_wb) + goto out_free; + + /* while holding I_WB_SWITCH, no one else can update the association */ + spin_lock(&inode->i_lock); + if (inode->i_state & (I_WB_SWITCH | I_FREEING) || + inode_to_wb(inode) == isw->new_wb) { + spin_unlock(&inode->i_lock); + goto out_free; + } + inode->i_state |= I_WB_SWITCH; + spin_unlock(&inode->i_lock); + + ihold(inode); + isw->inode = inode; + + /* + * In addition to synchronizing among switchers, I_WB_SWITCH tells + * the RCU protected stat update paths to grab the mapping's + * tree_lock so that stat transfer can synchronize against them. + * Let's continue after I_WB_SWITCH is guaranteed to be visible. + */ + call_rcu(&isw->rcu_head, inode_switch_wbs_rcu_fn); + return; + +out_free: + if (isw->new_wb) + wb_put(isw->new_wb); + kfree(isw); +} + +/** + * wbc_attach_and_unlock_inode - associate wbc with target inode and unlock it + * @wbc: writeback_control of interest + * @inode: target inode + * + * @inode is locked and about to be written back under the control of @wbc. + * Record @inode's writeback context into @wbc and unlock the i_lock. On + * writeback completion, wbc_detach_inode() should be called. This is used + * to track the cgroup writeback context. + */ +void wbc_attach_and_unlock_inode(struct writeback_control *wbc, + struct inode *inode) +{ + if (!inode_cgwb_enabled(inode)) { + spin_unlock(&inode->i_lock); + return; + } + + wbc->wb = inode_to_wb(inode); + wbc->inode = inode; + + wbc->wb_id = wbc->wb->memcg_css->id; + wbc->wb_lcand_id = inode->i_wb_frn_winner; + wbc->wb_tcand_id = 0; + wbc->wb_bytes = 0; + wbc->wb_lcand_bytes = 0; + wbc->wb_tcand_bytes = 0; + + wb_get(wbc->wb); + spin_unlock(&inode->i_lock); + + /* + * A dying wb indicates that the memcg-blkcg mapping has changed + * and a new wb is already serving the memcg. Switch immediately. + */ + if (unlikely(wb_dying(wbc->wb))) + inode_switch_wbs(inode, wbc->wb_id); +} + +/** + * wbc_detach_inode - disassociate wbc from inode and perform foreign detection + * @wbc: writeback_control of the just finished writeback + * + * To be called after a writeback attempt of an inode finishes and undoes + * wbc_attach_and_unlock_inode(). Can be called under any context. + * + * As concurrent write sharing of an inode is expected to be very rare and + * memcg only tracks page ownership on first-use basis severely confining + * the usefulness of such sharing, cgroup writeback tracks ownership + * per-inode. While the support for concurrent write sharing of an inode + * is deemed unnecessary, an inode being written to by different cgroups at + * different points in time is a lot more common, and, more importantly, + * charging only by first-use can too readily lead to grossly incorrect + * behaviors (single foreign page can lead to gigabytes of writeback to be + * incorrectly attributed). + * + * To resolve this issue, cgroup writeback detects the majority dirtier of + * an inode and transfers the ownership to it. To avoid unnnecessary + * oscillation, the detection mechanism keeps track of history and gives + * out the switch verdict only if the foreign usage pattern is stable over + * a certain amount of time and/or writeback attempts. + * + * On each writeback attempt, @wbc tries to detect the majority writer + * using Boyer-Moore majority vote algorithm. In addition to the byte + * count from the majority voting, it also counts the bytes written for the + * current wb and the last round's winner wb (max of last round's current + * wb, the winner from two rounds ago, and the last round's majority + * candidate). Keeping track of the historical winner helps the algorithm + * to semi-reliably detect the most active writer even when it's not the + * absolute majority. + * + * Once the winner of the round is determined, whether the winner is + * foreign or not and how much IO time the round consumed is recorded in + * inode->i_wb_frn_history. If the amount of recorded foreign IO time is + * over a certain threshold, the switch verdict is given. + */ +void wbc_detach_inode(struct writeback_control *wbc) +{ + struct bdi_writeback *wb = wbc->wb; + struct inode *inode = wbc->inode; + unsigned long avg_time, max_bytes, max_time; + u16 history; + int max_id; + + if (!wb) + return; + + history = inode->i_wb_frn_history; + avg_time = inode->i_wb_frn_avg_time; + + /* pick the winner of this round */ + if (wbc->wb_bytes >= wbc->wb_lcand_bytes && + wbc->wb_bytes >= wbc->wb_tcand_bytes) { + max_id = wbc->wb_id; + max_bytes = wbc->wb_bytes; + } else if (wbc->wb_lcand_bytes >= wbc->wb_tcand_bytes) { + max_id = wbc->wb_lcand_id; + max_bytes = wbc->wb_lcand_bytes; + } else { + max_id = wbc->wb_tcand_id; + max_bytes = wbc->wb_tcand_bytes; + } + + /* + * Calculate the amount of IO time the winner consumed and fold it + * into the running average kept per inode. If the consumed IO + * time is lower than avag / WB_FRN_TIME_CUT_DIV, ignore it for + * deciding whether to switch or not. This is to prevent one-off + * small dirtiers from skewing the verdict. + */ + max_time = DIV_ROUND_UP((max_bytes >> PAGE_SHIFT) << WB_FRN_TIME_SHIFT, + wb->avg_write_bandwidth); + if (avg_time) + avg_time += (max_time >> WB_FRN_TIME_AVG_SHIFT) - + (avg_time >> WB_FRN_TIME_AVG_SHIFT); + else + avg_time = max_time; /* immediate catch up on first run */ + + if (max_time >= avg_time / WB_FRN_TIME_CUT_DIV) { + int slots; + + /* + * The switch verdict is reached if foreign wb's consume + * more than a certain proportion of IO time in a + * WB_FRN_TIME_PERIOD. This is loosely tracked by 16 slot + * history mask where each bit represents one sixteenth of + * the period. Determine the number of slots to shift into + * history from @max_time. + */ + slots = min(DIV_ROUND_UP(max_time, WB_FRN_HIST_UNIT), + (unsigned long)WB_FRN_HIST_MAX_SLOTS); + history <<= slots; + if (wbc->wb_id != max_id) + history |= (1U << slots) - 1; + + /* + * Switch if the current wb isn't the consistent winner. + * If there are multiple closely competing dirtiers, the + * inode may switch across them repeatedly over time, which + * is okay. The main goal is avoiding keeping an inode on + * the wrong wb for an extended period of time. + */ + if (hweight32(history) > WB_FRN_HIST_THR_SLOTS) + inode_switch_wbs(inode, max_id); + } + + /* + * Multiple instances of this function may race to update the + * following fields but we don't mind occassional inaccuracies. + */ + inode->i_wb_frn_winner = max_id; + inode->i_wb_frn_avg_time = min(avg_time, (unsigned long)U16_MAX); + inode->i_wb_frn_history = history; + + wb_put(wbc->wb); + wbc->wb = NULL; +} + +/** + * wbc_account_io - account IO issued during writeback + * @wbc: writeback_control of the writeback in progress + * @page: page being written out + * @bytes: number of bytes being written out + * + * @bytes from @page are about to written out during the writeback + * controlled by @wbc. Keep the book for foreign inode detection. See + * wbc_detach_inode(). + */ +void wbc_account_io(struct writeback_control *wbc, struct page *page, + size_t bytes) +{ + int id; + + /* + * pageout() path doesn't attach @wbc to the inode being written + * out. This is intentional as we don't want the function to block + * behind a slow cgroup. Ultimately, we want pageout() to kick off + * regular writeback instead of writing things out itself. + */ + if (!wbc->wb) + return; + + rcu_read_lock(); + id = mem_cgroup_css_from_page(page)->id; + rcu_read_unlock(); + + if (id == wbc->wb_id) { + wbc->wb_bytes += bytes; + return; + } + + if (id == wbc->wb_lcand_id) + wbc->wb_lcand_bytes += bytes; + + /* Boyer-Moore majority vote algorithm */ + if (!wbc->wb_tcand_bytes) + wbc->wb_tcand_id = id; + if (id == wbc->wb_tcand_id) + wbc->wb_tcand_bytes += bytes; + else + wbc->wb_tcand_bytes -= min(bytes, wbc->wb_tcand_bytes); } -static void -__bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages, - bool range_cyclic, enum wb_reason reason) +/** + * inode_congested - test whether an inode is congested + * @inode: inode to test for congestion + * @cong_bits: mask of WB_[a]sync_congested bits to test + * + * Tests whether @inode is congested. @cong_bits is the mask of congestion + * bits to test and the return value is the mask of set bits. + * + * If cgroup writeback is enabled for @inode, the congestion state is + * determined by whether the cgwb (cgroup bdi_writeback) for the blkcg + * associated with @inode is congested; otherwise, the root wb's congestion + * state is used. + */ +int inode_congested(struct inode *inode, int cong_bits) +{ + /* + * Once set, ->i_wb never becomes NULL while the inode is alive. + * Start transaction iff ->i_wb is visible. + */ + if (inode && inode_to_wb_is_valid(inode)) { + struct bdi_writeback *wb; + bool locked, congested; + + wb = unlocked_inode_to_wb_begin(inode, &locked); + congested = wb_congested(wb, cong_bits); + unlocked_inode_to_wb_end(inode, locked); + return congested; + } + + return wb_congested(&inode_to_bdi(inode)->wb, cong_bits); +} +EXPORT_SYMBOL_GPL(inode_congested); + +/** + * wb_wait_for_single_work - wait for completion of a single bdi_writeback_work + * @bdi: bdi the work item was issued to + * @work: work item to wait for + * + * Wait for the completion of @work which was issued to one of @bdi's + * bdi_writeback's. The caller must have set @work->single_wait before + * issuing it. This wait operates independently fo + * wb_wait_for_completion() and also disables automatic freeing of @work. + */ +static void wb_wait_for_single_work(struct backing_dev_info *bdi, + struct wb_writeback_work *work) +{ + if (WARN_ON_ONCE(!work->single_wait)) + return; + + wait_event(bdi->wb_waitq, work->single_done); + + /* + * Paired with smp_wmb() in wb_do_writeback() and ensures that all + * modifications to @work prior to assertion of ->single_done is + * visible to the caller once this function returns. + */ + smp_rmb(); +} + +/** + * wb_split_bdi_pages - split nr_pages to write according to bandwidth + * @wb: target bdi_writeback to split @nr_pages to + * @nr_pages: number of pages to write for the whole bdi + * + * Split @wb's portion of @nr_pages according to @wb's write bandwidth in + * relation to the total write bandwidth of all wb's w/ dirty inodes on + * @wb->bdi. + */ +static long wb_split_bdi_pages(struct bdi_writeback *wb, long nr_pages) +{ + unsigned long this_bw = wb->avg_write_bandwidth; + unsigned long tot_bw = atomic_long_read(&wb->bdi->tot_write_bandwidth); + + if (nr_pages == LONG_MAX) + return LONG_MAX; + + /* + * This may be called on clean wb's and proportional distribution + * may not make sense, just use the original @nr_pages in those + * cases. In general, we wanna err on the side of writing more. + */ + if (!tot_bw || this_bw >= tot_bw) + return nr_pages; + else + return DIV_ROUND_UP_ULL((u64)nr_pages * this_bw, tot_bw); +} + +/** + * wb_clone_and_queue_work - clone a wb_writeback_work and issue it to a wb + * @wb: target bdi_writeback + * @base_work: source wb_writeback_work + * + * Try to make a clone of @base_work and issue it to @wb. If cloning + * succeeds, %true is returned; otherwise, @base_work is issued directly + * and %false is returned. In the latter case, the caller is required to + * wait for @base_work's completion using wb_wait_for_single_work(). + * + * A clone is auto-freed on completion. @base_work never is. + */ +static bool wb_clone_and_queue_work(struct bdi_writeback *wb, + struct wb_writeback_work *base_work) { struct wb_writeback_work *work; + work = kmalloc(sizeof(*work), GFP_ATOMIC); + if (work) { + *work = *base_work; + work->auto_free = 1; + work->single_wait = 0; + } else { + work = base_work; + work->auto_free = 0; + work->single_wait = 1; + } + work->single_done = 0; + wb_queue_work(wb, work); + return work != base_work; +} + +/** + * bdi_split_work_to_wbs - split a wb_writeback_work to all wb's of a bdi + * @bdi: target backing_dev_info + * @base_work: wb_writeback_work to issue + * @skip_if_busy: skip wb's which already have writeback in progress + * + * Split and issue @base_work to all wb's (bdi_writeback's) of @bdi which + * have dirty inodes. If @base_work->nr_page isn't %LONG_MAX, it's + * distributed to the busy wbs according to each wb's proportion in the + * total active write bandwidth of @bdi. + */ +static void bdi_split_work_to_wbs(struct backing_dev_info *bdi, + struct wb_writeback_work *base_work, + bool skip_if_busy) +{ + long nr_pages = base_work->nr_pages; + int next_blkcg_id = 0; + struct bdi_writeback *wb; + struct wb_iter iter; + + might_sleep(); + + if (!bdi_has_dirty_io(bdi)) + return; +restart: + rcu_read_lock(); + bdi_for_each_wb(wb, bdi, &iter, next_blkcg_id) { + if (!wb_has_dirty_io(wb) || + (skip_if_busy && writeback_in_progress(wb))) + continue; + + base_work->nr_pages = wb_split_bdi_pages(wb, nr_pages); + if (!wb_clone_and_queue_work(wb, base_work)) { + next_blkcg_id = wb->blkcg_css->id + 1; + rcu_read_unlock(); + wb_wait_for_single_work(bdi, base_work); + goto restart; + } + } + rcu_read_unlock(); +} + +#else /* CONFIG_CGROUP_WRITEBACK */ + +static struct bdi_writeback * +locked_inode_to_wb_and_lock_list(struct inode *inode) + __releases(&inode->i_lock) + __acquires(&wb->list_lock) +{ + struct bdi_writeback *wb = inode_to_wb(inode); + + spin_unlock(&inode->i_lock); + spin_lock(&wb->list_lock); + return wb; +} + +static struct bdi_writeback *inode_to_wb_and_lock_list(struct inode *inode) + __acquires(&wb->list_lock) +{ + struct bdi_writeback *wb = inode_to_wb(inode); + + spin_lock(&wb->list_lock); + return wb; +} + +static long wb_split_bdi_pages(struct bdi_writeback *wb, long nr_pages) +{ + return nr_pages; +} + +static void bdi_split_work_to_wbs(struct backing_dev_info *bdi, + struct wb_writeback_work *base_work, + bool skip_if_busy) +{ + might_sleep(); + + if (bdi_has_dirty_io(bdi) && + (!skip_if_busy || !writeback_in_progress(&bdi->wb))) { + base_work->auto_free = 0; + base_work->single_wait = 0; + base_work->single_done = 0; + wb_queue_work(&bdi->wb, base_work); + } +} + +#endif /* CONFIG_CGROUP_WRITEBACK */ + +void wb_start_writeback(struct bdi_writeback *wb, long nr_pages, + bool range_cyclic, enum wb_reason reason) +{ + struct wb_writeback_work *work; + + if (!wb_has_dirty_io(wb)) + return; + /* * This is WB_SYNC_NONE writeback, so if allocation fails just * wakeup the thread for old dirty data writeback */ work = kzalloc(sizeof(*work), GFP_ATOMIC); if (!work) { - trace_writeback_nowork(bdi); - bdi_wakeup_thread(bdi); + trace_writeback_nowork(wb->bdi); + wb_wakeup(wb); return; } @@ -155,46 +932,29 @@ __bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages, work->nr_pages = nr_pages; work->range_cyclic = range_cyclic; work->reason = reason; + work->auto_free = 1; - bdi_queue_work(bdi, work); + wb_queue_work(wb, work); } /** - * bdi_start_writeback - start writeback - * @bdi: the backing device to write from - * @nr_pages: the number of pages to write - * @reason: reason why some writeback work was initiated - * - * Description: - * This does WB_SYNC_NONE opportunistic writeback. The IO is only - * started when this function returns, we make no guarantees on - * completion. Caller need not hold sb s_umount semaphore. - * - */ -void bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages, - enum wb_reason reason) -{ - __bdi_start_writeback(bdi, nr_pages, true, reason); -} - -/** - * bdi_start_background_writeback - start background writeback - * @bdi: the backing device to write from + * wb_start_background_writeback - start background writeback + * @wb: bdi_writback to write from * * Description: * This makes sure WB_SYNC_NONE background writeback happens. When - * this function returns, it is only guaranteed that for given BDI + * this function returns, it is only guaranteed that for given wb * some IO is happening if we are over background dirty threshold. * Caller need not hold sb s_umount semaphore. */ -void bdi_start_background_writeback(struct backing_dev_info *bdi) +void wb_start_background_writeback(struct bdi_writeback *wb) { /* * We just wake up the flusher thread. It will perform background * writeback as soon as there is no other work to do. */ - trace_writeback_wake_background(bdi); - bdi_wakeup_thread(bdi); + trace_writeback_wake_background(wb->bdi); + wb_wakeup(wb); } /* @@ -202,11 +962,11 @@ void bdi_start_background_writeback(struct backing_dev_info *bdi) */ void inode_wb_list_del(struct inode *inode) { - struct backing_dev_info *bdi = inode_to_bdi(inode); + struct bdi_writeback *wb; - spin_lock(&bdi->wb.list_lock); - list_del_init(&inode->i_wb_list); - spin_unlock(&bdi->wb.list_lock); + wb = inode_to_wb_and_lock_list(inode); + inode_wb_list_del_locked(inode, wb); + spin_unlock(&wb->list_lock); } /* @@ -220,7 +980,6 @@ void inode_wb_list_del(struct inode *inode) */ static void redirty_tail(struct inode *inode, struct bdi_writeback *wb) { - assert_spin_locked(&wb->list_lock); if (!list_empty(&wb->b_dirty)) { struct inode *tail; @@ -228,7 +987,7 @@ static void redirty_tail(struct inode *inode, struct bdi_writeback *wb) if (time_before(inode->dirtied_when, tail->dirtied_when)) inode->dirtied_when = jiffies; } - list_move(&inode->i_wb_list, &wb->b_dirty); + inode_wb_list_move_locked(inode, wb, &wb->b_dirty); } /* @@ -236,8 +995,7 @@ static void redirty_tail(struct inode *inode, struct bdi_writeback *wb) */ static void requeue_io(struct inode *inode, struct bdi_writeback *wb) { - assert_spin_locked(&wb->list_lock); - list_move(&inode->i_wb_list, &wb->b_more_io); + inode_wb_list_move_locked(inode, wb, &wb->b_more_io); } static void inode_sync_complete(struct inode *inode) @@ -346,6 +1104,8 @@ static void queue_io(struct bdi_writeback *wb, struct wb_writeback_work *work) moved = move_expired_inodes(&wb->b_dirty, &wb->b_io, 0, work); moved += move_expired_inodes(&wb->b_dirty_time, &wb->b_io, EXPIRE_DIRTY_ATIME, work); + if (moved) + wb_io_lists_populated(wb); trace_writeback_queue_io(wb, work, moved); } @@ -471,10 +1231,10 @@ static void requeue_inode(struct inode *inode, struct bdi_writeback *wb, redirty_tail(inode, wb); } else if (inode->i_state & I_DIRTY_TIME) { inode->dirtied_when = jiffies; - list_move(&inode->i_wb_list, &wb->b_dirty_time); + inode_wb_list_move_locked(inode, wb, &wb->b_dirty_time); } else { /* The inode is clean. Remove from writeback lists. */ - list_del_init(&inode->i_wb_list); + inode_wb_list_del_locked(inode, wb); } } @@ -605,10 +1365,11 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, !mapping_tagged(inode->i_mapping, PAGECACHE_TAG_WRITEBACK))) goto out; inode->i_state |= I_SYNC; - spin_unlock(&inode->i_lock); + wbc_attach_and_unlock_inode(wbc, inode); ret = __writeback_single_inode(inode, wbc); + wbc_detach_inode(wbc); spin_lock(&wb->list_lock); spin_lock(&inode->i_lock); /* @@ -616,7 +1377,7 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, * touch it. See comment above for explanation. */ if (!(inode->i_state & I_DIRTY_ALL)) - list_del_init(&inode->i_wb_list); + inode_wb_list_del_locked(inode, wb); spin_unlock(&wb->list_lock); inode_sync_complete(inode); out: @@ -624,7 +1385,7 @@ out: return ret; } -static long writeback_chunk_size(struct backing_dev_info *bdi, +static long writeback_chunk_size(struct bdi_writeback *wb, struct wb_writeback_work *work) { long pages; @@ -645,8 +1406,8 @@ static long writeback_chunk_size(struct backing_dev_info *bdi, if (work->sync_mode == WB_SYNC_ALL || work->tagged_writepages) pages = LONG_MAX; else { - pages = min(bdi->avg_write_bandwidth / 2, - global_dirty_limit / DIRTY_SCOPE); + pages = min(wb->avg_write_bandwidth / 2, + global_wb_domain.dirty_limit / DIRTY_SCOPE); pages = min(pages, work->nr_pages); pages = round_down(pages + MIN_WRITEBACK_PAGES, MIN_WRITEBACK_PAGES); @@ -741,9 +1502,9 @@ static long writeback_sb_inodes(struct super_block *sb, continue; } inode->i_state |= I_SYNC; - spin_unlock(&inode->i_lock); + wbc_attach_and_unlock_inode(&wbc, inode); - write_chunk = writeback_chunk_size(wb->bdi, work); + write_chunk = writeback_chunk_size(wb, work); wbc.nr_to_write = write_chunk; wbc.pages_skipped = 0; @@ -753,6 +1514,7 @@ static long writeback_sb_inodes(struct super_block *sb, */ __writeback_single_inode(inode, &wbc); + wbc_detach_inode(&wbc); work->nr_pages -= write_chunk - wbc.nr_to_write; wrote += write_chunk - wbc.nr_to_write; spin_lock(&wb->list_lock); @@ -830,33 +1592,6 @@ static long writeback_inodes_wb(struct bdi_writeback *wb, long nr_pages, return nr_pages - work.nr_pages; } -static bool over_bground_thresh(struct backing_dev_info *bdi) -{ - unsigned long background_thresh, dirty_thresh; - - global_dirty_limits(&background_thresh, &dirty_thresh); - - if (global_page_state(NR_FILE_DIRTY) + - global_page_state(NR_UNSTABLE_NFS) > background_thresh) - return true; - - if (bdi_stat(bdi, BDI_RECLAIMABLE) > - bdi_dirty_limit(bdi, background_thresh)) - return true; - - return false; -} - -/* - * Called under wb->list_lock. If there are multiple wb per bdi, - * only the flusher working on the first wb should do it. - */ -static void wb_update_bandwidth(struct bdi_writeback *wb, - unsigned long start_time) -{ - __bdi_update_bandwidth(wb->bdi, 0, 0, 0, 0, 0, start_time); -} - /* * Explicit flushing or periodic writeback of "old" data. * @@ -899,14 +1634,14 @@ static long wb_writeback(struct bdi_writeback *wb, * after the other works are all done. */ if ((work->for_background || work->for_kupdate) && - !list_empty(&wb->bdi->work_list)) + !list_empty(&wb->work_list)) break; /* * For background writeout, stop when we are below the * background dirty threshold */ - if (work->for_background && !over_bground_thresh(wb->bdi)) + if (work->for_background && !wb_over_bg_thresh(wb)) break; /* @@ -970,18 +1705,17 @@ static long wb_writeback(struct bdi_writeback *wb, /* * Return the next wb_writeback_work struct that hasn't been processed yet. */ -static struct wb_writeback_work * -get_next_work_item(struct backing_dev_info *bdi) +static struct wb_writeback_work *get_next_work_item(struct bdi_writeback *wb) { struct wb_writeback_work *work = NULL; - spin_lock_bh(&bdi->wb_lock); - if (!list_empty(&bdi->work_list)) { - work = list_entry(bdi->work_list.next, + spin_lock_bh(&wb->work_lock); + if (!list_empty(&wb->work_list)) { + work = list_entry(wb->work_list.next, struct wb_writeback_work, list); list_del_init(&work->list); } - spin_unlock_bh(&bdi->wb_lock); + spin_unlock_bh(&wb->work_lock); return work; } @@ -998,7 +1732,7 @@ static unsigned long get_nr_dirty_pages(void) static long wb_check_background_flush(struct bdi_writeback *wb) { - if (over_bground_thresh(wb->bdi)) { + if (wb_over_bg_thresh(wb)) { struct wb_writeback_work work = { .nr_pages = LONG_MAX, @@ -1053,25 +1787,33 @@ static long wb_check_old_data_flush(struct bdi_writeback *wb) */ static long wb_do_writeback(struct bdi_writeback *wb) { - struct backing_dev_info *bdi = wb->bdi; struct wb_writeback_work *work; long wrote = 0; - set_bit(BDI_writeback_running, &wb->bdi->state); - while ((work = get_next_work_item(bdi)) != NULL) { + set_bit(WB_writeback_running, &wb->state); + while ((work = get_next_work_item(wb)) != NULL) { + struct wb_completion *done = work->done; + bool need_wake_up = false; - trace_writeback_exec(bdi, work); + trace_writeback_exec(wb->bdi, work); wrote += wb_writeback(wb, work); - /* - * Notify the caller of completion if this is a synchronous - * work item, otherwise just free it. - */ - if (work->done) - complete(work->done); - else + if (work->single_wait) { + WARN_ON_ONCE(work->auto_free); + /* paired w/ rmb in wb_wait_for_single_work() */ + smp_wmb(); + work->single_done = 1; + need_wake_up = true; + } else if (work->auto_free) { kfree(work); + } + + if (done && atomic_dec_and_test(&done->cnt)) + need_wake_up = true; + + if (need_wake_up) + wake_up_all(&wb->bdi->wb_waitq); } /* @@ -1079,7 +1821,7 @@ static long wb_do_writeback(struct bdi_writeback *wb) */ wrote += wb_check_old_data_flush(wb); wrote += wb_check_background_flush(wb); - clear_bit(BDI_writeback_running, &wb->bdi->state); + clear_bit(WB_writeback_running, &wb->state); return wrote; } @@ -1088,43 +1830,42 @@ static long wb_do_writeback(struct bdi_writeback *wb) * Handle writeback of dirty data for the device backed by this bdi. Also * reschedules periodically and does kupdated style flushing. */ -void bdi_writeback_workfn(struct work_struct *work) +void wb_workfn(struct work_struct *work) { struct bdi_writeback *wb = container_of(to_delayed_work(work), struct bdi_writeback, dwork); - struct backing_dev_info *bdi = wb->bdi; long pages_written; - set_worker_desc("flush-%s", dev_name(bdi->dev)); + set_worker_desc("flush-%s", dev_name(wb->bdi->dev)); current->flags |= PF_SWAPWRITE; if (likely(!current_is_workqueue_rescuer() || - !test_bit(BDI_registered, &bdi->state))) { + !test_bit(WB_registered, &wb->state))) { /* - * The normal path. Keep writing back @bdi until its + * The normal path. Keep writing back @wb until its * work_list is empty. Note that this path is also taken - * if @bdi is shutting down even when we're running off the + * if @wb is shutting down even when we're running off the * rescuer as work_list needs to be drained. */ do { pages_written = wb_do_writeback(wb); trace_writeback_pages_written(pages_written); - } while (!list_empty(&bdi->work_list)); + } while (!list_empty(&wb->work_list)); } else { /* * bdi_wq can't get enough workers and we're running off * the emergency worker. Don't hog it. Hopefully, 1024 is * enough for efficient IO. */ - pages_written = writeback_inodes_wb(&bdi->wb, 1024, + pages_written = writeback_inodes_wb(wb, 1024, WB_REASON_FORKER_THREAD); trace_writeback_pages_written(pages_written); } - if (!list_empty(&bdi->work_list)) + if (!list_empty(&wb->work_list)) mod_delayed_work(bdi_wq, &wb->dwork, 0); else if (wb_has_dirty_io(wb) && dirty_writeback_interval) - bdi_wakeup_thread_delayed(bdi); + wb_wakeup_delayed(wb); current->flags &= ~PF_SWAPWRITE; } @@ -1142,9 +1883,15 @@ void wakeup_flusher_threads(long nr_pages, enum wb_reason reason) rcu_read_lock(); list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) { + struct bdi_writeback *wb; + struct wb_iter iter; + if (!bdi_has_dirty_io(bdi)) continue; - __bdi_start_writeback(bdi, nr_pages, false, reason); + + bdi_for_each_wb(wb, bdi, &iter, 0) + wb_start_writeback(wb, wb_split_bdi_pages(wb, nr_pages), + false, reason); } rcu_read_unlock(); } @@ -1173,9 +1920,12 @@ static void wakeup_dirtytime_writeback(struct work_struct *w) rcu_read_lock(); list_for_each_entry_rcu(bdi, &bdi_list, bdi_list) { - if (list_empty(&bdi->wb.b_dirty_time)) - continue; - bdi_wakeup_thread(bdi); + struct bdi_writeback *wb; + struct wb_iter iter; + + bdi_for_each_wb(wb, bdi, &iter, 0) + if (!list_empty(&bdi->wb.b_dirty_time)) + wb_wakeup(&bdi->wb); } rcu_read_unlock(); schedule_delayed_work(&dirtytime_work, dirtytime_expire_interval * HZ); @@ -1249,7 +1999,6 @@ static noinline void block_dump___mark_inode_dirty(struct inode *inode) void __mark_inode_dirty(struct inode *inode, int flags) { struct super_block *sb = inode->i_sb; - struct backing_dev_info *bdi = NULL; int dirtytime; trace_writeback_mark_inode_dirty(inode, flags); @@ -1289,6 +2038,8 @@ void __mark_inode_dirty(struct inode *inode, int flags) if ((inode->i_state & flags) != flags) { const int was_dirty = inode->i_state & I_DIRTY; + inode_attach_wb(inode, NULL); + if (flags & I_DIRTY_INODE) inode->i_state &= ~I_DIRTY_TIME; inode->i_state |= flags; @@ -1317,38 +2068,39 @@ void __mark_inode_dirty(struct inode *inode, int flags) * reposition it (that would break b_dirty time-ordering). */ if (!was_dirty) { + struct bdi_writeback *wb; + struct list_head *dirty_list; bool wakeup_bdi = false; - bdi = inode_to_bdi(inode); - spin_unlock(&inode->i_lock); - spin_lock(&bdi->wb.list_lock); - if (bdi_cap_writeback_dirty(bdi)) { - WARN(!test_bit(BDI_registered, &bdi->state), - "bdi-%s not registered\n", bdi->name); + wb = locked_inode_to_wb_and_lock_list(inode); - /* - * If this is the first dirty inode for this - * bdi, we have to wake-up the corresponding - * bdi thread to make sure background - * write-back happens later. - */ - if (!wb_has_dirty_io(&bdi->wb)) - wakeup_bdi = true; - } + WARN(bdi_cap_writeback_dirty(wb->bdi) && + !test_bit(WB_registered, &wb->state), + "bdi-%s not registered\n", wb->bdi->name); inode->dirtied_when = jiffies; if (dirtytime) inode->dirtied_time_when = jiffies; + if (inode->i_state & (I_DIRTY_INODE | I_DIRTY_PAGES)) - list_move(&inode->i_wb_list, &bdi->wb.b_dirty); + dirty_list = &wb->b_dirty; else - list_move(&inode->i_wb_list, - &bdi->wb.b_dirty_time); - spin_unlock(&bdi->wb.list_lock); + dirty_list = &wb->b_dirty_time; + + wakeup_bdi = inode_wb_list_move_locked(inode, wb, + dirty_list); + + spin_unlock(&wb->list_lock); trace_writeback_dirty_inode_enqueue(inode); - if (wakeup_bdi) - bdi_wakeup_thread_delayed(bdi); + /* + * If this is the first dirty inode for this bdi, + * we have to wake-up the corresponding bdi thread + * to make sure background write-back happens + * later. + */ + if (bdi_cap_writeback_dirty(wb->bdi) && wakeup_bdi) + wb_wakeup_delayed(wb); return; } } @@ -1411,6 +2163,28 @@ static void wait_sb_inodes(struct super_block *sb) iput(old_inode); } +static void __writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr, + enum wb_reason reason, bool skip_if_busy) +{ + DEFINE_WB_COMPLETION_ONSTACK(done); + struct wb_writeback_work work = { + .sb = sb, + .sync_mode = WB_SYNC_NONE, + .tagged_writepages = 1, + .done = &done, + .nr_pages = nr, + .reason = reason, + }; + struct backing_dev_info *bdi = sb->s_bdi; + + if (!bdi_has_dirty_io(bdi) || bdi == &noop_backing_dev_info) + return; + WARN_ON(!rwsem_is_locked(&sb->s_umount)); + + bdi_split_work_to_wbs(sb->s_bdi, &work, skip_if_busy); + wb_wait_for_completion(bdi, &done); +} + /** * writeback_inodes_sb_nr - writeback dirty inodes from given super_block * @sb: the superblock @@ -1425,21 +2199,7 @@ void writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr, enum wb_reason reason) { - DECLARE_COMPLETION_ONSTACK(done); - struct wb_writeback_work work = { - .sb = sb, - .sync_mode = WB_SYNC_NONE, - .tagged_writepages = 1, - .done = &done, - .nr_pages = nr, - .reason = reason, - }; - - if (sb->s_bdi == &noop_backing_dev_info) - return; - WARN_ON(!rwsem_is_locked(&sb->s_umount)); - bdi_queue_work(sb->s_bdi, &work); - wait_for_completion(&done); + __writeback_inodes_sb_nr(sb, nr, reason, false); } EXPORT_SYMBOL(writeback_inodes_sb_nr); @@ -1467,19 +2227,15 @@ EXPORT_SYMBOL(writeback_inodes_sb); * Invoke writeback_inodes_sb_nr if no writeback is currently underway. * Returns 1 if writeback was started, 0 if not. */ -int try_to_writeback_inodes_sb_nr(struct super_block *sb, - unsigned long nr, - enum wb_reason reason) +bool try_to_writeback_inodes_sb_nr(struct super_block *sb, unsigned long nr, + enum wb_reason reason) { - if (writeback_in_progress(sb->s_bdi)) - return 1; - if (!down_read_trylock(&sb->s_umount)) - return 0; + return false; - writeback_inodes_sb_nr(sb, nr, reason); + __writeback_inodes_sb_nr(sb, nr, reason, true); up_read(&sb->s_umount); - return 1; + return true; } EXPORT_SYMBOL(try_to_writeback_inodes_sb_nr); @@ -1491,7 +2247,7 @@ EXPORT_SYMBOL(try_to_writeback_inodes_sb_nr); * Implement by try_to_writeback_inodes_sb_nr() * Returns 1 if writeback was started, 0 if not. */ -int try_to_writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) +bool try_to_writeback_inodes_sb(struct super_block *sb, enum wb_reason reason) { return try_to_writeback_inodes_sb_nr(sb, get_nr_dirty_pages(), reason); } @@ -1506,7 +2262,7 @@ EXPORT_SYMBOL(try_to_writeback_inodes_sb); */ void sync_inodes_sb(struct super_block *sb) { - DECLARE_COMPLETION_ONSTACK(done); + DEFINE_WB_COMPLETION_ONSTACK(done); struct wb_writeback_work work = { .sb = sb, .sync_mode = WB_SYNC_ALL, @@ -1516,14 +2272,15 @@ void sync_inodes_sb(struct super_block *sb) .reason = WB_REASON_SYNC, .for_sync = 1, }; + struct backing_dev_info *bdi = sb->s_bdi; /* Nothing to do? */ - if (sb->s_bdi == &noop_backing_dev_info) + if (!bdi_has_dirty_io(bdi) || bdi == &noop_backing_dev_info) return; WARN_ON(!rwsem_is_locked(&sb->s_umount)); - bdi_queue_work(sb->s_bdi, &work); - wait_for_completion(&done); + bdi_split_work_to_wbs(bdi, &work, false); + wb_wait_for_completion(bdi, &done); wait_sb_inodes(sb); } diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c index e5bbf74..eae2c11 100644 --- a/fs/fuse/cuse.c +++ b/fs/fuse/cuse.c @@ -489,6 +489,7 @@ static void cuse_fc_release(struct fuse_conn *fc) */ static int cuse_channel_open(struct inode *inode, struct file *file) { + struct fuse_dev *fud; struct cuse_conn *cc; int rc; @@ -499,17 +500,22 @@ static int cuse_channel_open(struct inode *inode, struct file *file) fuse_conn_init(&cc->fc); + fud = fuse_dev_alloc(&cc->fc); + if (!fud) { + kfree(cc); + return -ENOMEM; + } + INIT_LIST_HEAD(&cc->list); cc->fc.release = cuse_fc_release; - cc->fc.connected = 1; cc->fc.initialized = 1; rc = cuse_send_init(cc); if (rc) { - fuse_conn_put(&cc->fc); + fuse_dev_free(fud); return rc; } - file->private_data = &cc->fc; /* channel owns base reference to cc */ + file->private_data = fud; return 0; } @@ -527,7 +533,8 @@ static int cuse_channel_open(struct inode *inode, struct file *file) */ static int cuse_channel_release(struct inode *inode, struct file *file) { - struct cuse_conn *cc = fc_to_cc(file->private_data); + struct fuse_dev *fud = file->private_data; + struct cuse_conn *cc = fc_to_cc(fud->fc); int rc; /* remove from the conntbl, no more access from this point on */ diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index c8b68ab..80cc1b3 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -25,13 +25,13 @@ MODULE_ALIAS("devname:fuse"); static struct kmem_cache *fuse_req_cachep; -static struct fuse_conn *fuse_get_conn(struct file *file) +static struct fuse_dev *fuse_get_dev(struct file *file) { /* * Lockless access is OK, because file->private data is set * once during mount and is valid until the file is released. */ - return file->private_data; + return ACCESS_ONCE(file->private_data); } static void fuse_request_init(struct fuse_req *req, struct page **pages, @@ -48,6 +48,7 @@ static void fuse_request_init(struct fuse_req *req, struct page **pages, req->pages = pages; req->page_descs = page_descs; req->max_pages = npages; + __set_bit(FR_PENDING, &req->flags); } static struct fuse_req *__fuse_request_alloc(unsigned npages, gfp_t flags) @@ -168,6 +169,10 @@ static struct fuse_req *__fuse_get_req(struct fuse_conn *fc, unsigned npages, if (!fc->connected) goto out; + err = -ECONNREFUSED; + if (fc->conn_error) + goto out; + req = fuse_request_alloc(npages); err = -ENOMEM; if (!req) { @@ -177,8 +182,10 @@ static struct fuse_req *__fuse_get_req(struct fuse_conn *fc, unsigned npages, } fuse_req_init_context(req); - req->waiting = 1; - req->background = for_background; + __set_bit(FR_WAITING, &req->flags); + if (for_background) + __set_bit(FR_BACKGROUND, &req->flags); + return req; out: @@ -268,15 +275,15 @@ struct fuse_req *fuse_get_req_nofail_nopages(struct fuse_conn *fc, req = get_reserved_req(fc, file); fuse_req_init_context(req); - req->waiting = 1; - req->background = 0; + __set_bit(FR_WAITING, &req->flags); + __clear_bit(FR_BACKGROUND, &req->flags); return req; } void fuse_put_request(struct fuse_conn *fc, struct fuse_req *req) { if (atomic_dec_and_test(&req->count)) { - if (unlikely(req->background)) { + if (test_bit(FR_BACKGROUND, &req->flags)) { /* * We get here in the unlikely case that a background * request was allocated but not sent @@ -287,8 +294,10 @@ void fuse_put_request(struct fuse_conn *fc, struct fuse_req *req) spin_unlock(&fc->lock); } - if (req->waiting) + if (test_bit(FR_WAITING, &req->flags)) { + __clear_bit(FR_WAITING, &req->flags); atomic_dec(&fc->num_waiting); + } if (req->stolen_file) put_reserved_req(fc, req); @@ -309,46 +318,38 @@ static unsigned len_args(unsigned numargs, struct fuse_arg *args) return nbytes; } -static u64 fuse_get_unique(struct fuse_conn *fc) +static u64 fuse_get_unique(struct fuse_iqueue *fiq) { - fc->reqctr++; - /* zero is special */ - if (fc->reqctr == 0) - fc->reqctr = 1; - - return fc->reqctr; + return ++fiq->reqctr; } -static void queue_request(struct fuse_conn *fc, struct fuse_req *req) +static void queue_request(struct fuse_iqueue *fiq, struct fuse_req *req) { req->in.h.len = sizeof(struct fuse_in_header) + len_args(req->in.numargs, (struct fuse_arg *) req->in.args); - list_add_tail(&req->list, &fc->pending); - req->state = FUSE_REQ_PENDING; - if (!req->waiting) { - req->waiting = 1; - atomic_inc(&fc->num_waiting); - } - wake_up(&fc->waitq); - kill_fasync(&fc->fasync, SIGIO, POLL_IN); + list_add_tail(&req->list, &fiq->pending); + wake_up_locked(&fiq->waitq); + kill_fasync(&fiq->fasync, SIGIO, POLL_IN); } void fuse_queue_forget(struct fuse_conn *fc, struct fuse_forget_link *forget, u64 nodeid, u64 nlookup) { + struct fuse_iqueue *fiq = &fc->iq; + forget->forget_one.nodeid = nodeid; forget->forget_one.nlookup = nlookup; - spin_lock(&fc->lock); - if (fc->connected) { - fc->forget_list_tail->next = forget; - fc->forget_list_tail = forget; - wake_up(&fc->waitq); - kill_fasync(&fc->fasync, SIGIO, POLL_IN); + spin_lock(&fiq->waitq.lock); + if (fiq->connected) { + fiq->forget_list_tail->next = forget; + fiq->forget_list_tail = forget; + wake_up_locked(&fiq->waitq); + kill_fasync(&fiq->fasync, SIGIO, POLL_IN); } else { kfree(forget); } - spin_unlock(&fc->lock); + spin_unlock(&fiq->waitq.lock); } static void flush_bg_queue(struct fuse_conn *fc) @@ -356,12 +357,15 @@ static void flush_bg_queue(struct fuse_conn *fc) while (fc->active_background < fc->max_background && !list_empty(&fc->bg_queue)) { struct fuse_req *req; + struct fuse_iqueue *fiq = &fc->iq; req = list_entry(fc->bg_queue.next, struct fuse_req, list); list_del(&req->list); fc->active_background++; - req->in.h.unique = fuse_get_unique(fc); - queue_request(fc, req); + spin_lock(&fiq->waitq.lock); + req->in.h.unique = fuse_get_unique(fiq); + queue_request(fiq, req); + spin_unlock(&fiq->waitq.lock); } } @@ -372,20 +376,22 @@ static void flush_bg_queue(struct fuse_conn *fc) * was closed. The requester thread is woken up (if still waiting), * the 'end' callback is called if given, else the reference to the * request is released - * - * Called with fc->lock, unlocks it */ static void request_end(struct fuse_conn *fc, struct fuse_req *req) -__releases(fc->lock) { - void (*end) (struct fuse_conn *, struct fuse_req *) = req->end; - req->end = NULL; - list_del(&req->list); - list_del(&req->intr_entry); - req->state = FUSE_REQ_FINISHED; - if (req->background) { - req->background = 0; + struct fuse_iqueue *fiq = &fc->iq; + + if (test_and_set_bit(FR_FINISHED, &req->flags)) + return; + spin_lock(&fiq->waitq.lock); + list_del_init(&req->intr_entry); + spin_unlock(&fiq->waitq.lock); + WARN_ON(test_bit(FR_PENDING, &req->flags)); + WARN_ON(test_bit(FR_SENT, &req->flags)); + if (test_bit(FR_BACKGROUND, &req->flags)) { + spin_lock(&fc->lock); + clear_bit(FR_BACKGROUND, &req->flags); if (fc->num_background == fc->max_background) fc->blocked = 0; @@ -401,122 +407,105 @@ __releases(fc->lock) fc->num_background--; fc->active_background--; flush_bg_queue(fc); + spin_unlock(&fc->lock); } - spin_unlock(&fc->lock); wake_up(&req->waitq); - if (end) - end(fc, req); + if (req->end) + req->end(fc, req); fuse_put_request(fc, req); } -static void wait_answer_interruptible(struct fuse_conn *fc, - struct fuse_req *req) -__releases(fc->lock) -__acquires(fc->lock) -{ - if (signal_pending(current)) - return; - - spin_unlock(&fc->lock); - wait_event_interruptible(req->waitq, req->state == FUSE_REQ_FINISHED); - spin_lock(&fc->lock); -} - -static void queue_interrupt(struct fuse_conn *fc, struct fuse_req *req) +static void queue_interrupt(struct fuse_iqueue *fiq, struct fuse_req *req) { - list_add_tail(&req->intr_entry, &fc->interrupts); - wake_up(&fc->waitq); - kill_fasync(&fc->fasync, SIGIO, POLL_IN); + spin_lock(&fiq->waitq.lock); + if (list_empty(&req->intr_entry)) { + list_add_tail(&req->intr_entry, &fiq->interrupts); + wake_up_locked(&fiq->waitq); + } + spin_unlock(&fiq->waitq.lock); + kill_fasync(&fiq->fasync, SIGIO, POLL_IN); } static void request_wait_answer(struct fuse_conn *fc, struct fuse_req *req) -__releases(fc->lock) -__acquires(fc->lock) { + struct fuse_iqueue *fiq = &fc->iq; + int err; + if (!fc->no_interrupt) { /* Any signal may interrupt this */ - wait_answer_interruptible(fc, req); - - if (req->aborted) - goto aborted; - if (req->state == FUSE_REQ_FINISHED) + err = wait_event_interruptible(req->waitq, + test_bit(FR_FINISHED, &req->flags)); + if (!err) return; - req->interrupted = 1; - if (req->state == FUSE_REQ_SENT) - queue_interrupt(fc, req); + set_bit(FR_INTERRUPTED, &req->flags); + /* matches barrier in fuse_dev_do_read() */ + smp_mb__after_atomic(); + if (test_bit(FR_SENT, &req->flags)) + queue_interrupt(fiq, req); } - if (!req->force) { + if (!test_bit(FR_FORCE, &req->flags)) { sigset_t oldset; /* Only fatal signals may interrupt this */ block_sigs(&oldset); - wait_answer_interruptible(fc, req); + err = wait_event_interruptible(req->waitq, + test_bit(FR_FINISHED, &req->flags)); restore_sigs(&oldset); - if (req->aborted) - goto aborted; - if (req->state == FUSE_REQ_FINISHED) + if (!err) return; + spin_lock(&fiq->waitq.lock); /* Request is not yet in userspace, bail out */ - if (req->state == FUSE_REQ_PENDING) { + if (test_bit(FR_PENDING, &req->flags)) { list_del(&req->list); + spin_unlock(&fiq->waitq.lock); __fuse_put_request(req); req->out.h.error = -EINTR; return; } + spin_unlock(&fiq->waitq.lock); } /* * Either request is already in userspace, or it was forced. * Wait it out. */ - spin_unlock(&fc->lock); - wait_event(req->waitq, req->state == FUSE_REQ_FINISHED); - spin_lock(&fc->lock); - - if (!req->aborted) - return; - - aborted: - BUG_ON(req->state != FUSE_REQ_FINISHED); - if (req->locked) { - /* This is uninterruptible sleep, because data is - being copied to/from the buffers of req. During - locked state, there mustn't be any filesystem - operation (e.g. page fault), since that could lead - to deadlock */ - spin_unlock(&fc->lock); - wait_event(req->waitq, !req->locked); - spin_lock(&fc->lock); - } + wait_event(req->waitq, test_bit(FR_FINISHED, &req->flags)); } static void __fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) { - BUG_ON(req->background); - spin_lock(&fc->lock); - if (!fc->connected) + struct fuse_iqueue *fiq = &fc->iq; + + BUG_ON(test_bit(FR_BACKGROUND, &req->flags)); + spin_lock(&fiq->waitq.lock); + if (!fiq->connected) { + spin_unlock(&fiq->waitq.lock); req->out.h.error = -ENOTCONN; - else if (fc->conn_error) - req->out.h.error = -ECONNREFUSED; - else { - req->in.h.unique = fuse_get_unique(fc); - queue_request(fc, req); + } else { + req->in.h.unique = fuse_get_unique(fiq); + queue_request(fiq, req); /* acquire extra reference, since request is still needed after request_end() */ __fuse_get_request(req); + spin_unlock(&fiq->waitq.lock); request_wait_answer(fc, req); + /* Pairs with smp_wmb() in request_end() */ + smp_rmb(); } - spin_unlock(&fc->lock); } void fuse_request_send(struct fuse_conn *fc, struct fuse_req *req) { - req->isreply = 1; + __set_bit(FR_ISREPLY, &req->flags); + if (!test_bit(FR_WAITING, &req->flags)) { + __set_bit(FR_WAITING, &req->flags); + atomic_inc(&fc->num_waiting); + } __fuse_request_send(fc, req); } EXPORT_SYMBOL_GPL(fuse_request_send); @@ -586,10 +575,20 @@ ssize_t fuse_simple_request(struct fuse_conn *fc, struct fuse_args *args) return ret; } -static void fuse_request_send_nowait_locked(struct fuse_conn *fc, - struct fuse_req *req) +/* + * Called under fc->lock + * + * fc->connected must have been checked previously + */ +void fuse_request_send_background_locked(struct fuse_conn *fc, + struct fuse_req *req) { - BUG_ON(!req->background); + BUG_ON(!test_bit(FR_BACKGROUND, &req->flags)); + if (!test_bit(FR_WAITING, &req->flags)) { + __set_bit(FR_WAITING, &req->flags); + atomic_inc(&fc->num_waiting); + } + __set_bit(FR_ISREPLY, &req->flags); fc->num_background++; if (fc->num_background == fc->max_background) fc->blocked = 1; @@ -602,54 +601,40 @@ static void fuse_request_send_nowait_locked(struct fuse_conn *fc, flush_bg_queue(fc); } -static void fuse_request_send_nowait(struct fuse_conn *fc, struct fuse_req *req) +void fuse_request_send_background(struct fuse_conn *fc, struct fuse_req *req) { + BUG_ON(!req->end); spin_lock(&fc->lock); if (fc->connected) { - fuse_request_send_nowait_locked(fc, req); + fuse_request_send_background_locked(fc, req); spin_unlock(&fc->lock); } else { + spin_unlock(&fc->lock); req->out.h.error = -ENOTCONN; - request_end(fc, req); + req->end(fc, req); + fuse_put_request(fc, req); } } - -void fuse_request_send_background(struct fuse_conn *fc, struct fuse_req *req) -{ - req->isreply = 1; - fuse_request_send_nowait(fc, req); -} EXPORT_SYMBOL_GPL(fuse_request_send_background); static int fuse_request_send_notify_reply(struct fuse_conn *fc, struct fuse_req *req, u64 unique) { int err = -ENODEV; + struct fuse_iqueue *fiq = &fc->iq; - req->isreply = 0; + __clear_bit(FR_ISREPLY, &req->flags); req->in.h.unique = unique; - spin_lock(&fc->lock); - if (fc->connected) { - queue_request(fc, req); + spin_lock(&fiq->waitq.lock); + if (fiq->connected) { + queue_request(fiq, req); err = 0; } - spin_unlock(&fc->lock); + spin_unlock(&fiq->waitq.lock); return err; } -/* - * Called under fc->lock - * - * fc->connected must have been checked previously - */ -void fuse_request_send_background_locked(struct fuse_conn *fc, - struct fuse_req *req) -{ - req->isreply = 1; - fuse_request_send_nowait_locked(fc, req); -} - void fuse_force_forget(struct file *file, u64 nodeid) { struct inode *inode = file_inode(file); @@ -665,7 +650,7 @@ void fuse_force_forget(struct file *file, u64 nodeid) req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; - req->isreply = 0; + __clear_bit(FR_ISREPLY, &req->flags); __fuse_request_send(fc, req); /* ignore errors */ fuse_put_request(fc, req); @@ -676,38 +661,39 @@ void fuse_force_forget(struct file *file, u64 nodeid) * anything that could cause a page-fault. If the request was already * aborted bail out. */ -static int lock_request(struct fuse_conn *fc, struct fuse_req *req) +static int lock_request(struct fuse_req *req) { int err = 0; if (req) { - spin_lock(&fc->lock); - if (req->aborted) + spin_lock(&req->waitq.lock); + if (test_bit(FR_ABORTED, &req->flags)) err = -ENOENT; else - req->locked = 1; - spin_unlock(&fc->lock); + set_bit(FR_LOCKED, &req->flags); + spin_unlock(&req->waitq.lock); } return err; } /* - * Unlock request. If it was aborted during being locked, the - * requester thread is currently waiting for it to be unlocked, so - * wake it up. + * Unlock request. If it was aborted while locked, caller is responsible + * for unlocking and ending the request. */ -static void unlock_request(struct fuse_conn *fc, struct fuse_req *req) +static int unlock_request(struct fuse_req *req) { + int err = 0; if (req) { - spin_lock(&fc->lock); - req->locked = 0; - if (req->aborted) - wake_up(&req->waitq); - spin_unlock(&fc->lock); + spin_lock(&req->waitq.lock); + if (test_bit(FR_ABORTED, &req->flags)) + err = -ENOENT; + else + clear_bit(FR_LOCKED, &req->flags); + spin_unlock(&req->waitq.lock); } + return err; } struct fuse_copy_state { - struct fuse_conn *fc; int write; struct fuse_req *req; struct iov_iter *iter; @@ -721,13 +707,10 @@ struct fuse_copy_state { unsigned move_pages:1; }; -static void fuse_copy_init(struct fuse_copy_state *cs, - struct fuse_conn *fc, - int write, +static void fuse_copy_init(struct fuse_copy_state *cs, int write, struct iov_iter *iter) { memset(cs, 0, sizeof(*cs)); - cs->fc = fc; cs->write = write; cs->iter = iter; } @@ -760,7 +743,10 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) struct page *page; int err; - unlock_request(cs->fc, cs->req); + err = unlock_request(cs->req); + if (err) + return err; + fuse_copy_finish(cs); if (cs->pipebufs) { struct pipe_buffer *buf = cs->pipebufs; @@ -809,7 +795,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) iov_iter_advance(cs->iter, err); } - return lock_request(cs->fc, cs->req); + return lock_request(cs->req); } /* Do as much copy to/from userspace buffer as we can */ @@ -860,7 +846,10 @@ static int fuse_try_move_page(struct fuse_copy_state *cs, struct page **pagep) struct page *newpage; struct pipe_buffer *buf = cs->pipebufs; - unlock_request(cs->fc, cs->req); + err = unlock_request(cs->req); + if (err) + return err; + fuse_copy_finish(cs); err = buf->ops->confirm(cs->pipe, buf); @@ -914,12 +903,12 @@ static int fuse_try_move_page(struct fuse_copy_state *cs, struct page **pagep) lru_cache_add_file(newpage); err = 0; - spin_lock(&cs->fc->lock); - if (cs->req->aborted) + spin_lock(&cs->req->waitq.lock); + if (test_bit(FR_ABORTED, &cs->req->flags)) err = -ENOENT; else *pagep = newpage; - spin_unlock(&cs->fc->lock); + spin_unlock(&cs->req->waitq.lock); if (err) { unlock_page(newpage); @@ -939,7 +928,7 @@ out_fallback: cs->pg = buf->page; cs->offset = buf->offset; - err = lock_request(cs->fc, cs->req); + err = lock_request(cs->req); if (err) return err; @@ -950,11 +939,15 @@ static int fuse_ref_page(struct fuse_copy_state *cs, struct page *page, unsigned offset, unsigned count) { struct pipe_buffer *buf; + int err; if (cs->nr_segs == cs->pipe->buffers) return -EIO; - unlock_request(cs->fc, cs->req); + err = unlock_request(cs->req); + if (err) + return err; + fuse_copy_finish(cs); buf = cs->pipebufs; @@ -1065,36 +1058,15 @@ static int fuse_copy_args(struct fuse_copy_state *cs, unsigned numargs, return err; } -static int forget_pending(struct fuse_conn *fc) +static int forget_pending(struct fuse_iqueue *fiq) { - return fc->forget_list_head.next != NULL; + return fiq->forget_list_head.next != NULL; } -static int request_pending(struct fuse_conn *fc) +static int request_pending(struct fuse_iqueue *fiq) { - return !list_empty(&fc->pending) || !list_empty(&fc->interrupts) || - forget_pending(fc); -} - -/* Wait until a request is available on the pending list */ -static void request_wait(struct fuse_conn *fc) -__releases(fc->lock) -__acquires(fc->lock) -{ - DECLARE_WAITQUEUE(wait, current); - - add_wait_queue_exclusive(&fc->waitq, &wait); - while (fc->connected && !request_pending(fc)) { - set_current_state(TASK_INTERRUPTIBLE); - if (signal_pending(current)) - break; - - spin_unlock(&fc->lock); - schedule(); - spin_lock(&fc->lock); - } - set_current_state(TASK_RUNNING); - remove_wait_queue(&fc->waitq, &wait); + return !list_empty(&fiq->pending) || !list_empty(&fiq->interrupts) || + forget_pending(fiq); } /* @@ -1103,11 +1075,12 @@ __acquires(fc->lock) * Unlike other requests this is assembled on demand, without a need * to allocate a separate fuse_req structure. * - * Called with fc->lock held, releases it + * Called with fiq->waitq.lock held, releases it */ -static int fuse_read_interrupt(struct fuse_conn *fc, struct fuse_copy_state *cs, +static int fuse_read_interrupt(struct fuse_iqueue *fiq, + struct fuse_copy_state *cs, size_t nbytes, struct fuse_req *req) -__releases(fc->lock) +__releases(fiq->waitq.lock) { struct fuse_in_header ih; struct fuse_interrupt_in arg; @@ -1115,7 +1088,7 @@ __releases(fc->lock) int err; list_del_init(&req->intr_entry); - req->intr_unique = fuse_get_unique(fc); + req->intr_unique = fuse_get_unique(fiq); memset(&ih, 0, sizeof(ih)); memset(&arg, 0, sizeof(arg)); ih.len = reqsize; @@ -1123,7 +1096,7 @@ __releases(fc->lock) ih.unique = req->intr_unique; arg.unique = req->in.h.unique; - spin_unlock(&fc->lock); + spin_unlock(&fiq->waitq.lock); if (nbytes < reqsize) return -EINVAL; @@ -1135,21 +1108,21 @@ __releases(fc->lock) return err ? err : reqsize; } -static struct fuse_forget_link *dequeue_forget(struct fuse_conn *fc, +static struct fuse_forget_link *dequeue_forget(struct fuse_iqueue *fiq, unsigned max, unsigned *countp) { - struct fuse_forget_link *head = fc->forget_list_head.next; + struct fuse_forget_link *head = fiq->forget_list_head.next; struct fuse_forget_link **newhead = &head; unsigned count; for (count = 0; *newhead != NULL && count < max; count++) newhead = &(*newhead)->next; - fc->forget_list_head.next = *newhead; + fiq->forget_list_head.next = *newhead; *newhead = NULL; - if (fc->forget_list_head.next == NULL) - fc->forget_list_tail = &fc->forget_list_head; + if (fiq->forget_list_head.next == NULL) + fiq->forget_list_tail = &fiq->forget_list_head; if (countp != NULL) *countp = count; @@ -1157,24 +1130,24 @@ static struct fuse_forget_link *dequeue_forget(struct fuse_conn *fc, return head; } -static int fuse_read_single_forget(struct fuse_conn *fc, +static int fuse_read_single_forget(struct fuse_iqueue *fiq, struct fuse_copy_state *cs, size_t nbytes) -__releases(fc->lock) +__releases(fiq->waitq.lock) { int err; - struct fuse_forget_link *forget = dequeue_forget(fc, 1, NULL); + struct fuse_forget_link *forget = dequeue_forget(fiq, 1, NULL); struct fuse_forget_in arg = { .nlookup = forget->forget_one.nlookup, }; struct fuse_in_header ih = { .opcode = FUSE_FORGET, .nodeid = forget->forget_one.nodeid, - .unique = fuse_get_unique(fc), + .unique = fuse_get_unique(fiq), .len = sizeof(ih) + sizeof(arg), }; - spin_unlock(&fc->lock); + spin_unlock(&fiq->waitq.lock); kfree(forget); if (nbytes < ih.len) return -EINVAL; @@ -1190,9 +1163,9 @@ __releases(fc->lock) return ih.len; } -static int fuse_read_batch_forget(struct fuse_conn *fc, +static int fuse_read_batch_forget(struct fuse_iqueue *fiq, struct fuse_copy_state *cs, size_t nbytes) -__releases(fc->lock) +__releases(fiq->waitq.lock) { int err; unsigned max_forgets; @@ -1201,18 +1174,18 @@ __releases(fc->lock) struct fuse_batch_forget_in arg = { .count = 0 }; struct fuse_in_header ih = { .opcode = FUSE_BATCH_FORGET, - .unique = fuse_get_unique(fc), + .unique = fuse_get_unique(fiq), .len = sizeof(ih) + sizeof(arg), }; if (nbytes < ih.len) { - spin_unlock(&fc->lock); + spin_unlock(&fiq->waitq.lock); return -EINVAL; } max_forgets = (nbytes - ih.len) / sizeof(struct fuse_forget_one); - head = dequeue_forget(fc, max_forgets, &count); - spin_unlock(&fc->lock); + head = dequeue_forget(fiq, max_forgets, &count); + spin_unlock(&fiq->waitq.lock); arg.count = count; ih.len += count * sizeof(struct fuse_forget_one); @@ -1239,14 +1212,15 @@ __releases(fc->lock) return ih.len; } -static int fuse_read_forget(struct fuse_conn *fc, struct fuse_copy_state *cs, +static int fuse_read_forget(struct fuse_conn *fc, struct fuse_iqueue *fiq, + struct fuse_copy_state *cs, size_t nbytes) -__releases(fc->lock) +__releases(fiq->waitq.lock) { - if (fc->minor < 16 || fc->forget_list_head.next->next == NULL) - return fuse_read_single_forget(fc, cs, nbytes); + if (fc->minor < 16 || fiq->forget_list_head.next->next == NULL) + return fuse_read_single_forget(fiq, cs, nbytes); else - return fuse_read_batch_forget(fc, cs, nbytes); + return fuse_read_batch_forget(fiq, cs, nbytes); } /* @@ -1258,46 +1232,51 @@ __releases(fc->lock) * request_end(). Otherwise add it to the processing list, and set * the 'sent' flag. */ -static ssize_t fuse_dev_do_read(struct fuse_conn *fc, struct file *file, +static ssize_t fuse_dev_do_read(struct fuse_dev *fud, struct file *file, struct fuse_copy_state *cs, size_t nbytes) { - int err; + ssize_t err; + struct fuse_conn *fc = fud->fc; + struct fuse_iqueue *fiq = &fc->iq; + struct fuse_pqueue *fpq = &fud->pq; struct fuse_req *req; struct fuse_in *in; unsigned reqsize; restart: - spin_lock(&fc->lock); + spin_lock(&fiq->waitq.lock); err = -EAGAIN; - if ((file->f_flags & O_NONBLOCK) && fc->connected && - !request_pending(fc)) + if ((file->f_flags & O_NONBLOCK) && fiq->connected && + !request_pending(fiq)) goto err_unlock; - request_wait(fc); - err = -ENODEV; - if (!fc->connected) + err = wait_event_interruptible_exclusive_locked(fiq->waitq, + !fiq->connected || request_pending(fiq)); + if (err) goto err_unlock; - err = -ERESTARTSYS; - if (!request_pending(fc)) + + err = -ENODEV; + if (!fiq->connected) goto err_unlock; - if (!list_empty(&fc->interrupts)) { - req = list_entry(fc->interrupts.next, struct fuse_req, + if (!list_empty(&fiq->interrupts)) { + req = list_entry(fiq->interrupts.next, struct fuse_req, intr_entry); - return fuse_read_interrupt(fc, cs, nbytes, req); + return fuse_read_interrupt(fiq, cs, nbytes, req); } - if (forget_pending(fc)) { - if (list_empty(&fc->pending) || fc->forget_batch-- > 0) - return fuse_read_forget(fc, cs, nbytes); + if (forget_pending(fiq)) { + if (list_empty(&fiq->pending) || fiq->forget_batch-- > 0) + return fuse_read_forget(fc, fiq, cs, nbytes); - if (fc->forget_batch <= -8) - fc->forget_batch = 16; + if (fiq->forget_batch <= -8) + fiq->forget_batch = 16; } - req = list_entry(fc->pending.next, struct fuse_req, list); - req->state = FUSE_REQ_READING; - list_move(&req->list, &fc->io); + req = list_entry(fiq->pending.next, struct fuse_req, list); + clear_bit(FR_PENDING, &req->flags); + list_del_init(&req->list); + spin_unlock(&fiq->waitq.lock); in = &req->in; reqsize = in->h.len; @@ -1310,37 +1289,48 @@ static ssize_t fuse_dev_do_read(struct fuse_conn *fc, struct file *file, request_end(fc, req); goto restart; } - spin_unlock(&fc->lock); + spin_lock(&fpq->lock); + list_add(&req->list, &fpq->io); + spin_unlock(&fpq->lock); cs->req = req; err = fuse_copy_one(cs, &in->h, sizeof(in->h)); if (!err) err = fuse_copy_args(cs, in->numargs, in->argpages, (struct fuse_arg *) in->args, 0); fuse_copy_finish(cs); - spin_lock(&fc->lock); - req->locked = 0; - if (req->aborted) { - request_end(fc, req); - return -ENODEV; + spin_lock(&fpq->lock); + clear_bit(FR_LOCKED, &req->flags); + if (!fpq->connected) { + err = -ENODEV; + goto out_end; } if (err) { req->out.h.error = -EIO; - request_end(fc, req); - return err; + goto out_end; } - if (!req->isreply) - request_end(fc, req); - else { - req->state = FUSE_REQ_SENT; - list_move_tail(&req->list, &fc->processing); - if (req->interrupted) - queue_interrupt(fc, req); - spin_unlock(&fc->lock); + if (!test_bit(FR_ISREPLY, &req->flags)) { + err = reqsize; + goto out_end; } + list_move_tail(&req->list, &fpq->processing); + spin_unlock(&fpq->lock); + set_bit(FR_SENT, &req->flags); + /* matches barrier in request_wait_answer() */ + smp_mb__after_atomic(); + if (test_bit(FR_INTERRUPTED, &req->flags)) + queue_interrupt(fiq, req); + return reqsize; +out_end: + if (!test_bit(FR_PRIVATE, &req->flags)) + list_del_init(&req->list); + spin_unlock(&fpq->lock); + request_end(fc, req); + return err; + err_unlock: - spin_unlock(&fc->lock); + spin_unlock(&fiq->waitq.lock); return err; } @@ -1359,16 +1349,17 @@ static ssize_t fuse_dev_read(struct kiocb *iocb, struct iov_iter *to) { struct fuse_copy_state cs; struct file *file = iocb->ki_filp; - struct fuse_conn *fc = fuse_get_conn(file); - if (!fc) + struct fuse_dev *fud = fuse_get_dev(file); + + if (!fud) return -EPERM; if (!iter_is_iovec(to)) return -EINVAL; - fuse_copy_init(&cs, fc, 1, to); + fuse_copy_init(&cs, 1, to); - return fuse_dev_do_read(fc, file, &cs, iov_iter_count(to)); + return fuse_dev_do_read(fud, file, &cs, iov_iter_count(to)); } static ssize_t fuse_dev_splice_read(struct file *in, loff_t *ppos, @@ -1380,18 +1371,19 @@ static ssize_t fuse_dev_splice_read(struct file *in, loff_t *ppos, int do_wakeup = 0; struct pipe_buffer *bufs; struct fuse_copy_state cs; - struct fuse_conn *fc = fuse_get_conn(in); - if (!fc) + struct fuse_dev *fud = fuse_get_dev(in); + + if (!fud) return -EPERM; bufs = kmalloc(pipe->buffers * sizeof(struct pipe_buffer), GFP_KERNEL); if (!bufs) return -ENOMEM; - fuse_copy_init(&cs, fc, 1, NULL); + fuse_copy_init(&cs, 1, NULL); cs.pipebufs = bufs; cs.pipe = pipe; - ret = fuse_dev_do_read(fc, in, &cs, len); + ret = fuse_dev_do_read(fud, in, &cs, len); if (ret < 0) goto out; @@ -1830,11 +1822,11 @@ static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, } /* Look up request on processing list by unique ID */ -static struct fuse_req *request_find(struct fuse_conn *fc, u64 unique) +static struct fuse_req *request_find(struct fuse_pqueue *fpq, u64 unique) { struct fuse_req *req; - list_for_each_entry(req, &fc->processing, list) { + list_for_each_entry(req, &fpq->processing, list) { if (req->in.h.unique == unique || req->intr_unique == unique) return req; } @@ -1871,10 +1863,12 @@ static int copy_out_args(struct fuse_copy_state *cs, struct fuse_out *out, * it from the list and copy the rest of the buffer to the request. * The request is finished by calling request_end() */ -static ssize_t fuse_dev_do_write(struct fuse_conn *fc, +static ssize_t fuse_dev_do_write(struct fuse_dev *fud, struct fuse_copy_state *cs, size_t nbytes) { int err; + struct fuse_conn *fc = fud->fc; + struct fuse_pqueue *fpq = &fud->pq; struct fuse_req *req; struct fuse_out_header oh; @@ -1902,63 +1896,60 @@ static ssize_t fuse_dev_do_write(struct fuse_conn *fc, if (oh.error <= -1000 || oh.error > 0) goto err_finish; - spin_lock(&fc->lock); + spin_lock(&fpq->lock); err = -ENOENT; - if (!fc->connected) - goto err_unlock; + if (!fpq->connected) + goto err_unlock_pq; - req = request_find(fc, oh.unique); + req = request_find(fpq, oh.unique); if (!req) - goto err_unlock; + goto err_unlock_pq; - if (req->aborted) { - spin_unlock(&fc->lock); - fuse_copy_finish(cs); - spin_lock(&fc->lock); - request_end(fc, req); - return -ENOENT; - } /* Is it an interrupt reply? */ if (req->intr_unique == oh.unique) { + spin_unlock(&fpq->lock); + err = -EINVAL; if (nbytes != sizeof(struct fuse_out_header)) - goto err_unlock; + goto err_finish; if (oh.error == -ENOSYS) fc->no_interrupt = 1; else if (oh.error == -EAGAIN) - queue_interrupt(fc, req); + queue_interrupt(&fc->iq, req); - spin_unlock(&fc->lock); fuse_copy_finish(cs); return nbytes; } - req->state = FUSE_REQ_WRITING; - list_move(&req->list, &fc->io); + clear_bit(FR_SENT, &req->flags); + list_move(&req->list, &fpq->io); req->out.h = oh; - req->locked = 1; + set_bit(FR_LOCKED, &req->flags); + spin_unlock(&fpq->lock); cs->req = req; if (!req->out.page_replace) cs->move_pages = 0; - spin_unlock(&fc->lock); err = copy_out_args(cs, &req->out, nbytes); fuse_copy_finish(cs); - spin_lock(&fc->lock); - req->locked = 0; - if (!err) { - if (req->aborted) - err = -ENOENT; - } else if (!req->aborted) + spin_lock(&fpq->lock); + clear_bit(FR_LOCKED, &req->flags); + if (!fpq->connected) + err = -ENOENT; + else if (err) req->out.h.error = -EIO; + if (!test_bit(FR_PRIVATE, &req->flags)) + list_del_init(&req->list); + spin_unlock(&fpq->lock); + request_end(fc, req); return err ? err : nbytes; - err_unlock: - spin_unlock(&fc->lock); + err_unlock_pq: + spin_unlock(&fpq->lock); err_finish: fuse_copy_finish(cs); return err; @@ -1967,16 +1958,17 @@ static ssize_t fuse_dev_do_write(struct fuse_conn *fc, static ssize_t fuse_dev_write(struct kiocb *iocb, struct iov_iter *from) { struct fuse_copy_state cs; - struct fuse_conn *fc = fuse_get_conn(iocb->ki_filp); - if (!fc) + struct fuse_dev *fud = fuse_get_dev(iocb->ki_filp); + + if (!fud) return -EPERM; if (!iter_is_iovec(from)) return -EINVAL; - fuse_copy_init(&cs, fc, 0, from); + fuse_copy_init(&cs, 0, from); - return fuse_dev_do_write(fc, &cs, iov_iter_count(from)); + return fuse_dev_do_write(fud, &cs, iov_iter_count(from)); } static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, @@ -1987,12 +1979,12 @@ static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, unsigned idx; struct pipe_buffer *bufs; struct fuse_copy_state cs; - struct fuse_conn *fc; + struct fuse_dev *fud; size_t rem; ssize_t ret; - fc = fuse_get_conn(out); - if (!fc) + fud = fuse_get_dev(out); + if (!fud) return -EPERM; bufs = kmalloc(pipe->buffers * sizeof(struct pipe_buffer), GFP_KERNEL); @@ -2039,7 +2031,7 @@ static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, } pipe_unlock(pipe); - fuse_copy_init(&cs, fc, 0, NULL); + fuse_copy_init(&cs, 0, NULL); cs.pipebufs = bufs; cs.nr_segs = nbuf; cs.pipe = pipe; @@ -2047,7 +2039,7 @@ static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, if (flags & SPLICE_F_MOVE) cs.move_pages = 1; - ret = fuse_dev_do_write(fc, &cs, len); + ret = fuse_dev_do_write(fud, &cs, len); for (idx = 0; idx < nbuf; idx++) { struct pipe_buffer *buf = &bufs[idx]; @@ -2061,18 +2053,21 @@ out: static unsigned fuse_dev_poll(struct file *file, poll_table *wait) { unsigned mask = POLLOUT | POLLWRNORM; - struct fuse_conn *fc = fuse_get_conn(file); - if (!fc) + struct fuse_iqueue *fiq; + struct fuse_dev *fud = fuse_get_dev(file); + + if (!fud) return POLLERR; - poll_wait(file, &fc->waitq, wait); + fiq = &fud->fc->iq; + poll_wait(file, &fiq->waitq, wait); - spin_lock(&fc->lock); - if (!fc->connected) + spin_lock(&fiq->waitq.lock); + if (!fiq->connected) mask = POLLERR; - else if (request_pending(fc)) + else if (request_pending(fiq)) mask |= POLLIN | POLLRDNORM; - spin_unlock(&fc->lock); + spin_unlock(&fiq->waitq.lock); return mask; } @@ -2083,67 +2078,18 @@ static unsigned fuse_dev_poll(struct file *file, poll_table *wait) * This function releases and reacquires fc->lock */ static void end_requests(struct fuse_conn *fc, struct list_head *head) -__releases(fc->lock) -__acquires(fc->lock) { while (!list_empty(head)) { struct fuse_req *req; req = list_entry(head->next, struct fuse_req, list); req->out.h.error = -ECONNABORTED; - request_end(fc, req); - spin_lock(&fc->lock); - } -} - -/* - * Abort requests under I/O - * - * The requests are set to aborted and finished, and the request - * waiter is woken up. This will make request_wait_answer() wait - * until the request is unlocked and then return. - * - * If the request is asynchronous, then the end function needs to be - * called after waiting for the request to be unlocked (if it was - * locked). - */ -static void end_io_requests(struct fuse_conn *fc) -__releases(fc->lock) -__acquires(fc->lock) -{ - while (!list_empty(&fc->io)) { - struct fuse_req *req = - list_entry(fc->io.next, struct fuse_req, list); - void (*end) (struct fuse_conn *, struct fuse_req *) = req->end; - - req->aborted = 1; - req->out.h.error = -ECONNABORTED; - req->state = FUSE_REQ_FINISHED; + clear_bit(FR_PENDING, &req->flags); + clear_bit(FR_SENT, &req->flags); list_del_init(&req->list); - wake_up(&req->waitq); - if (end) { - req->end = NULL; - __fuse_get_request(req); - spin_unlock(&fc->lock); - wait_event(req->waitq, !req->locked); - end(fc, req); - fuse_put_request(fc, req); - spin_lock(&fc->lock); - } + request_end(fc, req); } } -static void end_queued_requests(struct fuse_conn *fc) -__releases(fc->lock) -__acquires(fc->lock) -{ - fc->max_background = UINT_MAX; - flush_bg_queue(fc); - end_requests(fc, &fc->pending); - end_requests(fc, &fc->processing); - while (forget_pending(fc)) - kfree(dequeue_forget(fc, 1, NULL)); -} - static void end_polls(struct fuse_conn *fc) { struct rb_node *p; @@ -2162,67 +2108,156 @@ static void end_polls(struct fuse_conn *fc) /* * Abort all requests. * - * Emergency exit in case of a malicious or accidental deadlock, or - * just a hung filesystem. + * Emergency exit in case of a malicious or accidental deadlock, or just a hung + * filesystem. * - * The same effect is usually achievable through killing the - * filesystem daemon and all users of the filesystem. The exception - * is the combination of an asynchronous request and the tricky - * deadlock (see Documentation/filesystems/fuse.txt). + * The same effect is usually achievable through killing the filesystem daemon + * and all users of the filesystem. The exception is the combination of an + * asynchronous request and the tricky deadlock (see + * Documentation/filesystems/fuse.txt). * - * During the aborting, progression of requests from the pending and - * processing lists onto the io list, and progression of new requests - * onto the pending list is prevented by req->connected being false. - * - * Progression of requests under I/O to the processing list is - * prevented by the req->aborted flag being true for these requests. - * For this reason requests on the io list must be aborted first. + * Aborting requests under I/O goes as follows: 1: Separate out unlocked + * requests, they should be finished off immediately. Locked requests will be + * finished after unlock; see unlock_request(). 2: Finish off the unlocked + * requests. It is possible that some request will finish before we can. This + * is OK, the request will in that case be removed from the list before we touch + * it. */ void fuse_abort_conn(struct fuse_conn *fc) { + struct fuse_iqueue *fiq = &fc->iq; + spin_lock(&fc->lock); if (fc->connected) { + struct fuse_dev *fud; + struct fuse_req *req, *next; + LIST_HEAD(to_end1); + LIST_HEAD(to_end2); + fc->connected = 0; fc->blocked = 0; fuse_set_initialized(fc); - end_io_requests(fc); - end_queued_requests(fc); + list_for_each_entry(fud, &fc->devices, entry) { + struct fuse_pqueue *fpq = &fud->pq; + + spin_lock(&fpq->lock); + fpq->connected = 0; + list_for_each_entry_safe(req, next, &fpq->io, list) { + req->out.h.error = -ECONNABORTED; + spin_lock(&req->waitq.lock); + set_bit(FR_ABORTED, &req->flags); + if (!test_bit(FR_LOCKED, &req->flags)) { + set_bit(FR_PRIVATE, &req->flags); + list_move(&req->list, &to_end1); + } + spin_unlock(&req->waitq.lock); + } + list_splice_init(&fpq->processing, &to_end2); + spin_unlock(&fpq->lock); + } + fc->max_background = UINT_MAX; + flush_bg_queue(fc); + + spin_lock(&fiq->waitq.lock); + fiq->connected = 0; + list_splice_init(&fiq->pending, &to_end2); + while (forget_pending(fiq)) + kfree(dequeue_forget(fiq, 1, NULL)); + wake_up_all_locked(&fiq->waitq); + spin_unlock(&fiq->waitq.lock); + kill_fasync(&fiq->fasync, SIGIO, POLL_IN); end_polls(fc); - wake_up_all(&fc->waitq); wake_up_all(&fc->blocked_waitq); - kill_fasync(&fc->fasync, SIGIO, POLL_IN); + spin_unlock(&fc->lock); + + while (!list_empty(&to_end1)) { + req = list_first_entry(&to_end1, struct fuse_req, list); + __fuse_get_request(req); + list_del_init(&req->list); + request_end(fc, req); + } + end_requests(fc, &to_end2); + } else { + spin_unlock(&fc->lock); } - spin_unlock(&fc->lock); } EXPORT_SYMBOL_GPL(fuse_abort_conn); int fuse_dev_release(struct inode *inode, struct file *file) { - struct fuse_conn *fc = fuse_get_conn(file); - if (fc) { - spin_lock(&fc->lock); - fc->connected = 0; - fc->blocked = 0; - fuse_set_initialized(fc); - end_queued_requests(fc); - end_polls(fc); - wake_up_all(&fc->blocked_waitq); - spin_unlock(&fc->lock); - fuse_conn_put(fc); - } + struct fuse_dev *fud = fuse_get_dev(file); + if (fud) { + struct fuse_conn *fc = fud->fc; + struct fuse_pqueue *fpq = &fud->pq; + + WARN_ON(!list_empty(&fpq->io)); + end_requests(fc, &fpq->processing); + /* Are we the last open device? */ + if (atomic_dec_and_test(&fc->dev_count)) { + WARN_ON(fc->iq.fasync != NULL); + fuse_abort_conn(fc); + } + fuse_dev_free(fud); + } return 0; } EXPORT_SYMBOL_GPL(fuse_dev_release); static int fuse_dev_fasync(int fd, struct file *file, int on) { - struct fuse_conn *fc = fuse_get_conn(file); - if (!fc) + struct fuse_dev *fud = fuse_get_dev(file); + + if (!fud) return -EPERM; /* No locking - fasync_helper does its own locking */ - return fasync_helper(fd, file, on, &fc->fasync); + return fasync_helper(fd, file, on, &fud->fc->iq.fasync); +} + +static int fuse_device_clone(struct fuse_conn *fc, struct file *new) +{ + struct fuse_dev *fud; + + if (new->private_data) + return -EINVAL; + + fud = fuse_dev_alloc(fc); + if (!fud) + return -ENOMEM; + + new->private_data = fud; + atomic_inc(&fc->dev_count); + + return 0; +} + +static long fuse_dev_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + int err = -ENOTTY; + + if (cmd == FUSE_DEV_IOC_CLONE) { + int oldfd; + + err = -EFAULT; + if (!get_user(oldfd, (__u32 __user *) arg)) { + struct file *old = fget(oldfd); + + err = -EINVAL; + if (old) { + struct fuse_dev *fud = fuse_get_dev(old); + + if (fud) { + mutex_lock(&fuse_mutex); + err = fuse_device_clone(fud->fc, file); + mutex_unlock(&fuse_mutex); + } + fput(old); + } + } + } + return err; } const struct file_operations fuse_dev_operations = { @@ -2236,6 +2271,8 @@ const struct file_operations fuse_dev_operations = { .poll = fuse_dev_poll, .release = fuse_dev_release, .fasync = fuse_dev_fasync, + .unlocked_ioctl = fuse_dev_ioctl, + .compat_ioctl = fuse_dev_ioctl, }; EXPORT_SYMBOL_GPL(fuse_dev_operations); diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 1344647..f523f2f 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -96,17 +96,17 @@ static void fuse_file_put(struct fuse_file *ff, bool sync) * Drop the release request when client does not * implement 'open' */ - req->background = 0; + __clear_bit(FR_BACKGROUND, &req->flags); iput(req->misc.release.inode); fuse_put_request(ff->fc, req); } else if (sync) { - req->background = 0; + __clear_bit(FR_BACKGROUND, &req->flags); fuse_request_send(ff->fc, req); iput(req->misc.release.inode); fuse_put_request(ff->fc, req); } else { req->end = fuse_release_end; - req->background = 1; + __set_bit(FR_BACKGROUND, &req->flags); fuse_request_send_background(ff->fc, req); } kfree(ff); @@ -299,8 +299,8 @@ void fuse_sync_release(struct fuse_file *ff, int flags) { WARN_ON(atomic_read(&ff->count) > 1); fuse_prepare_release(ff, flags, FUSE_RELEASE); - ff->reserved_req->force = 1; - ff->reserved_req->background = 0; + __set_bit(FR_FORCE, &ff->reserved_req->flags); + __clear_bit(FR_BACKGROUND, &ff->reserved_req->flags); fuse_request_send(ff->fc, ff->reserved_req); fuse_put_request(ff->fc, ff->reserved_req); kfree(ff); @@ -426,7 +426,7 @@ static int fuse_flush(struct file *file, fl_owner_t id) req->in.numargs = 1; req->in.args[0].size = sizeof(inarg); req->in.args[0].value = &inarg; - req->force = 1; + __set_bit(FR_FORCE, &req->flags); fuse_request_send(fc, req); err = req->out.h.error; fuse_put_request(fc, req); @@ -1445,9 +1445,9 @@ static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req) list_del(&req->writepages_entry); for (i = 0; i < req->num_pages; i++) { - dec_bdi_stat(bdi, BDI_WRITEBACK); + dec_wb_stat(&bdi->wb, WB_WRITEBACK); dec_zone_page_state(req->pages[i], NR_WRITEBACK_TEMP); - bdi_writeout_inc(bdi); + wb_writeout_inc(&bdi->wb); } wake_up(&fi->page_waitq); } @@ -1611,7 +1611,8 @@ static int fuse_writepage_locked(struct page *page) if (!req) goto err; - req->background = 1; /* writeback always goes to bg_queue */ + /* writeback always goes to bg_queue */ + __set_bit(FR_BACKGROUND, &req->flags); tmp_page = alloc_page(GFP_NOFS | __GFP_HIGHMEM); if (!tmp_page) goto err_free; @@ -1634,7 +1635,7 @@ static int fuse_writepage_locked(struct page *page) req->end = fuse_writepage_end; req->inode = inode; - inc_bdi_stat(inode_to_bdi(inode), BDI_WRITEBACK); + inc_wb_stat(&inode_to_bdi(inode)->wb, WB_WRITEBACK); inc_zone_page_state(tmp_page, NR_WRITEBACK_TEMP); spin_lock(&fc->lock); @@ -1742,16 +1743,15 @@ static bool fuse_writepage_in_flight(struct fuse_req *new_req, } } - if (old_req->num_pages == 1 && (old_req->state == FUSE_REQ_INIT || - old_req->state == FUSE_REQ_PENDING)) { + if (old_req->num_pages == 1 && test_bit(FR_PENDING, &old_req->flags)) { struct backing_dev_info *bdi = inode_to_bdi(page->mapping->host); copy_highpage(old_req->pages[0], page); spin_unlock(&fc->lock); - dec_bdi_stat(bdi, BDI_WRITEBACK); + dec_wb_stat(&bdi->wb, WB_WRITEBACK); dec_zone_page_state(page, NR_WRITEBACK_TEMP); - bdi_writeout_inc(bdi); + wb_writeout_inc(&bdi->wb); fuse_writepage_free(fc, new_req); fuse_request_free(new_req); goto out; @@ -1830,7 +1830,7 @@ static int fuse_writepages_fill(struct page *page, req->misc.write.in.write_flags |= FUSE_WRITE_CACHE; req->misc.write.next = NULL; req->in.argpages = 1; - req->background = 1; + __set_bit(FR_BACKGROUND, &req->flags); req->num_pages = 0; req->end = fuse_writepage_end; req->inode = inode; @@ -1848,7 +1848,7 @@ static int fuse_writepages_fill(struct page *page, req->page_descs[req->num_pages].offset = 0; req->page_descs[req->num_pages].length = PAGE_SIZE; - inc_bdi_stat(inode_to_bdi(inode), BDI_WRITEBACK); + inc_wb_stat(&inode_to_bdi(inode)->wb, WB_WRITEBACK); inc_zone_page_state(tmp_page, NR_WRITEBACK_TEMP); err = 0; diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 7354dc1..4051131 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -241,16 +241,6 @@ struct fuse_args { #define FUSE_ARGS(args) struct fuse_args args = {} -/** The request state */ -enum fuse_req_state { - FUSE_REQ_INIT = 0, - FUSE_REQ_PENDING, - FUSE_REQ_READING, - FUSE_REQ_SENT, - FUSE_REQ_WRITING, - FUSE_REQ_FINISHED -}; - /** The request IO state (for asynchronous processing) */ struct fuse_io_priv { int async; @@ -267,7 +257,40 @@ struct fuse_io_priv { }; /** + * Request flags + * + * FR_ISREPLY: set if the request has reply + * FR_FORCE: force sending of the request even if interrupted + * FR_BACKGROUND: request is sent in the background + * FR_WAITING: request is counted as "waiting" + * FR_ABORTED: the request was aborted + * FR_INTERRUPTED: the request has been interrupted + * FR_LOCKED: data is being copied to/from the request + * FR_PENDING: request is not yet in userspace + * FR_SENT: request is in userspace, waiting for an answer + * FR_FINISHED: request is finished + * FR_PRIVATE: request is on private list + */ +enum fuse_req_flag { + FR_ISREPLY, + FR_FORCE, + FR_BACKGROUND, + FR_WAITING, + FR_ABORTED, + FR_INTERRUPTED, + FR_LOCKED, + FR_PENDING, + FR_SENT, + FR_FINISHED, + FR_PRIVATE, +}; + +/** * A request to the client + * + * .waitq.lock protects the following fields: + * - FR_ABORTED + * - FR_LOCKED (may also be modified under fc->lock, tested under both) */ struct fuse_req { /** This can be on either pending processing or io lists in @@ -283,35 +306,8 @@ struct fuse_req { /** Unique ID for the interrupt request */ u64 intr_unique; - /* - * The following bitfields are either set once before the - * request is queued or setting/clearing them is protected by - * fuse_conn->lock - */ - - /** True if the request has reply */ - unsigned isreply:1; - - /** Force sending of the request even if interrupted */ - unsigned force:1; - - /** The request was aborted */ - unsigned aborted:1; - - /** Request is sent in the background */ - unsigned background:1; - - /** The request has been interrupted */ - unsigned interrupted:1; - - /** Data is being copied to/from the request */ - unsigned locked:1; - - /** Request is counted as "waiting" */ - unsigned waiting:1; - - /** State of the request */ - enum fuse_req_state state; + /* Request flags, updated with test/set/clear_bit() */ + unsigned long flags; /** The request input */ struct fuse_in in; @@ -380,6 +376,61 @@ struct fuse_req { struct file *stolen_file; }; +struct fuse_iqueue { + /** Connection established */ + unsigned connected; + + /** Readers of the connection are waiting on this */ + wait_queue_head_t waitq; + + /** The next unique request id */ + u64 reqctr; + + /** The list of pending requests */ + struct list_head pending; + + /** Pending interrupts */ + struct list_head interrupts; + + /** Queue of pending forgets */ + struct fuse_forget_link forget_list_head; + struct fuse_forget_link *forget_list_tail; + + /** Batching of FORGET requests (positive indicates FORGET batch) */ + int forget_batch; + + /** O_ASYNC requests */ + struct fasync_struct *fasync; +}; + +struct fuse_pqueue { + /** Connection established */ + unsigned connected; + + /** Lock protecting accessess to members of this structure */ + spinlock_t lock; + + /** The list of requests being processed */ + struct list_head processing; + + /** The list of requests under I/O */ + struct list_head io; +}; + +/** + * Fuse device instance + */ +struct fuse_dev { + /** Fuse connection for this device */ + struct fuse_conn *fc; + + /** Processing queue */ + struct fuse_pqueue pq; + + /** list entry on fc->devices */ + struct list_head entry; +}; + /** * A Fuse connection. * @@ -394,6 +445,9 @@ struct fuse_conn { /** Refcount */ atomic_t count; + /** Number of fuse_dev's */ + atomic_t dev_count; + struct rcu_head rcu; /** The user id for this mount */ @@ -411,17 +465,8 @@ struct fuse_conn { /** Maximum write size */ unsigned max_write; - /** Readers of the connection are waiting on this */ - wait_queue_head_t waitq; - - /** The list of pending requests */ - struct list_head pending; - - /** The list of requests being processed */ - struct list_head processing; - - /** The list of requests under I/O */ - struct list_head io; + /** Input queue */ + struct fuse_iqueue iq; /** The next unique kernel file handle */ u64 khctr; @@ -444,16 +489,6 @@ struct fuse_conn { /** The list of background requests set aside for later queuing */ struct list_head bg_queue; - /** Pending interrupts */ - struct list_head interrupts; - - /** Queue of pending forgets */ - struct fuse_forget_link forget_list_head; - struct fuse_forget_link *forget_list_tail; - - /** Batching of FORGET requests (positive indicates FORGET batch) */ - int forget_batch; - /** Flag indicating that INIT reply has been received. Allocating * any fuse request will be suspended until the flag is set */ int initialized; @@ -469,9 +504,6 @@ struct fuse_conn { /** waitq for reserved requests */ wait_queue_head_t reserved_req_waitq; - /** The next unique request id */ - u64 reqctr; - /** Connection established, cleared on umount, connection abort and device release */ unsigned connected; @@ -594,9 +626,6 @@ struct fuse_conn { /** number of dentries used in the above array */ int ctl_ndents; - /** O_ASYNC requests */ - struct fasync_struct *fasync; - /** Key for lock owner ID scrambling */ u32 scramble_key[4]; @@ -614,6 +643,9 @@ struct fuse_conn { /** Read/write semaphore to hold when accessing sb. */ struct rw_semaphore killsb; + + /** List of device instances belonging to this connection */ + struct list_head devices; }; static inline struct fuse_conn *get_fuse_conn_super(struct super_block *sb) @@ -826,6 +858,9 @@ void fuse_conn_init(struct fuse_conn *fc); */ void fuse_conn_put(struct fuse_conn *fc); +struct fuse_dev *fuse_dev_alloc(struct fuse_conn *fc); +void fuse_dev_free(struct fuse_dev *fud); + /** * Add connection to control filesystem */ diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 082ac1c..2913db2 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -362,8 +362,8 @@ static void fuse_send_destroy(struct fuse_conn *fc) if (req && fc->conn_init) { fc->destroy_req = NULL; req->in.h.opcode = FUSE_DESTROY; - req->force = 1; - req->background = 0; + __set_bit(FR_FORCE, &req->flags); + __clear_bit(FR_BACKGROUND, &req->flags); fuse_request_send(fc, req); fuse_put_request(fc, req); } @@ -567,30 +567,46 @@ static int fuse_show_options(struct seq_file *m, struct dentry *root) return 0; } +static void fuse_iqueue_init(struct fuse_iqueue *fiq) +{ + memset(fiq, 0, sizeof(struct fuse_iqueue)); + init_waitqueue_head(&fiq->waitq); + INIT_LIST_HEAD(&fiq->pending); + INIT_LIST_HEAD(&fiq->interrupts); + fiq->forget_list_tail = &fiq->forget_list_head; + fiq->connected = 1; +} + +static void fuse_pqueue_init(struct fuse_pqueue *fpq) +{ + memset(fpq, 0, sizeof(struct fuse_pqueue)); + spin_lock_init(&fpq->lock); + INIT_LIST_HEAD(&fpq->processing); + INIT_LIST_HEAD(&fpq->io); + fpq->connected = 1; +} + void fuse_conn_init(struct fuse_conn *fc) { memset(fc, 0, sizeof(*fc)); spin_lock_init(&fc->lock); init_rwsem(&fc->killsb); atomic_set(&fc->count, 1); - init_waitqueue_head(&fc->waitq); + atomic_set(&fc->dev_count, 1); init_waitqueue_head(&fc->blocked_waitq); init_waitqueue_head(&fc->reserved_req_waitq); - INIT_LIST_HEAD(&fc->pending); - INIT_LIST_HEAD(&fc->processing); - INIT_LIST_HEAD(&fc->io); - INIT_LIST_HEAD(&fc->interrupts); + fuse_iqueue_init(&fc->iq); INIT_LIST_HEAD(&fc->bg_queue); INIT_LIST_HEAD(&fc->entry); - fc->forget_list_tail = &fc->forget_list_head; + INIT_LIST_HEAD(&fc->devices); atomic_set(&fc->num_waiting, 0); fc->max_background = FUSE_DEFAULT_MAX_BACKGROUND; fc->congestion_threshold = FUSE_DEFAULT_CONGESTION_THRESHOLD; fc->khctr = 0; fc->polled_files = RB_ROOT; - fc->reqctr = 0; fc->blocked = 0; fc->initialized = 0; + fc->connected = 1; fc->attr_version = 1; get_random_bytes(&fc->scramble_key, sizeof(fc->scramble_key)); } @@ -930,6 +946,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) static void fuse_free_conn(struct fuse_conn *fc) { + WARN_ON(!list_empty(&fc->devices)); kfree_rcu(fc, rcu); } @@ -975,8 +992,42 @@ static int fuse_bdi_init(struct fuse_conn *fc, struct super_block *sb) return 0; } +struct fuse_dev *fuse_dev_alloc(struct fuse_conn *fc) +{ + struct fuse_dev *fud; + + fud = kzalloc(sizeof(struct fuse_dev), GFP_KERNEL); + if (fud) { + fud->fc = fuse_conn_get(fc); + fuse_pqueue_init(&fud->pq); + + spin_lock(&fc->lock); + list_add_tail(&fud->entry, &fc->devices); + spin_unlock(&fc->lock); + } + + return fud; +} +EXPORT_SYMBOL_GPL(fuse_dev_alloc); + +void fuse_dev_free(struct fuse_dev *fud) +{ + struct fuse_conn *fc = fud->fc; + + if (fc) { + spin_lock(&fc->lock); + list_del(&fud->entry); + spin_unlock(&fc->lock); + + fuse_conn_put(fc); + } + kfree(fud); +} +EXPORT_SYMBOL_GPL(fuse_dev_free); + static int fuse_fill_super(struct super_block *sb, void *data, int silent) { + struct fuse_dev *fud; struct fuse_conn *fc; struct inode *root; struct fuse_mount_data d; @@ -1026,12 +1077,17 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) goto err_fput; fuse_conn_init(fc); + fc->release = fuse_free_conn; + + fud = fuse_dev_alloc(fc); + if (!fud) + goto err_put_conn; fc->dev = sb->s_dev; fc->sb = sb; err = fuse_bdi_init(fc, sb); if (err) - goto err_put_conn; + goto err_dev_free; sb->s_bdi = &fc->bdi; @@ -1040,7 +1096,6 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) fc->dont_mask = 1; sb->s_flags |= MS_POSIXACL; - fc->release = fuse_free_conn; fc->flags = d.flags; fc->user_id = d.user_id; fc->group_id = d.group_id; @@ -1053,14 +1108,14 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) root = fuse_get_root_inode(sb, d.rootmode); root_dentry = d_make_root(root); if (!root_dentry) - goto err_put_conn; + goto err_dev_free; /* only now - we want root dentry with NULL ->d_op */ sb->s_d_op = &fuse_dentry_operations; init_req = fuse_request_alloc(0); if (!init_req) goto err_put_root; - init_req->background = 1; + __set_bit(FR_BACKGROUND, &init_req->flags); if (is_bdev) { fc->destroy_req = fuse_request_alloc(0); @@ -1079,8 +1134,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) list_add_tail(&fc->entry, &fuse_conn_list); sb->s_root = root_dentry; - fc->connected = 1; - file->private_data = fuse_conn_get(fc); + file->private_data = fud; mutex_unlock(&fuse_mutex); /* * atomic_dec_and_test() in fput() provides the necessary @@ -1099,6 +1153,8 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) fuse_request_free(init_req); err_put_root: dput(root_dentry); + err_dev_free: + fuse_dev_free(fud); err_put_conn: fuse_bdi_destroy(fc); fuse_conn_put(fc); @@ -1238,7 +1294,6 @@ static void fuse_fs_cleanup(void) } static struct kobject *fuse_kobj; -static struct kobject *connections_kobj; static int fuse_sysfs_init(void) { @@ -1250,11 +1305,9 @@ static int fuse_sysfs_init(void) goto out_err; } - connections_kobj = kobject_create_and_add("connections", fuse_kobj); - if (!connections_kobj) { - err = -ENOMEM; + err = sysfs_create_mount_point(fuse_kobj, "connections"); + if (err) goto out_fuse_unregister; - } return 0; @@ -1266,7 +1319,7 @@ static int fuse_sysfs_init(void) static void fuse_sysfs_cleanup(void) { - kobject_put(connections_kobj); + sysfs_remove_mount_point(fuse_kobj, "connections"); kobject_put(fuse_kobj); } diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 5551fea..1caee05 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -171,6 +171,7 @@ static int __gfs2_jdata_writepage(struct page *page, struct writeback_control *w /** * gfs2_jdata_writepage - Write complete page * @page: Page to write + * @wbc: The writeback control * * Returns: errno * @@ -221,9 +222,10 @@ static int gfs2_writepages(struct address_space *mapping, * gfs2_write_jdata_pagevec - Write back a pagevec's worth of pages * @mapping: The mapping * @wbc: The writeback control - * @writepage: The writepage function to call for each page * @pvec: The vector of pages * @nr_pages: The number of pages to write + * @end: End position + * @done_index: Page index * * Returns: non-zero if loop should terminate, zero otherwise */ @@ -333,8 +335,6 @@ continue_unlock: * gfs2_write_cache_jdata - Like write_cache_pages but different * @mapping: The mapping to write * @wbc: The writeback control - * @writepage: The writepage function to call - * @data: The data to pass to writepage * * The reason that we use our own function here is that we need to * start transactions before we grab page locks. This allows us @@ -588,6 +588,10 @@ int gfs2_internal_read(struct gfs2_inode *ip, char *buf, loff_t *pos, /** * gfs2_readpages - Read a bunch of pages at once + * @file: The file to read from + * @mapping: Address space info + * @pages: List of pages to read + * @nr_pages: Number of pages to read * * Some notes: * 1. This is only for readahead, so we can simply ignore any things @@ -853,7 +857,7 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh, * @mapping: The address space to write to * @pos: The file position * @len: The length of the data - * @copied: + * @copied: How much was actually copied by the VFS * @page: The page that has been written * @fsdata: The fsdata (unused in GFS2) * diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 3189287..cf4ab89 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -180,7 +180,7 @@ void gfs2_set_inode_flags(struct inode *inode) flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC|S_NOSEC); if ((ip->i_eattr == 0) && !is_sxid(inode->i_mode)) - inode->i_flags |= S_NOSEC; + flags |= S_NOSEC; if (ip->i_diskflags & GFS2_DIF_IMMUTABLE) flags |= S_IMMUTABLE; if (ip->i_diskflags & GFS2_DIF_APPENDONLY) @@ -917,7 +917,7 @@ static long gfs2_fallocate(struct file *file, int mode, loff_t offset, loff_t le struct gfs2_holder gh; int ret; - if (mode & ~FALLOC_FL_KEEP_SIZE) + if ((mode & ~FALLOC_FL_KEEP_SIZE) || gfs2_is_jdata(ip)) return -EOPNOTSUPP; mutex_lock(&inode->i_mutex); diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 0fa8062..a38e38f 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -1076,7 +1076,8 @@ void gfs2_glock_dq(struct gfs2_holder *gh) !test_bit(GLF_DEMOTE, &gl->gl_flags)) fast_path = 1; } - if (!test_bit(GLF_LFLUSH, &gl->gl_flags) && demote_ok(gl)) + if (!test_bit(GLF_LFLUSH, &gl->gl_flags) && demote_ok(gl) && + (glops->go_flags & GLOF_LRU)) gfs2_glock_add_to_lru(gl); trace_gfs2_glock_queue(gh, 0); diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index fe91951..fa3fa5e 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -144,6 +144,12 @@ static void rgrp_go_sync(struct gfs2_glock *gl) struct gfs2_rgrpd *rgd; int error; + spin_lock(&gl->gl_spin); + rgd = gl->gl_object; + if (rgd) + gfs2_rgrp_brelse(rgd); + spin_unlock(&gl->gl_spin); + if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags)) return; GLOCK_BUG_ON(gl, gl->gl_state != LM_ST_EXCLUSIVE); @@ -175,15 +181,17 @@ static void rgrp_go_inval(struct gfs2_glock *gl, int flags) { struct gfs2_sbd *sdp = gl->gl_sbd; struct address_space *mapping = &sdp->sd_aspace; + struct gfs2_rgrpd *rgd = gl->gl_object; + + if (rgd) + gfs2_rgrp_brelse(rgd); WARN_ON_ONCE(!(flags & DIO_METADATA)); gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count)); truncate_inode_pages_range(mapping, gl->gl_vm.start, gl->gl_vm.end); - if (gl->gl_object) { - struct gfs2_rgrpd *rgd = (struct gfs2_rgrpd *)gl->gl_object; + if (rgd) rgd->rd_flags &= ~GFS2_RDF_UPTODATE; - } } /** @@ -561,7 +569,7 @@ const struct gfs2_glock_operations gfs2_inode_glops = { .go_lock = inode_go_lock, .go_dump = inode_go_dump, .go_type = LM_TYPE_INODE, - .go_flags = GLOF_ASPACE, + .go_flags = GLOF_ASPACE | GLOF_LRU, }; const struct gfs2_glock_operations gfs2_rgrp_glops = { @@ -584,10 +592,12 @@ const struct gfs2_glock_operations gfs2_freeze_glops = { const struct gfs2_glock_operations gfs2_iopen_glops = { .go_type = LM_TYPE_IOPEN, .go_callback = iopen_go_callback, + .go_flags = GLOF_LRU, }; const struct gfs2_glock_operations gfs2_flock_glops = { .go_type = LM_TYPE_FLOCK, + .go_flags = GLOF_LRU, }; const struct gfs2_glock_operations gfs2_nondisk_glops = { @@ -596,7 +606,7 @@ const struct gfs2_glock_operations gfs2_nondisk_glops = { const struct gfs2_glock_operations gfs2_quota_glops = { .go_type = LM_TYPE_QUOTA, - .go_flags = GLOF_LVB, + .go_flags = GLOF_LVB | GLOF_LRU, }; const struct gfs2_glock_operations gfs2_journal_glops = { diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 58b75ab..a1ec7c2 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -225,6 +225,7 @@ struct gfs2_glock_operations { const unsigned long go_flags; #define GLOF_ASPACE 1 #define GLOF_LVB 2 +#define GLOF_LRU 4 }; enum { @@ -432,6 +433,7 @@ enum { QDF_CHANGE = 1, QDF_LOCKED = 2, QDF_REFRESH = 3, + QDF_QMSG_QUIET = 4, }; struct gfs2_quota_data { diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 3a1461d..063fdfc 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -1227,8 +1227,8 @@ static int gfs2_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, */ static int gfs2_atomic_open(struct inode *dir, struct dentry *dentry, - struct file *file, unsigned flags, - umode_t mode, int *opened) + struct file *file, unsigned flags, + umode_t mode, int *opened) { struct dentry *d; bool excl = !!(flags & O_EXCL); @@ -1307,6 +1307,35 @@ static int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to) } /** + * update_moved_ino - Update an inode that's being moved + * @ip: The inode being moved + * @ndip: The parent directory of the new filename + * @dir_rename: True of ip is a directory + * + * Returns: errno + */ + +static int update_moved_ino(struct gfs2_inode *ip, struct gfs2_inode *ndip, + int dir_rename) +{ + int error; + struct buffer_head *dibh; + + if (dir_rename) + return gfs2_dir_mvino(ip, &gfs2_qdotdot, ndip, DT_DIR); + + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) + return error; + ip->i_inode.i_ctime = CURRENT_TIME; + gfs2_trans_add_meta(ip->i_gl, dibh); + gfs2_dinode_out(ip, dibh->b_data); + brelse(dibh); + return 0; +} + + +/** * gfs2_rename - Rename a file * @odir: Parent directory of old file name * @odentry: The old dentry of the file @@ -1354,7 +1383,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry, if (S_ISDIR(ip->i_inode.i_mode)) { dir_rename = 1; - /* don't move a dirctory into it's subdir */ + /* don't move a directory into its subdir */ error = gfs2_ok_to_move(ip, ndip); if (error) goto out_gunlock_r; @@ -1494,20 +1523,9 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry, if (nip) error = gfs2_unlink_inode(ndip, ndentry); - if (dir_rename) { - error = gfs2_dir_mvino(ip, &gfs2_qdotdot, ndip, DT_DIR); - if (error) - goto out_end_trans; - } else { - struct buffer_head *dibh; - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - goto out_end_trans; - ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_meta(ip->i_gl, dibh); - gfs2_dinode_out(ip, dibh->b_data); - brelse(dibh); - } + error = update_moved_ino(ip, ndip, dir_rename); + if (error) + goto out_end_trans; error = gfs2_dir_del(odip, odentry); if (error) @@ -1539,6 +1557,161 @@ out: } /** + * gfs2_exchange - exchange two files + * @odir: Parent directory of old file name + * @odentry: The old dentry of the file + * @ndir: Parent directory of new file name + * @ndentry: The new dentry of the file + * @flags: The rename flags + * + * Returns: errno + */ + +static int gfs2_exchange(struct inode *odir, struct dentry *odentry, + struct inode *ndir, struct dentry *ndentry, + unsigned int flags) +{ + struct gfs2_inode *odip = GFS2_I(odir); + struct gfs2_inode *ndip = GFS2_I(ndir); + struct gfs2_inode *oip = GFS2_I(odentry->d_inode); + struct gfs2_inode *nip = GFS2_I(ndentry->d_inode); + struct gfs2_sbd *sdp = GFS2_SB(odir); + struct gfs2_holder ghs[5], r_gh = { .gh_gl = NULL, }; + unsigned int num_gh; + unsigned int x; + umode_t old_mode = oip->i_inode.i_mode; + umode_t new_mode = nip->i_inode.i_mode; + int error; + + error = gfs2_rindex_update(sdp); + if (error) + return error; + + if (odip != ndip) { + error = gfs2_glock_nq_init(sdp->sd_rename_gl, LM_ST_EXCLUSIVE, + 0, &r_gh); + if (error) + goto out; + + if (S_ISDIR(old_mode)) { + /* don't move a directory into its subdir */ + error = gfs2_ok_to_move(oip, ndip); + if (error) + goto out_gunlock_r; + } + + if (S_ISDIR(new_mode)) { + /* don't move a directory into its subdir */ + error = gfs2_ok_to_move(nip, odip); + if (error) + goto out_gunlock_r; + } + } + + num_gh = 1; + gfs2_holder_init(odip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); + if (odip != ndip) { + gfs2_holder_init(ndip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + } + gfs2_holder_init(oip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + + gfs2_holder_init(nip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + + for (x = 0; x < num_gh; x++) { + error = gfs2_glock_nq(ghs + x); + if (error) + goto out_gunlock; + } + + error = -ENOENT; + if (oip->i_inode.i_nlink == 0 || nip->i_inode.i_nlink == 0) + goto out_gunlock; + + error = gfs2_unlink_ok(odip, &odentry->d_name, oip); + if (error) + goto out_gunlock; + error = gfs2_unlink_ok(ndip, &ndentry->d_name, nip); + if (error) + goto out_gunlock; + + if (S_ISDIR(old_mode)) { + error = gfs2_permission(odentry->d_inode, MAY_WRITE); + if (error) + goto out_gunlock; + } + if (S_ISDIR(new_mode)) { + error = gfs2_permission(ndentry->d_inode, MAY_WRITE); + if (error) + goto out_gunlock; + } + error = gfs2_trans_begin(sdp, 4 * RES_DINODE + 4 * RES_LEAF, 0); + if (error) + goto out_gunlock; + + error = update_moved_ino(oip, ndip, S_ISDIR(old_mode)); + if (error) + goto out_end_trans; + + error = update_moved_ino(nip, odip, S_ISDIR(new_mode)); + if (error) + goto out_end_trans; + + error = gfs2_dir_mvino(ndip, &ndentry->d_name, oip, + IF2DT(old_mode)); + if (error) + goto out_end_trans; + + error = gfs2_dir_mvino(odip, &odentry->d_name, nip, + IF2DT(new_mode)); + if (error) + goto out_end_trans; + + if (odip != ndip) { + if (S_ISDIR(new_mode) && !S_ISDIR(old_mode)) { + inc_nlink(&odip->i_inode); + drop_nlink(&ndip->i_inode); + } else if (S_ISDIR(old_mode) && !S_ISDIR(new_mode)) { + inc_nlink(&ndip->i_inode); + drop_nlink(&odip->i_inode); + } + } + mark_inode_dirty(&ndip->i_inode); + if (odip != ndip) + mark_inode_dirty(&odip->i_inode); + +out_end_trans: + gfs2_trans_end(sdp); +out_gunlock: + while (x--) { + gfs2_glock_dq(ghs + x); + gfs2_holder_uninit(ghs + x); + } +out_gunlock_r: + if (r_gh.gh_gl) + gfs2_glock_dq_uninit(&r_gh); +out: + return error; +} + +static int gfs2_rename2(struct inode *odir, struct dentry *odentry, + struct inode *ndir, struct dentry *ndentry, + unsigned int flags) +{ + flags &= ~RENAME_NOREPLACE; + + if (flags & ~RENAME_EXCHANGE) + return -EINVAL; + + if (flags & RENAME_EXCHANGE) + return gfs2_exchange(odir, odentry, ndir, ndentry, flags); + + return gfs2_rename(odir, odentry, ndir, ndentry); +} + +/** * gfs2_follow_link - Follow a symbolic link * @dentry: The dentry of the link * @nd: Data that we pass to vfs_follow_link() @@ -1716,7 +1889,7 @@ static int setattr_chown(struct inode *inode, struct iattr *attr) if (!uid_eq(ouid, NO_UID_QUOTA_CHANGE) || !gid_eq(ogid, NO_GID_QUOTA_CHANGE)) { - gfs2_quota_change(ip, -ap.target, ouid, ogid); + gfs2_quota_change(ip, -(s64)ap.target, ouid, ogid); gfs2_quota_change(ip, ap.target, nuid, ngid); } @@ -1943,7 +2116,7 @@ const struct inode_operations gfs2_dir_iops = { .mkdir = gfs2_mkdir, .rmdir = gfs2_unlink, .mknod = gfs2_mknod, - .rename = gfs2_rename, + .rename2 = gfs2_rename2, .permission = gfs2_permission, .setattr = gfs2_setattr, .getattr = gfs2_getattr, diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 35b49f4..1e3a93f 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -756,6 +756,7 @@ static int init_journal(struct gfs2_sbd *sdp, int undo) } } + sdp->sd_log_idle = 1; set_bit(SDF_JOURNAL_CHECKED, &sdp->sd_flags); gfs2_glock_dq_uninit(&ji_gh); jindex = 0; diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index e3065cb..9b61f92 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -649,9 +649,117 @@ static void do_qc(struct gfs2_quota_data *qd, s64 change) slot_hold(qd); } + if (change < 0) /* Reset quiet flag if we freed some blocks */ + clear_bit(QDF_QMSG_QUIET, &qd->qd_flags); mutex_unlock(&sdp->sd_quota_mutex); } +static int gfs2_write_buf_to_page(struct gfs2_inode *ip, unsigned long index, + unsigned off, void *buf, unsigned bytes) +{ + struct inode *inode = &ip->i_inode; + struct gfs2_sbd *sdp = GFS2_SB(inode); + struct address_space *mapping = inode->i_mapping; + struct page *page; + struct buffer_head *bh; + void *kaddr; + u64 blk; + unsigned bsize = sdp->sd_sb.sb_bsize, bnum = 0, boff = 0; + unsigned to_write = bytes, pg_off = off; + int done = 0; + + blk = index << (PAGE_CACHE_SHIFT - sdp->sd_sb.sb_bsize_shift); + boff = off % bsize; + + page = find_or_create_page(mapping, index, GFP_NOFS); + if (!page) + return -ENOMEM; + if (!page_has_buffers(page)) + create_empty_buffers(page, bsize, 0); + + bh = page_buffers(page); + while (!done) { + /* Find the beginning block within the page */ + if (pg_off >= ((bnum * bsize) + bsize)) { + bh = bh->b_this_page; + bnum++; + blk++; + continue; + } + if (!buffer_mapped(bh)) { + gfs2_block_map(inode, blk, bh, 1); + if (!buffer_mapped(bh)) + goto unlock_out; + /* If it's a newly allocated disk block, zero it */ + if (buffer_new(bh)) + zero_user(page, bnum * bsize, bh->b_size); + } + if (PageUptodate(page)) + set_buffer_uptodate(bh); + if (!buffer_uptodate(bh)) { + ll_rw_block(READ | REQ_META, 1, &bh); + wait_on_buffer(bh); + if (!buffer_uptodate(bh)) + goto unlock_out; + } + gfs2_trans_add_data(ip->i_gl, bh); + + /* If we need to write to the next block as well */ + if (to_write > (bsize - boff)) { + pg_off += (bsize - boff); + to_write -= (bsize - boff); + boff = pg_off % bsize; + continue; + } + done = 1; + } + + /* Write to the page, now that we have setup the buffer(s) */ + kaddr = kmap_atomic(page); + memcpy(kaddr + off, buf, bytes); + flush_dcache_page(page); + kunmap_atomic(kaddr); + unlock_page(page); + page_cache_release(page); + + return 0; + +unlock_out: + unlock_page(page); + page_cache_release(page); + return -EIO; +} + +static int gfs2_write_disk_quota(struct gfs2_inode *ip, struct gfs2_quota *qp, + loff_t loc) +{ + unsigned long pg_beg; + unsigned pg_off, nbytes, overflow = 0; + int pg_oflow = 0, error; + void *ptr; + + nbytes = sizeof(struct gfs2_quota); + + pg_beg = loc >> PAGE_CACHE_SHIFT; + pg_off = loc % PAGE_CACHE_SIZE; + + /* If the quota straddles a page boundary, split the write in two */ + if ((pg_off + nbytes) > PAGE_CACHE_SIZE) { + pg_oflow = 1; + overflow = (pg_off + nbytes) - PAGE_CACHE_SIZE; + } + + ptr = qp; + error = gfs2_write_buf_to_page(ip, pg_beg, pg_off, ptr, + nbytes - overflow); + /* If there's an overflow, write the remaining bytes to the next page */ + if (!error && pg_oflow) + error = gfs2_write_buf_to_page(ip, pg_beg + 1, 0, + ptr + nbytes - overflow, + overflow); + return error; +} + /** * gfs2_adjust_quota - adjust record of current block usage * @ip: The quota inode @@ -672,15 +780,8 @@ static int gfs2_adjust_quota(struct gfs2_inode *ip, loff_t loc, { struct inode *inode = &ip->i_inode; struct gfs2_sbd *sdp = GFS2_SB(inode); - struct address_space *mapping = inode->i_mapping; - unsigned long index = loc >> PAGE_CACHE_SHIFT; - unsigned offset = loc & (PAGE_CACHE_SIZE - 1); - unsigned blocksize, iblock, pos; - struct buffer_head *bh; - struct page *page; - void *kaddr, *ptr; struct gfs2_quota q; - int err, nbytes; + int err; u64 size; if (gfs2_is_stuffed(ip)) { @@ -694,8 +795,11 @@ static int gfs2_adjust_quota(struct gfs2_inode *ip, loff_t loc, if (err < 0) return err; + loc -= sizeof(q); /* gfs2_internal_read would've advanced the loc ptr */ err = -EIO; be64_add_cpu(&q.qu_value, change); + if (((s64)be64_to_cpu(q.qu_value)) < 0) + q.qu_value = 0; /* Never go negative on quota usage */ qd->qd_qb.qb_value = q.qu_value; if (fdq) { if (fdq->d_fieldmask & QC_SPC_SOFT) { @@ -712,79 +816,16 @@ static int gfs2_adjust_quota(struct gfs2_inode *ip, loff_t loc, } } - /* Write the quota into the quota file on disk */ - ptr = &q; - nbytes = sizeof(struct gfs2_quota); -get_a_page: - page = find_or_create_page(mapping, index, GFP_NOFS); - if (!page) - return -ENOMEM; - - blocksize = inode->i_sb->s_blocksize; - iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits); - - if (!page_has_buffers(page)) - create_empty_buffers(page, blocksize, 0); - - bh = page_buffers(page); - pos = blocksize; - while (offset >= pos) { - bh = bh->b_this_page; - iblock++; - pos += blocksize; - } - - if (!buffer_mapped(bh)) { - gfs2_block_map(inode, iblock, bh, 1); - if (!buffer_mapped(bh)) - goto unlock_out; - /* If it's a newly allocated disk block for quota, zero it */ - if (buffer_new(bh)) - zero_user(page, pos - blocksize, bh->b_size); - } - - if (PageUptodate(page)) - set_buffer_uptodate(bh); - - if (!buffer_uptodate(bh)) { - ll_rw_block(READ | REQ_META, 1, &bh); - wait_on_buffer(bh); - if (!buffer_uptodate(bh)) - goto unlock_out; - } - - gfs2_trans_add_data(ip->i_gl, bh); - - kaddr = kmap_atomic(page); - if (offset + sizeof(struct gfs2_quota) > PAGE_CACHE_SIZE) - nbytes = PAGE_CACHE_SIZE - offset; - memcpy(kaddr + offset, ptr, nbytes); - flush_dcache_page(page); - kunmap_atomic(kaddr); - unlock_page(page); - page_cache_release(page); - - /* If quota straddles page boundary, we need to update the rest of the - * quota at the beginning of the next page */ - if ((offset + sizeof(struct gfs2_quota)) > PAGE_CACHE_SIZE) { - ptr = ptr + nbytes; - nbytes = sizeof(struct gfs2_quota) - nbytes; - offset = 0; - index++; - goto get_a_page; + err = gfs2_write_disk_quota(ip, &q, loc); + if (!err) { + size = loc + sizeof(struct gfs2_quota); + if (size > inode->i_size) + i_size_write(inode, size); + inode->i_mtime = inode->i_atime = CURRENT_TIME; + mark_inode_dirty(inode); + set_bit(QDF_REFRESH, &qd->qd_flags); } - size = loc + sizeof(struct gfs2_quota); - if (size > inode->i_size) - i_size_write(inode, size); - inode->i_mtime = inode->i_atime = CURRENT_TIME; - mark_inode_dirty(inode); - set_bit(QDF_REFRESH, &qd->qd_flags); - return 0; - -unlock_out: - unlock_page(page); - page_cache_release(page); return err; } @@ -1148,10 +1189,13 @@ int gfs2_quota_check(struct gfs2_inode *ip, kuid_t uid, kgid_t gid, /* If no min_target specified or we don't meet * min_target, return -EDQUOT */ if (!ap->min_target || ap->min_target > ap->allowed) { - print_message(qd, "exceeded"); - quota_send_warning(qd->qd_id, - sdp->sd_vfs->s_dev, - QUOTA_NL_BHARDWARN); + if (!test_and_set_bit(QDF_QMSG_QUIET, + &qd->qd_flags)) { + print_message(qd, "exceeded"); + quota_send_warning(qd->qd_id, + sdp->sd_vfs->s_dev, + QUOTA_NL_BHARDWARN); + } error = -EDQUOT; break; } @@ -1648,6 +1692,8 @@ static int gfs2_set_dqblk(struct super_block *sb, struct kqid qid, /* Apply changes */ error = gfs2_adjust_quota(ip, offset, 0, qd, fdq); + if (!error) + clear_bit(QDF_QMSG_QUIET, &qd->qd_flags); gfs2_trans_end(sdp); out_release: diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 6af2396..c6c6232 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -978,10 +978,10 @@ static void set_rgrp_preferences(struct gfs2_sbd *sdp) rgd->rd_flags |= GFS2_RDF_PREFERRED; for (i = 0; i < sdp->sd_journals; i++) { rgd = gfs2_rgrpd_get_next(rgd); - if (rgd == first) + if (!rgd || rgd == first) break; } - } while (rgd != first); + } while (rgd && rgd != first); } /** @@ -1244,14 +1244,13 @@ int gfs2_rgrp_go_lock(struct gfs2_holder *gh) } /** - * gfs2_rgrp_go_unlock - Release RG bitmaps read in with gfs2_rgrp_bh_get() - * @gh: The glock holder for the resource group + * gfs2_rgrp_brelse - Release RG bitmaps read in with gfs2_rgrp_bh_get() + * @rgd: The resource group * */ -void gfs2_rgrp_go_unlock(struct gfs2_holder *gh) +void gfs2_rgrp_brelse(struct gfs2_rgrpd *rgd) { - struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object; int x, length = rgd->rd_length; for (x = 0; x < length; x++) { @@ -1264,6 +1263,22 @@ void gfs2_rgrp_go_unlock(struct gfs2_holder *gh) } +/** + * gfs2_rgrp_go_unlock - Unlock a rgrp glock + * @gh: The glock holder for the resource group + * + */ + +void gfs2_rgrp_go_unlock(struct gfs2_holder *gh) +{ + struct gfs2_rgrpd *rgd = gh->gh_gl->gl_object; + int demote_requested = test_bit(GLF_DEMOTE, &gh->gh_gl->gl_flags) | + test_bit(GLF_PENDING_DEMOTE, &gh->gh_gl->gl_flags); + + if (rgd && demote_requested) + gfs2_rgrp_brelse(rgd); +} + int gfs2_rgrp_send_discards(struct gfs2_sbd *sdp, u64 offset, struct buffer_head *bh, const struct gfs2_bitmap *bi, unsigned minlen, u64 *ptrimmed) @@ -1711,10 +1726,8 @@ static int gfs2_rbm_find(struct gfs2_rbm *rbm, u8 state, u32 *minext, return ret; bitmap_full: /* Mark bitmap as full and fall through */ - if ((state == GFS2_BLKST_FREE) && initial_offset == 0) { - struct gfs2_bitmap *bi = rbm_bi(rbm); + if ((state == GFS2_BLKST_FREE) && initial_offset == 0) set_bit(GBF_FULL, &bi->bi_flags); - } next_bitmap: /* Find next bitmap in the rgrp */ rbm->offset = 0; @@ -1850,14 +1863,23 @@ static bool gfs2_rgrp_congested(const struct gfs2_rgrpd *rgd, int loops) const struct gfs2_sbd *sdp = gl->gl_sbd; struct gfs2_lkstats *st; s64 r_dcount, l_dcount; - s64 r_srttb, l_srttb; + s64 l_srttb, a_srttb = 0; s64 srttb_diff; s64 sqr_diff; s64 var; + int cpu, nonzero = 0; preempt_disable(); + for_each_present_cpu(cpu) { + st = &per_cpu_ptr(sdp->sd_lkstats, cpu)->lkstats[LM_TYPE_RGRP]; + if (st->stats[GFS2_LKS_SRTTB]) { + a_srttb += st->stats[GFS2_LKS_SRTTB]; + nonzero++; + } + } st = &this_cpu_ptr(sdp->sd_lkstats)->lkstats[LM_TYPE_RGRP]; - r_srttb = st->stats[GFS2_LKS_SRTTB]; + if (nonzero) + do_div(a_srttb, nonzero); r_dcount = st->stats[GFS2_LKS_DCOUNT]; var = st->stats[GFS2_LKS_SRTTVARB] + gl->gl_stats.stats[GFS2_LKS_SRTTVARB]; @@ -1866,10 +1888,10 @@ static bool gfs2_rgrp_congested(const struct gfs2_rgrpd *rgd, int loops) l_srttb = gl->gl_stats.stats[GFS2_LKS_SRTTB]; l_dcount = gl->gl_stats.stats[GFS2_LKS_DCOUNT]; - if ((l_dcount < 1) || (r_dcount < 1) || (r_srttb == 0)) + if ((l_dcount < 1) || (r_dcount < 1) || (a_srttb == 0)) return false; - srttb_diff = r_srttb - l_srttb; + srttb_diff = a_srttb - l_srttb; sqr_diff = srttb_diff * srttb_diff; var *= 2; diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h index 68972ec..c0ab33f 100644 --- a/fs/gfs2/rgrp.h +++ b/fs/gfs2/rgrp.h @@ -36,6 +36,7 @@ extern void gfs2_clear_rgrpd(struct gfs2_sbd *sdp); extern int gfs2_rindex_update(struct gfs2_sbd *sdp); extern void gfs2_free_clones(struct gfs2_rgrpd *rgd); extern int gfs2_rgrp_go_lock(struct gfs2_holder *gh); +extern void gfs2_rgrp_brelse(struct gfs2_rgrpd *rgd); extern void gfs2_rgrp_go_unlock(struct gfs2_holder *gh); extern struct gfs2_alloc *gfs2_alloc_get(struct gfs2_inode *ip); diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 859c6ed..2982445 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -748,7 +748,7 @@ static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc) if (wbc->sync_mode == WB_SYNC_ALL) gfs2_log_flush(GFS2_SB(inode), ip->i_gl, NORMAL_FLUSH); - if (bdi->dirty_exceeded) + if (bdi->wb.dirty_exceeded) gfs2_ail1_flush(sdp, wbc); else filemap_fdatawrite(metamapping); diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index ae8e881..c9ff1cf 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -101,8 +101,11 @@ static ssize_t freeze_show(struct gfs2_sbd *sdp, char *buf) static ssize_t freeze_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { - int error; - int n = simple_strtol(buf, NULL, 0); + int error, n; + + error = kstrtoint(buf, 0, &n); + if (error) + return error; if (!capable(CAP_SYS_ADMIN)) return -EPERM; @@ -134,10 +137,16 @@ static ssize_t withdraw_show(struct gfs2_sbd *sdp, char *buf) static ssize_t withdraw_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { + int error, val; + if (!capable(CAP_SYS_ADMIN)) return -EPERM; - if (simple_strtol(buf, NULL, 0) != 1) + error = kstrtoint(buf, 0, &val); + if (error) + return error; + + if (val != 1) return -EINVAL; gfs2_lm_withdraw(sdp, "withdrawing from cluster at user's request\n"); @@ -148,10 +157,16 @@ static ssize_t withdraw_store(struct gfs2_sbd *sdp, const char *buf, size_t len) static ssize_t statfs_sync_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { + int error, val; + if (!capable(CAP_SYS_ADMIN)) return -EPERM; - if (simple_strtol(buf, NULL, 0) != 1) + error = kstrtoint(buf, 0, &val); + if (error) + return error; + + if (val != 1) return -EINVAL; gfs2_statfs_sync(sdp->sd_vfs, 0); @@ -161,10 +176,16 @@ static ssize_t statfs_sync_store(struct gfs2_sbd *sdp, const char *buf, static ssize_t quota_sync_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { + int error, val; + if (!capable(CAP_SYS_ADMIN)) return -EPERM; - if (simple_strtol(buf, NULL, 0) != 1) + error = kstrtoint(buf, 0, &val); + if (error) + return error; + + if (val != 1) return -EINVAL; gfs2_quota_sync(sdp->sd_vfs, 0); @@ -181,7 +202,9 @@ static ssize_t quota_refresh_user_store(struct gfs2_sbd *sdp, const char *buf, if (!capable(CAP_SYS_ADMIN)) return -EPERM; - id = simple_strtoul(buf, NULL, 0); + error = kstrtou32(buf, 0, &id); + if (error) + return error; qid = make_kqid(current_user_ns(), USRQUOTA, id); if (!qid_valid(qid)) @@ -201,7 +224,9 @@ static ssize_t quota_refresh_group_store(struct gfs2_sbd *sdp, const char *buf, if (!capable(CAP_SYS_ADMIN)) return -EPERM; - id = simple_strtoul(buf, NULL, 0); + error = kstrtou32(buf, 0, &id); + if (error) + return error; qid = make_kqid(current_user_ns(), GRPQUOTA, id); if (!qid_valid(qid)) @@ -324,10 +349,11 @@ static ssize_t block_show(struct gfs2_sbd *sdp, char *buf) static ssize_t block_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { struct lm_lockstruct *ls = &sdp->sd_lockstruct; - ssize_t ret = len; - int val; + int ret, val; - val = simple_strtol(buf, NULL, 0); + ret = kstrtoint(buf, 0, &val); + if (ret) + return ret; if (val == 1) set_bit(DFL_BLOCK_LOCKS, &ls->ls_recover_flags); @@ -336,9 +362,9 @@ static ssize_t block_store(struct gfs2_sbd *sdp, const char *buf, size_t len) smp_mb__after_atomic(); gfs2_glock_thaw(sdp); } else { - ret = -EINVAL; + return -EINVAL; } - return ret; + return len; } static ssize_t wdack_show(struct gfs2_sbd *sdp, char *buf) @@ -350,17 +376,18 @@ static ssize_t wdack_show(struct gfs2_sbd *sdp, char *buf) static ssize_t wdack_store(struct gfs2_sbd *sdp, const char *buf, size_t len) { - ssize_t ret = len; - int val; + int ret, val; - val = simple_strtol(buf, NULL, 0); + ret = kstrtoint(buf, 0, &val); + if (ret) + return ret; if ((val == 1) && !strcmp(sdp->sd_lockstruct.ls_ops->lm_proto_name, "lock_dlm")) complete(&sdp->sd_wdack); else - ret = -EINVAL; - return ret; + return -EINVAL; + return len; } static ssize_t lkfirst_show(struct gfs2_sbd *sdp, char *buf) @@ -553,11 +580,14 @@ static ssize_t tune_set(struct gfs2_sbd *sdp, unsigned int *field, { struct gfs2_tune *gt = &sdp->sd_tune; unsigned int x; + int error; if (!capable(CAP_SYS_ADMIN)) return -EPERM; - x = simple_strtoul(buf, NULL, 0); + error = kstrtouint(buf, 0, &x); + if (error) + return error; if (check_zero && !x) return -EINVAL; diff --git a/fs/hfs/super.c b/fs/hfs/super.c index eee7206..55c03b9 100644 --- a/fs/hfs/super.c +++ b/fs/hfs/super.c @@ -14,6 +14,7 @@ #include <linux/module.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <linux/mount.h> #include <linux/init.h> #include <linux/nls.h> diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 593af2f..7302d96 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -11,6 +11,7 @@ #include <linux/init.h> #include <linux/pagemap.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <linux/fs.h> #include <linux/slab.h> #include <linux/vfs.h> diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index 4a437ab..059597b 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -581,7 +581,7 @@ static int hostfs_create(struct inode *dir, struct dentry *dentry, umode_t mode, if (name == NULL) goto out_put; - fd = file_create(name, mode & S_IFMT); + fd = file_create(name, mode & 0777); if (fd < 0) error = fd; else diff --git a/fs/hppfs/Makefile b/fs/hppfs/Makefile deleted file mode 100644 index 3a982bd..0000000 --- a/fs/hppfs/Makefile +++ /dev/null @@ -1,6 +0,0 @@ -# -# Copyright (C) 2002 - 2008 Jeff Dike (jdike@{addtoit,linux.intel}.com) -# Licensed under the GPL -# - -obj-$(CONFIG_HPPFS) += hppfs.o diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c deleted file mode 100644 index 2867837..0000000 --- a/fs/hppfs/hppfs.c +++ /dev/null @@ -1,765 +0,0 @@ -/* - * Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) - * Licensed under the GPL - */ - -#include <linux/ctype.h> -#include <linux/dcache.h> -#include <linux/file.h> -#include <linux/fs.h> -#include <linux/init.h> -#include <linux/kernel.h> -#include <linux/list.h> -#include <linux/module.h> -#include <linux/mount.h> -#include <linux/slab.h> -#include <linux/statfs.h> -#include <linux/types.h> -#include <linux/pid_namespace.h> -#include <linux/namei.h> -#include <asm/uaccess.h> -#include <os.h> - -static struct inode *get_inode(struct super_block *, struct dentry *); - -struct hppfs_data { - struct list_head list; - char contents[PAGE_SIZE - sizeof(struct list_head)]; -}; - -struct hppfs_private { - struct file *proc_file; - int host_fd; - loff_t len; - struct hppfs_data *contents; -}; - -struct hppfs_inode_info { - struct dentry *proc_dentry; - struct inode vfs_inode; -}; - -static inline struct hppfs_inode_info *HPPFS_I(struct inode *inode) -{ - return container_of(inode, struct hppfs_inode_info, vfs_inode); -} - -#define HPPFS_SUPER_MAGIC 0xb00000ee - -static const struct super_operations hppfs_sbops; - -static int is_pid(struct dentry *dentry) -{ - struct super_block *sb; - int i; - - sb = dentry->d_sb; - if (dentry->d_parent != sb->s_root) - return 0; - - for (i = 0; i < dentry->d_name.len; i++) { - if (!isdigit(dentry->d_name.name[i])) - return 0; - } - return 1; -} - -static char *dentry_name(struct dentry *dentry, int extra) -{ - struct dentry *parent; - char *root, *name; - const char *seg_name; - int len, seg_len, root_len; - - len = 0; - parent = dentry; - while (parent->d_parent != parent) { - if (is_pid(parent)) - len += strlen("pid") + 1; - else len += parent->d_name.len + 1; - parent = parent->d_parent; - } - - root = "proc"; - root_len = strlen(root); - len += root_len; - name = kmalloc(len + extra + 1, GFP_KERNEL); - if (name == NULL) - return NULL; - - name[len] = '\0'; - parent = dentry; - while (parent->d_parent != parent) { - if (is_pid(parent)) { - seg_name = "pid"; - seg_len = strlen(seg_name); - } - else { - seg_name = parent->d_name.name; - seg_len = parent->d_name.len; - } - - len -= seg_len + 1; - name[len] = '/'; - memcpy(&name[len + 1], seg_name, seg_len); - parent = parent->d_parent; - } - memcpy(name, root, root_len); - return name; -} - -static int file_removed(struct dentry *dentry, const char *file) -{ - char *host_file; - int extra, fd; - - extra = 0; - if (file != NULL) - extra += strlen(file) + 1; - - host_file = dentry_name(dentry, extra + strlen("/remove")); - if (host_file == NULL) { - printk(KERN_ERR "file_removed : allocation failed\n"); - return -ENOMEM; - } - - if (file != NULL) { - strcat(host_file, "/"); - strcat(host_file, file); - } - strcat(host_file, "/remove"); - - fd = os_open_file(host_file, of_read(OPENFLAGS()), 0); - kfree(host_file); - if (fd > 0) { - os_close_file(fd); - return 1; - } - return 0; -} - -static struct dentry *hppfs_lookup(struct inode *ino, struct dentry *dentry, - unsigned int flags) -{ - struct dentry *proc_dentry, *parent; - struct qstr *name = &dentry->d_name; - struct inode *inode; - int err, deleted; - - deleted = file_removed(dentry, NULL); - if (deleted < 0) - return ERR_PTR(deleted); - else if (deleted) - return ERR_PTR(-ENOENT); - - parent = HPPFS_I(ino)->proc_dentry; - mutex_lock(&d_inode(parent)->i_mutex); - proc_dentry = lookup_one_len(name->name, parent, name->len); - mutex_unlock(&d_inode(parent)->i_mutex); - - if (IS_ERR(proc_dentry)) - return proc_dentry; - - err = -ENOMEM; - inode = get_inode(ino->i_sb, proc_dentry); - if (!inode) - goto out; - - d_add(dentry, inode); - return NULL; - - out: - return ERR_PTR(err); -} - -static const struct inode_operations hppfs_file_iops = { -}; - -static ssize_t read_proc(struct file *file, char __user *buf, ssize_t count, - loff_t *ppos, int is_user) -{ - ssize_t (*read)(struct file *, char __user *, size_t, loff_t *); - ssize_t n; - - read = file_inode(file)->i_fop->read; - - if (!is_user) - set_fs(KERNEL_DS); - - n = (*read)(file, buf, count, &file->f_pos); - - if (!is_user) - set_fs(USER_DS); - - if (ppos) - *ppos = file->f_pos; - return n; -} - -static ssize_t hppfs_read_file(int fd, char __user *buf, ssize_t count) -{ - ssize_t n; - int cur, err; - char *new_buf; - - n = -ENOMEM; - new_buf = kmalloc(PAGE_SIZE, GFP_KERNEL); - if (new_buf == NULL) { - printk(KERN_ERR "hppfs_read_file : kmalloc failed\n"); - goto out; - } - n = 0; - while (count > 0) { - cur = min_t(ssize_t, count, PAGE_SIZE); - err = os_read_file(fd, new_buf, cur); - if (err < 0) { - printk(KERN_ERR "hppfs_read : read failed, " - "errno = %d\n", err); - n = err; - goto out_free; - } else if (err == 0) - break; - - if (copy_to_user(buf, new_buf, err)) { - n = -EFAULT; - goto out_free; - } - n += err; - count -= err; - } - out_free: - kfree(new_buf); - out: - return n; -} - -static ssize_t hppfs_read(struct file *file, char __user *buf, size_t count, - loff_t *ppos) -{ - struct hppfs_private *hppfs = file->private_data; - struct hppfs_data *data; - loff_t off; - int err; - - if (hppfs->contents != NULL) { - int rem; - - if (*ppos >= hppfs->len) - return 0; - - data = hppfs->contents; - off = *ppos; - while (off >= sizeof(data->contents)) { - data = list_entry(data->list.next, struct hppfs_data, - list); - off -= sizeof(data->contents); - } - - if (off + count > hppfs->len) - count = hppfs->len - off; - rem = copy_to_user(buf, &data->contents[off], count); - *ppos += count - rem; - if (rem > 0) - return -EFAULT; - } else if (hppfs->host_fd != -1) { - err = os_seek_file(hppfs->host_fd, *ppos); - if (err) { - printk(KERN_ERR "hppfs_read : seek failed, " - "errno = %d\n", err); - return err; - } - err = hppfs_read_file(hppfs->host_fd, buf, count); - if (err < 0) { - printk(KERN_ERR "hppfs_read: read failed: %d\n", err); - return err; - } - count = err; - if (count > 0) - *ppos += count; - } - else count = read_proc(hppfs->proc_file, buf, count, ppos, 1); - - return count; -} - -static ssize_t hppfs_write(struct file *file, const char __user *buf, - size_t len, loff_t *ppos) -{ - struct hppfs_private *data = file->private_data; - struct file *proc_file = data->proc_file; - ssize_t (*write)(struct file *, const char __user *, size_t, loff_t *); - - write = file_inode(proc_file)->i_fop->write; - return (*write)(proc_file, buf, len, ppos); -} - -static int open_host_sock(char *host_file, int *filter_out) -{ - char *end; - int fd; - - end = &host_file[strlen(host_file)]; - strcpy(end, "/rw"); - *filter_out = 1; - fd = os_connect_socket(host_file); - if (fd > 0) - return fd; - - strcpy(end, "/r"); - *filter_out = 0; - fd = os_connect_socket(host_file); - return fd; -} - -static void free_contents(struct hppfs_data *head) -{ - struct hppfs_data *data; - struct list_head *ele, *next; - - if (head == NULL) - return; - - list_for_each_safe(ele, next, &head->list) { - data = list_entry(ele, struct hppfs_data, list); - kfree(data); - } - kfree(head); -} - -static struct hppfs_data *hppfs_get_data(int fd, int filter, - struct file *proc_file, - struct file *hppfs_file, - loff_t *size_out) -{ - struct hppfs_data *data, *new, *head; - int n, err; - - err = -ENOMEM; - data = kmalloc(sizeof(*data), GFP_KERNEL); - if (data == NULL) { - printk(KERN_ERR "hppfs_get_data : head allocation failed\n"); - goto failed; - } - - INIT_LIST_HEAD(&data->list); - - head = data; - *size_out = 0; - - if (filter) { - while ((n = read_proc(proc_file, data->contents, - sizeof(data->contents), NULL, 0)) > 0) - os_write_file(fd, data->contents, n); - err = os_shutdown_socket(fd, 0, 1); - if (err) { - printk(KERN_ERR "hppfs_get_data : failed to shut down " - "socket\n"); - goto failed_free; - } - } - while (1) { - n = os_read_file(fd, data->contents, sizeof(data->contents)); - if (n < 0) { - err = n; - printk(KERN_ERR "hppfs_get_data : read failed, " - "errno = %d\n", err); - goto failed_free; - } else if (n == 0) - break; - - *size_out += n; - - if (n < sizeof(data->contents)) - break; - - new = kmalloc(sizeof(*data), GFP_KERNEL); - if (new == 0) { - printk(KERN_ERR "hppfs_get_data : data allocation " - "failed\n"); - err = -ENOMEM; - goto failed_free; - } - - INIT_LIST_HEAD(&new->list); - list_add(&new->list, &data->list); - data = new; - } - return head; - - failed_free: - free_contents(head); - failed: - return ERR_PTR(err); -} - -static struct hppfs_private *hppfs_data(void) -{ - struct hppfs_private *data; - - data = kmalloc(sizeof(*data), GFP_KERNEL); - if (data == NULL) - return data; - - *data = ((struct hppfs_private ) { .host_fd = -1, - .len = -1, - .contents = NULL } ); - return data; -} - -static int file_mode(int fmode) -{ - if (fmode == (FMODE_READ | FMODE_WRITE)) - return O_RDWR; - if (fmode == FMODE_READ) - return O_RDONLY; - if (fmode == FMODE_WRITE) - return O_WRONLY; - return 0; -} - -static int hppfs_open(struct inode *inode, struct file *file) -{ - const struct cred *cred = file->f_cred; - struct hppfs_private *data; - struct path path; - char *host_file; - int err, fd, type, filter; - - err = -ENOMEM; - data = hppfs_data(); - if (data == NULL) - goto out; - - host_file = dentry_name(file->f_path.dentry, strlen("/rw")); - if (host_file == NULL) - goto out_free2; - - path.mnt = inode->i_sb->s_fs_info; - path.dentry = HPPFS_I(inode)->proc_dentry; - - data->proc_file = dentry_open(&path, file_mode(file->f_mode), cred); - err = PTR_ERR(data->proc_file); - if (IS_ERR(data->proc_file)) - goto out_free1; - - type = os_file_type(host_file); - if (type == OS_TYPE_FILE) { - fd = os_open_file(host_file, of_read(OPENFLAGS()), 0); - if (fd >= 0) - data->host_fd = fd; - else - printk(KERN_ERR "hppfs_open : failed to open '%s', " - "errno = %d\n", host_file, -fd); - - data->contents = NULL; - } else if (type == OS_TYPE_DIR) { - fd = open_host_sock(host_file, &filter); - if (fd > 0) { - data->contents = hppfs_get_data(fd, filter, - data->proc_file, - file, &data->len); - if (!IS_ERR(data->contents)) - data->host_fd = fd; - } else - printk(KERN_ERR "hppfs_open : failed to open a socket " - "in '%s', errno = %d\n", host_file, -fd); - } - kfree(host_file); - - file->private_data = data; - return 0; - - out_free1: - kfree(host_file); - out_free2: - free_contents(data->contents); - kfree(data); - out: - return err; -} - -static int hppfs_dir_open(struct inode *inode, struct file *file) -{ - const struct cred *cred = file->f_cred; - struct hppfs_private *data; - struct path path; - int err; - - err = -ENOMEM; - data = hppfs_data(); - if (data == NULL) - goto out; - - path.mnt = inode->i_sb->s_fs_info; - path.dentry = HPPFS_I(inode)->proc_dentry; - data->proc_file = dentry_open(&path, file_mode(file->f_mode), cred); - err = PTR_ERR(data->proc_file); - if (IS_ERR(data->proc_file)) - goto out_free; - - file->private_data = data; - return 0; - - out_free: - kfree(data); - out: - return err; -} - -static loff_t hppfs_llseek(struct file *file, loff_t off, int where) -{ - struct hppfs_private *data = file->private_data; - struct file *proc_file = data->proc_file; - loff_t (*llseek)(struct file *, loff_t, int); - loff_t ret; - - llseek = file_inode(proc_file)->i_fop->llseek; - if (llseek != NULL) { - ret = (*llseek)(proc_file, off, where); - if (ret < 0) - return ret; - } - - return default_llseek(file, off, where); -} - -static int hppfs_release(struct inode *inode, struct file *file) -{ - struct hppfs_private *data = file->private_data; - struct file *proc_file = data->proc_file; - if (proc_file) - fput(proc_file); - kfree(data); - return 0; -} - -static const struct file_operations hppfs_file_fops = { - .owner = NULL, - .llseek = hppfs_llseek, - .read = hppfs_read, - .write = hppfs_write, - .open = hppfs_open, - .release = hppfs_release, -}; - -struct hppfs_dirent { - struct dir_context ctx; - struct dir_context *caller; - struct dentry *dentry; -}; - -static int hppfs_filldir(struct dir_context *ctx, const char *name, int size, - loff_t offset, u64 inode, unsigned int type) -{ - struct hppfs_dirent *dirent = - container_of(ctx, struct hppfs_dirent, ctx); - - if (file_removed(dirent->dentry, name)) - return 0; - - dirent->caller->pos = dirent->ctx.pos; - return !dir_emit(dirent->caller, name, size, inode, type); -} - -static int hppfs_readdir(struct file *file, struct dir_context *ctx) -{ - struct hppfs_private *data = file->private_data; - struct file *proc_file = data->proc_file; - struct hppfs_dirent d = { - .ctx.actor = hppfs_filldir, - .caller = ctx, - .dentry = file->f_path.dentry - }; - int err; - proc_file->f_pos = ctx->pos; - err = iterate_dir(proc_file, &d.ctx); - ctx->pos = d.ctx.pos; - return err; -} - -static const struct file_operations hppfs_dir_fops = { - .owner = NULL, - .iterate = hppfs_readdir, - .open = hppfs_dir_open, - .llseek = default_llseek, - .release = hppfs_release, -}; - -static int hppfs_statfs(struct dentry *dentry, struct kstatfs *sf) -{ - sf->f_blocks = 0; - sf->f_bfree = 0; - sf->f_bavail = 0; - sf->f_files = 0; - sf->f_ffree = 0; - sf->f_type = HPPFS_SUPER_MAGIC; - return 0; -} - -static struct inode *hppfs_alloc_inode(struct super_block *sb) -{ - struct hppfs_inode_info *hi; - - hi = kmalloc(sizeof(*hi), GFP_KERNEL); - if (!hi) - return NULL; - - hi->proc_dentry = NULL; - inode_init_once(&hi->vfs_inode); - return &hi->vfs_inode; -} - -void hppfs_evict_inode(struct inode *ino) -{ - clear_inode(ino); - dput(HPPFS_I(ino)->proc_dentry); - mntput(ino->i_sb->s_fs_info); -} - -static void hppfs_i_callback(struct rcu_head *head) -{ - struct inode *inode = container_of(head, struct inode, i_rcu); - kfree(HPPFS_I(inode)); -} - -static void hppfs_destroy_inode(struct inode *inode) -{ - call_rcu(&inode->i_rcu, hppfs_i_callback); -} - -static const struct super_operations hppfs_sbops = { - .alloc_inode = hppfs_alloc_inode, - .destroy_inode = hppfs_destroy_inode, - .evict_inode = hppfs_evict_inode, - .statfs = hppfs_statfs, -}; - -static int hppfs_readlink(struct dentry *dentry, char __user *buffer, - int buflen) -{ - struct dentry *proc_dentry = HPPFS_I(d_inode(dentry))->proc_dentry; - return d_inode(proc_dentry)->i_op->readlink(proc_dentry, buffer, - buflen); -} - -static const char *hppfs_follow_link(struct dentry *dentry, void **cookie) -{ - struct dentry *proc_dentry = HPPFS_I(d_inode(dentry))->proc_dentry; - - return d_inode(proc_dentry)->i_op->follow_link(proc_dentry, cookie); -} - -static void hppfs_put_link(struct inode *inode, void *cookie) -{ - struct inode *proc_inode = d_inode(HPPFS_I(inode)->proc_dentry); - - if (proc_inode->i_op->put_link) - proc_inode->i_op->put_link(proc_inode, cookie); -} - -static const struct inode_operations hppfs_dir_iops = { - .lookup = hppfs_lookup, -}; - -static const struct inode_operations hppfs_link_iops = { - .readlink = hppfs_readlink, - .follow_link = hppfs_follow_link, - .put_link = hppfs_put_link, -}; - -static struct inode *get_inode(struct super_block *sb, struct dentry *dentry) -{ - struct inode *proc_ino = d_inode(dentry); - struct inode *inode = new_inode(sb); - - if (!inode) { - dput(dentry); - return NULL; - } - - if (d_is_dir(dentry)) { - inode->i_op = &hppfs_dir_iops; - inode->i_fop = &hppfs_dir_fops; - } else if (d_is_symlink(dentry)) { - inode->i_op = &hppfs_link_iops; - inode->i_fop = &hppfs_file_fops; - } else { - inode->i_op = &hppfs_file_iops; - inode->i_fop = &hppfs_file_fops; - } - - HPPFS_I(inode)->proc_dentry = dentry; - - inode->i_uid = proc_ino->i_uid; - inode->i_gid = proc_ino->i_gid; - inode->i_atime = proc_ino->i_atime; - inode->i_mtime = proc_ino->i_mtime; - inode->i_ctime = proc_ino->i_ctime; - inode->i_ino = proc_ino->i_ino; - inode->i_mode = proc_ino->i_mode; - set_nlink(inode, proc_ino->i_nlink); - inode->i_size = proc_ino->i_size; - inode->i_blocks = proc_ino->i_blocks; - - return inode; -} - -static int hppfs_fill_super(struct super_block *sb, void *d, int silent) -{ - struct inode *root_inode; - struct vfsmount *proc_mnt; - int err = -ENOENT; - - proc_mnt = mntget(task_active_pid_ns(current)->proc_mnt); - if (IS_ERR(proc_mnt)) - goto out; - - sb->s_blocksize = 1024; - sb->s_blocksize_bits = 10; - sb->s_magic = HPPFS_SUPER_MAGIC; - sb->s_op = &hppfs_sbops; - sb->s_fs_info = proc_mnt; - - err = -ENOMEM; - root_inode = get_inode(sb, dget(proc_mnt->mnt_root)); - sb->s_root = d_make_root(root_inode); - if (!sb->s_root) - goto out_mntput; - - return 0; - - out_mntput: - mntput(proc_mnt); - out: - return(err); -} - -static struct dentry *hppfs_read_super(struct file_system_type *type, - int flags, const char *dev_name, - void *data) -{ - return mount_nodev(type, flags, data, hppfs_fill_super); -} - -static struct file_system_type hppfs_type = { - .owner = THIS_MODULE, - .name = "hppfs", - .mount = hppfs_read_super, - .kill_sb = kill_anon_super, - .fs_flags = 0, -}; -MODULE_ALIAS_FS("hppfs"); - -static int __init init_hppfs(void) -{ - return register_filesystem(&hppfs_type); -} - -static void __exit exit_hppfs(void) -{ - unregister_filesystem(&hppfs_type); -} - -module_init(init_hppfs) -module_exit(exit_hppfs) -MODULE_LICENSE("GPL"); diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 87724c1..0cf74df 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -130,7 +130,6 @@ static int hugetlbfs_file_mmap(struct file *file, struct vm_area_struct *vma) goto out; ret = 0; - hugetlb_prefault_arch_hook(vma->vm_mm); if (vma->vm_flags & VM_WRITE && inode->i_size < len) inode->i_size = len; out: @@ -224,6 +224,7 @@ EXPORT_SYMBOL(free_inode_nonrcu); void __destroy_inode(struct inode *inode) { BUG_ON(inode_has_buffers(inode)); + inode_detach_wb(inode); security_inode_free(inode); fsnotify_inode_delete(inode); locks_free_lock_context(inode->i_flctx); diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 988b32e..4227dc4 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -390,7 +390,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal) unsigned long blocknr; if (is_journal_aborted(journal)) - return 1; + return -EIO; if (!jbd2_journal_get_log_tail(journal, &first_tid, &blocknr)) return 1; @@ -405,10 +405,9 @@ int jbd2_cleanup_journal_tail(journal_t *journal) * jbd2_cleanup_journal_tail() doesn't get called all that often. */ if (journal->j_flags & JBD2_BARRIER) - blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL); + blkdev_issue_flush(journal->j_fs_dev, GFP_NOFS, NULL); - __jbd2_update_log_tail(journal, first_tid, blocknr); - return 0; + return __jbd2_update_log_tail(journal, first_tid, blocknr); } diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index b96bd80..4ff3fad 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -371,16 +371,7 @@ int jbd2_journal_write_metadata_buffer(transaction_t *transaction, */ J_ASSERT_BH(bh_in, buffer_jbddirty(bh_in)); -retry_alloc: - new_bh = alloc_buffer_head(GFP_NOFS); - if (!new_bh) { - /* - * Failure is not an option, but __GFP_NOFAIL is going - * away; so we retry ourselves here. - */ - congestion_wait(BLK_RW_ASYNC, HZ/50); - goto retry_alloc; - } + new_bh = alloc_buffer_head(GFP_NOFS|__GFP_NOFAIL); /* keep subsequent assertions sane */ atomic_set(&new_bh->b_count, 1); @@ -885,9 +876,10 @@ int jbd2_journal_get_log_tail(journal_t *journal, tid_t *tid, * * Requires j_checkpoint_mutex */ -void __jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block) +int __jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block) { unsigned long freed; + int ret; BUG_ON(!mutex_is_locked(&journal->j_checkpoint_mutex)); @@ -897,7 +889,10 @@ void __jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block) * space and if we lose sb update during power failure we'd replay * old transaction with possibly newly overwritten data. */ - jbd2_journal_update_sb_log_tail(journal, tid, block, WRITE_FUA); + ret = jbd2_journal_update_sb_log_tail(journal, tid, block, WRITE_FUA); + if (ret) + goto out; + write_lock(&journal->j_state_lock); freed = block - journal->j_tail; if (block < journal->j_tail) @@ -913,6 +908,9 @@ void __jbd2_update_log_tail(journal_t *journal, tid_t tid, unsigned long block) journal->j_tail_sequence = tid; journal->j_tail = block; write_unlock(&journal->j_state_lock); + +out: + return ret; } /* @@ -1137,7 +1135,6 @@ journal_t * jbd2_journal_init_dev(struct block_device *bdev, { journal_t *journal = journal_init_common(); struct buffer_head *bh; - char *p; int n; if (!journal) @@ -1150,9 +1147,7 @@ journal_t * jbd2_journal_init_dev(struct block_device *bdev, journal->j_blk_offset = start; journal->j_maxlen = len; bdevname(journal->j_dev, journal->j_devname); - p = journal->j_devname; - while ((p = strchr(p, '/'))) - *p = '!'; + strreplace(journal->j_devname, '/', '!'); jbd2_stats_proc_init(journal); n = journal->j_blocksize / sizeof(journal_block_tag_t); journal->j_wbufsize = n; @@ -1204,10 +1199,7 @@ journal_t * jbd2_journal_init_inode (struct inode *inode) journal->j_dev = journal->j_fs_dev = inode->i_sb->s_bdev; journal->j_inode = inode; bdevname(journal->j_dev, journal->j_devname); - p = journal->j_devname; - while ((p = strchr(p, '/'))) - *p = '!'; - p = journal->j_devname + strlen(journal->j_devname); + p = strreplace(journal->j_devname, '/', '!'); sprintf(p, "-%lu", journal->j_inode->i_ino); jbd_debug(1, "journal %p: inode %s/%ld, size %Ld, bits %d, blksize %ld\n", @@ -1331,7 +1323,7 @@ static int journal_reset(journal_t *journal) return jbd2_journal_start_thread(journal); } -static void jbd2_write_superblock(journal_t *journal, int write_op) +static int jbd2_write_superblock(journal_t *journal, int write_op) { struct buffer_head *bh = journal->j_sb_buffer; journal_superblock_t *sb = journal->j_superblock; @@ -1370,7 +1362,10 @@ static void jbd2_write_superblock(journal_t *journal, int write_op) printk(KERN_ERR "JBD2: Error %d detected when updating " "journal superblock for %s.\n", ret, journal->j_devname); + jbd2_journal_abort(journal, ret); } + + return ret; } /** @@ -1383,10 +1378,11 @@ static void jbd2_write_superblock(journal_t *journal, int write_op) * Update a journal's superblock information about log tail and write it to * disk, waiting for the IO to complete. */ -void jbd2_journal_update_sb_log_tail(journal_t *journal, tid_t tail_tid, +int jbd2_journal_update_sb_log_tail(journal_t *journal, tid_t tail_tid, unsigned long tail_block, int write_op) { journal_superblock_t *sb = journal->j_superblock; + int ret; BUG_ON(!mutex_is_locked(&journal->j_checkpoint_mutex)); jbd_debug(1, "JBD2: updating superblock (start %lu, seq %u)\n", @@ -1395,13 +1391,18 @@ void jbd2_journal_update_sb_log_tail(journal_t *journal, tid_t tail_tid, sb->s_sequence = cpu_to_be32(tail_tid); sb->s_start = cpu_to_be32(tail_block); - jbd2_write_superblock(journal, write_op); + ret = jbd2_write_superblock(journal, write_op); + if (ret) + goto out; /* Log is no longer empty */ write_lock(&journal->j_state_lock); WARN_ON(!sb->s_sequence); journal->j_flags &= ~JBD2_FLUSHED; write_unlock(&journal->j_state_lock); + +out: + return ret; } /** @@ -1950,7 +1951,14 @@ int jbd2_journal_flush(journal_t *journal) return -EIO; mutex_lock(&journal->j_checkpoint_mutex); - jbd2_cleanup_journal_tail(journal); + if (!err) { + err = jbd2_cleanup_journal_tail(journal); + if (err < 0) { + mutex_unlock(&journal->j_checkpoint_mutex); + goto out; + } + err = 0; + } /* Finally, mark the journal as really needing no recovery. * This sets s_start==0 in the underlying superblock, which is @@ -1966,7 +1974,8 @@ int jbd2_journal_flush(journal_t *journal) J_ASSERT(journal->j_head == journal->j_tail); J_ASSERT(journal->j_tail_sequence == journal->j_transaction_sequence); write_unlock(&journal->j_state_lock); - return 0; +out: + return err; } /** @@ -2330,7 +2339,7 @@ static int jbd2_journal_init_journal_head_cache(void) jbd2_journal_head_cache = kmem_cache_create("jbd2_journal_head", sizeof(struct journal_head), 0, /* offset */ - SLAB_TEMPORARY, /* flags */ + SLAB_TEMPORARY | SLAB_DESTROY_BY_RCU, NULL); /* ctor */ retval = 0; if (!jbd2_journal_head_cache) { @@ -2362,10 +2371,8 @@ static struct journal_head *journal_alloc_journal_head(void) if (!ret) { jbd_debug(1, "out of memory for journal_head\n"); pr_notice_ratelimited("ENOMEM in %s, retrying.\n", __func__); - while (!ret) { - yield(); - ret = kmem_cache_zalloc(jbd2_journal_head_cache, GFP_NOFS); - } + ret = kmem_cache_zalloc(jbd2_journal_head_cache, + GFP_NOFS | __GFP_NOFAIL); } return ret; } diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c index b5128c6..a9079d0 100644 --- a/fs/jbd2/recovery.c +++ b/fs/jbd2/recovery.c @@ -842,15 +842,23 @@ static int scan_revoke_records(journal_t *journal, struct buffer_head *bh, { jbd2_journal_revoke_header_t *header; int offset, max; + int csum_size = 0; + __u32 rcount; int record_len = 4; header = (jbd2_journal_revoke_header_t *) bh->b_data; offset = sizeof(jbd2_journal_revoke_header_t); - max = be32_to_cpu(header->r_count); + rcount = be32_to_cpu(header->r_count); if (!jbd2_revoke_block_csum_verify(journal, header)) return -EINVAL; + if (jbd2_journal_has_csum_v2or3(journal)) + csum_size = sizeof(struct jbd2_journal_revoke_tail); + if (rcount > journal->j_blocksize - csum_size) + return -EINVAL; + max = rcount; + if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT)) record_len = 8; diff --git a/fs/jbd2/revoke.c b/fs/jbd2/revoke.c index c6cbaef..0abf2e7f7 100644 --- a/fs/jbd2/revoke.c +++ b/fs/jbd2/revoke.c @@ -141,11 +141,13 @@ static int insert_revoke_hash(journal_t *journal, unsigned long long blocknr, { struct list_head *hash_list; struct jbd2_revoke_record_s *record; + gfp_t gfp_mask = GFP_NOFS; -repeat: - record = kmem_cache_alloc(jbd2_revoke_record_cache, GFP_NOFS); + if (journal_oom_retry) + gfp_mask |= __GFP_NOFAIL; + record = kmem_cache_alloc(jbd2_revoke_record_cache, gfp_mask); if (!record) - goto oom; + return -ENOMEM; record->sequence = seq; record->blocknr = blocknr; @@ -154,13 +156,6 @@ repeat: list_add(&record->hash, hash_list); spin_unlock(&journal->j_revoke_lock); return 0; - -oom: - if (!journal_oom_retry) - return -ENOMEM; - jbd_debug(1, "ENOMEM in %s, retrying\n", __func__); - yield(); - goto repeat; } /* Find a revoke record in the journal's hash table. */ @@ -577,7 +572,7 @@ static void write_one_revoke_record(journal_t *journal, { int csum_size = 0; struct buffer_head *descriptor; - int offset; + int sz, offset; journal_header_t *header; /* If we are already aborting, this all becomes a noop. We @@ -594,9 +589,14 @@ static void write_one_revoke_record(journal_t *journal, if (jbd2_journal_has_csum_v2or3(journal)) csum_size = sizeof(struct jbd2_journal_revoke_tail); + if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT)) + sz = 8; + else + sz = 4; + /* Make sure we have a descriptor with space left for the record */ if (descriptor) { - if (offset >= journal->j_blocksize - csum_size) { + if (offset + sz > journal->j_blocksize - csum_size) { flush_descriptor(journal, descriptor, offset, write_op); descriptor = NULL; } @@ -619,16 +619,13 @@ static void write_one_revoke_record(journal_t *journal, *descriptorp = descriptor; } - if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT)) { + if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_64BIT)) * ((__be64 *)(&descriptor->b_data[offset])) = cpu_to_be64(record->blocknr); - offset += 8; - - } else { + else * ((__be32 *)(&descriptor->b_data[offset])) = cpu_to_be32(record->blocknr); - offset += 4; - } + offset += sz; *offsetp = offset; } diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index 5f09370..f3d0617 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -278,22 +278,16 @@ static int start_this_handle(journal_t *journal, handle_t *handle, alloc_transaction: if (!journal->j_running_transaction) { + /* + * If __GFP_FS is not present, then we may be being called from + * inside the fs writeback layer, so we MUST NOT fail. + */ + if ((gfp_mask & __GFP_FS) == 0) + gfp_mask |= __GFP_NOFAIL; new_transaction = kmem_cache_zalloc(transaction_cache, gfp_mask); - if (!new_transaction) { - /* - * If __GFP_FS is not present, then we may be - * being called from inside the fs writeback - * layer, so we MUST NOT fail. Since - * __GFP_NOFAIL is going away, we will arrange - * to retry the allocation ourselves. - */ - if ((gfp_mask & __GFP_FS) == 0) { - congestion_wait(BLK_RW_ASYNC, HZ/50); - goto alloc_transaction; - } + if (!new_transaction) return -ENOMEM; - } } jbd_debug(3, "New handle %p going live.\n", handle); @@ -551,7 +545,6 @@ int jbd2_journal_extend(handle_t *handle, int nblocks) int result; int wanted; - WARN_ON(!transaction); if (is_handle_aborted(handle)) return -EROFS; journal = transaction->t_journal; @@ -627,7 +620,6 @@ int jbd2__journal_restart(handle_t *handle, int nblocks, gfp_t gfp_mask) tid_t tid; int need_to_start, ret; - WARN_ON(!transaction); /* If we've had an abort of any type, don't even think about * actually doing the restart! */ if (is_handle_aborted(handle)) @@ -763,6 +755,30 @@ static void warn_dirty_buffer(struct buffer_head *bh) bdevname(bh->b_bdev, b), (unsigned long long)bh->b_blocknr); } +/* Call t_frozen trigger and copy buffer data into jh->b_frozen_data. */ +static void jbd2_freeze_jh_data(struct journal_head *jh) +{ + struct page *page; + int offset; + char *source; + struct buffer_head *bh = jh2bh(jh); + + J_EXPECT_JH(jh, buffer_uptodate(bh), "Possible IO failure.\n"); + page = bh->b_page; + offset = offset_in_page(bh->b_data); + source = kmap_atomic(page); + /* Fire data frozen trigger just before we copy the data */ + jbd2_buffer_frozen_trigger(jh, source + offset, jh->b_triggers); + memcpy(jh->b_frozen_data, source + offset, bh->b_size); + kunmap_atomic(source); + + /* + * Now that the frozen data is saved off, we need to store any matching + * triggers. + */ + jh->b_frozen_triggers = jh->b_triggers; +} + /* * If the buffer is already part of the current transaction, then there * is nothing we need to do. If it is already part of a prior @@ -782,10 +798,8 @@ do_get_write_access(handle_t *handle, struct journal_head *jh, journal_t *journal; int error; char *frozen_buffer = NULL; - int need_copy = 0; unsigned long start_lock, time_lock; - WARN_ON(!transaction); if (is_handle_aborted(handle)) return -EROFS; journal = transaction->t_journal; @@ -870,119 +884,96 @@ repeat: jh->b_modified = 0; /* + * If the buffer is not journaled right now, we need to make sure it + * doesn't get written to disk before the caller actually commits the + * new data + */ + if (!jh->b_transaction) { + JBUFFER_TRACE(jh, "no transaction"); + J_ASSERT_JH(jh, !jh->b_next_transaction); + JBUFFER_TRACE(jh, "file as BJ_Reserved"); + /* + * Make sure all stores to jh (b_modified, b_frozen_data) are + * visible before attaching it to the running transaction. + * Paired with barrier in jbd2_write_access_granted() + */ + smp_wmb(); + spin_lock(&journal->j_list_lock); + __jbd2_journal_file_buffer(jh, transaction, BJ_Reserved); + spin_unlock(&journal->j_list_lock); + goto done; + } + /* * If there is already a copy-out version of this buffer, then we don't * need to make another one */ if (jh->b_frozen_data) { JBUFFER_TRACE(jh, "has frozen data"); J_ASSERT_JH(jh, jh->b_next_transaction == NULL); - jh->b_next_transaction = transaction; - goto done; + goto attach_next; } - /* Is there data here we need to preserve? */ + JBUFFER_TRACE(jh, "owned by older transaction"); + J_ASSERT_JH(jh, jh->b_next_transaction == NULL); + J_ASSERT_JH(jh, jh->b_transaction == journal->j_committing_transaction); - if (jh->b_transaction && jh->b_transaction != transaction) { - JBUFFER_TRACE(jh, "owned by older transaction"); - J_ASSERT_JH(jh, jh->b_next_transaction == NULL); - J_ASSERT_JH(jh, jh->b_transaction == - journal->j_committing_transaction); + /* + * There is one case we have to be very careful about. If the + * committing transaction is currently writing this buffer out to disk + * and has NOT made a copy-out, then we cannot modify the buffer + * contents at all right now. The essence of copy-out is that it is + * the extra copy, not the primary copy, which gets journaled. If the + * primary copy is already going to disk then we cannot do copy-out + * here. + */ + if (buffer_shadow(bh)) { + JBUFFER_TRACE(jh, "on shadow: sleep"); + jbd_unlock_bh_state(bh); + wait_on_bit_io(&bh->b_state, BH_Shadow, TASK_UNINTERRUPTIBLE); + goto repeat; + } - /* There is one case we have to be very careful about. - * If the committing transaction is currently writing - * this buffer out to disk and has NOT made a copy-out, - * then we cannot modify the buffer contents at all - * right now. The essence of copy-out is that it is the - * extra copy, not the primary copy, which gets - * journaled. If the primary copy is already going to - * disk then we cannot do copy-out here. */ - - if (buffer_shadow(bh)) { - JBUFFER_TRACE(jh, "on shadow: sleep"); + /* + * Only do the copy if the currently-owning transaction still needs it. + * If buffer isn't on BJ_Metadata list, the committing transaction is + * past that stage (here we use the fact that BH_Shadow is set under + * bh_state lock together with refiling to BJ_Shadow list and at this + * point we know the buffer doesn't have BH_Shadow set). + * + * Subtle point, though: if this is a get_undo_access, then we will be + * relying on the frozen_data to contain the new value of the + * committed_data record after the transaction, so we HAVE to force the + * frozen_data copy in that case. + */ + if (jh->b_jlist == BJ_Metadata || force_copy) { + JBUFFER_TRACE(jh, "generate frozen data"); + if (!frozen_buffer) { + JBUFFER_TRACE(jh, "allocate memory for buffer"); jbd_unlock_bh_state(bh); - wait_on_bit_io(&bh->b_state, BH_Shadow, - TASK_UNINTERRUPTIBLE); - goto repeat; - } - - /* - * Only do the copy if the currently-owning transaction still - * needs it. If buffer isn't on BJ_Metadata list, the - * committing transaction is past that stage (here we use the - * fact that BH_Shadow is set under bh_state lock together with - * refiling to BJ_Shadow list and at this point we know the - * buffer doesn't have BH_Shadow set). - * - * Subtle point, though: if this is a get_undo_access, - * then we will be relying on the frozen_data to contain - * the new value of the committed_data record after the - * transaction, so we HAVE to force the frozen_data copy - * in that case. - */ - if (jh->b_jlist == BJ_Metadata || force_copy) { - JBUFFER_TRACE(jh, "generate frozen data"); + frozen_buffer = jbd2_alloc(jh2bh(jh)->b_size, GFP_NOFS); if (!frozen_buffer) { - JBUFFER_TRACE(jh, "allocate memory for buffer"); - jbd_unlock_bh_state(bh); - frozen_buffer = - jbd2_alloc(jh2bh(jh)->b_size, - GFP_NOFS); - if (!frozen_buffer) { - printk(KERN_ERR - "%s: OOM for frozen_buffer\n", - __func__); - JBUFFER_TRACE(jh, "oom!"); - error = -ENOMEM; - jbd_lock_bh_state(bh); - goto done; - } - goto repeat; + printk(KERN_ERR "%s: OOM for frozen_buffer\n", + __func__); + JBUFFER_TRACE(jh, "oom!"); + error = -ENOMEM; + goto out; } - jh->b_frozen_data = frozen_buffer; - frozen_buffer = NULL; - need_copy = 1; + goto repeat; } - jh->b_next_transaction = transaction; + jh->b_frozen_data = frozen_buffer; + frozen_buffer = NULL; + jbd2_freeze_jh_data(jh); } - - +attach_next: /* - * Finally, if the buffer is not journaled right now, we need to make - * sure it doesn't get written to disk before the caller actually - * commits the new data + * Make sure all stores to jh (b_modified, b_frozen_data) are visible + * before attaching it to the running transaction. Paired with barrier + * in jbd2_write_access_granted() */ - if (!jh->b_transaction) { - JBUFFER_TRACE(jh, "no transaction"); - J_ASSERT_JH(jh, !jh->b_next_transaction); - JBUFFER_TRACE(jh, "file as BJ_Reserved"); - spin_lock(&journal->j_list_lock); - __jbd2_journal_file_buffer(jh, transaction, BJ_Reserved); - spin_unlock(&journal->j_list_lock); - } + smp_wmb(); + jh->b_next_transaction = transaction; done: - if (need_copy) { - struct page *page; - int offset; - char *source; - - J_EXPECT_JH(jh, buffer_uptodate(jh2bh(jh)), - "Possible IO failure.\n"); - page = jh2bh(jh)->b_page; - offset = offset_in_page(jh2bh(jh)->b_data); - source = kmap_atomic(page); - /* Fire data frozen trigger just before we copy the data */ - jbd2_buffer_frozen_trigger(jh, source + offset, - jh->b_triggers); - memcpy(jh->b_frozen_data, source+offset, jh2bh(jh)->b_size); - kunmap_atomic(source); - - /* - * Now that the frozen data is saved off, we need to store - * any matching triggers. - */ - jh->b_frozen_triggers = jh->b_triggers; - } jbd_unlock_bh_state(bh); /* @@ -999,6 +990,55 @@ out: return error; } +/* Fast check whether buffer is already attached to the required transaction */ +static bool jbd2_write_access_granted(handle_t *handle, struct buffer_head *bh) +{ + struct journal_head *jh; + bool ret = false; + + /* Dirty buffers require special handling... */ + if (buffer_dirty(bh)) + return false; + + /* + * RCU protects us from dereferencing freed pages. So the checks we do + * are guaranteed not to oops. However the jh slab object can get freed + * & reallocated while we work with it. So we have to be careful. When + * we see jh attached to the running transaction, we know it must stay + * so until the transaction is committed. Thus jh won't be freed and + * will be attached to the same bh while we run. However it can + * happen jh gets freed, reallocated, and attached to the transaction + * just after we get pointer to it from bh. So we have to be careful + * and recheck jh still belongs to our bh before we return success. + */ + rcu_read_lock(); + if (!buffer_jbd(bh)) + goto out; + /* This should be bh2jh() but that doesn't work with inline functions */ + jh = READ_ONCE(bh->b_private); + if (!jh) + goto out; + if (jh->b_transaction != handle->h_transaction && + jh->b_next_transaction != handle->h_transaction) + goto out; + /* + * There are two reasons for the barrier here: + * 1) Make sure to fetch b_bh after we did previous checks so that we + * detect when jh went through free, realloc, attach to transaction + * while we were checking. Paired with implicit barrier in that path. + * 2) So that access to bh done after jbd2_write_access_granted() + * doesn't get reordered and see inconsistent state of concurrent + * do_get_write_access(). + */ + smp_mb(); + if (unlikely(jh->b_bh != bh)) + goto out; + ret = true; +out: + rcu_read_unlock(); + return ret; +} + /** * int jbd2_journal_get_write_access() - notify intent to modify a buffer for metadata (not data) update. * @handle: transaction to add buffer modifications to @@ -1012,9 +1052,13 @@ out: int jbd2_journal_get_write_access(handle_t *handle, struct buffer_head *bh) { - struct journal_head *jh = jbd2_journal_add_journal_head(bh); + struct journal_head *jh; int rc; + if (jbd2_write_access_granted(handle, bh)) + return 0; + + jh = jbd2_journal_add_journal_head(bh); /* We do not want to get caught playing with fields which the * log thread also manipulates. Make sure that the buffer * completes any outstanding IO before proceeding. */ @@ -1051,7 +1095,6 @@ int jbd2_journal_get_create_access(handle_t *handle, struct buffer_head *bh) int err; jbd_debug(5, "journal_head %p\n", jh); - WARN_ON(!transaction); err = -EROFS; if (is_handle_aborted(handle)) goto out; @@ -1145,11 +1188,14 @@ out: int jbd2_journal_get_undo_access(handle_t *handle, struct buffer_head *bh) { int err; - struct journal_head *jh = jbd2_journal_add_journal_head(bh); + struct journal_head *jh; char *committed_data = NULL; JBUFFER_TRACE(jh, "entry"); + if (jbd2_write_access_granted(handle, bh)) + return 0; + jh = jbd2_journal_add_journal_head(bh); /* * Do this first --- it can drop the journal lock, so we want to * make sure that obtaining the committed_data is done @@ -1266,7 +1312,6 @@ int jbd2_journal_dirty_metadata(handle_t *handle, struct buffer_head *bh) struct journal_head *jh; int ret = 0; - WARN_ON(!transaction); if (is_handle_aborted(handle)) return -EROFS; journal = transaction->t_journal; @@ -1397,7 +1442,6 @@ int jbd2_journal_forget (handle_t *handle, struct buffer_head *bh) int err = 0; int was_modified = 0; - WARN_ON(!transaction); if (is_handle_aborted(handle)) return -EROFS; journal = transaction->t_journal; @@ -1530,8 +1574,22 @@ int jbd2_journal_stop(handle_t *handle) tid_t tid; pid_t pid; - if (!transaction) - goto free_and_exit; + if (!transaction) { + /* + * Handle is already detached from the transaction so + * there is nothing to do other than decrease a refcount, + * or free the handle if refcount drops to zero + */ + if (--handle->h_ref > 0) { + jbd_debug(4, "h_ref %d -> %d\n", handle->h_ref + 1, + handle->h_ref); + return err; + } else { + if (handle->h_rsv_handle) + jbd2_free_handle(handle->h_rsv_handle); + goto free_and_exit; + } + } journal = transaction->t_journal; J_ASSERT(journal_current_handle() == handle); @@ -2373,7 +2431,6 @@ int jbd2_journal_file_inode(handle_t *handle, struct jbd2_inode *jinode) transaction_t *transaction = handle->h_transaction; journal_t *journal; - WARN_ON(!transaction); if (is_handle_aborted(handle)) return -EROFS; journal = transaction->t_journal; diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c index 60d86e8..2caf168 100644 --- a/fs/jffs2/fs.c +++ b/fs/jffs2/fs.c @@ -272,12 +272,9 @@ struct inode *jffs2_iget(struct super_block *sb, unsigned long ino) mutex_lock(&f->sem); ret = jffs2_do_read_inode(c, f, inode->i_ino, &latest_node); + if (ret) + goto error; - if (ret) { - mutex_unlock(&f->sem); - iget_failed(inode); - return ERR_PTR(ret); - } inode->i_mode = jemode_to_cpu(latest_node.mode); i_uid_write(inode, je16_to_cpu(latest_node.uid)); i_gid_write(inode, je16_to_cpu(latest_node.gid)); diff --git a/fs/jffs2/readinode.c b/fs/jffs2/readinode.c index dddbde4..28e0aab 100644 --- a/fs/jffs2/readinode.c +++ b/fs/jffs2/readinode.c @@ -1203,17 +1203,13 @@ static int jffs2_do_read_inode_internal(struct jffs2_sb_info *c, JFFS2_ERROR("failed to read from flash: error %d, %zd of %zd bytes read\n", ret, retlen, sizeof(*latest_node)); /* FIXME: If this fails, there seems to be a memory leak. Find it. */ - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); - return ret?ret:-EIO; + return ret ? ret : -EIO; } crc = crc32(0, latest_node, sizeof(*latest_node)-8); if (crc != je32_to_cpu(latest_node->node_crc)) { JFFS2_ERROR("CRC failed for read_inode of inode %u at physical location 0x%x\n", f->inocache->ino, ref_offset(rii.latest_ref)); - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); return -EIO; } @@ -1250,16 +1246,11 @@ static int jffs2_do_read_inode_internal(struct jffs2_sb_info *c, * keep in RAM to facilitate quick follow symlink * operation. */ uint32_t csize = je32_to_cpu(latest_node->csize); - if (csize > JFFS2_MAX_NAME_LEN) { - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); + if (csize > JFFS2_MAX_NAME_LEN) return -ENAMETOOLONG; - } f->target = kmalloc(csize + 1, GFP_KERNEL); if (!f->target) { JFFS2_ERROR("can't allocate %u bytes of memory for the symlink target path cache\n", csize); - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); return -ENOMEM; } @@ -1271,8 +1262,6 @@ static int jffs2_do_read_inode_internal(struct jffs2_sb_info *c, ret = -EIO; kfree(f->target); f->target = NULL; - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); return ret; } @@ -1289,15 +1278,11 @@ static int jffs2_do_read_inode_internal(struct jffs2_sb_info *c, if (f->metadata) { JFFS2_ERROR("Argh. Special inode #%u with mode 0%o had metadata node\n", f->inocache->ino, jemode_to_cpu(latest_node->mode)); - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); return -EIO; } if (!frag_first(&f->fragtree)) { JFFS2_ERROR("Argh. Special inode #%u with mode 0%o has no fragments\n", f->inocache->ino, jemode_to_cpu(latest_node->mode)); - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); return -EIO; } /* ASSERT: f->fraglist != NULL */ @@ -1305,8 +1290,6 @@ static int jffs2_do_read_inode_internal(struct jffs2_sb_info *c, JFFS2_ERROR("Argh. Special inode #%u with mode 0x%x had more than one node\n", f->inocache->ino, jemode_to_cpu(latest_node->mode)); /* FIXME: Deal with it - check crc32, check for duplicate node, check times and discard the older one */ - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); return -EIO; } /* OK. We're happy */ @@ -1400,10 +1383,8 @@ int jffs2_do_crccheck_inode(struct jffs2_sb_info *c, struct jffs2_inode_cache *i f->inocache = ic; ret = jffs2_do_read_inode_internal(c, f, &n); - if (!ret) { - mutex_unlock(&f->sem); - jffs2_do_clear_inode(c, f); - } + mutex_unlock(&f->sem); + jffs2_do_clear_inode(c, f); jffs2_xattr_do_crccheck_inode(c, ic); kfree (f); return ret; diff --git a/fs/kernfs/dir.c b/fs/kernfs/dir.c index f131fc2..2d48d28 100644 --- a/fs/kernfs/dir.c +++ b/fs/kernfs/dir.c @@ -518,7 +518,14 @@ static struct kernfs_node *__kernfs_new_node(struct kernfs_root *root, if (!kn) goto err_out1; - ret = ida_simple_get(&root->ino_ida, 1, 0, GFP_KERNEL); + /* + * If the ino of the sysfs entry created for a kmem cache gets + * allocated from an ida layer, which is accounted to the memcg that + * owns the cache, the memcg will get pinned forever. So do not account + * ino ida allocations. + */ + ret = ida_simple_get(&root->ino_ida, 1, 0, + GFP_KERNEL | __GFP_NOACCOUNT); if (ret < 0) goto err_out2; kn->ino = ret; @@ -585,6 +592,9 @@ int kernfs_add_one(struct kernfs_node *kn) goto out_unlock; ret = -ENOENT; + if (parent->flags & KERNFS_EMPTY_DIR) + goto out_unlock; + if ((parent->flags & KERNFS_ACTIVATED) && !kernfs_active(parent)) goto out_unlock; @@ -776,6 +786,38 @@ struct kernfs_node *kernfs_create_dir_ns(struct kernfs_node *parent, return ERR_PTR(rc); } +/** + * kernfs_create_empty_dir - create an always empty directory + * @parent: parent in which to create a new directory + * @name: name of the new directory + * + * Returns the created node on success, ERR_PTR() value on failure. + */ +struct kernfs_node *kernfs_create_empty_dir(struct kernfs_node *parent, + const char *name) +{ + struct kernfs_node *kn; + int rc; + + /* allocate */ + kn = kernfs_new_node(parent, name, S_IRUGO|S_IXUGO|S_IFDIR, KERNFS_DIR); + if (!kn) + return ERR_PTR(-ENOMEM); + + kn->flags |= KERNFS_EMPTY_DIR; + kn->dir.root = parent->dir.root; + kn->ns = NULL; + kn->priv = NULL; + + /* link in */ + rc = kernfs_add_one(kn); + if (!rc) + return kn; + + kernfs_put(kn); + return ERR_PTR(rc); +} + static struct dentry *kernfs_iop_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) @@ -1247,7 +1289,8 @@ int kernfs_rename_ns(struct kernfs_node *kn, struct kernfs_node *new_parent, mutex_lock(&kernfs_mutex); error = -ENOENT; - if (!kernfs_active(kn) || !kernfs_active(new_parent)) + if (!kernfs_active(kn) || !kernfs_active(new_parent) || + (new_parent->flags & KERNFS_EMPTY_DIR)) goto out; error = 0; diff --git a/fs/kernfs/file.c b/fs/kernfs/file.c index 2bacb99..7247252 100644 --- a/fs/kernfs/file.c +++ b/fs/kernfs/file.c @@ -785,7 +785,6 @@ static unsigned int kernfs_fop_poll(struct file *filp, poll_table *wait) struct kernfs_node *kn = filp->f_path.dentry->d_fsdata; struct kernfs_open_node *on = kn->attr.open; - /* need parent for the kobj, grab both */ if (!kernfs_get_active(kn)) goto trigger; diff --git a/fs/kernfs/inode.c b/fs/kernfs/inode.c index 2da8493..756dd56 100644 --- a/fs/kernfs/inode.c +++ b/fs/kernfs/inode.c @@ -296,6 +296,8 @@ static void kernfs_init_inode(struct kernfs_node *kn, struct inode *inode) case KERNFS_DIR: inode->i_op = &kernfs_dir_iops; inode->i_fop = &kernfs_dir_fops; + if (kn->flags & KERNFS_EMPTY_DIR) + make_empty_dir_inode(inode); break; case KERNFS_FILE: inode->i_size = kn->attr.size; diff --git a/fs/kernfs/kernfs-internal.h b/fs/kernfs/kernfs-internal.h index af9fa74..6762bfb 100644 --- a/fs/kernfs/kernfs-internal.h +++ b/fs/kernfs/kernfs-internal.h @@ -76,7 +76,6 @@ extern struct kmem_cache *kernfs_node_cache; /* * inode.c */ -struct inode *kernfs_get_inode(struct super_block *sb, struct kernfs_node *kn); void kernfs_evict_inode(struct inode *inode); int kernfs_iop_permission(struct inode *inode, int mask); int kernfs_iop_setattr(struct dentry *dentry, struct iattr *iattr); @@ -1103,3 +1103,98 @@ const struct inode_operations simple_symlink_inode_operations = { .readlink = generic_readlink }; EXPORT_SYMBOL(simple_symlink_inode_operations); + +/* + * Operations for a permanently empty directory. + */ +static struct dentry *empty_dir_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) +{ + return ERR_PTR(-ENOENT); +} + +static int empty_dir_getattr(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat) +{ + struct inode *inode = d_inode(dentry); + generic_fillattr(inode, stat); + return 0; +} + +static int empty_dir_setattr(struct dentry *dentry, struct iattr *attr) +{ + return -EPERM; +} + +static int empty_dir_setxattr(struct dentry *dentry, const char *name, + const void *value, size_t size, int flags) +{ + return -EOPNOTSUPP; +} + +static ssize_t empty_dir_getxattr(struct dentry *dentry, const char *name, + void *value, size_t size) +{ + return -EOPNOTSUPP; +} + +static int empty_dir_removexattr(struct dentry *dentry, const char *name) +{ + return -EOPNOTSUPP; +} + +static ssize_t empty_dir_listxattr(struct dentry *dentry, char *list, size_t size) +{ + return -EOPNOTSUPP; +} + +static const struct inode_operations empty_dir_inode_operations = { + .lookup = empty_dir_lookup, + .permission = generic_permission, + .setattr = empty_dir_setattr, + .getattr = empty_dir_getattr, + .setxattr = empty_dir_setxattr, + .getxattr = empty_dir_getxattr, + .removexattr = empty_dir_removexattr, + .listxattr = empty_dir_listxattr, +}; + +static loff_t empty_dir_llseek(struct file *file, loff_t offset, int whence) +{ + /* An empty directory has two entries . and .. at offsets 0 and 1 */ + return generic_file_llseek_size(file, offset, whence, 2, 2); +} + +static int empty_dir_readdir(struct file *file, struct dir_context *ctx) +{ + dir_emit_dots(file, ctx); + return 0; +} + +static const struct file_operations empty_dir_operations = { + .llseek = empty_dir_llseek, + .read = generic_read_dir, + .iterate = empty_dir_readdir, + .fsync = noop_fsync, +}; + + +void make_empty_dir_inode(struct inode *inode) +{ + set_nlink(inode, 2); + inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO; + inode->i_uid = GLOBAL_ROOT_UID; + inode->i_gid = GLOBAL_ROOT_GID; + inode->i_rdev = 0; + inode->i_size = 2; + inode->i_blkbits = PAGE_SHIFT; + inode->i_blocks = 0; + + inode->i_op = &empty_dir_inode_operations; + inode->i_fop = &empty_dir_operations; +} + +bool is_empty_dir_inode(struct inode *inode) +{ + return (inode->i_fop == &empty_dir_operations) && + (inode->i_op == &empty_dir_inode_operations); +} diff --git a/fs/minix/inode.c b/fs/minix/inode.c index 1182d1e..086cd0a 100644 --- a/fs/minix/inode.c +++ b/fs/minix/inode.c @@ -62,7 +62,7 @@ static struct kmem_cache * minix_inode_cachep; static struct inode *minix_alloc_inode(struct super_block *sb) { struct minix_inode_info *ei; - ei = (struct minix_inode_info *)kmem_cache_alloc(minix_inode_cachep, GFP_KERNEL); + ei = kmem_cache_alloc(minix_inode_cachep, GFP_KERNEL); if (!ei) return NULL; return &ei->vfs_inode; @@ -118,7 +118,6 @@ static inline void unlock_mount_hash(void) } struct proc_mounts { - struct seq_file m; struct mnt_namespace *ns; struct path root; int (*show)(struct seq_file *, struct vfsmount *); @@ -127,8 +126,6 @@ struct proc_mounts { loff_t cached_index; }; -#define proc_mounts(p) (container_of((p), struct proc_mounts, m)) - extern const struct seq_operations mounts_op; extern bool __is_local_mountpoint(struct dentry *dentry); @@ -605,6 +605,8 @@ alloc_new: bio_get_nr_vecs(bdev), GFP_NOFS|__GFP_HIGH); if (bio == NULL) goto confused; + + wbc_init_bio(wbc, bio); } /* @@ -612,6 +614,7 @@ alloc_new: * the confused fail path above (OOM) will be very confused when * it finds all bh marked clean (i.e. it will not write anything) */ + wbc_account_io(wbc, page, PAGE_SIZE); length = first_unmapped << blkbits; if (bio_add_page(bio, page, length, 0) < length) { bio = mpage_bio_submit(WRITE, bio); diff --git a/fs/namespace.c b/fs/namespace.c index 9c1c43d..c7cb8a5 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1226,7 +1226,7 @@ EXPORT_SYMBOL(replace_mount_options); /* iterator; we want it to have access to namespace_sem, thus here... */ static void *m_start(struct seq_file *m, loff_t *pos) { - struct proc_mounts *p = proc_mounts(m); + struct proc_mounts *p = m->private; down_read(&namespace_sem); if (p->cached_event == p->ns->event) { @@ -1247,7 +1247,7 @@ static void *m_start(struct seq_file *m, loff_t *pos) static void *m_next(struct seq_file *m, void *v, loff_t *pos) { - struct proc_mounts *p = proc_mounts(m); + struct proc_mounts *p = m->private; p->cached_mount = seq_list_next(v, &p->ns->list, pos); p->cached_index = *pos; @@ -1261,7 +1261,7 @@ static void m_stop(struct seq_file *m, void *v) static int m_show(struct seq_file *m, void *v) { - struct proc_mounts *p = proc_mounts(m); + struct proc_mounts *p = m->private; struct mount *r = list_entry(v, struct mount, mnt_list); return p->show(m, &r->mnt); } @@ -2343,6 +2343,8 @@ unlock: return err; } +static bool fs_fully_visible(struct file_system_type *fs_type, int *new_mnt_flags); + /* * create a new mount for userspace and request it to be added into the * namespace's tree @@ -2374,6 +2376,10 @@ static int do_new_mount(struct path *path, const char *fstype, int flags, flags |= MS_NODEV; mnt_flags |= MNT_NODEV | MNT_LOCK_NODEV; } + if (type->fs_flags & FS_USERNS_VISIBLE) { + if (!fs_fully_visible(type, &mnt_flags)) + return -EPERM; + } } mnt = vfs_kern_mount(type, flags, name, data); @@ -3175,9 +3181,10 @@ bool current_chrooted(void) return chrooted; } -bool fs_fully_visible(struct file_system_type *type) +static bool fs_fully_visible(struct file_system_type *type, int *new_mnt_flags) { struct mnt_namespace *ns = current->nsproxy->mnt_ns; + int new_flags = *new_mnt_flags; struct mount *mnt; bool visible = false; @@ -3196,16 +3203,36 @@ bool fs_fully_visible(struct file_system_type *type) if (mnt->mnt.mnt_root != mnt->mnt.mnt_sb->s_root) continue; - /* This mount is not fully visible if there are any child mounts - * that cover anything except for empty directories. + /* Verify the mount flags are equal to or more permissive + * than the proposed new mount. + */ + if ((mnt->mnt.mnt_flags & MNT_LOCK_READONLY) && + !(new_flags & MNT_READONLY)) + continue; + if ((mnt->mnt.mnt_flags & MNT_LOCK_NODEV) && + !(new_flags & MNT_NODEV)) + continue; + if ((mnt->mnt.mnt_flags & MNT_LOCK_ATIME) && + ((mnt->mnt.mnt_flags & MNT_ATIME_MASK) != (new_flags & MNT_ATIME_MASK))) + continue; + + /* This mount is not fully visible if there are any + * locked child mounts that cover anything except for + * empty directories. */ list_for_each_entry(child, &mnt->mnt_mounts, mnt_child) { struct inode *inode = child->mnt_mountpoint->d_inode; - if (!S_ISDIR(inode->i_mode)) - goto next; - if (inode->i_nlink > 2) + /* Only worry about locked mounts */ + if (!(mnt->mnt.mnt_flags & MNT_LOCKED)) + continue; + /* Is the directory permanetly empty? */ + if (!is_empty_dir_inode(inode)) goto next; } + /* Preserve the locked attributes */ + *new_mnt_flags |= mnt->mnt.mnt_flags & (MNT_LOCK_READONLY | \ + MNT_LOCK_NODEV | \ + MNT_LOCK_ATIME); visible = true; goto found; next: ; diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 8d129bb..682529c 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -458,7 +458,7 @@ check_gss_callback_principal(struct nfs_client *clp, struct svc_rqst *rqstp) * pg_authenticate method for nfsv4 callback threads. * * The authflavor has been negotiated, so an incorrect flavor is a server - * bug. Drop packets with incorrect authflavor. + * bug. Deny packets with incorrect authflavor. * * All other checking done after NFS decoding where the nfs_client can be * found in nfs4_callback_compound @@ -468,12 +468,12 @@ static int nfs_callback_authenticate(struct svc_rqst *rqstp) switch (rqstp->rq_authop->flavour) { case RPC_AUTH_NULL: if (rqstp->rq_proc != CB_NULL) - return SVC_DROP; + return SVC_DENIED; break; case RPC_AUTH_GSS: /* No RPC_AUTH_GSS support yet in NFSv4.1 */ if (svc_is_backchannel(rqstp)) - return SVC_DROP; + return SVC_DENIED; } return SVC_OK; } diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index 197806f..29e3c1b 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -327,10 +327,8 @@ validate_seqid(struct nfs4_slot_table *tbl, struct cb_sequenceargs * args) dprintk("%s slot table seqid: %u\n", __func__, slot->seq_nr); /* Normal */ - if (likely(args->csa_sequenceid == slot->seq_nr + 1)) { - slot->seq_nr++; + if (likely(args->csa_sequenceid == slot->seq_nr + 1)) goto out_ok; - } /* Replay */ if (args->csa_sequenceid == slot->seq_nr) { @@ -418,6 +416,7 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args, struct cb_process_state *cps) { struct nfs4_slot_table *tbl; + struct nfs4_slot *slot; struct nfs_client *clp; int i; __be32 status = htonl(NFS4ERR_BADSESSION); @@ -429,25 +428,32 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args, if (!(clp->cl_session->flags & SESSION4_BACK_CHAN)) goto out; + tbl = &clp->cl_session->bc_slot_table; + slot = tbl->slots + args->csa_slotid; spin_lock(&tbl->slot_tbl_lock); /* state manager is resetting the session */ if (test_bit(NFS4_SLOT_TBL_DRAINING, &tbl->slot_tbl_state)) { - spin_unlock(&tbl->slot_tbl_lock); status = htonl(NFS4ERR_DELAY); /* Return NFS4ERR_BADSESSION if we're draining the session * in order to reset it. */ if (test_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state)) status = htonl(NFS4ERR_BADSESSION); - goto out; + goto out_unlock; } - status = validate_seqid(&clp->cl_session->bc_slot_table, args); - spin_unlock(&tbl->slot_tbl_lock); + memcpy(&res->csr_sessionid, &args->csa_sessionid, + sizeof(res->csr_sessionid)); + res->csr_sequenceid = args->csa_sequenceid; + res->csr_slotid = args->csa_slotid; + res->csr_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; + res->csr_target_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; + + status = validate_seqid(tbl, args); if (status) - goto out; + goto out_unlock; cps->slotid = args->csa_slotid; @@ -458,15 +464,17 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args, */ if (referring_call_exists(clp, args->csa_nrclists, args->csa_rclists)) { status = htonl(NFS4ERR_DELAY); - goto out; + goto out_unlock; } - memcpy(&res->csr_sessionid, &args->csa_sessionid, - sizeof(res->csr_sessionid)); - res->csr_sequenceid = args->csa_sequenceid; - res->csr_slotid = args->csa_slotid; - res->csr_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; - res->csr_target_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; + /* + * RFC5661 20.9.3 + * If CB_SEQUENCE returns an error, then the state of the slot + * (sequence ID, cached reply) MUST NOT change. + */ + slot->seq_nr++; +out_unlock: + spin_unlock(&tbl->slot_tbl_lock); out: cps->clp = clp; /* put in nfs4_callback_compound */ diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 19ca95c..6b1697a 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -909,7 +909,7 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r xdr_init_encode(&xdr_out, &rqstp->rq_res, p); status = decode_compound_hdr_arg(&xdr_in, &hdr_arg); - if (status == __constant_htonl(NFS4ERR_RESOURCE)) + if (status == htonl(NFS4ERR_RESOURCE)) return rpc_garbage_args; if (hdr_arg.minorversion == 0) { diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 892aeff..ecebb40 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -825,7 +825,6 @@ error: * Load up the server record from information gained in an fsinfo record */ static void nfs_server_set_fsinfo(struct nfs_server *server, - struct nfs_fh *mntfh, struct nfs_fsinfo *fsinfo) { unsigned long max_rpc_payload; @@ -901,7 +900,7 @@ int nfs_probe_fsinfo(struct nfs_server *server, struct nfs_fh *mntfh, struct nfs if (error < 0) goto out_error; - nfs_server_set_fsinfo(server, mntfh, &fsinfo); + nfs_server_set_fsinfo(server, &fsinfo); /* Get some general file system info */ if (server->namelen == 0) { @@ -1193,8 +1192,6 @@ void nfs_clients_init(struct net *net) } #ifdef CONFIG_PROC_FS -static struct proc_dir_entry *proc_fs_nfs; - static int nfs_server_list_open(struct inode *inode, struct file *file); static void *nfs_server_list_start(struct seq_file *p, loff_t *pos); static void *nfs_server_list_next(struct seq_file *p, void *v, loff_t *pos); @@ -1364,27 +1361,29 @@ static int nfs_volume_list_show(struct seq_file *m, void *v) { struct nfs_server *server; struct nfs_client *clp; - char dev[8], fsid[17]; + char dev[13]; // 8 for 2^24, 1 for ':', 3 for 2^8, 1 for '\0' + char fsid[34]; // 2 * 16 for %llx, 1 for ':', 1 for '\0' struct nfs_net *nn = net_generic(seq_file_net(m), nfs_net_id); /* display header on line 1 */ if (v == &nn->nfs_volume_list) { - seq_puts(m, "NV SERVER PORT DEV FSID FSC\n"); + seq_puts(m, "NV SERVER PORT DEV FSID" + " FSC\n"); return 0; } /* display one transport per line on subsequent lines */ server = list_entry(v, struct nfs_server, master_link); clp = server->nfs_client; - snprintf(dev, 8, "%u:%u", + snprintf(dev, sizeof(dev), "%u:%u", MAJOR(server->s_dev), MINOR(server->s_dev)); - snprintf(fsid, 17, "%llx:%llx", + snprintf(fsid, sizeof(fsid), "%llx:%llx", (unsigned long long) server->fsid.major, (unsigned long long) server->fsid.minor); rcu_read_lock(); - seq_printf(m, "v%u %s %s %-7s %-17s %s\n", + seq_printf(m, "v%u %s %s %-12s %-33s %s\n", clp->rpc_ops->version, rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_ADDR), rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_HEX_PORT), @@ -1434,27 +1433,20 @@ void nfs_fs_proc_net_exit(struct net *net) */ int __init nfs_fs_proc_init(void) { - struct proc_dir_entry *p; - - proc_fs_nfs = proc_mkdir("fs/nfsfs", NULL); - if (!proc_fs_nfs) + if (!proc_mkdir("fs/nfsfs", NULL)) goto error_0; /* a file of servers with which we're dealing */ - p = proc_symlink("servers", proc_fs_nfs, "../../net/nfsfs/servers"); - if (!p) + if (!proc_symlink("fs/nfsfs/servers", NULL, "../../net/nfsfs/servers")) goto error_1; /* a file of volumes that we have mounted */ - p = proc_symlink("volumes", proc_fs_nfs, "../../net/nfsfs/volumes"); - if (!p) - goto error_2; - return 0; + if (!proc_symlink("fs/nfsfs/volumes", NULL, "../../net/nfsfs/volumes")) + goto error_1; -error_2: - remove_proc_entry("servers", proc_fs_nfs); + return 0; error_1: - remove_proc_entry("fs/nfsfs", NULL); + remove_proc_subtree("fs/nfsfs", NULL); error_0: return -ENOMEM; } @@ -1464,9 +1456,7 @@ error_0: */ void nfs_fs_proc_exit(void) { - remove_proc_entry("volumes", proc_fs_nfs); - remove_proc_entry("servers", proc_fs_nfs); - remove_proc_entry("fs/nfsfs", NULL); + remove_proc_subtree("fs/nfsfs", NULL); } #endif /* CONFIG_PROC_FS */ diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index b9108f4..547308a 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1470,9 +1470,6 @@ static int nfs_finish_open(struct nfs_open_context *ctx, { int err; - if ((open_flags & (O_CREAT | O_EXCL)) == (O_CREAT | O_EXCL)) - *opened |= FILE_CREATED; - err = finish_open(file, dentry, do_open, opened); if (err) goto out; diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 8b8d83a..cc4fa1e 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -555,31 +555,22 @@ static int nfs_launder_page(struct page *page) return nfs_wb_page(inode, page); } -#ifdef CONFIG_NFS_SWAP static int nfs_swap_activate(struct swap_info_struct *sis, struct file *file, sector_t *span) { - int ret; struct rpc_clnt *clnt = NFS_CLIENT(file->f_mapping->host); *span = sis->pages; - rcu_read_lock(); - ret = xs_swapper(rcu_dereference(clnt->cl_xprt), 1); - rcu_read_unlock(); - - return ret; + return rpc_clnt_swap_activate(clnt); } static void nfs_swap_deactivate(struct file *file) { struct rpc_clnt *clnt = NFS_CLIENT(file->f_mapping->host); - rcu_read_lock(); - xs_swapper(rcu_dereference(clnt->cl_xprt), 0); - rcu_read_unlock(); + rpc_clnt_swap_deactivate(clnt); } -#endif const struct address_space_operations nfs_file_aops = { .readpage = nfs_readpage, @@ -596,10 +587,8 @@ const struct address_space_operations nfs_file_aops = { .launder_page = nfs_launder_page, .is_dirty_writeback = nfs_check_dirty_writeback, .error_remove_page = generic_error_remove_page, -#ifdef CONFIG_NFS_SWAP .swap_activate = nfs_swap_activate, .swap_deactivate = nfs_swap_deactivate, -#endif }; /* diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c index a46bf6d..b34f2e2 100644 --- a/fs/nfs/filelayout/filelayout.c +++ b/fs/nfs/filelayout/filelayout.c @@ -32,6 +32,7 @@ #include <linux/nfs_fs.h> #include <linux/nfs_page.h> #include <linux/module.h> +#include <linux/backing-dev.h> #include <linux/sunrpc/metrics.h> diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index 7d05089..c12951b 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -20,6 +20,7 @@ #include "../nfs4trace.h" #include "../iostat.h" #include "../nfs.h" +#include "../nfs42.h" #define NFSDBG_FACILITY NFSDBG_PNFS_LD @@ -182,17 +183,14 @@ static void _ff_layout_free_lseg(struct nfs4_ff_layout_segment *fls) static void ff_layout_sort_mirrors(struct nfs4_ff_layout_segment *fls) { - struct nfs4_ff_layout_mirror *tmp; int i, j; for (i = 0; i < fls->mirror_array_cnt - 1; i++) { for (j = i + 1; j < fls->mirror_array_cnt; j++) if (fls->mirror_array[i]->efficiency < - fls->mirror_array[j]->efficiency) { - tmp = fls->mirror_array[i]; - fls->mirror_array[i] = fls->mirror_array[j]; - fls->mirror_array[j] = tmp; - } + fls->mirror_array[j]->efficiency) + swap(fls->mirror_array[i], + fls->mirror_array[j]); } } @@ -274,6 +272,7 @@ ff_layout_alloc_lseg(struct pnfs_layout_hdr *lh, spin_lock_init(&fls->mirror_array[i]->lock); fls->mirror_array[i]->ds_count = ds_count; + fls->mirror_array[i]->lseg = &fls->generic_hdr; /* deviceid */ rc = decode_deviceid(&stream, &devid); @@ -344,6 +343,10 @@ ff_layout_alloc_lseg(struct pnfs_layout_hdr *lh, fls->mirror_array[i]->gid); } + p = xdr_inline_decode(&stream, 4); + if (p) + fls->flags = be32_to_cpup(p); + ff_layout_sort_mirrors(fls); rc = ff_layout_check_layout(lgr); if (rc) @@ -415,6 +418,146 @@ ff_layout_get_lseg_count(struct nfs4_ff_layout_segment *fls) return 1; } +static void +nfs4_ff_start_busy_timer(struct nfs4_ff_busy_timer *timer) +{ + /* first IO request? */ + if (atomic_inc_return(&timer->n_ops) == 1) { + timer->start_time = ktime_get(); + } +} + +static ktime_t +nfs4_ff_end_busy_timer(struct nfs4_ff_busy_timer *timer) +{ + ktime_t start, now; + + if (atomic_dec_return(&timer->n_ops) < 0) + WARN_ON_ONCE(1); + + now = ktime_get(); + start = timer->start_time; + timer->start_time = now; + return ktime_sub(now, start); +} + +static ktime_t +nfs4_ff_layout_calc_completion_time(struct rpc_task *task) +{ + return ktime_sub(ktime_get(), task->tk_start); +} + +static bool +nfs4_ff_layoutstat_start_io(struct nfs4_ff_layout_mirror *mirror, + struct nfs4_ff_layoutstat *layoutstat) +{ + static const ktime_t notime = {0}; + ktime_t now = ktime_get(); + + nfs4_ff_start_busy_timer(&layoutstat->busy_timer); + if (ktime_equal(mirror->start_time, notime)) + mirror->start_time = now; + if (ktime_equal(mirror->last_report_time, notime)) + mirror->last_report_time = now; + if (ktime_to_ms(ktime_sub(now, mirror->last_report_time)) >= + FF_LAYOUTSTATS_REPORT_INTERVAL) { + mirror->last_report_time = now; + return true; + } + + return false; +} + +static void +nfs4_ff_layout_stat_io_update_requested(struct nfs4_ff_layoutstat *layoutstat, + __u64 requested) +{ + struct nfs4_ff_io_stat *iostat = &layoutstat->io_stat; + + iostat->ops_requested++; + iostat->bytes_requested += requested; +} + +static void +nfs4_ff_layout_stat_io_update_completed(struct nfs4_ff_layoutstat *layoutstat, + __u64 requested, + __u64 completed, + ktime_t time_completed) +{ + struct nfs4_ff_io_stat *iostat = &layoutstat->io_stat; + ktime_t timer; + + iostat->ops_completed++; + iostat->bytes_completed += completed; + iostat->bytes_not_delivered += requested - completed; + + timer = nfs4_ff_end_busy_timer(&layoutstat->busy_timer); + iostat->total_busy_time = + ktime_add(iostat->total_busy_time, timer); + iostat->aggregate_completion_time = + ktime_add(iostat->aggregate_completion_time, time_completed); +} + +static void +nfs4_ff_layout_stat_io_start_read(struct nfs4_ff_layout_mirror *mirror, + __u64 requested) +{ + bool report; + + spin_lock(&mirror->lock); + report = nfs4_ff_layoutstat_start_io(mirror, &mirror->read_stat); + nfs4_ff_layout_stat_io_update_requested(&mirror->read_stat, requested); + spin_unlock(&mirror->lock); + + if (report) + pnfs_report_layoutstat(mirror->lseg->pls_layout->plh_inode); +} + +static void +nfs4_ff_layout_stat_io_end_read(struct rpc_task *task, + struct nfs4_ff_layout_mirror *mirror, + __u64 requested, + __u64 completed) +{ + spin_lock(&mirror->lock); + nfs4_ff_layout_stat_io_update_completed(&mirror->read_stat, + requested, completed, + nfs4_ff_layout_calc_completion_time(task)); + spin_unlock(&mirror->lock); +} + +static void +nfs4_ff_layout_stat_io_start_write(struct nfs4_ff_layout_mirror *mirror, + __u64 requested) +{ + bool report; + + spin_lock(&mirror->lock); + report = nfs4_ff_layoutstat_start_io(mirror , &mirror->write_stat); + nfs4_ff_layout_stat_io_update_requested(&mirror->write_stat, requested); + spin_unlock(&mirror->lock); + + if (report) + pnfs_report_layoutstat(mirror->lseg->pls_layout->plh_inode); +} + +static void +nfs4_ff_layout_stat_io_end_write(struct rpc_task *task, + struct nfs4_ff_layout_mirror *mirror, + __u64 requested, + __u64 completed, + enum nfs3_stable_how committed) +{ + if (committed == NFS_UNSTABLE) + requested = completed = 0; + + spin_lock(&mirror->lock); + nfs4_ff_layout_stat_io_update_completed(&mirror->write_stat, + requested, completed, + nfs4_ff_layout_calc_completion_time(task)); + spin_unlock(&mirror->lock); +} + static int ff_layout_alloc_commit_info(struct pnfs_layout_segment *lseg, struct nfs_commit_info *cinfo, @@ -631,7 +774,7 @@ static void ff_layout_reset_write(struct nfs_pgio_header *hdr, bool retry_pnfs) nfs_direct_set_resched_writes(hdr->dreq); /* fake unstable write to let common nfs resend pages */ hdr->verf.committed = NFS_UNSTABLE; - hdr->good_bytes = 0; + hdr->good_bytes = hdr->args.count; } return; } @@ -879,6 +1022,12 @@ static int ff_layout_read_done_cb(struct rpc_task *task, return 0; } +static bool +ff_layout_need_layoutcommit(struct pnfs_layout_segment *lseg) +{ + return !(FF_LAYOUT_LSEG(lseg)->flags & FF_FLAGS_NO_LAYOUTCOMMIT); +} + /* * We reference the rpc_cred of the first WRITE that triggers the need for * a LAYOUTCOMMIT, and use it to send the layoutcommit compound. @@ -891,6 +1040,9 @@ static int ff_layout_read_done_cb(struct rpc_task *task, static void ff_layout_set_layoutcommit(struct nfs_pgio_header *hdr) { + if (!ff_layout_need_layoutcommit(hdr->lseg)) + return; + pnfs_set_layoutcommit(hdr->inode, hdr->lseg, hdr->mds_offset + hdr->res.count); dprintk("%s inode %lu pls_end_pos %lu\n", __func__, hdr->inode->i_ino, @@ -909,6 +1061,10 @@ ff_layout_reset_to_mds(struct pnfs_layout_segment *lseg, int idx) static int ff_layout_read_prepare_common(struct rpc_task *task, struct nfs_pgio_header *hdr) { + nfs4_ff_layout_stat_io_start_read( + FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx), + hdr->args.count); + if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags))) { rpc_exit(task, -EIO); return -EIO; @@ -962,15 +1118,15 @@ static void ff_layout_read_prepare_v4(struct rpc_task *task, void *data) { struct nfs_pgio_header *hdr = data; - if (ff_layout_read_prepare_common(task, hdr)) - return; - if (ff_layout_setup_sequence(hdr->ds_clp, &hdr->args.seq_args, &hdr->res.seq_res, task)) return; + if (ff_layout_read_prepare_common(task, hdr)) + return; + if (nfs4_set_rw_stateid(&hdr->args.stateid, hdr->args.context, hdr->args.lock_context, FMODE_READ) == -EIO) rpc_exit(task, -EIO); /* lost lock, terminate I/O */ @@ -982,6 +1138,10 @@ static void ff_layout_read_call_done(struct rpc_task *task, void *data) dprintk("--> %s task->tk_status %d\n", __func__, task->tk_status); + nfs4_ff_layout_stat_io_end_read(task, + FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx), + hdr->args.count, hdr->res.count); + if (test_bit(NFS_IOHDR_REDO, &hdr->flags) && task->tk_status == 0) { nfs4_sequence_done(task, &hdr->res.seq_res); @@ -1074,7 +1234,8 @@ static int ff_layout_commit_done_cb(struct rpc_task *task, return -EAGAIN; } - if (data->verf.committed == NFS_UNSTABLE) + if (data->verf.committed == NFS_UNSTABLE + && ff_layout_need_layoutcommit(data->lseg)) pnfs_set_layoutcommit(data->inode, data->lseg, data->lwb); return 0; @@ -1083,6 +1244,10 @@ static int ff_layout_commit_done_cb(struct rpc_task *task, static int ff_layout_write_prepare_common(struct rpc_task *task, struct nfs_pgio_header *hdr) { + nfs4_ff_layout_stat_io_start_write( + FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx), + hdr->args.count); + if (unlikely(test_bit(NFS_CONTEXT_BAD, &hdr->args.context->flags))) { rpc_exit(task, -EIO); return -EIO; @@ -1116,15 +1281,15 @@ static void ff_layout_write_prepare_v4(struct rpc_task *task, void *data) { struct nfs_pgio_header *hdr = data; - if (ff_layout_write_prepare_common(task, hdr)) - return; - if (ff_layout_setup_sequence(hdr->ds_clp, &hdr->args.seq_args, &hdr->res.seq_res, task)) return; + if (ff_layout_write_prepare_common(task, hdr)) + return; + if (nfs4_set_rw_stateid(&hdr->args.stateid, hdr->args.context, hdr->args.lock_context, FMODE_WRITE) == -EIO) rpc_exit(task, -EIO); /* lost lock, terminate I/O */ @@ -1134,6 +1299,11 @@ static void ff_layout_write_call_done(struct rpc_task *task, void *data) { struct nfs_pgio_header *hdr = data; + nfs4_ff_layout_stat_io_end_write(task, + FF_LAYOUT_COMP(hdr->lseg, hdr->pgio_mirror_idx), + hdr->args.count, hdr->res.count, + hdr->res.verf->committed); + if (test_bit(NFS_IOHDR_REDO, &hdr->flags) && task->tk_status == 0) { nfs4_sequence_done(task, &hdr->res.seq_res); @@ -1152,8 +1322,17 @@ static void ff_layout_write_count_stats(struct rpc_task *task, void *data) &NFS_CLIENT(hdr->inode)->cl_metrics[NFSPROC4_CLNT_WRITE]); } +static void ff_layout_commit_prepare_common(struct rpc_task *task, + struct nfs_commit_data *cdata) +{ + nfs4_ff_layout_stat_io_start_write( + FF_LAYOUT_COMP(cdata->lseg, cdata->ds_commit_index), + 0); +} + static void ff_layout_commit_prepare_v3(struct rpc_task *task, void *data) { + ff_layout_commit_prepare_common(task, data); rpc_call_start(task); } @@ -1161,10 +1340,30 @@ static void ff_layout_commit_prepare_v4(struct rpc_task *task, void *data) { struct nfs_commit_data *wdata = data; - ff_layout_setup_sequence(wdata->ds_clp, + if (ff_layout_setup_sequence(wdata->ds_clp, &wdata->args.seq_args, &wdata->res.seq_res, - task); + task)) + return; + ff_layout_commit_prepare_common(task, data); +} + +static void ff_layout_commit_done(struct rpc_task *task, void *data) +{ + struct nfs_commit_data *cdata = data; + struct nfs_page *req; + __u64 count = 0; + + if (task->tk_status == 0) { + list_for_each_entry(req, &cdata->pages, wb_list) + count += req->wb_bytes; + } + + nfs4_ff_layout_stat_io_end_write(task, + FF_LAYOUT_COMP(cdata->lseg, cdata->ds_commit_index), + count, count, NFS_FILE_SYNC); + + pnfs_generic_write_commit_done(task, data); } static void ff_layout_commit_count_stats(struct rpc_task *task, void *data) @@ -1205,14 +1404,14 @@ static const struct rpc_call_ops ff_layout_write_call_ops_v4 = { static const struct rpc_call_ops ff_layout_commit_call_ops_v3 = { .rpc_call_prepare = ff_layout_commit_prepare_v3, - .rpc_call_done = pnfs_generic_write_commit_done, + .rpc_call_done = ff_layout_commit_done, .rpc_count_stats = ff_layout_commit_count_stats, .rpc_release = pnfs_generic_commit_release, }; static const struct rpc_call_ops ff_layout_commit_call_ops_v4 = { .rpc_call_prepare = ff_layout_commit_prepare_v4, - .rpc_call_done = pnfs_generic_write_commit_done, + .rpc_call_done = ff_layout_commit_done, .rpc_count_stats = ff_layout_commit_count_stats, .rpc_release = pnfs_generic_commit_release, }; @@ -1256,7 +1455,6 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr) fh = nfs4_ff_layout_select_ds_fh(lseg, idx); if (fh) hdr->args.fh = fh; - /* * Note that if we ever decide to split across DSes, * then we may need to handle dense-like offsets. @@ -1385,6 +1583,7 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how) fh = select_ds_fh_from_commit(lseg, data->ds_commit_index); if (fh) data->args.fh = fh; + return nfs_initiate_commit(ds_clnt, data, ds->ds_clp->rpc_ops, vers == 3 ? &ff_layout_commit_call_ops_v3 : &ff_layout_commit_call_ops_v4, @@ -1488,6 +1687,247 @@ out: dprintk("%s: Return\n", __func__); } +static int +ff_layout_ntop4(const struct sockaddr *sap, char *buf, const size_t buflen) +{ + const struct sockaddr_in *sin = (struct sockaddr_in *)sap; + + return snprintf(buf, buflen, "%pI4", &sin->sin_addr); +} + +static size_t +ff_layout_ntop6_noscopeid(const struct sockaddr *sap, char *buf, + const int buflen) +{ + const struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *)sap; + const struct in6_addr *addr = &sin6->sin6_addr; + + /* + * RFC 4291, Section 2.2.2 + * + * Shorthanded ANY address + */ + if (ipv6_addr_any(addr)) + return snprintf(buf, buflen, "::"); + + /* + * RFC 4291, Section 2.2.2 + * + * Shorthanded loopback address + */ + if (ipv6_addr_loopback(addr)) + return snprintf(buf, buflen, "::1"); + + /* + * RFC 4291, Section 2.2.3 + * + * Special presentation address format for mapped v4 + * addresses. + */ + if (ipv6_addr_v4mapped(addr)) + return snprintf(buf, buflen, "::ffff:%pI4", + &addr->s6_addr32[3]); + + /* + * RFC 4291, Section 2.2.1 + */ + return snprintf(buf, buflen, "%pI6c", addr); +} + +/* Derived from rpc_sockaddr2uaddr */ +static void +ff_layout_encode_netaddr(struct xdr_stream *xdr, struct nfs4_pnfs_ds_addr *da) +{ + struct sockaddr *sap = (struct sockaddr *)&da->da_addr; + char portbuf[RPCBIND_MAXUADDRPLEN]; + char addrbuf[RPCBIND_MAXUADDRLEN]; + char *netid; + unsigned short port; + int len, netid_len; + __be32 *p; + + switch (sap->sa_family) { + case AF_INET: + if (ff_layout_ntop4(sap, addrbuf, sizeof(addrbuf)) == 0) + return; + port = ntohs(((struct sockaddr_in *)sap)->sin_port); + netid = "tcp"; + netid_len = 3; + break; + case AF_INET6: + if (ff_layout_ntop6_noscopeid(sap, addrbuf, sizeof(addrbuf)) == 0) + return; + port = ntohs(((struct sockaddr_in6 *)sap)->sin6_port); + netid = "tcp6"; + netid_len = 4; + break; + default: + /* we only support tcp and tcp6 */ + WARN_ON_ONCE(1); + return; + } + + snprintf(portbuf, sizeof(portbuf), ".%u.%u", port >> 8, port & 0xff); + len = strlcat(addrbuf, portbuf, sizeof(addrbuf)); + + p = xdr_reserve_space(xdr, 4 + netid_len); + xdr_encode_opaque(p, netid, netid_len); + + p = xdr_reserve_space(xdr, 4 + len); + xdr_encode_opaque(p, addrbuf, len); +} + +static void +ff_layout_encode_nfstime(struct xdr_stream *xdr, + ktime_t t) +{ + struct timespec64 ts; + __be32 *p; + + p = xdr_reserve_space(xdr, 12); + ts = ktime_to_timespec64(t); + p = xdr_encode_hyper(p, ts.tv_sec); + *p++ = cpu_to_be32(ts.tv_nsec); +} + +static void +ff_layout_encode_io_latency(struct xdr_stream *xdr, + struct nfs4_ff_io_stat *stat) +{ + __be32 *p; + + p = xdr_reserve_space(xdr, 5 * 8); + p = xdr_encode_hyper(p, stat->ops_requested); + p = xdr_encode_hyper(p, stat->bytes_requested); + p = xdr_encode_hyper(p, stat->ops_completed); + p = xdr_encode_hyper(p, stat->bytes_completed); + p = xdr_encode_hyper(p, stat->bytes_not_delivered); + ff_layout_encode_nfstime(xdr, stat->total_busy_time); + ff_layout_encode_nfstime(xdr, stat->aggregate_completion_time); +} + +static void +ff_layout_encode_layoutstats(struct xdr_stream *xdr, + struct nfs42_layoutstat_args *args, + struct nfs42_layoutstat_devinfo *devinfo) +{ + struct nfs4_ff_layout_mirror *mirror = devinfo->layout_private; + struct nfs4_pnfs_ds_addr *da; + struct nfs4_pnfs_ds *ds = mirror->mirror_ds->ds; + struct nfs_fh *fh = &mirror->fh_versions[0]; + __be32 *p, *start; + + da = list_first_entry(&ds->ds_addrs, struct nfs4_pnfs_ds_addr, da_node); + dprintk("%s: DS %s: encoding address %s\n", + __func__, ds->ds_remotestr, da->da_remotestr); + /* layoutupdate length */ + start = xdr_reserve_space(xdr, 4); + /* netaddr4 */ + ff_layout_encode_netaddr(xdr, da); + /* nfs_fh4 */ + p = xdr_reserve_space(xdr, 4 + fh->size); + xdr_encode_opaque(p, fh->data, fh->size); + /* ff_io_latency4 read */ + spin_lock(&mirror->lock); + ff_layout_encode_io_latency(xdr, &mirror->read_stat.io_stat); + /* ff_io_latency4 write */ + ff_layout_encode_io_latency(xdr, &mirror->write_stat.io_stat); + spin_unlock(&mirror->lock); + /* nfstime4 */ + ff_layout_encode_nfstime(xdr, ktime_sub(ktime_get(), mirror->start_time)); + /* bool */ + p = xdr_reserve_space(xdr, 4); + *p = cpu_to_be32(false); + + *start = cpu_to_be32((xdr->p - start - 1) * 4); +} + +static bool +ff_layout_mirror_prepare_stats(struct nfs42_layoutstat_args *args, + struct pnfs_layout_segment *pls, + int *dev_count, int dev_limit) +{ + struct nfs4_ff_layout_mirror *mirror; + struct nfs4_deviceid_node *dev; + struct nfs42_layoutstat_devinfo *devinfo; + int i; + + for (i = 0; i <= FF_LAYOUT_MIRROR_COUNT(pls); i++) { + if (*dev_count >= dev_limit) + break; + mirror = FF_LAYOUT_COMP(pls, i); + if (!mirror || !mirror->mirror_ds) + continue; + dev = FF_LAYOUT_DEVID_NODE(pls, i); + devinfo = &args->devinfo[*dev_count]; + memcpy(&devinfo->dev_id, &dev->deviceid, NFS4_DEVICEID4_SIZE); + devinfo->offset = pls->pls_range.offset; + devinfo->length = pls->pls_range.length; + /* well, we don't really know if IO is continuous or not! */ + devinfo->read_count = mirror->read_stat.io_stat.bytes_completed; + devinfo->read_bytes = mirror->read_stat.io_stat.bytes_completed; + devinfo->write_count = mirror->write_stat.io_stat.bytes_completed; + devinfo->write_bytes = mirror->write_stat.io_stat.bytes_completed; + devinfo->layout_type = LAYOUT_FLEX_FILES; + devinfo->layoutstats_encode = ff_layout_encode_layoutstats; + devinfo->layout_private = mirror; + /* lseg refcount put in cleanup_layoutstats */ + pnfs_get_lseg(pls); + + ++(*dev_count); + } + + return *dev_count < dev_limit; +} + +static int +ff_layout_prepare_layoutstats(struct nfs42_layoutstat_args *args) +{ + struct pnfs_layout_segment *pls; + int dev_count = 0; + + spin_lock(&args->inode->i_lock); + list_for_each_entry(pls, &NFS_I(args->inode)->layout->plh_segs, pls_list) { + dev_count += FF_LAYOUT_MIRROR_COUNT(pls); + } + spin_unlock(&args->inode->i_lock); + /* For now, send at most PNFS_LAYOUTSTATS_MAXDEV statistics */ + if (dev_count > PNFS_LAYOUTSTATS_MAXDEV) { + dprintk("%s: truncating devinfo to limit (%d:%d)\n", + __func__, dev_count, PNFS_LAYOUTSTATS_MAXDEV); + dev_count = PNFS_LAYOUTSTATS_MAXDEV; + } + args->devinfo = kmalloc(dev_count * sizeof(*args->devinfo), GFP_KERNEL); + if (!args->devinfo) + return -ENOMEM; + + dev_count = 0; + spin_lock(&args->inode->i_lock); + list_for_each_entry(pls, &NFS_I(args->inode)->layout->plh_segs, pls_list) { + if (!ff_layout_mirror_prepare_stats(args, pls, &dev_count, + PNFS_LAYOUTSTATS_MAXDEV)) { + break; + } + } + spin_unlock(&args->inode->i_lock); + args->num_dev = dev_count; + + return 0; +} + +static void +ff_layout_cleanup_layoutstats(struct nfs42_layoutstat_data *data) +{ + struct nfs4_ff_layout_mirror *mirror; + int i; + + for (i = 0; i < data->args.num_dev; i++) { + mirror = data->args.devinfo[i].layout_private; + data->args.devinfo[i].layout_private = NULL; + pnfs_put_lseg(mirror->lseg); + } +} + static struct pnfs_layoutdriver_type flexfilelayout_type = { .id = LAYOUT_FLEX_FILES, .name = "LAYOUT_FLEX_FILES", @@ -1510,6 +1950,8 @@ static struct pnfs_layoutdriver_type flexfilelayout_type = { .alloc_deviceid_node = ff_layout_alloc_deviceid_node, .encode_layoutreturn = ff_layout_encode_layoutreturn, .sync = pnfs_nfs_generic_sync, + .prepare_layoutstats = ff_layout_prepare_layoutstats, + .cleanup_layoutstats = ff_layout_cleanup_layoutstats, }; static int __init nfs4flexfilelayout_init(void) diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h index 070f204..f92f9a0 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.h +++ b/fs/nfs/flexfilelayout/flexfilelayout.h @@ -9,12 +9,17 @@ #ifndef FS_NFS_NFS4FLEXFILELAYOUT_H #define FS_NFS_NFS4FLEXFILELAYOUT_H +#define FF_FLAGS_NO_LAYOUTCOMMIT 1 + #include "../pnfs.h" /* XXX: Let's filter out insanely large mirror count for now to avoid oom * due to network error etc. */ #define NFS4_FLEXFILE_LAYOUT_MAX_MIRROR_CNT 4096 +/* LAYOUTSTATS report interval in ms */ +#define FF_LAYOUTSTATS_REPORT_INTERVAL (60000L) + struct nfs4_ff_ds_version { u32 version; u32 minor_version; @@ -41,24 +46,48 @@ struct nfs4_ff_layout_ds_err { struct nfs4_deviceid deviceid; }; +struct nfs4_ff_io_stat { + __u64 ops_requested; + __u64 bytes_requested; + __u64 ops_completed; + __u64 bytes_completed; + __u64 bytes_not_delivered; + ktime_t total_busy_time; + ktime_t aggregate_completion_time; +}; + +struct nfs4_ff_busy_timer { + ktime_t start_time; + atomic_t n_ops; +}; + +struct nfs4_ff_layoutstat { + struct nfs4_ff_io_stat io_stat; + struct nfs4_ff_busy_timer busy_timer; +}; + struct nfs4_ff_layout_mirror { + struct pnfs_layout_segment *lseg; /* back pointer */ u32 ds_count; u32 efficiency; struct nfs4_ff_layout_ds *mirror_ds; u32 fh_versions_cnt; struct nfs_fh *fh_versions; nfs4_stateid stateid; - struct nfs4_string user_name; - struct nfs4_string group_name; u32 uid; u32 gid; struct rpc_cred *cred; spinlock_t lock; + struct nfs4_ff_layoutstat read_stat; + struct nfs4_ff_layoutstat write_stat; + ktime_t start_time; + ktime_t last_report_time; }; struct nfs4_ff_layout_segment { struct pnfs_layout_segment generic_hdr; u64 stripe_unit; + u32 flags; u32 mirror_array_cnt; struct nfs4_ff_layout_mirror **mirror_array; }; diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c index 77a2d02..f13e196 100644 --- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c +++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c @@ -324,7 +324,8 @@ static int ff_layout_update_mirror_cred(struct nfs4_ff_layout_mirror *mirror, __func__, PTR_ERR(cred)); return PTR_ERR(cred); } else { - mirror->cred = cred; + if (cmpxchg(&mirror->cred, NULL, cred)) + put_rpccred(cred); } } return 0; @@ -386,7 +387,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx, /* matching smp_wmb() in _nfs4_pnfs_v3/4_ds_connect */ smp_rmb(); if (ds->ds_clp) - goto out; + goto out_update_creds; flavor = nfs4_ff_layout_choose_authflavor(mirror); @@ -430,7 +431,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx, } } } - +out_update_creds: if (ff_layout_update_mirror_cred(mirror, ds)) ds = NULL; out: diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index f734562..b77b328 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -678,6 +678,8 @@ int nfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) if (!err) { generic_fillattr(inode, stat); stat->ino = nfs_compat_user_ino64(NFS_FILEID(inode)); + if (S_ISDIR(inode->i_mode)) + stat->blksize = NFS_SERVER(inode)->dtsize; } out: trace_nfs_getattr_exit(inode, err); @@ -2008,17 +2010,15 @@ static int __init init_nfs_fs(void) if (err) goto out1; -#ifdef CONFIG_PROC_FS rpc_proc_register(&init_net, &nfs_rpcstat); -#endif - if ((err = register_nfs_fs()) != 0) + + err = register_nfs_fs(); + if (err) goto out0; return 0; out0: -#ifdef CONFIG_PROC_FS rpc_proc_unregister(&init_net, "nfs"); -#endif nfs_destroy_directcache(); out1: nfs_destroy_writepagecache(); @@ -2049,9 +2049,7 @@ static void __exit exit_nfs_fs(void) nfs_destroy_nfspagecache(); nfs_fscache_unregister(); unregister_pernet_subsys(&nfs_net_ops); -#ifdef CONFIG_PROC_FS rpc_proc_unregister(&init_net, "nfs"); -#endif unregister_nfs_fs(); nfs_fs_proc_exit(); nfsiod_stop(); diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 9e6475b..7e3c460 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -607,7 +607,7 @@ void nfs_mark_page_unstable(struct page *page) struct inode *inode = page_file_mapping(page)->host; inc_zone_page_state(page, NR_UNSTABLE_NFS); - inc_bdi_stat(inode_to_bdi(inode), BDI_RECLAIMABLE); + inc_wb_stat(&inode_to_bdi(inode)->wb, WB_RECLAIMABLE); __mark_inode_dirty(inode, I_DIRTY_DATASYNC); } diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 53852a4..9b04c2e 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -1342,7 +1342,7 @@ static void nfs3_xdr_enc_setacl3args(struct rpc_rqst *req, if (args->npages != 0) xdr_write_pages(xdr, args->pages, 0, args->len); else - xdr_reserve_space(xdr, NFS_ACL_INLINE_BUFSIZE); + xdr_reserve_space(xdr, args->len); error = nfsacl_encode(xdr->buf, base, args->inode, (args->mask & NFS_ACL) ? diff --git a/fs/nfs/nfs42.h b/fs/nfs/nfs42.h index 7afb894..ff66ae7 100644 --- a/fs/nfs/nfs42.h +++ b/fs/nfs/nfs42.h @@ -5,11 +5,18 @@ #ifndef __LINUX_FS_NFS_NFS4_2_H #define __LINUX_FS_NFS_NFS4_2_H +/* + * FIXME: four LAYOUTSTATS calls per compound at most! Do we need to support + * more? Need to consider not to pre-alloc too much for a compound. + */ +#define PNFS_LAYOUTSTATS_MAXDEV (4) + /* nfs4.2proc.c */ int nfs42_proc_allocate(struct file *, loff_t, loff_t); int nfs42_proc_deallocate(struct file *, loff_t, loff_t); loff_t nfs42_proc_llseek(struct file *, loff_t, int); - +int nfs42_proc_layoutstats_generic(struct nfs_server *, + struct nfs42_layoutstat_data *); /* nfs4.2xdr.h */ extern struct rpc_procinfo nfs4_2_procedures[]; diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c index 3a9e752..f486b80 100644 --- a/fs/nfs/nfs42proc.c +++ b/fs/nfs/nfs42proc.c @@ -10,6 +10,11 @@ #include <linux/nfs_fs.h> #include "nfs4_fs.h" #include "nfs42.h" +#include "iostat.h" +#include "pnfs.h" +#include "internal.h" + +#define NFSDBG_FACILITY NFSDBG_PNFS static int nfs42_set_rw_stateid(nfs4_stateid *dst, struct file *file, fmode_t fmode) @@ -165,3 +170,85 @@ loff_t nfs42_proc_llseek(struct file *filep, loff_t offset, int whence) return vfs_setpos(filep, res.sr_offset, inode->i_sb->s_maxbytes); } + +static void +nfs42_layoutstat_prepare(struct rpc_task *task, void *calldata) +{ + struct nfs42_layoutstat_data *data = calldata; + struct nfs_server *server = NFS_SERVER(data->args.inode); + + nfs41_setup_sequence(nfs4_get_session(server), &data->args.seq_args, + &data->res.seq_res, task); +} + +static void +nfs42_layoutstat_done(struct rpc_task *task, void *calldata) +{ + struct nfs42_layoutstat_data *data = calldata; + + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return; + + switch (task->tk_status) { + case 0: + break; + case -ENOTSUPP: + case -EOPNOTSUPP: + NFS_SERVER(data->inode)->caps &= ~NFS_CAP_LAYOUTSTATS; + default: + dprintk("%s server returns %d\n", __func__, task->tk_status); + } +} + +static void +nfs42_layoutstat_release(void *calldata) +{ + struct nfs42_layoutstat_data *data = calldata; + struct nfs_server *nfss = NFS_SERVER(data->args.inode); + + if (nfss->pnfs_curr_ld->cleanup_layoutstats) + nfss->pnfs_curr_ld->cleanup_layoutstats(data); + + pnfs_put_layout_hdr(NFS_I(data->args.inode)->layout); + smp_mb__before_atomic(); + clear_bit(NFS_INO_LAYOUTSTATS, &NFS_I(data->args.inode)->flags); + smp_mb__after_atomic(); + nfs_iput_and_deactive(data->inode); + kfree(data->args.devinfo); + kfree(data); +} + +static const struct rpc_call_ops nfs42_layoutstat_ops = { + .rpc_call_prepare = nfs42_layoutstat_prepare, + .rpc_call_done = nfs42_layoutstat_done, + .rpc_release = nfs42_layoutstat_release, +}; + +int nfs42_proc_layoutstats_generic(struct nfs_server *server, + struct nfs42_layoutstat_data *data) +{ + struct rpc_message msg = { + .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LAYOUTSTATS], + .rpc_argp = &data->args, + .rpc_resp = &data->res, + }; + struct rpc_task_setup task_setup = { + .rpc_client = server->client, + .rpc_message = &msg, + .callback_ops = &nfs42_layoutstat_ops, + .callback_data = data, + .flags = RPC_TASK_ASYNC, + }; + struct rpc_task *task; + + data->inode = nfs_igrab_and_active(data->args.inode); + if (!data->inode) { + nfs42_layoutstat_release(data); + return -EAGAIN; + } + nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 0); + task = rpc_run_task(&task_setup); + if (IS_ERR(task)) + return PTR_ERR(task); + return 0; +} diff --git a/fs/nfs/nfs42xdr.c b/fs/nfs/nfs42xdr.c index 1a25b27..a6bd27d 100644 --- a/fs/nfs/nfs42xdr.c +++ b/fs/nfs/nfs42xdr.c @@ -4,6 +4,8 @@ #ifndef __LINUX_FS_NFS_NFS4_2XDR_H #define __LINUX_FS_NFS_NFS4_2XDR_H +#include "nfs42.h" + #define encode_fallocate_maxsz (encode_stateid_maxsz + \ 2 /* offset */ + \ 2 /* length */) @@ -22,6 +24,16 @@ 1 /* whence */ + \ 2 /* offset */ + \ 2 /* length */) +#define encode_io_info_maxsz 4 +#define encode_layoutstats_maxsz (op_decode_hdr_maxsz + \ + 2 /* offset */ + \ + 2 /* length */ + \ + encode_stateid_maxsz + \ + encode_io_info_maxsz + \ + encode_io_info_maxsz + \ + 1 /* opaque devaddr4 length */ + \ + XDR_QUADLEN(PNFS_LAYOUTSTATS_MAXSIZE)) +#define decode_layoutstats_maxsz (op_decode_hdr_maxsz) #define NFS4_enc_allocate_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ @@ -45,6 +57,14 @@ #define NFS4_dec_seek_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ decode_seek_maxsz) +#define NFS4_enc_layoutstats_sz (compound_encode_hdr_maxsz + \ + encode_sequence_maxsz + \ + encode_putfh_maxsz + \ + PNFS_LAYOUTSTATS_MAXDEV * encode_layoutstats_maxsz) +#define NFS4_dec_layoutstats_sz (compound_decode_hdr_maxsz + \ + decode_sequence_maxsz + \ + decode_putfh_maxsz + \ + PNFS_LAYOUTSTATS_MAXDEV * decode_layoutstats_maxsz) static void encode_fallocate(struct xdr_stream *xdr, @@ -81,6 +101,33 @@ static void encode_seek(struct xdr_stream *xdr, encode_uint32(xdr, args->sa_what); } +static void encode_layoutstats(struct xdr_stream *xdr, + struct nfs42_layoutstat_args *args, + struct nfs42_layoutstat_devinfo *devinfo, + struct compound_hdr *hdr) +{ + __be32 *p; + + encode_op_hdr(xdr, OP_LAYOUTSTATS, decode_layoutstats_maxsz, hdr); + p = reserve_space(xdr, 8 + 8); + p = xdr_encode_hyper(p, devinfo->offset); + p = xdr_encode_hyper(p, devinfo->length); + encode_nfs4_stateid(xdr, &args->stateid); + p = reserve_space(xdr, 4*8 + NFS4_DEVICEID4_SIZE + 4); + p = xdr_encode_hyper(p, devinfo->read_count); + p = xdr_encode_hyper(p, devinfo->read_bytes); + p = xdr_encode_hyper(p, devinfo->write_count); + p = xdr_encode_hyper(p, devinfo->write_bytes); + p = xdr_encode_opaque_fixed(p, devinfo->dev_id.data, + NFS4_DEVICEID4_SIZE); + /* Encode layoutupdate4 */ + *p++ = cpu_to_be32(devinfo->layout_type); + if (devinfo->layoutstats_encode != NULL) + devinfo->layoutstats_encode(xdr, args, devinfo); + else + encode_uint32(xdr, 0); +} + /* * Encode ALLOCATE request */ @@ -137,6 +184,28 @@ static void nfs4_xdr_enc_seek(struct rpc_rqst *req, encode_nops(&hdr); } +/* + * Encode LAYOUTSTATS request + */ +static void nfs4_xdr_enc_layoutstats(struct rpc_rqst *req, + struct xdr_stream *xdr, + struct nfs42_layoutstat_args *args) +{ + int i; + + struct compound_hdr hdr = { + .minorversion = nfs4_xdr_minorversion(&args->seq_args), + }; + + encode_compound_hdr(xdr, req, &hdr); + encode_sequence(xdr, &args->seq_args, &hdr); + encode_putfh(xdr, args->fh, &hdr); + WARN_ON(args->num_dev > PNFS_LAYOUTSTATS_MAXDEV); + for (i = 0; i < args->num_dev; i++) + encode_layoutstats(xdr, args, &args->devinfo[i], &hdr); + encode_nops(&hdr); +} + static int decode_allocate(struct xdr_stream *xdr, struct nfs42_falloc_res *res) { return decode_op_hdr(xdr, OP_ALLOCATE); @@ -169,6 +238,12 @@ out_overflow: return -EIO; } +static int decode_layoutstats(struct xdr_stream *xdr, + struct nfs42_layoutstat_res *res) +{ + return decode_op_hdr(xdr, OP_LAYOUTSTATS); +} + /* * Decode ALLOCATE request */ @@ -246,4 +321,35 @@ static int nfs4_xdr_dec_seek(struct rpc_rqst *rqstp, out: return status; } + +/* + * Decode LAYOUTSTATS request + */ +static int nfs4_xdr_dec_layoutstats(struct rpc_rqst *rqstp, + struct xdr_stream *xdr, + struct nfs42_layoutstat_res *res) +{ + struct compound_hdr hdr; + int status, i; + + status = decode_compound_hdr(xdr, &hdr); + if (status) + goto out; + status = decode_sequence(xdr, &res->seq_res, rqstp); + if (status) + goto out; + status = decode_putfh(xdr); + if (status) + goto out; + WARN_ON(res->num_dev > PNFS_LAYOUTSTATS_MAXDEV); + for (i = 0; i < res->num_dev; i++) { + status = decode_layoutstats(xdr, res); + if (status) + goto out; + } +out: + res->rpc_status = status; + return status; +} + #endif /* __LINUX_FS_NFS_NFS4_2XDR_H */ diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index fdef424..ea3bee9 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -233,6 +233,7 @@ extern int nfs4_handle_exception(struct nfs_server *, int, struct nfs4_exception extern int nfs4_call_sync(struct rpc_clnt *, struct nfs_server *, struct rpc_message *, struct nfs4_sequence_args *, struct nfs4_sequence_res *, int); +extern void nfs4_init_sequence(struct nfs4_sequence_args *, struct nfs4_sequence_res *, int); extern int nfs4_proc_setclientid(struct nfs_client *, u32, unsigned short, struct rpc_cred *, struct nfs4_setclientid_res *); extern int nfs4_proc_setclientid_confirm(struct nfs_client *, struct nfs4_setclientid_res *arg, struct rpc_cred *); extern int nfs4_proc_get_rootfh(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *, bool); diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index e42be52..3aa6a9b 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -676,7 +676,6 @@ found: break; } - /* No matching nfs_client found. */ spin_unlock(&nn->nfs_client_lock); dprintk("NFS: <-- %s status = %d\n", __func__, status); nfs_put_client(prev); diff --git a/fs/nfs/nfs4file.c b/fs/nfs/nfs4file.c index f58c17b..dcd39d4 100644 --- a/fs/nfs/nfs4file.c +++ b/fs/nfs/nfs4file.c @@ -41,6 +41,10 @@ nfs4_file_open(struct inode *inode, struct file *filp) dprintk("NFS: open file(%pd2)\n", dentry); + err = nfs_check_flags(openflags); + if (err) + return err; + if ((openflags & O_ACCMODE) == 3) openflags--; diff --git a/fs/nfs/nfs4getroot.c b/fs/nfs/nfs4getroot.c index c0b3a16..039b3eb 100644 --- a/fs/nfs/nfs4getroot.c +++ b/fs/nfs/nfs4getroot.c @@ -35,13 +35,6 @@ int nfs4_get_rootfh(struct nfs_server *server, struct nfs_fh *mntfh, bool auth_p goto out; } - if (fsinfo.fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL) { - printk(KERN_ERR "nfs4_get_rootfh:" - " getroot obtained referral\n"); - ret = -EREMOTE; - goto out; - } - memcpy(&server->fsid, &fsinfo.fattr->fsid, sizeof(server->fsid)); out: nfs_free_fattr(fsinfo.fattr); diff --git a/fs/nfs/nfs4idmap.c b/fs/nfs/nfs4idmap.c index 2e1737c..535dfc6 100644 --- a/fs/nfs/nfs4idmap.c +++ b/fs/nfs/nfs4idmap.c @@ -494,12 +494,7 @@ nfs_idmap_delete(struct nfs_client *clp) int nfs_idmap_init(void) { - int ret; - ret = nfs_idmap_init_keyring(); - if (ret != 0) - goto out; -out: - return ret; + return nfs_idmap_init_keyring(); } void nfs_idmap_quit(void) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 45b35b9..6f228b5 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -38,6 +38,7 @@ #include <linux/mm.h> #include <linux/delay.h> #include <linux/errno.h> +#include <linux/file.h> #include <linux/string.h> #include <linux/ratelimit.h> #include <linux/printk.h> @@ -355,6 +356,9 @@ int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_ case 0: return 0; case -NFS4ERR_OPENMODE: + case -NFS4ERR_DELEG_REVOKED: + case -NFS4ERR_ADMIN_REVOKED: + case -NFS4ERR_BAD_STATEID: if (inode && nfs4_have_delegation(inode, FMODE_READ)) { nfs4_inode_return_delegation(inode); exception->retry = 1; @@ -366,15 +370,6 @@ int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_ if (ret < 0) break; goto wait_on_recovery; - case -NFS4ERR_DELEG_REVOKED: - case -NFS4ERR_ADMIN_REVOKED: - case -NFS4ERR_BAD_STATEID: - if (state == NULL) - break; - ret = nfs4_schedule_stateid_recovery(server, state); - if (ret < 0) - break; - goto wait_on_recovery; case -NFS4ERR_EXPIRED: if (state != NULL) { ret = nfs4_schedule_stateid_recovery(server, state); @@ -481,8 +476,8 @@ struct nfs4_call_sync_data { struct nfs4_sequence_res *seq_res; }; -static void nfs4_init_sequence(struct nfs4_sequence_args *args, - struct nfs4_sequence_res *res, int cache_reply) +void nfs4_init_sequence(struct nfs4_sequence_args *args, + struct nfs4_sequence_res *res, int cache_reply) { args->sa_slot = NULL; args->sa_cache_this = cache_reply; @@ -1552,6 +1547,13 @@ static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, fmode_t fmod struct nfs4_state *newstate; int ret; + if ((opendata->o_arg.claim == NFS4_OPEN_CLAIM_DELEGATE_CUR || + opendata->o_arg.claim == NFS4_OPEN_CLAIM_DELEG_CUR_FH) && + (opendata->o_arg.u.delegation_type & fmode) != fmode) + /* This mode can't have been delegated, so we must have + * a valid open_stateid to cover it - not need to reclaim. + */ + return 0; opendata->o_arg.open_flags = 0; opendata->o_arg.fmode = fmode; opendata->o_arg.share_access = nfs4_map_atomic_open_share( @@ -1683,6 +1685,7 @@ static int nfs4_handle_delegation_recall_error(struct nfs_server *server, struct "%d.\n", __func__, err); case 0: case -ENOENT: + case -EAGAIN: case -ESTALE: break; case -NFS4ERR_BADSESSION: @@ -3354,6 +3357,8 @@ static int nfs4_proc_lookup_common(struct rpc_clnt **clnt, struct inode *dir, goto out; case -NFS4ERR_MOVED: err = nfs4_get_referral(client, dir, name, fattr, fhandle); + if (err == -NFS4ERR_MOVED) + err = nfs4_handle_exception(NFS_SERVER(dir), err, &exception); goto out; case -NFS4ERR_WRONGSEC: err = -EPERM; @@ -4954,49 +4959,128 @@ static void nfs4_init_boot_verifier(const struct nfs_client *clp, memcpy(bootverf->data, verf, sizeof(bootverf->data)); } -static unsigned int -nfs4_init_nonuniform_client_string(struct nfs_client *clp, - char *buf, size_t len) +static int +nfs4_init_nonuniform_client_string(struct nfs_client *clp) { - unsigned int result; + int result; + size_t len; + char *str; + bool retried = false; if (clp->cl_owner_id != NULL) - return strlcpy(buf, clp->cl_owner_id, len); + return 0; +retry: + rcu_read_lock(); + len = 10 + strlen(clp->cl_ipaddr) + 1 + + strlen(rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR)) + + 1 + + strlen(rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_PROTO)) + + 1; + rcu_read_unlock(); + + if (len > NFS4_OPAQUE_LIMIT + 1) + return -EINVAL; + + /* + * Since this string is allocated at mount time, and held until the + * nfs_client is destroyed, we can use GFP_KERNEL here w/o worrying + * about a memory-reclaim deadlock. + */ + str = kmalloc(len, GFP_KERNEL); + if (!str) + return -ENOMEM; rcu_read_lock(); - result = scnprintf(buf, len, "Linux NFSv4.0 %s/%s %s", - clp->cl_ipaddr, - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR), - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_PROTO)); + result = scnprintf(str, len, "Linux NFSv4.0 %s/%s %s", + clp->cl_ipaddr, + rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_ADDR), + rpc_peeraddr2str(clp->cl_rpcclient, RPC_DISPLAY_PROTO)); rcu_read_unlock(); - clp->cl_owner_id = kstrdup(buf, GFP_KERNEL); - return result; + + /* Did something change? */ + if (result >= len) { + kfree(str); + if (retried) + return -EINVAL; + retried = true; + goto retry; + } + clp->cl_owner_id = str; + return 0; } -static unsigned int -nfs4_init_uniform_client_string(struct nfs_client *clp, - char *buf, size_t len) +static int +nfs4_init_uniquifier_client_string(struct nfs_client *clp) { - const char *nodename = clp->cl_rpcclient->cl_nodename; - unsigned int result; + int result; + size_t len; + char *str; + + len = 10 + 10 + 1 + 10 + 1 + + strlen(nfs4_client_id_uniquifier) + 1 + + strlen(clp->cl_rpcclient->cl_nodename) + 1; + + if (len > NFS4_OPAQUE_LIMIT + 1) + return -EINVAL; + + /* + * Since this string is allocated at mount time, and held until the + * nfs_client is destroyed, we can use GFP_KERNEL here w/o worrying + * about a memory-reclaim deadlock. + */ + str = kmalloc(len, GFP_KERNEL); + if (!str) + return -ENOMEM; + + result = scnprintf(str, len, "Linux NFSv%u.%u %s/%s", + clp->rpc_ops->version, clp->cl_minorversion, + nfs4_client_id_uniquifier, + clp->cl_rpcclient->cl_nodename); + if (result >= len) { + kfree(str); + return -EINVAL; + } + clp->cl_owner_id = str; + return 0; +} + +static int +nfs4_init_uniform_client_string(struct nfs_client *clp) +{ + int result; + size_t len; + char *str; if (clp->cl_owner_id != NULL) - return strlcpy(buf, clp->cl_owner_id, len); + return 0; if (nfs4_client_id_uniquifier[0] != '\0') - result = scnprintf(buf, len, "Linux NFSv%u.%u %s/%s", - clp->rpc_ops->version, - clp->cl_minorversion, - nfs4_client_id_uniquifier, - nodename); - else - result = scnprintf(buf, len, "Linux NFSv%u.%u %s", - clp->rpc_ops->version, clp->cl_minorversion, - nodename); - clp->cl_owner_id = kstrdup(buf, GFP_KERNEL); - return result; + return nfs4_init_uniquifier_client_string(clp); + + len = 10 + 10 + 1 + 10 + 1 + + strlen(clp->cl_rpcclient->cl_nodename) + 1; + + if (len > NFS4_OPAQUE_LIMIT + 1) + return -EINVAL; + + /* + * Since this string is allocated at mount time, and held until the + * nfs_client is destroyed, we can use GFP_KERNEL here w/o worrying + * about a memory-reclaim deadlock. + */ + str = kmalloc(len, GFP_KERNEL); + if (!str) + return -ENOMEM; + + result = scnprintf(str, len, "Linux NFSv%u.%u %s", + clp->rpc_ops->version, clp->cl_minorversion, + clp->cl_rpcclient->cl_nodename); + if (result >= len) { + kfree(str); + return -EINVAL; + } + clp->cl_owner_id = str; + return 0; } /* @@ -5043,7 +5127,7 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, struct nfs4_setclientid setclientid = { .sc_verifier = &sc_verifier, .sc_prog = program, - .sc_cb_ident = clp->cl_cb_ident, + .sc_clnt = clp, }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SETCLIENTID], @@ -5063,16 +5147,15 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, /* nfs_client_id4 */ nfs4_init_boot_verifier(clp, &sc_verifier); + if (test_bit(NFS_CS_MIGRATION, &clp->cl_flags)) - setclientid.sc_name_len = - nfs4_init_uniform_client_string(clp, - setclientid.sc_name, - sizeof(setclientid.sc_name)); + status = nfs4_init_uniform_client_string(clp); else - setclientid.sc_name_len = - nfs4_init_nonuniform_client_string(clp, - setclientid.sc_name, - sizeof(setclientid.sc_name)); + status = nfs4_init_nonuniform_client_string(clp); + + if (status) + goto out; + /* cb_client4 */ setclientid.sc_netid_len = nfs4_init_callback_netid(clp, @@ -5082,9 +5165,9 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, sizeof(setclientid.sc_uaddr), "%s.%u.%u", clp->cl_ipaddr, port >> 8, port & 255); - dprintk("NFS call setclientid auth=%s, '%.*s'\n", + dprintk("NFS call setclientid auth=%s, '%s'\n", clp->cl_rpcclient->cl_auth->au_ops->au_name, - setclientid.sc_name_len, setclientid.sc_name); + clp->cl_owner_id); task = rpc_run_task(&task_setup_data); if (IS_ERR(task)) { status = PTR_ERR(task); @@ -5401,6 +5484,7 @@ static struct nfs4_unlockdata *nfs4_alloc_unlockdata(struct file_lock *fl, atomic_inc(&lsp->ls_count); /* Ensure we don't close file until we're done freeing locks! */ p->ctx = get_nfs_open_context(ctx); + get_file(fl->fl_file); memcpy(&p->fl, fl, sizeof(p->fl)); p->server = NFS_SERVER(inode); return p; @@ -5412,6 +5496,7 @@ static void nfs4_locku_release_calldata(void *data) nfs_free_seqid(calldata->arg.seqid); nfs4_put_lock_state(calldata->lsp); put_nfs_open_context(calldata->ctx); + fput(calldata->fl.fl_file); kfree(calldata); } @@ -5604,6 +5689,7 @@ static struct nfs4_lockdata *nfs4_alloc_lockdata(struct file_lock *fl, p->server = server; atomic_inc(&lsp->ls_count); p->ctx = get_nfs_open_context(ctx); + get_file(fl->fl_file); memcpy(&p->fl, fl, sizeof(p->fl)); return p; out_free_seqid: @@ -5716,6 +5802,7 @@ static void nfs4_lock_release(void *calldata) nfs_free_seqid(data->arg.lock_seqid); nfs4_put_lock_state(data->lsp); put_nfs_open_context(data->ctx); + fput(data->fl.fl_file); kfree(data); dprintk("%s: done!\n", __func__); } @@ -6843,11 +6930,14 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, }; nfs4_init_boot_verifier(clp, &verifier); - args.id_len = nfs4_init_uniform_client_string(clp, args.id, - sizeof(args.id)); - dprintk("NFS call exchange_id auth=%s, '%.*s'\n", + + status = nfs4_init_uniform_client_string(clp); + if (status) + goto out; + + dprintk("NFS call exchange_id auth=%s, '%s'\n", clp->cl_rpcclient->cl_auth->au_ops->au_name, - args.id_len, args.id); + clp->cl_owner_id); res.server_owner = kzalloc(sizeof(struct nfs41_server_owner), GFP_NOFS); @@ -6882,7 +6972,7 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, /* unsupported! */ WARN_ON_ONCE(1); status = -EINVAL; - goto out_server_scope; + goto out_impl_id; } status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); @@ -6910,6 +7000,7 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, /* use the most recent implementation id */ kfree(clp->cl_implid); clp->cl_implid = res.impl_id; + res.impl_id = NULL; if (clp->cl_serverscope != NULL && !nfs41_same_server_scope(clp->cl_serverscope, @@ -6923,15 +7014,16 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, if (clp->cl_serverscope == NULL) { clp->cl_serverscope = res.server_scope; - goto out; + res.server_scope = NULL; } - } else - kfree(res.impl_id); + } -out_server_owner: - kfree(res.server_owner); +out_impl_id: + kfree(res.impl_id); out_server_scope: kfree(res.server_scope); +out_server_owner: + kfree(res.server_owner); out: if (clp->cl_implid != NULL) dprintk("NFS reply exchange_id: Server Implementation ID: " @@ -8058,9 +8150,8 @@ nfs4_proc_layoutcommit(struct nfs4_layoutcommit_data *data, bool sync) struct rpc_task *task; int status = 0; - dprintk("NFS: %4d initiating layoutcommit call. sync %d " - "lbw: %llu inode %lu\n", - data->task.tk_pid, sync, + dprintk("NFS: initiating layoutcommit call. sync %d " + "lbw: %llu inode %lu\n", sync, data->args.lastbytewritten, data->args.inode->i_ino); @@ -8554,7 +8645,8 @@ static const struct nfs4_minor_version_ops nfs_v4_2_minor_ops = { | NFS_CAP_ATOMIC_OPEN_V1 | NFS_CAP_ALLOCATE | NFS_CAP_DEALLOCATE - | NFS_CAP_SEEK, + | NFS_CAP_SEEK + | NFS_CAP_LAYOUTSTATS, .init_client = nfs41_init_client, .shutdown_client = nfs41_shutdown_client, .match_stateid = nfs41_match_stateid, diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 2782cfc..605840d 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -309,7 +309,6 @@ int nfs41_init_clientid(struct nfs_client *clp, struct rpc_cred *cred) if (test_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state)) goto do_confirm; - nfs4_begin_drain_session(clp); status = nfs4_proc_exchange_id(clp, cred); if (status != 0) goto out; @@ -1482,6 +1481,8 @@ restart: spin_unlock(&state->state_lock); } nfs4_put_open_state(state); + clear_bit(NFS4CLNT_RECLAIM_NOGRACE, + &state->flags); spin_lock(&sp->so_lock); goto restart; } @@ -1830,6 +1831,7 @@ static int nfs4_establish_lease(struct nfs_client *clp) clp->cl_mvops->reboot_recovery_ops; int status; + nfs4_begin_drain_session(clp); cred = nfs4_get_clid_cred(clp); if (cred == NULL) return -ENOENT; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 0aea978..558cd65d 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -139,7 +139,8 @@ static int nfs4_stat_to_errno(int); #define encode_setclientid_maxsz \ (op_encode_hdr_maxsz + \ XDR_QUADLEN(NFS4_VERIFIER_SIZE) + \ - XDR_QUADLEN(NFS4_SETCLIENTID_NAMELEN) + \ + /* client name */ \ + 1 + XDR_QUADLEN(NFS4_OPAQUE_LIMIT) + \ 1 /* sc_prog */ + \ 1 + XDR_QUADLEN(RPCBIND_MAXNETIDLEN) + \ 1 + XDR_QUADLEN(RPCBIND_MAXUADDRLEN) + \ @@ -288,7 +289,8 @@ static int nfs4_stat_to_errno(int); #define encode_exchange_id_maxsz (op_encode_hdr_maxsz + \ encode_verifier_maxsz + \ 1 /* co_ownerid.len */ + \ - XDR_QUADLEN(NFS4_EXCHANGE_ID_LEN) + \ + /* eia_clientowner */ \ + 1 + XDR_QUADLEN(NFS4_OPAQUE_LIMIT) + \ 1 /* flags */ + \ 1 /* spa_how */ + \ /* max is SP4_MACH_CRED (for now) */ + \ @@ -1667,13 +1669,14 @@ static void encode_setclientid(struct xdr_stream *xdr, const struct nfs4_setclie encode_op_hdr(xdr, OP_SETCLIENTID, decode_setclientid_maxsz, hdr); encode_nfs4_verifier(xdr, setclientid->sc_verifier); - encode_string(xdr, setclientid->sc_name_len, setclientid->sc_name); + encode_string(xdr, strlen(setclientid->sc_clnt->cl_owner_id), + setclientid->sc_clnt->cl_owner_id); p = reserve_space(xdr, 4); *p = cpu_to_be32(setclientid->sc_prog); encode_string(xdr, setclientid->sc_netid_len, setclientid->sc_netid); encode_string(xdr, setclientid->sc_uaddr_len, setclientid->sc_uaddr); p = reserve_space(xdr, 4); - *p = cpu_to_be32(setclientid->sc_cb_ident); + *p = cpu_to_be32(setclientid->sc_clnt->cl_cb_ident); } static void encode_setclientid_confirm(struct xdr_stream *xdr, const struct nfs4_setclientid_res *arg, struct compound_hdr *hdr) @@ -1747,7 +1750,8 @@ static void encode_exchange_id(struct xdr_stream *xdr, encode_op_hdr(xdr, OP_EXCHANGE_ID, decode_exchange_id_maxsz, hdr); encode_nfs4_verifier(xdr, args->verifier); - encode_string(xdr, args->id_len, args->id); + encode_string(xdr, strlen(args->client->cl_owner_id), + args->client->cl_owner_id); encode_uint32(xdr, args->flags); encode_uint32(xdr, args->state_protect.how); @@ -7427,6 +7431,7 @@ struct rpc_procinfo nfs4_procedures[] = { PROC(SEEK, enc_seek, dec_seek), PROC(ALLOCATE, enc_allocate, dec_allocate), PROC(DEALLOCATE, enc_deallocate, dec_deallocate), + PROC(LAYOUTSTATS, enc_layoutstats, dec_layoutstats), #endif /* CONFIG_NFS_V4_2 */ }; diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index 282b393..1da68d3 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -636,9 +636,8 @@ int nfs_initiate_pgio(struct rpc_clnt *clnt, struct nfs_pgio_header *hdr, hdr->rw_ops->rw_initiate(hdr, &msg, rpc_ops, &task_setup_data, how); - dprintk("NFS: %5u initiated pgio call " + dprintk("NFS: initiated pgio call " "(req %s/%llu, %u bytes @ offset %llu)\n", - hdr->task.tk_pid, hdr->inode->i_sb->s_id, (unsigned long long)NFS_FILEID(hdr->inode), hdr->args.count, @@ -690,8 +689,6 @@ static int nfs_pgio_error(struct nfs_pageio_descriptor *desc, static void nfs_pgio_release(void *calldata) { struct nfs_pgio_header *hdr = calldata; - if (hdr->rw_ops->rw_release) - hdr->rw_ops->rw_release(hdr); nfs_pgio_data_destroy(hdr); hdr->completion_ops->completion(hdr); } @@ -711,7 +708,9 @@ static void nfs_pageio_mirror_init(struct nfs_pgio_mirror *mirror, * nfs_pageio_init - initialise a page io descriptor * @desc: pointer to descriptor * @inode: pointer to inode - * @doio: pointer to io function + * @pg_ops: pointer to pageio operations + * @compl_ops: pointer to pageio completion operations + * @rw_ops: pointer to nfs read/write operations * @bsize: io block size * @io_flags: extra parameters for the io function */ @@ -1186,6 +1185,7 @@ int nfs_pageio_add_request(struct nfs_pageio_descriptor *desc, * nfs_pageio_complete_mirror - Complete I/O on the current mirror of an * nfs_pageio_descriptor * @desc: pointer to io descriptor + * @mirror_idx: pointer to mirror index */ static void nfs_pageio_complete_mirror(struct nfs_pageio_descriptor *desc, u32 mirror_idx) diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index 2306062..0ba9a02 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -35,6 +35,7 @@ #include "iostat.h" #include "nfs4trace.h" #include "delegation.h" +#include "nfs42.h" #define NFSDBG_FACILITY NFSDBG_PNFS #define PNFS_LAYOUTGET_RETRY_TIMEOUT (120*HZ) @@ -1821,6 +1822,7 @@ int pnfs_write_done_resend_to_mds(struct nfs_pgio_header *hdr) /* Resend all requests through the MDS */ nfs_pageio_init_write(&pgio, hdr->inode, FLUSH_STABLE, true, hdr->completion_ops); + set_bit(NFS_CONTEXT_RESEND_WRITES, &hdr->args.context->flags); return nfs_pageio_resend(&pgio, hdr); } EXPORT_SYMBOL_GPL(pnfs_write_done_resend_to_mds); @@ -1865,6 +1867,7 @@ pnfs_write_through_mds(struct nfs_pageio_descriptor *desc, mirror->pg_recoalesce = 1; } nfs_pgio_data_destroy(hdr); + hdr->release(hdr); } static enum pnfs_try_status @@ -1979,6 +1982,7 @@ pnfs_read_through_mds(struct nfs_pageio_descriptor *desc, mirror->pg_recoalesce = 1; } nfs_pgio_data_destroy(hdr); + hdr->release(hdr); } /* @@ -2247,3 +2251,63 @@ struct nfs4_threshold *pnfs_mdsthreshold_alloc(void) } return thp; } + +#if IS_ENABLED(CONFIG_NFS_V4_2) +int +pnfs_report_layoutstat(struct inode *inode) +{ + struct pnfs_layoutdriver_type *ld = NFS_SERVER(inode)->pnfs_curr_ld; + struct nfs_server *server = NFS_SERVER(inode); + struct nfs_inode *nfsi = NFS_I(inode); + struct nfs42_layoutstat_data *data; + struct pnfs_layout_hdr *hdr; + int status = 0; + + if (!pnfs_enabled_sb(server) || !ld->prepare_layoutstats) + goto out; + + if (!nfs_server_capable(inode, NFS_CAP_LAYOUTSTATS)) + goto out; + + if (test_and_set_bit(NFS_INO_LAYOUTSTATS, &nfsi->flags)) + goto out; + + spin_lock(&inode->i_lock); + if (!NFS_I(inode)->layout) { + spin_unlock(&inode->i_lock); + goto out; + } + hdr = NFS_I(inode)->layout; + pnfs_get_layout_hdr(hdr); + spin_unlock(&inode->i_lock); + + data = kzalloc(sizeof(*data), GFP_KERNEL); + if (!data) { + status = -ENOMEM; + goto out_put; + } + + data->args.fh = NFS_FH(inode); + data->args.inode = inode; + nfs4_stateid_copy(&data->args.stateid, &hdr->plh_stateid); + status = ld->prepare_layoutstats(&data->args); + if (status) + goto out_free; + + status = nfs42_proc_layoutstats_generic(NFS_SERVER(inode), data); + +out: + dprintk("%s returns %d\n", __func__, status); + return status; + +out_free: + kfree(data); +out_put: + pnfs_put_layout_hdr(hdr); + smp_mb__before_atomic(); + clear_bit(NFS_INO_LAYOUTSTATS, &nfsi->flags); + smp_mb__after_atomic(); + goto out; +} +EXPORT_SYMBOL_GPL(pnfs_report_layoutstat); +#endif diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index 1e6308f..3e6ab7b 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -178,6 +178,8 @@ struct pnfs_layoutdriver_type { void (*encode_layoutcommit) (struct pnfs_layout_hdr *lo, struct xdr_stream *xdr, const struct nfs4_layoutcommit_args *args); + int (*prepare_layoutstats) (struct nfs42_layoutstat_args *args); + void (*cleanup_layoutstats) (struct nfs42_layoutstat_data *data); }; struct pnfs_layout_hdr { @@ -290,7 +292,6 @@ int pnfs_write_done_resend_to_mds(struct nfs_pgio_header *); struct nfs4_threshold *pnfs_mdsthreshold_alloc(void); void pnfs_error_mark_layout_for_return(struct inode *inode, struct pnfs_layout_segment *lseg); - /* nfs4_deviceid_flags */ enum { NFS_DEVICEID_INVALID = 0, /* set when MDS clientid recalled */ @@ -689,4 +690,14 @@ static inline void nfs4_pnfs_v3_ds_connect_unload(void) #endif /* CONFIG_NFS_V4_1 */ +#if IS_ENABLED(CONFIG_NFS_V4_2) +int pnfs_report_layoutstat(struct inode *inode); +#else +static inline int +pnfs_report_layoutstat(struct inode *inode) +{ + return 0; +} +#endif + #endif /* FS_NFS_PNFS_H */ diff --git a/fs/nfs/super.c b/fs/nfs/super.c index f175b83..aa62004 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -2847,7 +2847,7 @@ static int param_set_portnr(const char *val, const struct kernel_param *kp) *((unsigned int *)kp->arg) = num; return 0; } -static struct kernel_param_ops param_ops_portnr = { +static const struct kernel_param_ops param_ops_portnr = { .set = param_set_portnr, .get = param_get_uint, }; diff --git a/fs/nfs/write.c b/fs/nfs/write.c index d12a4be..65869ca 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -853,7 +853,8 @@ static void nfs_clear_page_commit(struct page *page) { dec_zone_page_state(page, NR_UNSTABLE_NFS); - dec_bdi_stat(inode_to_bdi(page_file_mapping(page)->host), BDI_RECLAIMABLE); + dec_wb_stat(&inode_to_bdi(page_file_mapping(page)->host)->wb, + WB_RECLAIMABLE); } /* Called holding inode (/cinfo) lock */ @@ -1289,6 +1290,7 @@ static void nfs_initiate_write(struct nfs_pgio_header *hdr, static void nfs_redirty_request(struct nfs_page *req) { nfs_mark_request_dirty(req); + set_bit(NFS_CONTEXT_RESEND_WRITES, &req->wb_context->flags); nfs_unlock_request(req); nfs_end_page_writeback(req); nfs_release_request(req); @@ -1347,11 +1349,6 @@ void nfs_commit_prepare(struct rpc_task *task, void *calldata) NFS_PROTO(data->inode)->commit_rpc_prepare(task, data); } -static void nfs_writeback_release_common(struct nfs_pgio_header *hdr) -{ - /* do nothing! */ -} - /* * Special version of should_remove_suid() that ignores capabilities. */ @@ -1555,7 +1552,7 @@ int nfs_initiate_commit(struct rpc_clnt *clnt, struct nfs_commit_data *data, /* Set up the initial task struct. */ nfs_ops->commit_setup(data, &msg); - dprintk("NFS: %5u initiated commit call\n", data->task.tk_pid); + dprintk("NFS: initiated commit call\n"); nfs4_state_protect(NFS_SERVER(data->inode)->nfs_client, NFS_SP4_MACH_CRED_COMMIT, &task_setup_data.rpc_client, &msg); @@ -1845,12 +1842,15 @@ int nfs_wb_all(struct inode *inode) trace_nfs_writeback_inode_enter(inode); ret = filemap_write_and_wait(inode->i_mapping); - if (!ret) { - ret = nfs_commit_inode(inode, FLUSH_SYNC); - if (!ret) - pnfs_sync_inode(inode, true); - } + if (ret) + goto out; + ret = nfs_commit_inode(inode, FLUSH_SYNC); + if (ret < 0) + goto out; + pnfs_sync_inode(inode, true); + ret = 0; +out: trace_nfs_writeback_inode_exit(inode, ret); return ret; } @@ -2009,7 +2009,6 @@ static const struct nfs_rw_ops nfs_rw_write_ops = { .rw_mode = FMODE_WRITE, .rw_alloc_header = nfs_writehdr_alloc, .rw_free_header = nfs_writehdr_free, - .rw_release = nfs_writeback_release_common, .rw_done = nfs_writeback_done, .rw_result = nfs_writeback_result, .rw_initiate = nfs_initiate_write, diff --git a/fs/nfsd/blocklayout.c b/fs/nfsd/blocklayout.c index 03d647b..cdefaa3 100644 --- a/fs/nfsd/blocklayout.c +++ b/fs/nfsd/blocklayout.c @@ -181,6 +181,17 @@ nfsd4_block_proc_layoutcommit(struct inode *inode, } const struct nfsd4_layout_ops bl_layout_ops = { + /* + * Pretend that we send notification to the client. This is a blatant + * lie to force recent Linux clients to cache our device IDs. + * We rarely ever change the device ID, so the harm of leaking deviceids + * for a while isn't too bad. Unfortunately RFC5661 is a complete mess + * in this regard, but I filed errata 4119 for this a while ago, and + * hopefully the Linux client will eventually start caching deviceids + * without this again. + */ + .notify_types = + NOTIFY_DEVICEID4_DELETE | NOTIFY_DEVICEID4_CHANGE, .proc_getdeviceinfo = nfsd4_block_proc_getdeviceinfo, .encode_getdeviceinfo = nfsd4_block_encode_getdeviceinfo, .proc_layoutget = nfsd4_block_proc_layoutget, diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c index e4b2b43..f6e7cba 100644 --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -805,7 +805,7 @@ encode_entry_baggage(struct nfsd3_readdirres *cd, __be32 *p, const char *name, static __be32 compose_entry_fh(struct nfsd3_readdirres *cd, struct svc_fh *fhp, - const char *name, int namlen) + const char *name, int namlen, u64 ino) { struct svc_export *exp; struct dentry *dparent, *dchild; @@ -830,19 +830,21 @@ compose_entry_fh(struct nfsd3_readdirres *cd, struct svc_fh *fhp, goto out; if (d_really_is_negative(dchild)) goto out; + if (dchild->d_inode->i_ino != ino) + goto out; rv = fh_compose(fhp, exp, dchild, &cd->fh); out: dput(dchild); return rv; } -static __be32 *encode_entryplus_baggage(struct nfsd3_readdirres *cd, __be32 *p, const char *name, int namlen) +static __be32 *encode_entryplus_baggage(struct nfsd3_readdirres *cd, __be32 *p, const char *name, int namlen, u64 ino) { struct svc_fh *fh = &cd->scratch; __be32 err; fh_init(fh, NFS3_FHSIZE); - err = compose_entry_fh(cd, fh, name, namlen); + err = compose_entry_fh(cd, fh, name, namlen, ino); if (err) { *p++ = 0; *p++ = 0; @@ -927,7 +929,7 @@ encode_entry(struct readdir_cd *ccd, const char *name, int namlen, p = encode_entry_baggage(cd, p, name, namlen, ino); if (plus) - p = encode_entryplus_baggage(cd, p, name, namlen); + p = encode_entryplus_baggage(cd, p, name, namlen, ino); num_entry_words = p - cd->buffer; } else if (*(page+1) != NULL) { /* temporarily encode entry into next page, then move back to @@ -941,7 +943,7 @@ encode_entry(struct readdir_cd *ccd, const char *name, int namlen, p1 = encode_entry_baggage(cd, p1, name, namlen, ino); if (plus) - p1 = encode_entryplus_baggage(cd, p1, name, namlen); + p1 = encode_entryplus_baggage(cd, p1, name, namlen, ino); /* determine entry word length and lengths to go in pages */ num_entry_words = p1 - tmp; diff --git a/fs/nfsd/nfs4acl.c b/fs/nfsd/nfs4acl.c index 67242bf..eb5accf 100644 --- a/fs/nfsd/nfs4acl.c +++ b/fs/nfsd/nfs4acl.c @@ -52,10 +52,6 @@ #define NFS4_ANYONE_MODE (NFS4_ACE_READ_ATTRIBUTES | NFS4_ACE_READ_ACL | NFS4_ACE_SYNCHRONIZE) #define NFS4_OWNER_MODE (NFS4_ACE_WRITE_ATTRIBUTES | NFS4_ACE_WRITE_ACL) -/* We don't support these bits; insist they be neither allowed nor denied */ -#define NFS4_MASK_UNSUPP (NFS4_ACE_DELETE | NFS4_ACE_WRITE_OWNER \ - | NFS4_ACE_READ_NAMED_ATTRS | NFS4_ACE_WRITE_NAMED_ATTRS) - /* flags used to simulate posix default ACLs */ #define NFS4_INHERITANCE_FLAGS (NFS4_ACE_FILE_INHERIT_ACE \ | NFS4_ACE_DIRECTORY_INHERIT_ACE) @@ -64,9 +60,6 @@ | NFS4_ACE_INHERIT_ONLY_ACE \ | NFS4_ACE_IDENTIFIER_GROUP) -#define MASK_EQUAL(mask1, mask2) \ - ( ((mask1) & NFS4_ACE_MASK_ALL) == ((mask2) & NFS4_ACE_MASK_ALL) ) - static u32 mask_from_posix(unsigned short perm, unsigned int flags) { @@ -126,11 +119,6 @@ low_mode_from_nfs4(u32 perm, unsigned short *mode, unsigned int flags) *mode |= ACL_EXECUTE; } -struct ace_container { - struct nfs4_ace *ace; - struct list_head ace_l; -}; - static short ace2type(struct nfs4_ace *); static void _posix_to_nfsv4_one(struct posix_acl *, struct nfs4_acl *, unsigned int); @@ -384,7 +372,6 @@ pace_gt(struct posix_acl_entry *pace1, struct posix_acl_entry *pace2) static void sort_pacl_range(struct posix_acl *pacl, int start, int end) { int sorted = 0, i; - struct posix_acl_entry tmp; /* We just do a bubble sort; easy to do in place, and we're not * expecting acl's to be long enough to justify anything more. */ @@ -394,9 +381,8 @@ sort_pacl_range(struct posix_acl *pacl, int start, int end) { if (pace_gt(&pacl->a_entries[i], &pacl->a_entries[i+1])) { sorted = 0; - tmp = pacl->a_entries[i]; - pacl->a_entries[i] = pacl->a_entries[i+1]; - pacl->a_entries[i+1] = tmp; + swap(pacl->a_entries[i], + pacl->a_entries[i + 1]); } } } diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 5827785..a492018 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -224,7 +224,7 @@ static int nfs_cb_stat_to_errno(int status) } static int decode_cb_op_status(struct xdr_stream *xdr, enum nfs_opnum4 expected, - enum nfsstat4 *status) + int *status) { __be32 *p; u32 op; @@ -235,7 +235,7 @@ static int decode_cb_op_status(struct xdr_stream *xdr, enum nfs_opnum4 expected, op = be32_to_cpup(p++); if (unlikely(op != expected)) goto out_unexpected; - *status = be32_to_cpup(p); + *status = nfs_cb_stat_to_errno(be32_to_cpup(p)); return 0; out_overflow: print_overflow_msg(__func__, xdr); @@ -446,22 +446,17 @@ out_overflow: static int decode_cb_sequence4res(struct xdr_stream *xdr, struct nfsd4_callback *cb) { - enum nfsstat4 nfserr; int status; if (cb->cb_minorversion == 0) return 0; - status = decode_cb_op_status(xdr, OP_CB_SEQUENCE, &nfserr); - if (unlikely(status)) - goto out; - if (unlikely(nfserr != NFS4_OK)) - goto out_default; - status = decode_cb_sequence4resok(xdr, cb); -out: - return status; -out_default: - return nfs_cb_stat_to_errno(nfserr); + status = decode_cb_op_status(xdr, OP_CB_SEQUENCE, &cb->cb_status); + if (unlikely(status || cb->cb_status)) + return status; + + cb->cb_update_seq_nr = true; + return decode_cb_sequence4resok(xdr, cb); } /* @@ -524,26 +519,19 @@ static int nfs4_xdr_dec_cb_recall(struct rpc_rqst *rqstp, struct nfsd4_callback *cb) { struct nfs4_cb_compound_hdr hdr; - enum nfsstat4 nfserr; int status; status = decode_cb_compound4res(xdr, &hdr); if (unlikely(status)) - goto out; + return status; if (cb != NULL) { status = decode_cb_sequence4res(xdr, cb); - if (unlikely(status)) - goto out; + if (unlikely(status || cb->cb_status)) + return status; } - status = decode_cb_op_status(xdr, OP_CB_RECALL, &nfserr); - if (unlikely(status)) - goto out; - if (unlikely(nfserr != NFS4_OK)) - status = nfs_cb_stat_to_errno(nfserr); -out: - return status; + return decode_cb_op_status(xdr, OP_CB_RECALL, &cb->cb_status); } #ifdef CONFIG_NFSD_PNFS @@ -621,24 +609,18 @@ static int nfs4_xdr_dec_cb_layout(struct rpc_rqst *rqstp, struct nfsd4_callback *cb) { struct nfs4_cb_compound_hdr hdr; - enum nfsstat4 nfserr; int status; status = decode_cb_compound4res(xdr, &hdr); if (unlikely(status)) - goto out; + return status; + if (cb) { status = decode_cb_sequence4res(xdr, cb); - if (unlikely(status)) - goto out; + if (unlikely(status || cb->cb_status)) + return status; } - status = decode_cb_op_status(xdr, OP_CB_LAYOUTRECALL, &nfserr); - if (unlikely(status)) - goto out; - if (unlikely(nfserr != NFS4_OK)) - status = nfs_cb_stat_to_errno(nfserr); -out: - return status; + return decode_cb_op_status(xdr, OP_CB_LAYOUTRECALL, &cb->cb_status); } #endif /* CONFIG_NFSD_PNFS */ @@ -894,17 +876,12 @@ static void nfsd4_cb_prepare(struct rpc_task *task, void *calldata) u32 minorversion = clp->cl_minorversion; cb->cb_minorversion = minorversion; + cb->cb_update_seq_nr = false; + cb->cb_status = 0; if (minorversion) { if (!nfsd41_cb_get_slot(clp, task)) return; } - spin_lock(&clp->cl_lock); - if (list_empty(&cb->cb_per_client)) { - /* This is the first call, not a restart */ - cb->cb_done = false; - list_add(&cb->cb_per_client, &clp->cl_callbacks); - } - spin_unlock(&clp->cl_lock); rpc_call_start(task); } @@ -917,23 +894,41 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata) clp->cl_minorversion); if (clp->cl_minorversion) { - /* No need for lock, access serialized in nfsd4_cb_prepare */ - ++clp->cl_cb_session->se_cb_seq_nr; + /* + * No need for lock, access serialized in nfsd4_cb_prepare + * + * RFC5661 20.9.3 + * If CB_SEQUENCE returns an error, then the state of the slot + * (sequence ID, cached reply) MUST NOT change. + */ + if (cb->cb_update_seq_nr) + ++clp->cl_cb_session->se_cb_seq_nr; + clear_bit(0, &clp->cl_cb_slot_busy); rpc_wake_up_next(&clp->cl_cb_waitq); dprintk("%s: freed slot, new seqid=%d\n", __func__, clp->cl_cb_session->se_cb_seq_nr); } - if (clp->cl_cb_client != task->tk_client) { - /* We're shutting down or changing cl_cb_client; leave - * it to nfsd4_process_cb_update to restart the call if - * necessary. */ + /* + * If the backchannel connection was shut down while this + * task was queued, we need to resubmit it after setting up + * a new backchannel connection. + * + * Note that if we lost our callback connection permanently + * the submission code will error out, so we don't need to + * handle that case here. + */ + if (task->tk_flags & RPC_TASK_KILLED) { + task->tk_status = 0; + cb->cb_need_restart = true; return; } - if (cb->cb_done) - return; + if (cb->cb_status) { + WARN_ON_ONCE(task->tk_status); + task->tk_status = cb->cb_status; + } switch (cb->cb_ops->done(cb, task)) { case 0: @@ -949,21 +944,17 @@ static void nfsd4_cb_done(struct rpc_task *task, void *calldata) default: BUG(); } - cb->cb_done = true; } static void nfsd4_cb_release(void *calldata) { struct nfsd4_callback *cb = calldata; - struct nfs4_client *clp = cb->cb_clp; - - if (cb->cb_done) { - spin_lock(&clp->cl_lock); - list_del(&cb->cb_per_client); - spin_unlock(&clp->cl_lock); + if (cb->cb_need_restart) + nfsd4_run_cb(cb); + else cb->cb_ops->release(cb); - } + } static const struct rpc_call_ops nfsd4_cb_ops = { @@ -1058,9 +1049,6 @@ static void nfsd4_process_cb_update(struct nfsd4_callback *cb) nfsd4_mark_cb_down(clp, err); return; } - /* Yay, the callback channel's back! Restart any callbacks: */ - list_for_each_entry(cb, &clp->cl_callbacks, cb_per_client) - queue_work(callback_wq, &cb->cb_work); } static void @@ -1071,8 +1059,12 @@ nfsd4_run_cb_work(struct work_struct *work) struct nfs4_client *clp = cb->cb_clp; struct rpc_clnt *clnt; - if (cb->cb_ops && cb->cb_ops->prepare) - cb->cb_ops->prepare(cb); + if (cb->cb_need_restart) { + cb->cb_need_restart = false; + } else { + if (cb->cb_ops && cb->cb_ops->prepare) + cb->cb_ops->prepare(cb); + } if (clp->cl_flags & NFSD4_CLIENT_CB_FLAG_MASK) nfsd4_process_cb_update(cb); @@ -1084,6 +1076,15 @@ nfsd4_run_cb_work(struct work_struct *work) cb->cb_ops->release(cb); return; } + + /* + * Don't send probe messages for 4.1 or later. + */ + if (!cb->cb_ops && clp->cl_minorversion) { + clp->cl_cb_state = NFSD4_CB_UP; + return; + } + cb->cb_msg.rpc_cred = clp->cl_cb_cred; rpc_call_async(clnt, &cb->cb_msg, RPC_TASK_SOFT | RPC_TASK_SOFTCONN, cb->cb_ops ? &nfsd4_cb_ops : &nfsd4_cb_probe_ops, cb); @@ -1098,8 +1099,9 @@ void nfsd4_init_cb(struct nfsd4_callback *cb, struct nfs4_client *clp, cb->cb_msg.rpc_resp = cb; cb->cb_ops = ops; INIT_WORK(&cb->cb_work, nfsd4_run_cb_work); - INIT_LIST_HEAD(&cb->cb_per_client); - cb->cb_done = true; + cb->cb_status = 0; + cb->cb_update_seq_nr = false; + cb->cb_need_restart = false; } void nfsd4_run_cb(struct nfsd4_callback *cb) diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 864e200..90cfda7 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -760,8 +760,6 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, { __be32 status; - /* no need to check permission - this will be done in nfsd_read() */ - read->rd_filp = NULL; if (read->rd_offset >= OFFSET_MAX) return nfserr_inval; @@ -778,9 +776,9 @@ nfsd4_read(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, clear_bit(RQ_SPLICE_OK, &rqstp->rq_flags); /* check stateid */ - if ((status = nfs4_preprocess_stateid_op(SVC_NET(rqstp), - cstate, &read->rd_stateid, - RD_STATE, &read->rd_filp))) { + status = nfs4_preprocess_stateid_op(rqstp, cstate, &read->rd_stateid, + RD_STATE, &read->rd_filp, &read->rd_tmp_file); + if (status) { dprintk("NFSD: nfsd4_read: couldn't process stateid!\n"); goto out; } @@ -924,8 +922,8 @@ nfsd4_setattr(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, int err; if (setattr->sa_iattr.ia_valid & ATTR_SIZE) { - status = nfs4_preprocess_stateid_op(SVC_NET(rqstp), cstate, - &setattr->sa_stateid, WR_STATE, NULL); + status = nfs4_preprocess_stateid_op(rqstp, cstate, + &setattr->sa_stateid, WR_STATE, NULL, NULL); if (status) { dprintk("NFSD: nfsd4_setattr: couldn't process stateid!\n"); return status; @@ -986,13 +984,11 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, unsigned long cnt; int nvecs; - /* no need to check permission - this will be done in nfsd_write() */ - if (write->wr_offset >= OFFSET_MAX) return nfserr_inval; - status = nfs4_preprocess_stateid_op(SVC_NET(rqstp), - cstate, stateid, WR_STATE, &filp); + status = nfs4_preprocess_stateid_op(rqstp, cstate, stateid, WR_STATE, + &filp, NULL); if (status) { dprintk("NFSD: nfsd4_write: couldn't process stateid!\n"); return status; @@ -1005,11 +1001,10 @@ nfsd4_write(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, nvecs = fill_in_write_vector(rqstp->rq_vec, write); WARN_ON_ONCE(nvecs > ARRAY_SIZE(rqstp->rq_vec)); - status = nfsd_write(rqstp, &cstate->current_fh, filp, - write->wr_offset, rqstp->rq_vec, nvecs, - &cnt, &write->wr_how_written); - if (filp) - fput(filp); + status = nfsd_vfs_write(rqstp, &cstate->current_fh, filp, + write->wr_offset, rqstp->rq_vec, nvecs, &cnt, + &write->wr_how_written); + fput(filp); write->wr_bytes_written = cnt; @@ -1023,15 +1018,13 @@ nfsd4_fallocate(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, __be32 status = nfserr_notsupp; struct file *file; - status = nfs4_preprocess_stateid_op(SVC_NET(rqstp), cstate, + status = nfs4_preprocess_stateid_op(rqstp, cstate, &fallocate->falloc_stateid, - WR_STATE, &file); + WR_STATE, &file, NULL); if (status != nfs_ok) { dprintk("NFSD: nfsd4_fallocate: couldn't process stateid!\n"); return status; } - if (!file) - return nfserr_bad_stateid; status = nfsd4_vfs_fallocate(rqstp, &cstate->current_fh, file, fallocate->falloc_offset, @@ -1064,15 +1057,13 @@ nfsd4_seek(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, __be32 status; struct file *file; - status = nfs4_preprocess_stateid_op(SVC_NET(rqstp), cstate, + status = nfs4_preprocess_stateid_op(rqstp, cstate, &seek->seek_stateid, - RD_STATE, &file); + RD_STATE, &file, NULL); if (status) { dprintk("NFSD: nfsd4_seek: couldn't process stateid!\n"); return status; } - if (!file) - return nfserr_bad_stateid; switch (seek->seek_whence) { case NFS4_CONTENT_DATA: @@ -1732,10 +1723,6 @@ encode_op: be32_to_cpu(status)); nfsd4_cstate_clear_replay(cstate); - /* XXX Ugh, we need to get rid of this kind of special case: */ - if (op->opnum == OP_READ && op->u.read.rd_filp) - fput(op->u.read.rd_filp); - nfsd4_increment_op_stats(op->opnum); } diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 38f2d7a..61dfb33 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -94,6 +94,7 @@ static struct kmem_cache *lockowner_slab; static struct kmem_cache *file_slab; static struct kmem_cache *stateid_slab; static struct kmem_cache *deleg_slab; +static struct kmem_cache *odstate_slab; static void free_session(struct nfsd4_session *); @@ -281,6 +282,7 @@ put_nfs4_file(struct nfs4_file *fi) if (atomic_dec_and_lock(&fi->fi_ref, &state_lock)) { hlist_del_rcu(&fi->fi_hash); spin_unlock(&state_lock); + WARN_ON_ONCE(!list_empty(&fi->fi_clnt_odstate)); WARN_ON_ONCE(!list_empty(&fi->fi_delegations)); call_rcu(&fi->fi_rcu, nfsd4_free_file_rcu); } @@ -471,6 +473,86 @@ static void nfs4_file_put_access(struct nfs4_file *fp, u32 access) __nfs4_file_put_access(fp, O_RDONLY); } +/* + * Allocate a new open/delegation state counter. This is needed for + * pNFS for proper return on close semantics. + * + * Note that we only allocate it for pNFS-enabled exports, otherwise + * all pointers to struct nfs4_clnt_odstate are always NULL. + */ +static struct nfs4_clnt_odstate * +alloc_clnt_odstate(struct nfs4_client *clp) +{ + struct nfs4_clnt_odstate *co; + + co = kmem_cache_zalloc(odstate_slab, GFP_KERNEL); + if (co) { + co->co_client = clp; + atomic_set(&co->co_odcount, 1); + } + return co; +} + +static void +hash_clnt_odstate_locked(struct nfs4_clnt_odstate *co) +{ + struct nfs4_file *fp = co->co_file; + + lockdep_assert_held(&fp->fi_lock); + list_add(&co->co_perfile, &fp->fi_clnt_odstate); +} + +static inline void +get_clnt_odstate(struct nfs4_clnt_odstate *co) +{ + if (co) + atomic_inc(&co->co_odcount); +} + +static void +put_clnt_odstate(struct nfs4_clnt_odstate *co) +{ + struct nfs4_file *fp; + + if (!co) + return; + + fp = co->co_file; + if (atomic_dec_and_lock(&co->co_odcount, &fp->fi_lock)) { + list_del(&co->co_perfile); + spin_unlock(&fp->fi_lock); + + nfsd4_return_all_file_layouts(co->co_client, fp); + kmem_cache_free(odstate_slab, co); + } +} + +static struct nfs4_clnt_odstate * +find_or_hash_clnt_odstate(struct nfs4_file *fp, struct nfs4_clnt_odstate *new) +{ + struct nfs4_clnt_odstate *co; + struct nfs4_client *cl; + + if (!new) + return NULL; + + cl = new->co_client; + + spin_lock(&fp->fi_lock); + list_for_each_entry(co, &fp->fi_clnt_odstate, co_perfile) { + if (co->co_client == cl) { + get_clnt_odstate(co); + goto out; + } + } + co = new; + co->co_file = fp; + hash_clnt_odstate_locked(new); +out: + spin_unlock(&fp->fi_lock); + return co; +} + struct nfs4_stid *nfs4_alloc_stid(struct nfs4_client *cl, struct kmem_cache *slab) { @@ -606,7 +688,8 @@ static void block_delegations(struct knfsd_fh *fh) } static struct nfs4_delegation * -alloc_init_deleg(struct nfs4_client *clp, struct svc_fh *current_fh) +alloc_init_deleg(struct nfs4_client *clp, struct svc_fh *current_fh, + struct nfs4_clnt_odstate *odstate) { struct nfs4_delegation *dp; long n; @@ -631,6 +714,8 @@ alloc_init_deleg(struct nfs4_client *clp, struct svc_fh *current_fh) INIT_LIST_HEAD(&dp->dl_perfile); INIT_LIST_HEAD(&dp->dl_perclnt); INIT_LIST_HEAD(&dp->dl_recall_lru); + dp->dl_clnt_odstate = odstate; + get_clnt_odstate(odstate); dp->dl_type = NFS4_OPEN_DELEGATE_READ; dp->dl_retries = 1; nfsd4_init_cb(&dp->dl_recall, dp->dl_stid.sc_client, @@ -714,6 +799,7 @@ static void destroy_delegation(struct nfs4_delegation *dp) spin_lock(&state_lock); unhash_delegation_locked(dp); spin_unlock(&state_lock); + put_clnt_odstate(dp->dl_clnt_odstate); nfs4_put_deleg_lease(dp->dl_stid.sc_file); nfs4_put_stid(&dp->dl_stid); } @@ -724,6 +810,7 @@ static void revoke_delegation(struct nfs4_delegation *dp) WARN_ON(!list_empty(&dp->dl_recall_lru)); + put_clnt_odstate(dp->dl_clnt_odstate); nfs4_put_deleg_lease(dp->dl_stid.sc_file); if (clp->cl_minorversion == 0) @@ -933,6 +1020,7 @@ static void nfs4_free_ol_stateid(struct nfs4_stid *stid) { struct nfs4_ol_stateid *stp = openlockstateid(stid); + put_clnt_odstate(stp->st_clnt_odstate); release_all_access(stp); if (stp->st_stateowner) nfs4_put_stateowner(stp->st_stateowner); @@ -1538,7 +1626,6 @@ static struct nfs4_client *alloc_client(struct xdr_netobj name) INIT_LIST_HEAD(&clp->cl_openowners); INIT_LIST_HEAD(&clp->cl_delegations); INIT_LIST_HEAD(&clp->cl_lru); - INIT_LIST_HEAD(&clp->cl_callbacks); INIT_LIST_HEAD(&clp->cl_revoked); #ifdef CONFIG_NFSD_PNFS INIT_LIST_HEAD(&clp->cl_lo_states); @@ -1634,6 +1721,7 @@ __destroy_client(struct nfs4_client *clp) while (!list_empty(&reaplist)) { dp = list_entry(reaplist.next, struct nfs4_delegation, dl_recall_lru); list_del_init(&dp->dl_recall_lru); + put_clnt_odstate(dp->dl_clnt_odstate); nfs4_put_deleg_lease(dp->dl_stid.sc_file); nfs4_put_stid(&dp->dl_stid); } @@ -3057,6 +3145,7 @@ static void nfsd4_init_file(struct knfsd_fh *fh, unsigned int hashval, spin_lock_init(&fp->fi_lock); INIT_LIST_HEAD(&fp->fi_stateids); INIT_LIST_HEAD(&fp->fi_delegations); + INIT_LIST_HEAD(&fp->fi_clnt_odstate); fh_copy_shallow(&fp->fi_fhandle, fh); fp->fi_deleg_file = NULL; fp->fi_had_conflict = false; @@ -3073,6 +3162,7 @@ static void nfsd4_init_file(struct knfsd_fh *fh, unsigned int hashval, void nfsd4_free_slabs(void) { + kmem_cache_destroy(odstate_slab); kmem_cache_destroy(openowner_slab); kmem_cache_destroy(lockowner_slab); kmem_cache_destroy(file_slab); @@ -3103,8 +3193,14 @@ nfsd4_init_slabs(void) sizeof(struct nfs4_delegation), 0, 0, NULL); if (deleg_slab == NULL) goto out_free_stateid_slab; + odstate_slab = kmem_cache_create("nfsd4_odstate", + sizeof(struct nfs4_clnt_odstate), 0, 0, NULL); + if (odstate_slab == NULL) + goto out_free_deleg_slab; return 0; +out_free_deleg_slab: + kmem_cache_destroy(deleg_slab); out_free_stateid_slab: kmem_cache_destroy(stateid_slab); out_free_file_slab: @@ -3581,6 +3677,14 @@ alloc_stateid: open->op_stp = nfs4_alloc_open_stateid(clp); if (!open->op_stp) return nfserr_jukebox; + + if (nfsd4_has_session(cstate) && + (cstate->current_fh.fh_export->ex_flags & NFSEXP_PNFS)) { + open->op_odstate = alloc_clnt_odstate(clp); + if (!open->op_odstate) + return nfserr_jukebox; + } + return nfs_ok; } @@ -3757,7 +3861,7 @@ static __be32 nfs4_upgrade_open(struct svc_rqst *rqstp, struct nfs4_file *fp, struct svc_fh *cur_fh, struct nfs4_ol_stateid *stp, struct nfsd4_open *open) { __be32 status; - unsigned char old_deny_bmap; + unsigned char old_deny_bmap = stp->st_deny_bmap; if (!test_access(open->op_share_access, stp)) return nfs4_get_vfs_file(rqstp, fp, cur_fh, stp, open); @@ -3766,7 +3870,6 @@ nfs4_upgrade_open(struct svc_rqst *rqstp, struct nfs4_file *fp, struct svc_fh *c spin_lock(&fp->fi_lock); status = nfs4_file_check_deny(fp, open->op_share_deny); if (status == nfs_ok) { - old_deny_bmap = stp->st_deny_bmap; set_deny(open->op_share_deny, stp); fp->fi_share_deny |= (open->op_share_deny & NFS4_SHARE_DENY_BOTH); @@ -3869,7 +3972,7 @@ out_fput: static struct nfs4_delegation * nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh, - struct nfs4_file *fp) + struct nfs4_file *fp, struct nfs4_clnt_odstate *odstate) { int status; struct nfs4_delegation *dp; @@ -3877,7 +3980,7 @@ nfs4_set_delegation(struct nfs4_client *clp, struct svc_fh *fh, if (fp->fi_had_conflict) return ERR_PTR(-EAGAIN); - dp = alloc_init_deleg(clp, fh); + dp = alloc_init_deleg(clp, fh, odstate); if (!dp) return ERR_PTR(-ENOMEM); @@ -3903,6 +4006,7 @@ out_unlock: spin_unlock(&state_lock); out: if (status) { + put_clnt_odstate(dp->dl_clnt_odstate); nfs4_put_stid(&dp->dl_stid); return ERR_PTR(status); } @@ -3980,7 +4084,7 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, default: goto out_no_deleg; } - dp = nfs4_set_delegation(clp, fh, stp->st_stid.sc_file); + dp = nfs4_set_delegation(clp, fh, stp->st_stid.sc_file, stp->st_clnt_odstate); if (IS_ERR(dp)) goto out_no_deleg; @@ -4069,6 +4173,11 @@ nfsd4_process_open2(struct svc_rqst *rqstp, struct svc_fh *current_fh, struct nf release_open_stateid(stp); goto out; } + + stp->st_clnt_odstate = find_or_hash_clnt_odstate(fp, + open->op_odstate); + if (stp->st_clnt_odstate == open->op_odstate) + open->op_odstate = NULL; } update_stateid(&stp->st_stid.sc_stateid); memcpy(&open->op_stateid, &stp->st_stid.sc_stateid, sizeof(stateid_t)); @@ -4129,6 +4238,8 @@ void nfsd4_cleanup_open_state(struct nfsd4_compound_state *cstate, kmem_cache_free(file_slab, open->op_file); if (open->op_stp) nfs4_put_stid(&open->op_stp->st_stid); + if (open->op_odstate) + kmem_cache_free(odstate_slab, open->op_odstate); } __be32 @@ -4385,10 +4496,17 @@ static __be32 check_stateid_generation(stateid_t *in, stateid_t *ref, bool has_s return nfserr_old_stateid; } +static __be32 nfsd4_check_openowner_confirmed(struct nfs4_ol_stateid *ols) +{ + if (ols->st_stateowner->so_is_open_owner && + !(openowner(ols->st_stateowner)->oo_flags & NFS4_OO_CONFIRMED)) + return nfserr_bad_stateid; + return nfs_ok; +} + static __be32 nfsd4_validate_stateid(struct nfs4_client *cl, stateid_t *stateid) { struct nfs4_stid *s; - struct nfs4_ol_stateid *ols; __be32 status = nfserr_bad_stateid; if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) @@ -4418,13 +4536,7 @@ static __be32 nfsd4_validate_stateid(struct nfs4_client *cl, stateid_t *stateid) break; case NFS4_OPEN_STID: case NFS4_LOCK_STID: - ols = openlockstateid(s); - if (ols->st_stateowner->so_is_open_owner - && !(openowner(ols->st_stateowner)->oo_flags - & NFS4_OO_CONFIRMED)) - status = nfserr_bad_stateid; - else - status = nfs_ok; + status = nfsd4_check_openowner_confirmed(openlockstateid(s)); break; default: printk("unknown stateid type %x\n", s->sc_type); @@ -4461,85 +4573,130 @@ nfsd4_lookup_stateid(struct nfsd4_compound_state *cstate, return nfs_ok; } +static struct file * +nfs4_find_file(struct nfs4_stid *s, int flags) +{ + if (!s) + return NULL; + + switch (s->sc_type) { + case NFS4_DELEG_STID: + if (WARN_ON_ONCE(!s->sc_file->fi_deleg_file)) + return NULL; + return get_file(s->sc_file->fi_deleg_file); + case NFS4_OPEN_STID: + case NFS4_LOCK_STID: + if (flags & RD_STATE) + return find_readable_file(s->sc_file); + else + return find_writeable_file(s->sc_file); + break; + } + + return NULL; +} + +static __be32 +nfs4_check_olstateid(struct svc_fh *fhp, struct nfs4_ol_stateid *ols, int flags) +{ + __be32 status; + + status = nfs4_check_fh(fhp, ols); + if (status) + return status; + status = nfsd4_check_openowner_confirmed(ols); + if (status) + return status; + return nfs4_check_openmode(ols, flags); +} + +static __be32 +nfs4_check_file(struct svc_rqst *rqstp, struct svc_fh *fhp, struct nfs4_stid *s, + struct file **filpp, bool *tmp_file, int flags) +{ + int acc = (flags & RD_STATE) ? NFSD_MAY_READ : NFSD_MAY_WRITE; + struct file *file; + __be32 status; + + file = nfs4_find_file(s, flags); + if (file) { + status = nfsd_permission(rqstp, fhp->fh_export, fhp->fh_dentry, + acc | NFSD_MAY_OWNER_OVERRIDE); + if (status) { + fput(file); + return status; + } + + *filpp = file; + } else { + status = nfsd_open(rqstp, fhp, S_IFREG, acc, filpp); + if (status) + return status; + + if (tmp_file) + *tmp_file = true; + } + + return 0; +} + /* -* Checks for stateid operations -*/ + * Checks for stateid operations + */ __be32 -nfs4_preprocess_stateid_op(struct net *net, struct nfsd4_compound_state *cstate, - stateid_t *stateid, int flags, struct file **filpp) +nfs4_preprocess_stateid_op(struct svc_rqst *rqstp, + struct nfsd4_compound_state *cstate, stateid_t *stateid, + int flags, struct file **filpp, bool *tmp_file) { - struct nfs4_stid *s; - struct nfs4_ol_stateid *stp = NULL; - struct nfs4_delegation *dp = NULL; - struct svc_fh *current_fh = &cstate->current_fh; - struct inode *ino = d_inode(current_fh->fh_dentry); + struct svc_fh *fhp = &cstate->current_fh; + struct inode *ino = d_inode(fhp->fh_dentry); + struct net *net = SVC_NET(rqstp); struct nfsd_net *nn = net_generic(net, nfsd_net_id); - struct file *file = NULL; + struct nfs4_stid *s = NULL; __be32 status; if (filpp) *filpp = NULL; + if (tmp_file) + *tmp_file = false; if (grace_disallows_io(net, ino)) return nfserr_grace; - if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) - return check_special_stateids(net, current_fh, stateid, flags); + if (ZERO_STATEID(stateid) || ONE_STATEID(stateid)) { + status = check_special_stateids(net, fhp, stateid, flags); + goto done; + } status = nfsd4_lookup_stateid(cstate, stateid, NFS4_DELEG_STID|NFS4_OPEN_STID|NFS4_LOCK_STID, &s, nn); if (status) return status; - status = check_stateid_generation(stateid, &s->sc_stateid, nfsd4_has_session(cstate)); + status = check_stateid_generation(stateid, &s->sc_stateid, + nfsd4_has_session(cstate)); if (status) goto out; + switch (s->sc_type) { case NFS4_DELEG_STID: - dp = delegstateid(s); - status = nfs4_check_delegmode(dp, flags); - if (status) - goto out; - if (filpp) { - file = dp->dl_stid.sc_file->fi_deleg_file; - if (!file) { - WARN_ON_ONCE(1); - status = nfserr_serverfault; - goto out; - } - get_file(file); - } + status = nfs4_check_delegmode(delegstateid(s), flags); break; case NFS4_OPEN_STID: case NFS4_LOCK_STID: - stp = openlockstateid(s); - status = nfs4_check_fh(current_fh, stp); - if (status) - goto out; - if (stp->st_stateowner->so_is_open_owner - && !(openowner(stp->st_stateowner)->oo_flags & NFS4_OO_CONFIRMED)) - goto out; - status = nfs4_check_openmode(stp, flags); - if (status) - goto out; - if (filpp) { - struct nfs4_file *fp = stp->st_stid.sc_file; - - if (flags & RD_STATE) - file = find_readable_file(fp); - else - file = find_writeable_file(fp); - } + status = nfs4_check_olstateid(fhp, openlockstateid(s), flags); break; default: status = nfserr_bad_stateid; - goto out; + break; } - status = nfs_ok; - if (file) - *filpp = file; + +done: + if (!status && filpp) + status = nfs4_check_file(rqstp, fhp, s, filpp, tmp_file, flags); out: - nfs4_put_stid(s); + if (s) + nfs4_put_stid(s); return status; } @@ -4852,9 +5009,6 @@ nfsd4_close(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, update_stateid(&stp->st_stid.sc_stateid); memcpy(&close->cl_stateid, &stp->st_stid.sc_stateid, sizeof(stateid_t)); - nfsd4_return_all_file_layouts(stp->st_stateowner->so_client, - stp->st_stid.sc_file); - nfsd4_close_open_stateid(stp); /* put reference from nfs4_preprocess_seqid_op */ @@ -5395,7 +5549,7 @@ static __be32 nfsd_test_lock(struct svc_rqst *rqstp, struct svc_fh *fhp, struct __be32 err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file); if (!err) { err = nfserrno(vfs_test_lock(file, lock)); - nfsd_close(file); + fput(file); } return err; } @@ -6488,6 +6642,7 @@ nfs4_state_shutdown_net(struct net *net) list_for_each_safe(pos, next, &reaplist) { dp = list_entry (pos, struct nfs4_delegation, dl_recall_lru); list_del_init(&dp->dl_recall_lru); + put_clnt_odstate(dp->dl_clnt_odstate); nfs4_put_deleg_lease(dp->dl_stid.sc_file); nfs4_put_stid(&dp->dl_stid); } diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 158badf..5463385 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -33,6 +33,7 @@ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ +#include <linux/file.h> #include <linux/slab.h> #include <linux/namei.h> #include <linux/statfs.h> @@ -2227,7 +2228,6 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp, u32 rdattr_err = 0; __be32 status; int err; - int aclsupport = 0; struct nfs4_acl *acl = NULL; void *context = NULL; int contextlen; @@ -2274,19 +2274,15 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp, goto out; fhp = tempfh; } - if (bmval0 & (FATTR4_WORD0_ACL | FATTR4_WORD0_ACLSUPPORT - | FATTR4_WORD0_SUPPORTED_ATTRS)) { + if (bmval0 & FATTR4_WORD0_ACL) { err = nfsd4_get_nfs4_acl(rqstp, dentry, &acl); - aclsupport = (err == 0); - if (bmval0 & FATTR4_WORD0_ACL) { - if (err == -EOPNOTSUPP) - bmval0 &= ~FATTR4_WORD0_ACL; - else if (err == -EINVAL) { - status = nfserr_attrnotsupp; - goto out; - } else if (err != 0) - goto out_nfserr; - } + if (err == -EOPNOTSUPP) + bmval0 &= ~FATTR4_WORD0_ACL; + else if (err == -EINVAL) { + status = nfserr_attrnotsupp; + goto out; + } else if (err != 0) + goto out_nfserr; } #ifdef CONFIG_NFSD_V4_SECURITY_LABEL @@ -2338,7 +2334,7 @@ nfsd4_encode_fattr(struct xdr_stream *xdr, struct svc_fh *fhp, u32 word1 = nfsd_suppattrs1(minorversion); u32 word2 = nfsd_suppattrs2(minorversion); - if (!aclsupport) + if (!IS_POSIXACL(dentry->d_inode)) word0 &= ~FATTR4_WORD0_ACL; if (!contextsupport) word2 &= ~FATTR4_WORD2_SECURITY_LABEL; @@ -2486,7 +2482,7 @@ out_acl: p = xdr_reserve_space(xdr, 4); if (!p) goto out_resource; - *p++ = cpu_to_be32(aclsupport ? + *p++ = cpu_to_be32(IS_POSIXACL(dentry->d_inode) ? ACL4_SUPPORT_ALLOW_ACL|ACL4_SUPPORT_DENY_ACL : 0); } if (bmval0 & FATTR4_WORD0_CANSETTIME) { @@ -3422,52 +3418,51 @@ nfsd4_encode_read(struct nfsd4_compoundres *resp, __be32 nfserr, unsigned long maxcount; struct xdr_stream *xdr = &resp->xdr; struct file *file = read->rd_filp; - struct svc_fh *fhp = read->rd_fhp; int starting_len = xdr->buf->len; - struct raparms *ra; + struct raparms *ra = NULL; __be32 *p; - __be32 err; if (nfserr) - return nfserr; + goto out; p = xdr_reserve_space(xdr, 8); /* eof flag and byte count */ if (!p) { WARN_ON_ONCE(test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags)); - return nfserr_resource; + nfserr = nfserr_resource; + goto out; } - if (resp->xdr.buf->page_len && test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags)) { + if (resp->xdr.buf->page_len && + test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags)) { WARN_ON_ONCE(1); - return nfserr_resource; + nfserr = nfserr_resource; + goto out; } xdr_commit_encode(xdr); maxcount = svc_max_payload(resp->rqstp); - maxcount = min_t(unsigned long, maxcount, (xdr->buf->buflen - xdr->buf->len)); + maxcount = min_t(unsigned long, maxcount, + (xdr->buf->buflen - xdr->buf->len)); maxcount = min_t(unsigned long, maxcount, read->rd_length); - if (read->rd_filp) - err = nfsd_permission(resp->rqstp, fhp->fh_export, - fhp->fh_dentry, - NFSD_MAY_READ|NFSD_MAY_OWNER_OVERRIDE); - else - err = nfsd_get_tmp_read_open(resp->rqstp, read->rd_fhp, - &file, &ra); - if (err) - goto err_truncate; + if (read->rd_tmp_file) + ra = nfsd_init_raparms(file); - if (file->f_op->splice_read && test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags)) - err = nfsd4_encode_splice_read(resp, read, file, maxcount); + if (file->f_op->splice_read && + test_bit(RQ_SPLICE_OK, &resp->rqstp->rq_flags)) + nfserr = nfsd4_encode_splice_read(resp, read, file, maxcount); else - err = nfsd4_encode_readv(resp, read, file, maxcount); + nfserr = nfsd4_encode_readv(resp, read, file, maxcount); - if (!read->rd_filp) - nfsd_put_tmp_read_open(file, ra); + if (ra) + nfsd_put_raparams(file, ra); -err_truncate: - if (err) + if (nfserr) xdr_truncate_encode(xdr, starting_len); - return err; + +out: + if (file) + fput(file); + return nfserr; } static __be32 diff --git a/fs/nfsd/nfsproc.c b/fs/nfsd/nfsproc.c index aecbcd3..4cd78ef 100644 --- a/fs/nfsd/nfsproc.c +++ b/fs/nfsd/nfsproc.c @@ -59,13 +59,61 @@ static __be32 nfsd_proc_setattr(struct svc_rqst *rqstp, struct nfsd_sattrargs *argp, struct nfsd_attrstat *resp) { + struct iattr *iap = &argp->attrs; + struct svc_fh *fhp; __be32 nfserr; + dprintk("nfsd: SETATTR %s, valid=%x, size=%ld\n", SVCFH_fmt(&argp->fh), argp->attrs.ia_valid, (long) argp->attrs.ia_size); - fh_copy(&resp->fh, &argp->fh); - nfserr = nfsd_setattr(rqstp, &resp->fh, &argp->attrs,0, (time_t)0); + fhp = fh_copy(&resp->fh, &argp->fh); + + /* + * NFSv2 does not differentiate between "set-[ac]time-to-now" + * which only requires access, and "set-[ac]time-to-X" which + * requires ownership. + * So if it looks like it might be "set both to the same time which + * is close to now", and if inode_change_ok fails, then we + * convert to "set to now" instead of "set to explicit time" + * + * We only call inode_change_ok as the last test as technically + * it is not an interface that we should be using. + */ +#define BOTH_TIME_SET (ATTR_ATIME_SET | ATTR_MTIME_SET) +#define MAX_TOUCH_TIME_ERROR (30*60) + if ((iap->ia_valid & BOTH_TIME_SET) == BOTH_TIME_SET && + iap->ia_mtime.tv_sec == iap->ia_atime.tv_sec) { + /* + * Looks probable. + * + * Now just make sure time is in the right ballpark. + * Solaris, at least, doesn't seem to care what the time + * request is. We require it be within 30 minutes of now. + */ + time_t delta = iap->ia_atime.tv_sec - get_seconds(); + struct inode *inode; + + nfserr = fh_verify(rqstp, fhp, 0, NFSD_MAY_NOP); + if (nfserr) + goto done; + inode = d_inode(fhp->fh_dentry); + + if (delta < 0) + delta = -delta; + if (delta < MAX_TOUCH_TIME_ERROR && + inode_change_ok(inode, iap) != 0) { + /* + * Turn off ATTR_[AM]TIME_SET but leave ATTR_[AM]TIME. + * This will cause notify_change to set these times + * to "now" + */ + iap->ia_valid &= ~BOTH_TIME_SET; + } + } + + nfserr = nfsd_setattr(rqstp, fhp, iap, 0, (time_t)0); +done: return nfsd_return_attrs(nfserr, resp); } diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h index 4f3bfeb..4874ce5 100644 --- a/fs/nfsd/state.h +++ b/fs/nfsd/state.h @@ -63,12 +63,13 @@ typedef struct { struct nfsd4_callback { struct nfs4_client *cb_clp; - struct list_head cb_per_client; u32 cb_minorversion; struct rpc_message cb_msg; struct nfsd4_callback_ops *cb_ops; struct work_struct cb_work; - bool cb_done; + int cb_status; + bool cb_update_seq_nr; + bool cb_need_restart; }; struct nfsd4_callback_ops { @@ -126,6 +127,7 @@ struct nfs4_delegation { struct list_head dl_perfile; struct list_head dl_perclnt; struct list_head dl_recall_lru; /* delegation recalled */ + struct nfs4_clnt_odstate *dl_clnt_odstate; u32 dl_type; time_t dl_time; /* For recall: */ @@ -332,7 +334,6 @@ struct nfs4_client { int cl_cb_state; struct nfsd4_callback cl_cb_null; struct nfsd4_session *cl_cb_session; - struct list_head cl_callbacks; /* list of in-progress callbacks */ /* for all client information that callback code might need: */ spinlock_t cl_lock; @@ -465,6 +466,17 @@ static inline struct nfs4_lockowner * lockowner(struct nfs4_stateowner *so) } /* + * Per-client state indicating no. of opens and outstanding delegations + * on a file from a particular client.'od' stands for 'open & delegation' + */ +struct nfs4_clnt_odstate { + struct nfs4_client *co_client; + struct nfs4_file *co_file; + struct list_head co_perfile; + atomic_t co_odcount; +}; + +/* * nfs4_file: a file opened by some number of (open) nfs4_stateowners. * * These objects are global. nfsd keeps one instance of a nfs4_file per @@ -485,6 +497,7 @@ struct nfs4_file { struct list_head fi_delegations; struct rcu_head fi_rcu; }; + struct list_head fi_clnt_odstate; /* One each for O_RDONLY, O_WRONLY, O_RDWR: */ struct file * fi_fds[3]; /* @@ -526,6 +539,7 @@ struct nfs4_ol_stateid { struct list_head st_perstateowner; struct list_head st_locks; struct nfs4_stateowner * st_stateowner; + struct nfs4_clnt_odstate * st_clnt_odstate; unsigned char st_access_bmap; unsigned char st_deny_bmap; struct nfs4_ol_stateid * st_openstp; @@ -569,9 +583,9 @@ enum nfsd4_cb_op { struct nfsd4_compound_state; struct nfsd_net; -extern __be32 nfs4_preprocess_stateid_op(struct net *net, - struct nfsd4_compound_state *cstate, - stateid_t *stateid, int flags, struct file **filp); +extern __be32 nfs4_preprocess_stateid_op(struct svc_rqst *rqstp, + struct nfsd4_compound_state *cstate, stateid_t *stateid, + int flags, struct file **filp, bool *tmp_file); __be32 nfsd4_lookup_stateid(struct nfsd4_compound_state *cstate, stateid_t *stateid, unsigned char typemask, struct nfs4_stid **s, struct nfsd_net *nn); diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 84d770b..b5e077a 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -302,42 +302,6 @@ commit_metadata(struct svc_fh *fhp) static void nfsd_sanitize_attrs(struct inode *inode, struct iattr *iap) { - /* - * NFSv2 does not differentiate between "set-[ac]time-to-now" - * which only requires access, and "set-[ac]time-to-X" which - * requires ownership. - * So if it looks like it might be "set both to the same time which - * is close to now", and if inode_change_ok fails, then we - * convert to "set to now" instead of "set to explicit time" - * - * We only call inode_change_ok as the last test as technically - * it is not an interface that we should be using. - */ -#define BOTH_TIME_SET (ATTR_ATIME_SET | ATTR_MTIME_SET) -#define MAX_TOUCH_TIME_ERROR (30*60) - if ((iap->ia_valid & BOTH_TIME_SET) == BOTH_TIME_SET && - iap->ia_mtime.tv_sec == iap->ia_atime.tv_sec) { - /* - * Looks probable. - * - * Now just make sure time is in the right ballpark. - * Solaris, at least, doesn't seem to care what the time - * request is. We require it be within 30 minutes of now. - */ - time_t delta = iap->ia_atime.tv_sec - get_seconds(); - if (delta < 0) - delta = -delta; - if (delta < MAX_TOUCH_TIME_ERROR && - inode_change_ok(inode, iap) != 0) { - /* - * Turn off ATTR_[AM]TIME_SET but leave ATTR_[AM]TIME. - * This will cause notify_change to set these times - * to "now" - */ - iap->ia_valid &= ~BOTH_TIME_SET; - } - } - /* sanitize the mode change */ if (iap->ia_valid & ATTR_MODE) { iap->ia_mode &= S_IALLUGO; @@ -538,16 +502,11 @@ __be32 nfsd4_vfs_fallocate(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, loff_t offset, loff_t len, int flags) { - __be32 err; int error; if (!S_ISREG(file_inode(file)->i_mode)) return nfserr_inval; - err = nfsd_permission(rqstp, fhp->fh_export, fhp->fh_dentry, NFSD_MAY_WRITE); - if (err) - return err; - error = vfs_fallocate(file, flags, offset, len); if (!error) error = commit_metadata(fhp); @@ -744,7 +703,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, umode_t type, host_err = ima_file_check(file, may_flags, 0); if (host_err) { - nfsd_close(file); + fput(file); goto out_nfserr; } @@ -761,23 +720,12 @@ out: return err; } -/* - * Close a file. - */ -void -nfsd_close(struct file *filp) -{ - fput(filp); -} - -/* - * Obtain the readahead parameters for the file - * specified by (dev, ino). - */ - -static inline struct raparms * -nfsd_get_raparms(dev_t dev, ino_t ino) +struct raparms * +nfsd_init_raparms(struct file *file) { + struct inode *inode = file_inode(file); + dev_t dev = inode->i_sb->s_dev; + ino_t ino = inode->i_ino; struct raparms *ra, **rap, **frap = NULL; int depth = 0; unsigned int hash; @@ -814,9 +762,23 @@ found: ra->p_count++; nfsdstats.ra_depth[depth*10/nfsdstats.ra_size]++; spin_unlock(&rab->pb_lock); + + if (ra->p_set) + file->f_ra = ra->p_ra; return ra; } +void nfsd_put_raparams(struct file *file, struct raparms *ra) +{ + struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex]; + + spin_lock(&rab->pb_lock); + ra->p_ra = file->f_ra; + ra->p_set = 1; + ra->p_count--; + spin_unlock(&rab->pb_lock); +} + /* * Grab and keep cached pages associated with a file in the svc_rqst * so that they can be passed to the network sendmsg/sendpage routines @@ -945,7 +907,7 @@ static int wait_for_concurrent_writes(struct file *file) return err; } -static __be32 +__be32 nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, loff_t offset, struct kvec *vec, int vlen, unsigned long *cnt, int *stablep) @@ -1009,40 +971,6 @@ out_nfserr: return err; } -__be32 nfsd_get_tmp_read_open(struct svc_rqst *rqstp, struct svc_fh *fhp, - struct file **file, struct raparms **ra) -{ - struct inode *inode; - __be32 err; - - err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, file); - if (err) - return err; - - inode = file_inode(*file); - - /* Get readahead parameters */ - *ra = nfsd_get_raparms(inode->i_sb->s_dev, inode->i_ino); - - if (*ra && (*ra)->p_set) - (*file)->f_ra = (*ra)->p_ra; - return nfs_ok; -} - -void nfsd_put_tmp_read_open(struct file *file, struct raparms *ra) -{ - /* Write back readahead params */ - if (ra) { - struct raparm_hbucket *rab = &raparm_hash[ra->p_hindex]; - spin_lock(&rab->pb_lock); - ra->p_ra = file->f_ra; - ra->p_set = 1; - ra->p_count--; - spin_unlock(&rab->pb_lock); - } - nfsd_close(file); -} - /* * Read data from a file. count must contain the requested read count * on entry. On return, *count contains the number of bytes actually read. @@ -1055,13 +983,15 @@ __be32 nfsd_read(struct svc_rqst *rqstp, struct svc_fh *fhp, struct raparms *ra; __be32 err; - err = nfsd_get_tmp_read_open(rqstp, fhp, &file, &ra); + err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_READ, &file); if (err) return err; + ra = nfsd_init_raparms(file); err = nfsd_vfs_read(rqstp, file, offset, vec, vlen, count); - - nfsd_put_tmp_read_open(file, ra); + if (ra) + nfsd_put_raparams(file, ra); + fput(file); return err; } @@ -1093,7 +1023,7 @@ nfsd_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file, if (cnt) err = nfsd_vfs_write(rqstp, fhp, file, offset, vec, vlen, cnt, stablep); - nfsd_close(file); + fput(file); } out: return err; @@ -1138,7 +1068,7 @@ nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp, err = nfserr_notsupp; } - nfsd_close(file); + fput(file); out: return err; } @@ -1977,7 +1907,7 @@ nfsd_readdir(struct svc_rqst *rqstp, struct svc_fh *fhp, loff_t *offsetp, if (err == nfserr_eof || err == nfserr_toosmall) err = nfs_ok; /* can still be found in ->err */ out_close: - nfsd_close(file); + fput(file); out: return err; } diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h index 2050cb0..5be875e 100644 --- a/fs/nfsd/vfs.h +++ b/fs/nfsd/vfs.h @@ -71,11 +71,7 @@ __be32 nfsd_commit(struct svc_rqst *, struct svc_fh *, #endif /* CONFIG_NFSD_V3 */ __be32 nfsd_open(struct svc_rqst *, struct svc_fh *, umode_t, int, struct file **); -void nfsd_close(struct file *); struct raparms; -__be32 nfsd_get_tmp_read_open(struct svc_rqst *, struct svc_fh *, - struct file **, struct raparms **); -void nfsd_put_tmp_read_open(struct file *, struct raparms *); __be32 nfsd_splice_read(struct svc_rqst *, struct file *, loff_t, unsigned long *); __be32 nfsd_readv(struct file *, loff_t, struct kvec *, int, @@ -84,6 +80,10 @@ __be32 nfsd_read(struct svc_rqst *, struct svc_fh *, loff_t, struct kvec *, int, unsigned long *); __be32 nfsd_write(struct svc_rqst *, struct svc_fh *,struct file *, loff_t, struct kvec *,int, unsigned long *, int *); +__be32 nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, + struct file *file, loff_t offset, + struct kvec *vec, int vlen, unsigned long *cnt, + int *stablep); __be32 nfsd_readlink(struct svc_rqst *, struct svc_fh *, char *, int *); __be32 nfsd_symlink(struct svc_rqst *, struct svc_fh *, @@ -104,6 +104,9 @@ __be32 nfsd_statfs(struct svc_rqst *, struct svc_fh *, __be32 nfsd_permission(struct svc_rqst *, struct svc_export *, struct dentry *, int); +struct raparms *nfsd_init_raparms(struct file *file); +void nfsd_put_raparams(struct file *file, struct raparms *ra); + static inline int fh_want_write(struct svc_fh *fh) { int ret = mnt_want_write(fh->fh_export->ex_path.mnt); diff --git a/fs/nfsd/xdr4.h b/fs/nfsd/xdr4.h index f982ae8..9f99100 100644 --- a/fs/nfsd/xdr4.h +++ b/fs/nfsd/xdr4.h @@ -247,6 +247,7 @@ struct nfsd4_open { struct nfs4_openowner *op_openowner; /* used during processing */ struct nfs4_file *op_file; /* used during processing */ struct nfs4_ol_stateid *op_stp; /* used during processing */ + struct nfs4_clnt_odstate *op_odstate; /* used during processing */ struct nfs4_acl *op_acl; struct xdr_netobj op_label; }; @@ -272,6 +273,7 @@ struct nfsd4_read { u32 rd_length; /* request */ int rd_vlen; struct file *rd_filp; + bool rd_tmp_file; struct svc_rqst *rd_rqstp; /* response */ struct svc_fh * rd_fhp; /* response */ diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c index 2218083..37dd6b0 100644 --- a/fs/nilfs2/namei.c +++ b/fs/nilfs2/namei.c @@ -496,8 +496,7 @@ static struct dentry *nilfs_fh_to_dentry(struct super_block *sb, struct fid *fh, { struct nilfs_fid *fid = (struct nilfs_fid *)fh; - if ((fh_len != NILFS_FID_SIZE_NON_CONNECTABLE && - fh_len != NILFS_FID_SIZE_CONNECTABLE) || + if (fh_len < NILFS_FID_SIZE_NON_CONNECTABLE || (fh_type != FILEID_NILFS_WITH_PARENT && fh_type != FILEID_NILFS_WITHOUT_PARENT)) return NULL; @@ -510,7 +509,7 @@ static struct dentry *nilfs_fh_to_parent(struct super_block *sb, struct fid *fh, { struct nilfs_fid *fid = (struct nilfs_fid *)fh; - if (fh_len != NILFS_FID_SIZE_CONNECTABLE || + if (fh_len < NILFS_FID_SIZE_CONNECTABLE || fh_type != FILEID_NILFS_WITH_PARENT) return NULL; diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c index dc3a9efd..42468e5 100644 --- a/fs/nilfs2/segbuf.c +++ b/fs/nilfs2/segbuf.c @@ -343,11 +343,6 @@ static void nilfs_end_bio_write(struct bio *bio, int err) const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); struct nilfs_segment_buffer *segbuf = bio->bi_private; - if (err == -EOPNOTSUPP) { - set_bit(BIO_EOPNOTSUPP, &bio->bi_flags); - /* to be detected by nilfs_segbuf_submit_bio() */ - } - if (!uptodate) atomic_inc(&segbuf->sb_err); @@ -374,15 +369,8 @@ static int nilfs_segbuf_submit_bio(struct nilfs_segment_buffer *segbuf, bio->bi_end_io = nilfs_end_bio_write; bio->bi_private = segbuf; - bio_get(bio); submit_bio(mode, bio); segbuf->sb_nbio++; - if (bio_flagged(bio, BIO_EOPNOTSUPP)) { - bio_put(bio); - err = -EOPNOTSUPP; - goto failed; - } - bio_put(bio); wi->bio = NULL; wi->rest_blocks -= wi->end - wi->start; diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 45064869..5b1e2a4 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -26,7 +26,7 @@ #include <linux/fs.h> /* struct inode */ #include <linux/fsnotify_backend.h> #include <linux/idr.h> -#include <linux/init.h> /* module_init */ +#include <linux/init.h> /* fs_initcall */ #include <linux/inotify.h> #include <linux/kernel.h> /* roundup() */ #include <linux/namei.h> /* LOOKUP_FOLLOW */ @@ -812,4 +812,4 @@ static int __init inotify_user_setup(void) return 0; } -module_init(inotify_user_setup); +fs_initcall(inotify_user_setup); diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c index 182bb93..262561f 100644 --- a/fs/ntfs/file.c +++ b/fs/ntfs/file.c @@ -525,7 +525,8 @@ static inline int __ntfs_grab_cache_pages(struct address_space *mapping, } } err = add_to_page_cache_lru(*cached_page, mapping, - index, GFP_KERNEL); + index, + GFP_KERNEL & mapping_gfp_mask(mapping)); if (unlikely(err)) { if (err == -EEXIST) continue; diff --git a/fs/ntfs/malloc.h b/fs/ntfs/malloc.h index a44b14c..ab172e5 100644 --- a/fs/ntfs/malloc.h +++ b/fs/ntfs/malloc.h @@ -85,12 +85,7 @@ static inline void *ntfs_malloc_nofs_nofail(unsigned long size) static inline void ntfs_free(void *addr) { - if (!is_vmalloc_addr(addr)) { - kfree(addr); - /* free_page((unsigned long)addr); */ - return; - } - vfree(addr); + kvfree(addr); } #endif /* _LINUX_NTFS_MALLOC_H */ diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 2d7f76e..5997c00 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -2925,7 +2925,8 @@ static int __ocfs2_rotate_tree_left(handle_t *handle, struct ocfs2_path *right_path = NULL; struct super_block *sb = ocfs2_metadata_cache_get_super(et->et_ci); - BUG_ON(!ocfs2_is_empty_extent(&(path_leaf_el(path)->l_recs[0]))); + if (!ocfs2_is_empty_extent(&(path_leaf_el(path)->l_recs[0]))) + return 0; *empty_extent_path = NULL; @@ -4311,13 +4312,13 @@ out: return ret; } -static enum ocfs2_contig_type -ocfs2_figure_merge_contig_type(struct ocfs2_extent_tree *et, +static int ocfs2_figure_merge_contig_type(struct ocfs2_extent_tree *et, struct ocfs2_path *path, struct ocfs2_extent_list *el, int index, - struct ocfs2_extent_rec *split_rec) + struct ocfs2_extent_rec *split_rec, + struct ocfs2_merge_ctxt *ctxt) { - int status; + int status = 0; enum ocfs2_contig_type ret = CONTIG_NONE; u32 left_cpos, right_cpos; struct ocfs2_extent_rec *rec = NULL; @@ -4336,8 +4337,11 @@ ocfs2_figure_merge_contig_type(struct ocfs2_extent_tree *et, if (left_cpos != 0) { left_path = ocfs2_new_path_from_path(path); - if (!left_path) + if (!left_path) { + status = -ENOMEM; + mlog_errno(status); goto exit; + } status = ocfs2_find_path(et->et_ci, left_path, left_cpos); @@ -4392,8 +4396,11 @@ ocfs2_figure_merge_contig_type(struct ocfs2_extent_tree *et, goto free_left_path; right_path = ocfs2_new_path_from_path(path); - if (!right_path) + if (!right_path) { + status = -ENOMEM; + mlog_errno(status); goto free_left_path; + } status = ocfs2_find_path(et->et_ci, right_path, right_cpos); if (status) @@ -4433,7 +4440,10 @@ free_right_path: free_left_path: ocfs2_free_path(left_path); exit: - return ret; + if (status == 0) + ctxt->c_contig_type = ret; + + return status; } static void ocfs2_figure_contig_type(struct ocfs2_extent_tree *et, @@ -5039,9 +5049,14 @@ int ocfs2_split_extent(handle_t *handle, goto out; } - ctxt.c_contig_type = ocfs2_figure_merge_contig_type(et, path, el, - split_index, - split_rec); + ret = ocfs2_figure_merge_contig_type(et, path, el, + split_index, + split_rec, + &ctxt); + if (ret) { + mlog_errno(ret); + goto out; + } /* * The core merge / split code wants to know how much room is diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index f906a25..1a35c61 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -523,7 +523,7 @@ static int ocfs2_direct_IO_get_blocks(struct inode *inode, sector_t iblock, unsigned char blocksize_bits = inode->i_sb->s_blocksize_bits; unsigned long max_blocks = bh_result->b_size >> inode->i_blkbits; unsigned long len = bh_result->b_size; - unsigned int clusters_to_alloc = 0; + unsigned int clusters_to_alloc = 0, contig_clusters = 0; cpos = ocfs2_blocks_to_clusters(inode->i_sb, iblock); @@ -560,8 +560,10 @@ static int ocfs2_direct_IO_get_blocks(struct inode *inode, sector_t iblock, /* fill hole, allocate blocks can't be larger than the size * of the hole */ clusters_to_alloc = ocfs2_clusters_for_bytes(inode->i_sb, len); - if (clusters_to_alloc > contig_blocks) - clusters_to_alloc = contig_blocks; + contig_clusters = ocfs2_clusters_for_blocks(inode->i_sb, + contig_blocks); + if (clusters_to_alloc > contig_clusters) + clusters_to_alloc = contig_clusters; /* allocate extent and insert them into the extent tree */ ret = ocfs2_extend_allocation(inode, cpos, @@ -619,9 +621,6 @@ static void ocfs2_dio_end_io(struct kiocb *iocb, /* this io's submitter should not have unlocked this before we could */ BUG_ON(!ocfs2_iocb_is_rw_locked(iocb)); - if (ocfs2_iocb_is_sem_locked(iocb)) - ocfs2_iocb_clear_sem_locked(iocb); - if (ocfs2_iocb_is_unaligned_aio(iocb)) { ocfs2_iocb_clear_unaligned_aio(iocb); @@ -925,13 +924,23 @@ clean_orphan: int update_isize = written > 0 ? 1 : 0; loff_t end = update_isize ? offset + written : 0; - tmp_ret = ocfs2_del_inode_from_orphan(osb, inode, + tmp_ret = ocfs2_inode_lock(inode, &di_bh, 1); + if (tmp_ret < 0) { + ret = tmp_ret; + mlog_errno(ret); + goto out; + } + + tmp_ret = ocfs2_del_inode_from_orphan(osb, inode, di_bh, update_isize, end); if (tmp_ret < 0) { ret = tmp_ret; + mlog_errno(ret); goto out; } + ocfs2_inode_unlock(inode, 1); + tmp_ret = jbd2_journal_force_commit(journal); if (tmp_ret < 0) { ret = tmp_ret; diff --git a/fs/ocfs2/aops.h b/fs/ocfs2/aops.h index dd59599..24e496d 100644 --- a/fs/ocfs2/aops.h +++ b/fs/ocfs2/aops.h @@ -79,7 +79,6 @@ static inline void ocfs2_iocb_set_rw_locked(struct kiocb *iocb, int level) enum ocfs2_iocb_lock_bits { OCFS2_IOCB_RW_LOCK = 0, OCFS2_IOCB_RW_LOCK_LEVEL, - OCFS2_IOCB_SEM, OCFS2_IOCB_UNALIGNED_IO, OCFS2_IOCB_NUM_LOCKS }; @@ -88,12 +87,6 @@ enum ocfs2_iocb_lock_bits { clear_bit(OCFS2_IOCB_RW_LOCK, (unsigned long *)&iocb->private) #define ocfs2_iocb_rw_locked_level(iocb) \ test_bit(OCFS2_IOCB_RW_LOCK_LEVEL, (unsigned long *)&iocb->private) -#define ocfs2_iocb_set_sem_locked(iocb) \ - set_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private) -#define ocfs2_iocb_clear_sem_locked(iocb) \ - clear_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private) -#define ocfs2_iocb_is_sem_locked(iocb) \ - test_bit(OCFS2_IOCB_SEM, (unsigned long *)&iocb->private) #define ocfs2_iocb_set_unaligned_aio(iocb) \ set_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private) diff --git a/fs/ocfs2/cluster/masklog.c b/fs/ocfs2/cluster/masklog.c index af7598b..dfe162f 100644 --- a/fs/ocfs2/cluster/masklog.c +++ b/fs/ocfs2/cluster/masklog.c @@ -64,6 +64,40 @@ static ssize_t mlog_mask_store(u64 mask, const char *buf, size_t count) return count; } +void __mlog_printk(const u64 *mask, const char *func, int line, + const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + const char *level; + const char *prefix = ""; + + if (!__mlog_test_u64(*mask, mlog_and_bits) || + __mlog_test_u64(*mask, mlog_not_bits)) + return; + + if (*mask & ML_ERROR) { + level = KERN_ERR; + prefix = "ERROR: "; + } else if (*mask & ML_NOTICE) { + level = KERN_NOTICE; + } else { + level = KERN_INFO; + } + + va_start(args, fmt); + + vaf.fmt = fmt; + vaf.va = &args; + + printk("%s(%s,%u,%u):%s:%d %s%pV", + level, current->comm, task_pid_nr(current), + raw_smp_processor_id(), func, line, prefix, &vaf); + + va_end(args); +} +EXPORT_SYMBOL_GPL(__mlog_printk); + struct mlog_attribute { struct attribute attr; u64 mask; diff --git a/fs/ocfs2/cluster/masklog.h b/fs/ocfs2/cluster/masklog.h index 7fdc25a..308ea0e 100644 --- a/fs/ocfs2/cluster/masklog.h +++ b/fs/ocfs2/cluster/masklog.h @@ -162,38 +162,20 @@ extern struct mlog_bits mlog_and_bits, mlog_not_bits; #endif -/* - * smp_processor_id() "helpfully" screams when called outside preemptible - * regions in current kernels. sles doesn't have the variants that don't - * scream. just do this instead of trying to guess which we're building - * against.. *sigh*. - */ -#define __mlog_cpu_guess ({ \ - unsigned long _cpu = get_cpu(); \ - put_cpu(); \ - _cpu; \ -}) +__printf(4, 5) +void __mlog_printk(const u64 *m, const char *func, int line, + const char *fmt, ...); -/* In the following two macros, the whitespace after the ',' just - * before ##args is intentional. Otherwise, gcc 2.95 will eat the - * previous token if args expands to nothing. +/* + * Testing before the __mlog_printk call lets the compiler eliminate the + * call completely when (m & ML_ALLOWED_BITS) is 0. */ -#define __mlog_printk(level, fmt, args...) \ - printk(level "(%s,%u,%lu):%s:%d " fmt, current->comm, \ - task_pid_nr(current), __mlog_cpu_guess, \ - __PRETTY_FUNCTION__, __LINE__ , ##args) - -#define mlog(mask, fmt, args...) do { \ - u64 __m = MLOG_MASK_PREFIX | (mask); \ - if ((__m & ML_ALLOWED_BITS) && \ - __mlog_test_u64(__m, mlog_and_bits) && \ - !__mlog_test_u64(__m, mlog_not_bits)) { \ - if (__m & ML_ERROR) \ - __mlog_printk(KERN_ERR, "ERROR: "fmt , ##args); \ - else if (__m & ML_NOTICE) \ - __mlog_printk(KERN_NOTICE, fmt , ##args); \ - else __mlog_printk(KERN_INFO, fmt , ##args); \ - } \ +#define mlog(mask, fmt, ...) \ +do { \ + u64 _m = MLOG_MASK_PREFIX | (mask); \ + if (_m & ML_ALLOWED_BITS) \ + __mlog_printk(&_m, __func__, __LINE__, fmt, \ + ##__VA_ARGS__); \ } while (0) #define mlog_errno(st) ({ \ diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 56c403a..2d0acd6 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -2204,7 +2204,7 @@ out: kfree(o2net_hand); kfree(o2net_keep_req); kfree(o2net_keep_resp); - + o2net_debugfs_exit(); o2quo_exit(); return -ENOMEM; } diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index ccd4dcf..02878a8 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -1617,7 +1617,7 @@ int __ocfs2_add_entry(handle_t *handle, struct ocfs2_dir_entry *de, *de1; struct ocfs2_dinode *di = (struct ocfs2_dinode *)parent_fe_bh->b_data; struct super_block *sb = dir->i_sb; - int retval, status; + int retval; unsigned int size = sb->s_blocksize; struct buffer_head *insert_bh = lookup->dl_leaf_bh; char *data_start = insert_bh->b_data; @@ -1695,25 +1695,25 @@ int __ocfs2_add_entry(handle_t *handle, } if (insert_bh == parent_fe_bh) - status = ocfs2_journal_access_di(handle, + retval = ocfs2_journal_access_di(handle, INODE_CACHE(dir), insert_bh, OCFS2_JOURNAL_ACCESS_WRITE); else { - status = ocfs2_journal_access_db(handle, + retval = ocfs2_journal_access_db(handle, INODE_CACHE(dir), insert_bh, OCFS2_JOURNAL_ACCESS_WRITE); - if (ocfs2_dir_indexed(dir)) { - status = ocfs2_dx_dir_insert(dir, + if (!retval && ocfs2_dir_indexed(dir)) + retval = ocfs2_dx_dir_insert(dir, handle, lookup); - if (status) { - mlog_errno(status); - goto bail; - } - } + } + + if (retval) { + mlog_errno(retval); + goto bail; } /* By now the buffer is marked for journaling */ @@ -3543,13 +3543,10 @@ static void dx_leaf_sort_swap(void *a, void *b, int size) { struct ocfs2_dx_entry *entry1 = a; struct ocfs2_dx_entry *entry2 = b; - struct ocfs2_dx_entry tmp; BUG_ON(size != sizeof(*entry1)); - tmp = *entry1; - *entry1 = *entry2; - *entry2 = tmp; + swap(*entry1, *entry2); } static int ocfs2_dx_leaf_same_major(struct ocfs2_dx_leaf *dx_leaf) diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h index fae17c6..e88ccf8 100644 --- a/fs/ocfs2/dlm/dlmcommon.h +++ b/fs/ocfs2/dlm/dlmcommon.h @@ -1014,7 +1014,6 @@ void dlm_move_lockres_to_recovery_list(struct dlm_ctxt *dlm, /* will exit holding res->spinlock, but may drop in function */ void __dlm_wait_on_lockres_flags(struct dlm_lock_resource *res, int flags); -void __dlm_wait_on_lockres_flags_set(struct dlm_lock_resource *res, int flags); /* will exit holding res->spinlock, but may drop in function */ static inline void __dlm_wait_on_lockres(struct dlm_lock_resource *res) diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index d8b670c..719f7f4 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -37,6 +37,7 @@ #include <linux/falloc.h> #include <linux/quotaops.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <cluster/masklog.h> @@ -2250,7 +2251,7 @@ out: static ssize_t ocfs2_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { - int direct_io, appending, rw_level, have_alloc_sem = 0; + int direct_io, appending, rw_level; int can_do_direct, has_refcount = 0; ssize_t written = 0; ssize_t ret; @@ -2279,16 +2280,7 @@ static ssize_t ocfs2_file_write_iter(struct kiocb *iocb, mutex_lock(&inode->i_mutex); - ocfs2_iocb_clear_sem_locked(iocb); - relock: - /* to match setattr's i_mutex -> rw_lock ordering */ - if (direct_io) { - have_alloc_sem = 1; - /* communicate with ocfs2_dio_end_io */ - ocfs2_iocb_set_sem_locked(iocb); - } - /* * Concurrent O_DIRECT writes are allowed with * mount_option "coherency=buffered". @@ -2298,7 +2290,7 @@ relock: ret = ocfs2_rw_lock(inode, rw_level); if (ret < 0) { mlog_errno(ret); - goto out_sems; + goto out_mutex; } /* @@ -2347,7 +2339,6 @@ relock: if (direct_io && !can_do_direct) { ocfs2_rw_unlock(inode, rw_level); - have_alloc_sem = 0; rw_level = -1; direct_io = 0; @@ -2416,7 +2407,6 @@ no_sync: */ if ((ret == -EIOCBQUEUED) || (!ocfs2_iocb_is_rw_locked(iocb))) { rw_level = -1; - have_alloc_sem = 0; unaligned_dio = 0; } @@ -2429,10 +2419,7 @@ out: if (rw_level != -1) ocfs2_rw_unlock(inode, rw_level); -out_sems: - if (have_alloc_sem) - ocfs2_iocb_clear_sem_locked(iocb); - +out_mutex: mutex_unlock(&inode->i_mutex); if (written) @@ -2473,7 +2460,7 @@ bail: static ssize_t ocfs2_file_read_iter(struct kiocb *iocb, struct iov_iter *to) { - int ret = 0, rw_level = -1, have_alloc_sem = 0, lock_level = 0; + int ret = 0, rw_level = -1, lock_level = 0; struct file *filp = iocb->ki_filp; struct inode *inode = file_inode(filp); @@ -2490,16 +2477,11 @@ static ssize_t ocfs2_file_read_iter(struct kiocb *iocb, goto bail; } - ocfs2_iocb_clear_sem_locked(iocb); - /* * buffered reads protect themselves in ->readpage(). O_DIRECT reads * need locks to protect pending reads from racing with truncate. */ if (iocb->ki_flags & IOCB_DIRECT) { - have_alloc_sem = 1; - ocfs2_iocb_set_sem_locked(iocb); - ret = ocfs2_rw_lock(inode, 0); if (ret < 0) { mlog_errno(ret); @@ -2535,13 +2517,9 @@ static ssize_t ocfs2_file_read_iter(struct kiocb *iocb, /* see ocfs2_file_write_iter */ if (ret == -EIOCBQUEUED || !ocfs2_iocb_is_rw_locked(iocb)) { rw_level = -1; - have_alloc_sem = 0; } bail: - if (have_alloc_sem) - ocfs2_iocb_clear_sem_locked(iocb); - if (rw_level != -1) ocfs2_rw_unlock(inode, rw_level); diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index ff53192..7c099f7 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -108,7 +108,7 @@ struct ocfs2_replay_map { unsigned char rm_replay_slots[0]; }; -void ocfs2_replay_map_set_state(struct ocfs2_super *osb, int state) +static void ocfs2_replay_map_set_state(struct ocfs2_super *osb, int state) { if (!osb->replay_map) return; @@ -153,7 +153,7 @@ int ocfs2_compute_replay_slots(struct ocfs2_super *osb) return 0; } -void ocfs2_queue_replay_slots(struct ocfs2_super *osb, +static void ocfs2_queue_replay_slots(struct ocfs2_super *osb, enum ocfs2_orphan_reco_type orphan_reco_type) { struct ocfs2_replay_map *replay_map = osb->replay_map; @@ -173,7 +173,7 @@ void ocfs2_queue_replay_slots(struct ocfs2_super *osb, replay_map->rm_state = REPLAY_DONE; } -void ocfs2_free_replay_slots(struct ocfs2_super *osb) +static void ocfs2_free_replay_slots(struct ocfs2_super *osb) { struct ocfs2_replay_map *replay_map = osb->replay_map; @@ -571,9 +571,7 @@ static void ocfs2_abort_trigger(struct jbd2_buffer_trigger_type *triggers, (unsigned long)bh, (unsigned long long)bh->b_blocknr); - /* We aren't guaranteed to have the superblock here - but if we - * don't, it'll just crash. */ - ocfs2_error(bh->b_assoc_map->host->i_sb, + ocfs2_error(bh->b_bdev->bd_super, "JBD2 has aborted our journal, ocfs2 cannot continue\n"); } @@ -775,7 +773,20 @@ void ocfs2_journal_dirty(handle_t *handle, struct buffer_head *bh) trace_ocfs2_journal_dirty((unsigned long long)bh->b_blocknr); status = jbd2_journal_dirty_metadata(handle, bh); - BUG_ON(status); + if (status) { + mlog_errno(status); + if (!is_handle_aborted(handle)) { + journal_t *journal = handle->h_transaction->t_journal; + struct super_block *sb = bh->b_bdev->bd_super; + + mlog(ML_ERROR, "jbd2_journal_dirty_metadata failed. " + "Aborting transaction and journal.\n"); + handle->h_err = status; + jbd2_journal_abort_handle(handle); + jbd2_journal_abort(journal, status); + ocfs2_abort(sb, "Journal already aborted.\n"); + } + } } #define OCFS2_DEFAULT_COMMIT_INTERVAL (HZ * JBD2_DEFAULT_MAX_COMMIT_AGE) @@ -1884,7 +1895,7 @@ static inline unsigned long ocfs2_orphan_scan_timeout(void) * hasn't happened. The node queues a scan and increments the * sequence number in the LVB. */ -void ocfs2_queue_orphan_scan(struct ocfs2_super *osb) +static void ocfs2_queue_orphan_scan(struct ocfs2_super *osb) { struct ocfs2_orphan_scan *os; int status, i; @@ -1933,7 +1944,7 @@ out: } /* Worker task that gets fired every ORPHAN_SCAN_SCHEDULE_TIMEOUT millsec */ -void ocfs2_orphan_scan_work(struct work_struct *work) +static void ocfs2_orphan_scan_work(struct work_struct *work) { struct ocfs2_orphan_scan *os; struct ocfs2_super *osb; @@ -2137,6 +2148,8 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb, struct inode *inode = NULL; struct inode *iter; struct ocfs2_inode_info *oi; + struct buffer_head *di_bh = NULL; + struct ocfs2_dinode *di = NULL; trace_ocfs2_recover_orphans(slot); @@ -2157,16 +2170,22 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb, iter = oi->ip_next_orphan; oi->ip_next_orphan = NULL; + ret = ocfs2_rw_lock(inode, 1); + if (ret < 0) { + mlog_errno(ret); + goto next; + } /* * We need to take and drop the inode lock to * force read inode from disk. */ - ret = ocfs2_inode_lock(inode, NULL, 0); + ret = ocfs2_inode_lock(inode, &di_bh, 1); if (ret) { mlog_errno(ret); - goto next; + goto unlock_rw; } - ocfs2_inode_unlock(inode, 0); + + di = (struct ocfs2_dinode *)di_bh->b_data; if (inode->i_nlink == 0) { spin_lock(&oi->ip_lock); @@ -2174,43 +2193,30 @@ static int ocfs2_recover_orphans(struct ocfs2_super *osb, * ocfs2_delete_inode. */ oi->ip_flags |= OCFS2_INODE_MAYBE_ORPHANED; spin_unlock(&oi->ip_lock); - } else if (orphan_reco_type == ORPHAN_NEED_TRUNCATE) { - struct buffer_head *di_bh = NULL; - - ret = ocfs2_rw_lock(inode, 1); - if (ret) { - mlog_errno(ret); - goto next; - } - - ret = ocfs2_inode_lock(inode, &di_bh, 1); - if (ret < 0) { - ocfs2_rw_unlock(inode, 1); - mlog_errno(ret); - goto next; - } - + } else if ((orphan_reco_type == ORPHAN_NEED_TRUNCATE) && + (di->i_flags & cpu_to_le32(OCFS2_DIO_ORPHANED_FL))) { ret = ocfs2_truncate_file(inode, di_bh, i_size_read(inode)); - ocfs2_inode_unlock(inode, 1); - ocfs2_rw_unlock(inode, 1); - brelse(di_bh); if (ret < 0) { if (ret != -ENOSPC) mlog_errno(ret); - goto next; + goto unlock_inode; } - ret = ocfs2_del_inode_from_orphan(osb, inode, 0, 0); + ret = ocfs2_del_inode_from_orphan(osb, inode, di_bh, 0, 0); if (ret) mlog_errno(ret); wake_up(&OCFS2_I(inode)->append_dio_wq); } /* else if ORPHAN_NO_NEED_TRUNCATE, do nothing */ - +unlock_inode: + ocfs2_inode_unlock(inode, 1); +unlock_rw: + ocfs2_rw_unlock(inode, 1); next: iput(inode); - + brelse(di_bh); + di_bh = NULL; inode = iter; } diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 176fe6a..6e6abb9 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -1116,8 +1116,6 @@ static int ocfs2_double_lock(struct ocfs2_super *osb, int inode1_is_ancestor, inode2_is_ancestor; struct ocfs2_inode_info *oi1 = OCFS2_I(inode1); struct ocfs2_inode_info *oi2 = OCFS2_I(inode2); - struct buffer_head **tmpbh; - struct inode *tmpinode; trace_ocfs2_double_lock((unsigned long long)oi1->ip_blkno, (unsigned long long)oi2->ip_blkno); @@ -1148,13 +1146,8 @@ static int ocfs2_double_lock(struct ocfs2_super *osb, (oi1->ip_blkno < oi2->ip_blkno && inode2_is_ancestor == 0)) { /* switch id1 and id2 around */ - tmpbh = bh2; - bh2 = bh1; - bh1 = tmpbh; - - tmpinode = inode2; - inode2 = inode1; - inode1 = tmpinode; + swap(bh2, bh1); + swap(inode2, inode1); } /* lock id2 */ status = ocfs2_inode_lock_nested(inode2, bh2, 1, @@ -2670,30 +2663,22 @@ bail: } int ocfs2_del_inode_from_orphan(struct ocfs2_super *osb, - struct inode *inode, int update_isize, - loff_t end) + struct inode *inode, struct buffer_head *di_bh, + int update_isize, loff_t end) { struct inode *orphan_dir_inode = NULL; struct buffer_head *orphan_dir_bh = NULL; - struct buffer_head *di_bh = NULL; - struct ocfs2_dinode *di = NULL; + struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; handle_t *handle = NULL; int status = 0; - status = ocfs2_inode_lock(inode, &di_bh, 1); - if (status < 0) { - mlog_errno(status); - goto bail; - } - di = (struct ocfs2_dinode *) di_bh->b_data; - orphan_dir_inode = ocfs2_get_system_file_inode(osb, ORPHAN_DIR_SYSTEM_INODE, le16_to_cpu(di->i_dio_orphaned_slot)); if (!orphan_dir_inode) { status = -ENOENT; mlog_errno(status); - goto bail_unlock_inode; + goto bail; } mutex_lock(&orphan_dir_inode->i_mutex); @@ -2702,7 +2687,7 @@ int ocfs2_del_inode_from_orphan(struct ocfs2_super *osb, mutex_unlock(&orphan_dir_inode->i_mutex); iput(orphan_dir_inode); mlog_errno(status); - goto bail_unlock_inode; + goto bail; } handle = ocfs2_start_trans(osb, @@ -2749,10 +2734,6 @@ bail_unlock_orphan: brelse(orphan_dir_bh); iput(orphan_dir_inode); -bail_unlock_inode: - ocfs2_inode_unlock(inode, 1); - brelse(di_bh); - bail: return status; } diff --git a/fs/ocfs2/namei.h b/fs/ocfs2/namei.h index 5ddecce..e173329 100644 --- a/fs/ocfs2/namei.h +++ b/fs/ocfs2/namei.h @@ -42,8 +42,8 @@ int ocfs2_create_inode_in_orphan(struct inode *dir, int ocfs2_add_inode_to_orphan(struct ocfs2_super *osb, struct inode *inode); int ocfs2_del_inode_from_orphan(struct ocfs2_super *osb, - struct inode *inode, int update_isize, - loff_t end); + struct inode *inode, struct buffer_head *di_bh, + int update_isize, loff_t end); int ocfs2_mv_orphaned_inode_to_new(struct inode *dir, struct inode *new_inode, struct dentry *new_dentry); diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 460c6c3..690ddc6 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -717,6 +717,16 @@ static inline u64 ocfs2_clusters_to_blocks(struct super_block *sb, return (u64)clusters << c_to_b_bits; } +static inline u32 ocfs2_clusters_for_blocks(struct super_block *sb, + u64 blocks) +{ + int b_to_c_bits = OCFS2_SB(sb)->s_clustersize_bits - + sb->s_blocksize_bits; + + blocks += (1 << b_to_c_bits) - 1; + return (u32)(blocks >> b_to_c_bits); +} + static inline u32 ocfs2_blocks_to_clusters(struct super_block *sb, u64 blocks) { diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index d8c6af1..b69dd14 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -1406,11 +1406,9 @@ static int cmp_refcount_rec_by_cpos(const void *a, const void *b) static void swap_refcount_rec(void *a, void *b, int size) { - struct ocfs2_refcount_rec *l = a, *r = b, tmp; + struct ocfs2_refcount_rec *l = a, *r = b; - tmp = *l; - *l = *r; - *r = tmp; + swap(*l, *r); } /* diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index d03bfbf..889f379 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -7271,7 +7271,7 @@ static int ocfs2_xattr_security_set(struct dentry *dentry, const char *name, name, value, size, flags); } -int ocfs2_initxattrs(struct inode *inode, const struct xattr *xattr_array, +static int ocfs2_initxattrs(struct inode *inode, const struct xattr *xattr_array, void *fs_info) { const struct xattr *xattr; diff --git a/fs/omfs/bitmap.c b/fs/omfs/bitmap.c index 0822345..83f4e76 100644 --- a/fs/omfs/bitmap.c +++ b/fs/omfs/bitmap.c @@ -159,7 +159,7 @@ int omfs_allocate_range(struct super_block *sb, goto out; found: - *return_block = i * bits_per_entry + bit; + *return_block = (u64) i * bits_per_entry + bit; *return_size = run; ret = set_run(sb, i, bits_per_entry, bit, run, 1); diff --git a/fs/omfs/inode.c b/fs/omfs/inode.c index 138321b..3d935c8 100644 --- a/fs/omfs/inode.c +++ b/fs/omfs/inode.c @@ -306,7 +306,8 @@ static const struct super_operations omfs_sops = { */ static int omfs_get_imap(struct super_block *sb) { - unsigned int bitmap_size, count, array_size; + unsigned int bitmap_size, array_size; + int count; struct omfs_sb_info *sbi = OMFS_SB(sb); struct buffer_head *bh; unsigned long **ptr; @@ -359,7 +360,7 @@ nomem: } enum { - Opt_uid, Opt_gid, Opt_umask, Opt_dmask, Opt_fmask + Opt_uid, Opt_gid, Opt_umask, Opt_dmask, Opt_fmask, Opt_err }; static const match_table_t tokens = { @@ -368,6 +369,7 @@ static const match_table_t tokens = { {Opt_umask, "umask=%o"}, {Opt_dmask, "dmask=%o"}, {Opt_fmask, "fmask=%o"}, + {Opt_err, NULL}, }; static int parse_options(char *options, struct omfs_sb_info *sbi) @@ -548,8 +550,10 @@ static int omfs_fill_super(struct super_block *sb, void *data, int silent) } sb->s_root = d_make_root(root); - if (!sb->s_root) + if (!sb->s_root) { + ret = -ENOMEM; goto out_brelse_bh2; + } printk(KERN_DEBUG "omfs: Mounted volume %s\n", omfs_rb->r_name); ret = 0; diff --git a/fs/overlayfs/copy_up.c b/fs/overlayfs/copy_up.c index 24f6404..84d693d 100644 --- a/fs/overlayfs/copy_up.c +++ b/fs/overlayfs/copy_up.c @@ -299,6 +299,9 @@ int ovl_copy_up_one(struct dentry *parent, struct dentry *dentry, struct cred *override_cred; char *link = NULL; + if (WARN_ON(!workdir)) + return -EROFS; + ovl_path_upper(parent, &parentpath); upperdir = parentpath.dentry; diff --git a/fs/overlayfs/dir.c b/fs/overlayfs/dir.c index d139405..692ceda 100644 --- a/fs/overlayfs/dir.c +++ b/fs/overlayfs/dir.c @@ -222,6 +222,9 @@ static struct dentry *ovl_clear_empty(struct dentry *dentry, struct kstat stat; int err; + if (WARN_ON(!workdir)) + return ERR_PTR(-EROFS); + err = ovl_lock_rename_workdir(workdir, upperdir); if (err) goto out; @@ -322,6 +325,9 @@ static int ovl_create_over_whiteout(struct dentry *dentry, struct inode *inode, struct dentry *newdentry; int err; + if (WARN_ON(!workdir)) + return -EROFS; + err = ovl_lock_rename_workdir(workdir, upperdir); if (err) goto out; @@ -506,11 +512,28 @@ static int ovl_remove_and_whiteout(struct dentry *dentry, bool is_dir) struct dentry *opaquedir = NULL; int err; - if (is_dir && OVL_TYPE_MERGE_OR_LOWER(ovl_path_type(dentry))) { - opaquedir = ovl_check_empty_and_clear(dentry); - err = PTR_ERR(opaquedir); - if (IS_ERR(opaquedir)) - goto out; + if (WARN_ON(!workdir)) + return -EROFS; + + if (is_dir) { + if (OVL_TYPE_MERGE_OR_LOWER(ovl_path_type(dentry))) { + opaquedir = ovl_check_empty_and_clear(dentry); + err = PTR_ERR(opaquedir); + if (IS_ERR(opaquedir)) + goto out; + } else { + LIST_HEAD(list); + + /* + * When removing an empty opaque directory, then it + * makes no sense to replace it with an exact replica of + * itself. But emptiness still needs to be checked. + */ + err = ovl_check_empty_dir(dentry, &list); + ovl_cache_free(&list); + if (err) + goto out; + } } err = ovl_lock_rename_workdir(workdir, upperdir); diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c index 907870e..70e9af5 100644 --- a/fs/overlayfs/readdir.c +++ b/fs/overlayfs/readdir.c @@ -23,6 +23,7 @@ struct ovl_cache_entry { u64 ino; struct list_head l_node; struct rb_node node; + struct ovl_cache_entry *next_maybe_whiteout; bool is_whiteout; char name[]; }; @@ -39,7 +40,7 @@ struct ovl_readdir_data { struct rb_root root; struct list_head *list; struct list_head middle; - struct dentry *dir; + struct ovl_cache_entry *first_maybe_whiteout; int count; int err; }; @@ -79,7 +80,7 @@ static struct ovl_cache_entry *ovl_cache_entry_find(struct rb_root *root, return NULL; } -static struct ovl_cache_entry *ovl_cache_entry_new(struct dentry *dir, +static struct ovl_cache_entry *ovl_cache_entry_new(struct ovl_readdir_data *rdd, const char *name, int len, u64 ino, unsigned int d_type) { @@ -98,29 +99,8 @@ static struct ovl_cache_entry *ovl_cache_entry_new(struct dentry *dir, p->is_whiteout = false; if (d_type == DT_CHR) { - struct dentry *dentry; - const struct cred *old_cred; - struct cred *override_cred; - - override_cred = prepare_creds(); - if (!override_cred) { - kfree(p); - return NULL; - } - - /* - * CAP_DAC_OVERRIDE for lookup - */ - cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE); - old_cred = override_creds(override_cred); - - dentry = lookup_one_len(name, dir, len); - if (!IS_ERR(dentry)) { - p->is_whiteout = ovl_is_whiteout(dentry); - dput(dentry); - } - revert_creds(old_cred); - put_cred(override_cred); + p->next_maybe_whiteout = rdd->first_maybe_whiteout; + rdd->first_maybe_whiteout = p; } return p; } @@ -148,7 +128,7 @@ static int ovl_cache_entry_add_rb(struct ovl_readdir_data *rdd, return 0; } - p = ovl_cache_entry_new(rdd->dir, name, len, ino, d_type); + p = ovl_cache_entry_new(rdd, name, len, ino, d_type); if (p == NULL) return -ENOMEM; @@ -169,7 +149,7 @@ static int ovl_fill_lower(struct ovl_readdir_data *rdd, if (p) { list_move_tail(&p->l_node, &rdd->middle); } else { - p = ovl_cache_entry_new(rdd->dir, name, namelen, ino, d_type); + p = ovl_cache_entry_new(rdd, name, namelen, ino, d_type); if (p == NULL) rdd->err = -ENOMEM; else @@ -219,6 +199,43 @@ static int ovl_fill_merge(struct dir_context *ctx, const char *name, return ovl_fill_lower(rdd, name, namelen, offset, ino, d_type); } +static int ovl_check_whiteouts(struct dentry *dir, struct ovl_readdir_data *rdd) +{ + int err; + struct ovl_cache_entry *p; + struct dentry *dentry; + const struct cred *old_cred; + struct cred *override_cred; + + override_cred = prepare_creds(); + if (!override_cred) + return -ENOMEM; + + /* + * CAP_DAC_OVERRIDE for lookup + */ + cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE); + old_cred = override_creds(override_cred); + + err = mutex_lock_killable(&dir->d_inode->i_mutex); + if (!err) { + while (rdd->first_maybe_whiteout) { + p = rdd->first_maybe_whiteout; + rdd->first_maybe_whiteout = p->next_maybe_whiteout; + dentry = lookup_one_len(p->name, dir, p->len); + if (!IS_ERR(dentry)) { + p->is_whiteout = ovl_is_whiteout(dentry); + dput(dentry); + } + } + mutex_unlock(&dir->d_inode->i_mutex); + } + revert_creds(old_cred); + put_cred(override_cred); + + return err; +} + static inline int ovl_dir_read(struct path *realpath, struct ovl_readdir_data *rdd) { @@ -229,7 +246,7 @@ static inline int ovl_dir_read(struct path *realpath, if (IS_ERR(realfile)) return PTR_ERR(realfile); - rdd->dir = realpath->dentry; + rdd->first_maybe_whiteout = NULL; rdd->ctx.pos = 0; do { rdd->count = 0; @@ -238,6 +255,10 @@ static inline int ovl_dir_read(struct path *realpath, if (err >= 0) err = rdd->err; } while (!err && rdd->count); + + if (!err && rdd->first_maybe_whiteout) + err = ovl_check_whiteouts(realpath->dentry, rdd); + fput(realfile); return err; diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index 84c5e27..7466ff3 100644 --- a/fs/overlayfs/super.c +++ b/fs/overlayfs/super.c @@ -273,11 +273,58 @@ static void ovl_dentry_release(struct dentry *dentry) } } +static int ovl_dentry_revalidate(struct dentry *dentry, unsigned int flags) +{ + struct ovl_entry *oe = dentry->d_fsdata; + unsigned int i; + int ret = 1; + + for (i = 0; i < oe->numlower; i++) { + struct dentry *d = oe->lowerstack[i].dentry; + + if (d->d_flags & DCACHE_OP_REVALIDATE) { + ret = d->d_op->d_revalidate(d, flags); + if (ret < 0) + return ret; + if (!ret) { + if (!(flags & LOOKUP_RCU)) + d_invalidate(d); + return -ESTALE; + } + } + } + return 1; +} + +static int ovl_dentry_weak_revalidate(struct dentry *dentry, unsigned int flags) +{ + struct ovl_entry *oe = dentry->d_fsdata; + unsigned int i; + int ret = 1; + + for (i = 0; i < oe->numlower; i++) { + struct dentry *d = oe->lowerstack[i].dentry; + + if (d->d_flags & DCACHE_OP_WEAK_REVALIDATE) { + ret = d->d_op->d_weak_revalidate(d, flags); + if (ret <= 0) + break; + } + } + return ret; +} + static const struct dentry_operations ovl_dentry_operations = { .d_release = ovl_dentry_release, .d_select_inode = ovl_d_select_inode, }; +static const struct dentry_operations ovl_reval_dentry_operations = { + .d_release = ovl_dentry_release, + .d_revalidate = ovl_dentry_revalidate, + .d_weak_revalidate = ovl_dentry_weak_revalidate, +}; + static struct ovl_entry *ovl_alloc_entry(unsigned int numlower) { size_t size = offsetof(struct ovl_entry, lowerstack[numlower]); @@ -289,6 +336,20 @@ static struct ovl_entry *ovl_alloc_entry(unsigned int numlower) return oe; } +static bool ovl_dentry_remote(struct dentry *dentry) +{ + return dentry->d_flags & + (DCACHE_OP_REVALIDATE | DCACHE_OP_WEAK_REVALIDATE); +} + +static bool ovl_dentry_weird(struct dentry *dentry) +{ + return dentry->d_flags & (DCACHE_NEED_AUTOMOUNT | + DCACHE_MANAGE_TRANSIT | + DCACHE_OP_HASH | + DCACHE_OP_COMPARE); +} + static inline struct dentry *ovl_lookup_real(struct dentry *dir, struct qstr *name) { @@ -304,6 +365,10 @@ static inline struct dentry *ovl_lookup_real(struct dentry *dir, } else if (!dentry->d_inode) { dput(dentry); dentry = NULL; + } else if (ovl_dentry_weird(dentry)) { + dput(dentry); + /* Don't support traversing automounts and other weirdness */ + dentry = ERR_PTR(-EREMOTE); } return dentry; } @@ -351,6 +416,11 @@ struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, goto out; if (this) { + if (unlikely(ovl_dentry_remote(this))) { + dput(this); + err = -EREMOTE; + goto out; + } if (ovl_is_whiteout(this)) { dput(this); this = NULL; @@ -530,7 +600,7 @@ static int ovl_remount(struct super_block *sb, int *flags, char *data) { struct ovl_fs *ufs = sb->s_fs_info; - if (!(*flags & MS_RDONLY) && !ufs->upper_mnt) + if (!(*flags & MS_RDONLY) && (!ufs->upper_mnt || !ufs->workdir)) return -EROFS; return 0; @@ -695,25 +765,6 @@ static void ovl_unescape(char *s) } } -static bool ovl_is_allowed_fs_type(struct dentry *root) -{ - const struct dentry_operations *dop = root->d_op; - - /* - * We don't support: - * - automount filesystems - * - filesystems with revalidate (FIXME for lower layer) - * - filesystems with case insensitive names - */ - if (dop && - (dop->d_manage || dop->d_automount || - dop->d_revalidate || dop->d_weak_revalidate || - dop->d_compare || dop->d_hash)) { - return false; - } - return true; -} - static int ovl_mount_dir_noesc(const char *name, struct path *path) { int err = -EINVAL; @@ -728,7 +779,7 @@ static int ovl_mount_dir_noesc(const char *name, struct path *path) goto out; } err = -EINVAL; - if (!ovl_is_allowed_fs_type(path->dentry)) { + if (ovl_dentry_weird(path->dentry)) { pr_err("overlayfs: filesystem on '%s' not supported\n", name); goto out_put; } @@ -752,13 +803,21 @@ static int ovl_mount_dir(const char *name, struct path *path) if (tmp) { ovl_unescape(tmp); err = ovl_mount_dir_noesc(tmp, path); + + if (!err) + if (ovl_dentry_remote(path->dentry)) { + pr_err("overlayfs: filesystem on '%s' not supported as upperdir\n", + tmp); + path_put(path); + err = -EINVAL; + } kfree(tmp); } return err; } static int ovl_lower_dir(const char *name, struct path *path, long *namelen, - int *stack_depth) + int *stack_depth, bool *remote) { int err; struct kstatfs statfs; @@ -775,6 +834,9 @@ static int ovl_lower_dir(const char *name, struct path *path, long *namelen, *namelen = max(*namelen, statfs.f_namelen); *stack_depth = max(*stack_depth, path->mnt->mnt_sb->s_stack_depth); + if (ovl_dentry_remote(path->dentry)) + *remote = true; + return 0; out_put: @@ -828,6 +890,7 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) unsigned int numlower; unsigned int stacklen = 0; unsigned int i; + bool remote = false; int err; err = -ENOMEM; @@ -901,7 +964,8 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) lower = lowertmp; for (numlower = 0; numlower < stacklen; numlower++) { err = ovl_lower_dir(lower, &stack[numlower], - &ufs->lower_namelen, &sb->s_stack_depth); + &ufs->lower_namelen, &sb->s_stack_depth, + &remote); if (err) goto out_put_lowerpath; @@ -926,9 +990,10 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) ufs->workdir = ovl_workdir_create(ufs->upper_mnt, workpath.dentry); err = PTR_ERR(ufs->workdir); if (IS_ERR(ufs->workdir)) { - pr_err("overlayfs: failed to create directory %s/%s\n", - ufs->config.workdir, OVL_WORKDIR_NAME); - goto out_put_upper_mnt; + pr_warn("overlayfs: failed to create directory %s/%s (errno: %i); mounting read-only\n", + ufs->config.workdir, OVL_WORKDIR_NAME, -err); + sb->s_flags |= MS_RDONLY; + ufs->workdir = NULL; } } @@ -958,7 +1023,10 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) if (!ufs->upper_mnt) sb->s_flags |= MS_RDONLY; - sb->s_d_op = &ovl_dentry_operations; + if (remote) + sb->s_d_op = &ovl_reval_dentry_operations; + else + sb->s_d_op = &ovl_dentry_operations; err = -ENOMEM; oe = ovl_alloc_entry(numlower); @@ -998,7 +1066,6 @@ out_put_lower_mnt: kfree(ufs->lower_mnt); out_put_workdir: dput(ufs->workdir); -out_put_upper_mnt: mntput(ufs->upper_mnt); out_put_lowerpath: for (i = 0; i < numlower; i++) diff --git a/fs/proc/Kconfig b/fs/proc/Kconfig index 2183fcf..d751fcb 100644 --- a/fs/proc/Kconfig +++ b/fs/proc/Kconfig @@ -71,3 +71,7 @@ config PROC_PAGE_MONITOR /proc/pid/smaps, /proc/pid/clear_refs, /proc/pid/pagemap, /proc/kpagecount, and /proc/kpageflags. Disabling these interfaces will reduce the size of the kernel by approximately 4kb. + +config PROC_CHILDREN + bool "Include /proc/<pid>/task/<tid>/children file" + default n diff --git a/fs/proc/array.c b/fs/proc/array.c index fd02a9e..ce065cf 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -126,6 +126,14 @@ static inline const char *get_task_state(struct task_struct *tsk) { unsigned int state = (tsk->state | tsk->exit_state) & TASK_REPORT; + /* + * Parked tasks do not run; they sit in __kthread_parkme(). + * Without this check, we would report them as running, which is + * clearly wrong, so we report them as sleeping instead. + */ + if (tsk->state == TASK_PARKED) + state = TASK_INTERRUPTIBLE; + BUILD_BUG_ON(1 + ilog2(TASK_REPORT) != ARRAY_SIZE(task_state_array)-1); return task_state_array[fls(state)]; @@ -569,7 +577,7 @@ int proc_pid_statm(struct seq_file *m, struct pid_namespace *ns, return 0; } -#ifdef CONFIG_CHECKPOINT_RESTORE +#ifdef CONFIG_PROC_CHILDREN static struct pid * get_children_pid(struct inode *inode, struct pid *pid_prev, loff_t pos) { @@ -692,4 +700,4 @@ const struct file_operations proc_tid_children_operations = { .llseek = seq_lseek, .release = children_seq_release, }; -#endif /* CONFIG_CHECKPOINT_RESTORE */ +#endif /* CONFIG_PROC_CHILDREN */ diff --git a/fs/proc/base.c b/fs/proc/base.c index 286a422..87782e8 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -196,18 +196,205 @@ static int proc_root_link(struct dentry *dentry, struct path *path) return result; } -static int proc_pid_cmdline(struct seq_file *m, struct pid_namespace *ns, - struct pid *pid, struct task_struct *task) +static ssize_t proc_pid_cmdline_read(struct file *file, char __user *buf, + size_t _count, loff_t *pos) { + struct task_struct *tsk; + struct mm_struct *mm; + char *page; + unsigned long count = _count; + unsigned long arg_start, arg_end, env_start, env_end; + unsigned long len1, len2, len; + unsigned long p; + char c; + ssize_t rv; + + BUG_ON(*pos < 0); + + tsk = get_proc_task(file_inode(file)); + if (!tsk) + return -ESRCH; + mm = get_task_mm(tsk); + put_task_struct(tsk); + if (!mm) + return 0; + /* Check if process spawned far enough to have cmdline. */ + if (!mm->env_end) { + rv = 0; + goto out_mmput; + } + + page = (char *)__get_free_page(GFP_TEMPORARY); + if (!page) { + rv = -ENOMEM; + goto out_mmput; + } + + down_read(&mm->mmap_sem); + arg_start = mm->arg_start; + arg_end = mm->arg_end; + env_start = mm->env_start; + env_end = mm->env_end; + up_read(&mm->mmap_sem); + + BUG_ON(arg_start > arg_end); + BUG_ON(env_start > env_end); + + len1 = arg_end - arg_start; + len2 = env_end - env_start; + /* - * Rely on struct seq_operations::show() being called once - * per internal buffer allocation. See single_open(), traverse(). + * Inherently racy -- command line shares address space + * with code and data. */ - BUG_ON(m->size < PAGE_SIZE); - m->count += get_cmdline(task, m->buf, PAGE_SIZE); - return 0; + rv = access_remote_vm(mm, arg_end - 1, &c, 1, 0); + if (rv <= 0) + goto out_free_page; + + rv = 0; + + if (c == '\0') { + /* Command line (set of strings) occupies whole ARGV. */ + if (len1 <= *pos) + goto out_free_page; + + p = arg_start + *pos; + len = len1 - *pos; + while (count > 0 && len > 0) { + unsigned int _count; + int nr_read; + + _count = min3(count, len, PAGE_SIZE); + nr_read = access_remote_vm(mm, p, page, _count, 0); + if (nr_read < 0) + rv = nr_read; + if (nr_read <= 0) + goto out_free_page; + + if (copy_to_user(buf, page, nr_read)) { + rv = -EFAULT; + goto out_free_page; + } + + p += nr_read; + len -= nr_read; + buf += nr_read; + count -= nr_read; + rv += nr_read; + } + } else { + /* + * Command line (1 string) occupies ARGV and maybe + * extends into ENVP. + */ + if (len1 + len2 <= *pos) + goto skip_argv_envp; + if (len1 <= *pos) + goto skip_argv; + + p = arg_start + *pos; + len = len1 - *pos; + while (count > 0 && len > 0) { + unsigned int _count, l; + int nr_read; + bool final; + + _count = min3(count, len, PAGE_SIZE); + nr_read = access_remote_vm(mm, p, page, _count, 0); + if (nr_read < 0) + rv = nr_read; + if (nr_read <= 0) + goto out_free_page; + + /* + * Command line can be shorter than whole ARGV + * even if last "marker" byte says it is not. + */ + final = false; + l = strnlen(page, nr_read); + if (l < nr_read) { + nr_read = l; + final = true; + } + + if (copy_to_user(buf, page, nr_read)) { + rv = -EFAULT; + goto out_free_page; + } + + p += nr_read; + len -= nr_read; + buf += nr_read; + count -= nr_read; + rv += nr_read; + + if (final) + goto out_free_page; + } +skip_argv: + /* + * Command line (1 string) occupies ARGV and + * extends into ENVP. + */ + if (len1 <= *pos) { + p = env_start + *pos - len1; + len = len1 + len2 - *pos; + } else { + p = env_start; + len = len2; + } + while (count > 0 && len > 0) { + unsigned int _count, l; + int nr_read; + bool final; + + _count = min3(count, len, PAGE_SIZE); + nr_read = access_remote_vm(mm, p, page, _count, 0); + if (nr_read < 0) + rv = nr_read; + if (nr_read <= 0) + goto out_free_page; + + /* Find EOS. */ + final = false; + l = strnlen(page, nr_read); + if (l < nr_read) { + nr_read = l; + final = true; + } + + if (copy_to_user(buf, page, nr_read)) { + rv = -EFAULT; + goto out_free_page; + } + + p += nr_read; + len -= nr_read; + buf += nr_read; + count -= nr_read; + rv += nr_read; + + if (final) + goto out_free_page; + } +skip_argv_envp: + ; + } + +out_free_page: + free_page((unsigned long)page); +out_mmput: + mmput(mm); + if (rv > 0) + *pos += rv; + return rv; } +static const struct file_operations proc_pid_cmdline_ops = { + .read = proc_pid_cmdline_read, + .llseek = generic_file_llseek, +}; + static int proc_pid_auxv(struct seq_file *m, struct pid_namespace *ns, struct pid *pid, struct task_struct *task) { @@ -304,14 +491,17 @@ static int proc_pid_stack(struct seq_file *m, struct pid_namespace *ns, } #endif -#ifdef CONFIG_SCHEDSTATS +#ifdef CONFIG_SCHED_INFO /* * Provides /proc/PID/schedstat */ static int proc_pid_schedstat(struct seq_file *m, struct pid_namespace *ns, struct pid *pid, struct task_struct *task) { - seq_printf(m, "%llu %llu %lu\n", + if (unlikely(!sched_info_on())) + seq_printf(m, "0 0 0\n"); + else + seq_printf(m, "%llu %llu %lu\n", (unsigned long long)task->se.sum_exec_runtime, (unsigned long long)task->sched_info.run_delay, task->sched_info.pcount); @@ -2572,7 +2762,7 @@ static const struct pid_entry tgid_base_stuff[] = { #ifdef CONFIG_HAVE_ARCH_TRACEHOOK ONE("syscall", S_IRUSR, proc_pid_syscall), #endif - ONE("cmdline", S_IRUGO, proc_pid_cmdline), + REG("cmdline", S_IRUGO, proc_pid_cmdline_ops), ONE("stat", S_IRUGO, proc_tgid_stat), ONE("statm", S_IRUGO, proc_pid_statm), REG("maps", S_IRUGO, proc_pid_maps_operations), @@ -2600,7 +2790,7 @@ static const struct pid_entry tgid_base_stuff[] = { #ifdef CONFIG_STACKTRACE ONE("stack", S_IRUSR, proc_pid_stack), #endif -#ifdef CONFIG_SCHEDSTATS +#ifdef CONFIG_SCHED_INFO ONE("schedstat", S_IRUGO, proc_pid_schedstat), #endif #ifdef CONFIG_LATENCYTOP @@ -2918,11 +3108,11 @@ static const struct pid_entry tid_base_stuff[] = { #ifdef CONFIG_HAVE_ARCH_TRACEHOOK ONE("syscall", S_IRUSR, proc_pid_syscall), #endif - ONE("cmdline", S_IRUGO, proc_pid_cmdline), + REG("cmdline", S_IRUGO, proc_pid_cmdline_ops), ONE("stat", S_IRUGO, proc_tid_stat), ONE("statm", S_IRUGO, proc_pid_statm), REG("maps", S_IRUGO, proc_tid_maps_operations), -#ifdef CONFIG_CHECKPOINT_RESTORE +#ifdef CONFIG_PROC_CHILDREN REG("children", S_IRUGO, proc_tid_children_operations), #endif #ifdef CONFIG_NUMA @@ -2948,7 +3138,7 @@ static const struct pid_entry tid_base_stuff[] = { #ifdef CONFIG_STACKTRACE ONE("stack", S_IRUSR, proc_pid_stack), #endif -#ifdef CONFIG_SCHEDSTATS +#ifdef CONFIG_SCHED_INFO ONE("schedstat", S_IRUGO, proc_pid_schedstat), #endif #ifdef CONFIG_LATENCYTOP diff --git a/fs/proc/generic.c b/fs/proc/generic.c index df6327a..e5dee5c 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -373,6 +373,10 @@ static struct proc_dir_entry *__proc_create(struct proc_dir_entry **parent, WARN(1, "create '/proc/%s' by hand\n", qstr.name); return NULL; } + if (is_empty_pde(*parent)) { + WARN(1, "attempt to add to permanently empty directory"); + return NULL; + } ent = kzalloc(sizeof(struct proc_dir_entry) + qstr.len + 1, GFP_KERNEL); if (!ent) @@ -455,6 +459,25 @@ struct proc_dir_entry *proc_mkdir(const char *name, } EXPORT_SYMBOL(proc_mkdir); +struct proc_dir_entry *proc_create_mount_point(const char *name) +{ + umode_t mode = S_IFDIR | S_IRUGO | S_IXUGO; + struct proc_dir_entry *ent, *parent = NULL; + + ent = __proc_create(&parent, name, mode, 2); + if (ent) { + ent->data = NULL; + ent->proc_fops = NULL; + ent->proc_iops = NULL; + if (proc_register(parent, ent) < 0) { + kfree(ent); + parent->nlink--; + ent = NULL; + } + } + return ent; +} + struct proc_dir_entry *proc_create_data(const char *name, umode_t mode, struct proc_dir_entry *parent, const struct file_operations *proc_fops, diff --git a/fs/proc/inode.c b/fs/proc/inode.c index afe232b..bd95b9f 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -422,6 +422,10 @@ struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; PROC_I(inode)->pde = de; + if (is_empty_pde(de)) { + make_empty_dir_inode(inode); + return inode; + } if (de->mode) { inode->i_mode = de->mode; inode->i_uid = de->uid; diff --git a/fs/proc/internal.h b/fs/proc/internal.h index c835b94..aa27810 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -191,6 +191,12 @@ static inline struct proc_dir_entry *pde_get(struct proc_dir_entry *pde) } extern void pde_put(struct proc_dir_entry *); +static inline bool is_empty_pde(const struct proc_dir_entry *pde) +{ + return S_ISDIR(pde->mode) && !pde->proc_iops; +} +struct proc_dir_entry *proc_create_mount_point(const char *name); + /* * inode.c */ diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index fea2561..fdda62e 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -19,6 +19,28 @@ static const struct inode_operations proc_sys_inode_operations; static const struct file_operations proc_sys_dir_file_operations; static const struct inode_operations proc_sys_dir_operations; +/* Support for permanently empty directories */ + +struct ctl_table sysctl_mount_point[] = { + { } +}; + +static bool is_empty_dir(struct ctl_table_header *head) +{ + return head->ctl_table[0].child == sysctl_mount_point; +} + +static void set_empty_dir(struct ctl_dir *dir) +{ + dir->header.ctl_table[0].child = sysctl_mount_point; +} + +static void clear_empty_dir(struct ctl_dir *dir) + +{ + dir->header.ctl_table[0].child = NULL; +} + void proc_sys_poll_notify(struct ctl_table_poll *poll) { if (!poll) @@ -187,6 +209,17 @@ static int insert_header(struct ctl_dir *dir, struct ctl_table_header *header) struct ctl_table *entry; int err; + /* Is this a permanently empty directory? */ + if (is_empty_dir(&dir->header)) + return -EROFS; + + /* Am I creating a permanently empty directory? */ + if (header->ctl_table == sysctl_mount_point) { + if (!RB_EMPTY_ROOT(&dir->root)) + return -EINVAL; + set_empty_dir(dir); + } + dir->header.nreg++; header->parent = dir; err = insert_links(header); @@ -202,6 +235,8 @@ fail: erase_header(header); put_links(header); fail_links: + if (header->ctl_table == sysctl_mount_point) + clear_empty_dir(dir); header->parent = NULL; drop_sysctl_table(&dir->header); return err; @@ -419,6 +454,8 @@ static struct inode *proc_sys_make_inode(struct super_block *sb, inode->i_mode |= S_IFDIR; inode->i_op = &proc_sys_dir_operations; inode->i_fop = &proc_sys_dir_file_operations; + if (is_empty_dir(head)) + make_empty_dir_inode(inode); } out: return inode; diff --git a/fs/proc/root.c b/fs/proc/root.c index b7fa4bf..68feb0f 100644 --- a/fs/proc/root.c +++ b/fs/proc/root.c @@ -112,9 +112,6 @@ static struct dentry *proc_mount(struct file_system_type *fs_type, ns = task_active_pid_ns(current); options = data; - if (!capable(CAP_SYS_ADMIN) && !fs_fully_visible(fs_type)) - return ERR_PTR(-EPERM); - /* Does the mounter have privilege over the pid namespace? */ if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN)) return ERR_PTR(-EPERM); @@ -159,7 +156,7 @@ static struct file_system_type proc_fs_type = { .name = "proc", .mount = proc_mount, .kill_sb = proc_kill_sb, - .fs_flags = FS_USERNS_MOUNT, + .fs_flags = FS_USERNS_VISIBLE | FS_USERNS_MOUNT, }; void __init proc_root_init(void) @@ -182,10 +179,10 @@ void __init proc_root_init(void) #endif proc_mkdir("fs", NULL); proc_mkdir("driver", NULL); - proc_mkdir("fs/nfsd", NULL); /* somewhere for the nfsd filesystem to be mounted */ + proc_create_mount_point("fs/nfsd"); /* somewhere for the nfsd filesystem to be mounted */ #if defined(CONFIG_SUN_OPENPROMFS) || defined(CONFIG_SUN_OPENPROMFS_MODULE) /* just give it a mountpoint */ - proc_mkdir("openprom", NULL); + proc_create_mount_point("openprom"); #endif proc_tty_init(); proc_mkdir("bus", NULL); diff --git a/fs/proc_namespace.c b/fs/proc_namespace.c index 8db932d..8ebd9a3 100644 --- a/fs/proc_namespace.c +++ b/fs/proc_namespace.c @@ -17,7 +17,8 @@ static unsigned mounts_poll(struct file *file, poll_table *wait) { - struct proc_mounts *p = proc_mounts(file->private_data); + struct seq_file *m = file->private_data; + struct proc_mounts *p = m->private; struct mnt_namespace *ns = p->ns; unsigned res = POLLIN | POLLRDNORM; int event; @@ -25,8 +26,8 @@ static unsigned mounts_poll(struct file *file, poll_table *wait) poll_wait(file, &p->ns->poll, wait); event = ACCESS_ONCE(ns->event); - if (p->m.poll_event != event) { - p->m.poll_event = event; + if (m->poll_event != event) { + m->poll_event = event; res |= POLLERR | POLLPRI; } @@ -92,7 +93,7 @@ static void show_type(struct seq_file *m, struct super_block *sb) static int show_vfsmnt(struct seq_file *m, struct vfsmount *mnt) { - struct proc_mounts *p = proc_mounts(m); + struct proc_mounts *p = m->private; struct mount *r = real_mount(mnt); int err = 0; struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt }; @@ -126,7 +127,7 @@ out: static int show_mountinfo(struct seq_file *m, struct vfsmount *mnt) { - struct proc_mounts *p = proc_mounts(m); + struct proc_mounts *p = m->private; struct mount *r = real_mount(mnt); struct super_block *sb = mnt->mnt_sb; struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt }; @@ -186,7 +187,7 @@ out: static int show_vfsstat(struct seq_file *m, struct vfsmount *mnt) { - struct proc_mounts *p = proc_mounts(m); + struct proc_mounts *p = m->private; struct mount *r = real_mount(mnt); struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt }; struct super_block *sb = mnt_path.dentry->d_sb; @@ -236,6 +237,7 @@ static int mounts_open_common(struct inode *inode, struct file *file, struct mnt_namespace *ns = NULL; struct path root; struct proc_mounts *p; + struct seq_file *m; int ret = -EINVAL; if (!task) @@ -260,26 +262,21 @@ static int mounts_open_common(struct inode *inode, struct file *file, task_unlock(task); put_task_struct(task); - ret = -ENOMEM; - p = kmalloc(sizeof(struct proc_mounts), GFP_KERNEL); - if (!p) + ret = seq_open_private(file, &mounts_op, sizeof(struct proc_mounts)); + if (ret) goto err_put_path; - file->private_data = &p->m; - ret = seq_open(file, &mounts_op); - if (ret) - goto err_free; + m = file->private_data; + m->poll_event = ns->event; + p = m->private; p->ns = ns; p->root = root; - p->m.poll_event = ns->event; p->show = show; p->cached_event = ~0ULL; return 0; - err_free: - kfree(p); err_put_path: path_put(&root); err_put_ns: @@ -290,10 +287,11 @@ static int mounts_open_common(struct inode *inode, struct file *file, static int mounts_release(struct inode *inode, struct file *file) { - struct proc_mounts *p = proc_mounts(file->private_data); + struct seq_file *m = file->private_data; + struct proc_mounts *p = m->private; path_put(&p->root); put_mnt_ns(p->ns); - return seq_release(inode, file); + return seq_release_private(inode, file); } static int mounts_open(struct inode *inode, struct file *file) diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c index dc43b5f..3adcc46 100644 --- a/fs/pstore/inode.c +++ b/fs/pstore/inode.c @@ -461,22 +461,18 @@ static struct file_system_type pstore_fs_type = { .kill_sb = pstore_kill_sb, }; -static struct kobject *pstore_kobj; - static int __init init_pstore_fs(void) { - int err = 0; + int err; /* Create a convenient mount point for people to access pstore */ - pstore_kobj = kobject_create_and_add("pstore", fs_kobj); - if (!pstore_kobj) { - err = -ENOMEM; + err = sysfs_create_mount_point(fs_kobj, "pstore"); + if (err) goto out; - } err = register_filesystem(&pstore_fs_type); if (err < 0) - kobject_put(pstore_kobj); + sysfs_remove_mount_point(fs_kobj, "pstore"); out: return err; diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c index c4c9a10..791743d 100644 --- a/fs/pstore/platform.c +++ b/fs/pstore/platform.c @@ -299,7 +299,7 @@ static void pstore_dump(struct kmsg_dumper *dumper, bool compressed; size_t total_len; - if (big_oops_buf) { + if (big_oops_buf && is_locked) { dst = big_oops_buf; hsize = sprintf(dst, "%s#%d Part%u\n", why, oopscount, part); @@ -456,6 +456,12 @@ int pstore_register(struct pstore_info *psi) add_timer(&pstore_timer); } + /* + * Update the module parameter backend, so it is visible + * through /sys/module/pstore/parameters/backend + */ + backend = psi->name; + pr_info("Registered %s as persistent store backend\n", psi->name); return 0; diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c index 44a549b..6c26c4d 100644 --- a/fs/pstore/ram.c +++ b/fs/pstore/ram.c @@ -186,12 +186,34 @@ static ssize_t ramoops_pstore_read(u64 *id, enum pstore_type_id *type, ssize_t size; ssize_t ecc_notice_size; struct ramoops_context *cxt = psi->data; - struct persistent_ram_zone *prz; - int header_length; + struct persistent_ram_zone *prz = NULL; + int header_length = 0; + + /* Ramoops headers provide time stamps for PSTORE_TYPE_DMESG, but + * PSTORE_TYPE_CONSOLE and PSTORE_TYPE_FTRACE don't currently have + * valid time stamps, so it is initialized to zero. + */ + time->tv_sec = 0; + time->tv_nsec = 0; + *compressed = false; + + /* Find the next valid persistent_ram_zone for DMESG */ + while (cxt->dump_read_cnt < cxt->max_dump_cnt && !prz) { + prz = ramoops_get_next_prz(cxt->przs, &cxt->dump_read_cnt, + cxt->max_dump_cnt, id, type, + PSTORE_TYPE_DMESG, 1); + if (!prz_ok(prz)) + continue; + header_length = ramoops_read_kmsg_hdr(persistent_ram_old(prz), + time, compressed); + /* Clear and skip this DMESG record if it has no valid header */ + if (!header_length) { + persistent_ram_free_old(prz); + persistent_ram_zap(prz); + prz = NULL; + } + } - prz = ramoops_get_next_prz(cxt->przs, &cxt->dump_read_cnt, - cxt->max_dump_cnt, id, type, - PSTORE_TYPE_DMESG, 1); if (!prz_ok(prz)) prz = ramoops_get_next_prz(&cxt->cprz, &cxt->console_read_cnt, 1, id, type, PSTORE_TYPE_CONSOLE, 0); @@ -204,13 +226,7 @@ static ssize_t ramoops_pstore_read(u64 *id, enum pstore_type_id *type, if (!prz_ok(prz)) return 0; - if (!persistent_ram_old(prz)) - return 0; - - size = persistent_ram_old_size(prz); - header_length = ramoops_read_kmsg_hdr(persistent_ram_old(prz), time, - compressed); - size -= header_length; + size = persistent_ram_old_size(prz) - header_length; /* ECC correction notice */ ecc_notice_size = persistent_ram_ecc_string(prz, NULL, 0); @@ -394,18 +410,16 @@ static int ramoops_init_przs(struct device *dev, struct ramoops_context *cxt, } for (i = 0; i < cxt->max_dump_cnt; i++) { - size_t sz = cxt->record_size; - - cxt->przs[i] = persistent_ram_new(*paddr, sz, 0, + cxt->przs[i] = persistent_ram_new(*paddr, cxt->record_size, 0, &cxt->ecc_info, cxt->memtype); if (IS_ERR(cxt->przs[i])) { err = PTR_ERR(cxt->przs[i]); dev_err(dev, "failed to request mem region (0x%zx@0x%llx): %d\n", - sz, (unsigned long long)*paddr, err); + cxt->record_size, (unsigned long long)*paddr, err); goto fail_prz; } - *paddr += sz; + *paddr += cxt->record_size; } return 0; @@ -608,7 +622,7 @@ static void ramoops_register_dummy(void) dummy_data->mem_size = mem_size; dummy_data->mem_address = mem_address; - dummy_data->mem_type = 0; + dummy_data->mem_type = mem_type; dummy_data->record_size = record_size; dummy_data->console_size = ramoops_console_size; dummy_data->ftrace_size = ramoops_ftrace_size; diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 0111ad0..0e4cf728 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -21,6 +21,7 @@ #include "xattr.h" #include <linux/init.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <linux/buffer_head.h> #include <linux/exportfs.h> #include <linux/quotaops.h> @@ -588,8 +589,7 @@ static struct kmem_cache *reiserfs_inode_cachep; static struct inode *reiserfs_alloc_inode(struct super_block *sb) { struct reiserfs_inode_info *ei; - ei = (struct reiserfs_inode_info *) - kmem_cache_alloc(reiserfs_inode_cachep, GFP_KERNEL); + ei = kmem_cache_alloc(reiserfs_inode_cachep, GFP_KERNEL); if (!ei) return NULL; atomic_set(&ei->openers, 0); diff --git a/fs/select.c b/fs/select.c index f684c75..0155473 100644 --- a/fs/select.c +++ b/fs/select.c @@ -189,7 +189,7 @@ static int __pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key) * doesn't imply write barrier and the users expect write * barrier semantics on wakeup functions. The following * smp_wmb() is equivalent to smp_wmb() in try_to_wake_up() - * and is paired with set_mb() in poll_schedule_timeout. + * and is paired with smp_store_mb() in poll_schedule_timeout. */ smp_wmb(); pwq->triggered = 1; @@ -244,7 +244,7 @@ int poll_schedule_timeout(struct poll_wqueues *pwq, int state, /* * Prepare for the next iteration. * - * The following set_mb() serves two purposes. First, it's + * The following smp_store_mb() serves two purposes. First, it's * the counterpart rmb of the wmb in pollwake() such that data * written before wake up is always visible after wake up. * Second, the full barrier guarantees that triggered clearing @@ -252,7 +252,7 @@ int poll_schedule_timeout(struct poll_wqueues *pwq, int state, * this problem doesn't exist for the first iteration as * add_wait_queue() has full barrier semantics. */ - set_mb(pwq->triggered, 0); + smp_store_mb(pwq->triggered, 0); return rc; } diff --git a/fs/seq_file.c b/fs/seq_file.c index d8a0545..ce9e39f 100644 --- a/fs/seq_file.c +++ b/fs/seq_file.c @@ -48,18 +48,21 @@ static void *seq_buf_alloc(unsigned long size) * ERR_PTR(error). In the end of sequence they return %NULL. ->show() * returns 0 in case of success and negative number in case of error. * Returning SEQ_SKIP means "discard this element and move on". + * Note: seq_open() will allocate a struct seq_file and store its + * pointer in @file->private_data. This pointer should not be modified. */ int seq_open(struct file *file, const struct seq_operations *op) { - struct seq_file *p = file->private_data; + struct seq_file *p; + + WARN_ON(file->private_data); + + p = kzalloc(sizeof(*p), GFP_KERNEL); + if (!p) + return -ENOMEM; + + file->private_data = p; - if (!p) { - p = kmalloc(sizeof(*p), GFP_KERNEL); - if (!p) - return -ENOMEM; - file->private_data = p; - } - memset(p, 0, sizeof(*p)); mutex_init(&p->lock); p->op = op; #ifdef CONFIG_USER_NS @@ -552,6 +555,7 @@ int seq_dentry(struct seq_file *m, struct dentry *dentry, const char *esc) return res; } +EXPORT_SYMBOL(seq_dentry); static void *single_start(struct seq_file *p, loff_t *pos) { diff --git a/fs/splice.c b/fs/splice.c index bfe62ae..5fc1e50 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -261,6 +261,7 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe, return ret; } +EXPORT_SYMBOL_GPL(splice_to_pipe); void spd_release_page(struct splice_pipe_desc *spd, unsigned int i) { @@ -359,7 +360,7 @@ __generic_file_splice_read(struct file *in, loff_t *ppos, break; error = add_to_page_cache_lru(page, mapping, index, - GFP_KERNEL); + GFP_KERNEL & mapping_gfp_mask(mapping)); if (unlikely(error)) { page_cache_release(page); if (error == -EEXIST) diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 0b45ff4..94374e4 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -121,3 +121,37 @@ int sysfs_move_dir_ns(struct kobject *kobj, struct kobject *new_parent_kobj, return kernfs_rename_ns(kn, new_parent, kn->name, new_ns); } + +/** + * sysfs_create_mount_point - create an always empty directory + * @parent_kobj: kobject that will contain this always empty directory + * @name: The name of the always empty directory to add + */ +int sysfs_create_mount_point(struct kobject *parent_kobj, const char *name) +{ + struct kernfs_node *kn, *parent = parent_kobj->sd; + + kn = kernfs_create_empty_dir(parent, name); + if (IS_ERR(kn)) { + if (PTR_ERR(kn) == -EEXIST) + sysfs_warn_dup(parent, name); + return PTR_ERR(kn); + } + + return 0; +} +EXPORT_SYMBOL_GPL(sysfs_create_mount_point); + +/** + * sysfs_remove_mount_point - remove an always empty directory. + * @parent_kobj: kobject that will contain this always empty directory + * @name: The name of the always empty directory to remove + * + */ +void sysfs_remove_mount_point(struct kobject *parent_kobj, const char *name) +{ + struct kernfs_node *parent = parent_kobj->sd; + + kernfs_remove_by_name_ns(parent, name, NULL); +} +EXPORT_SYMBOL_GPL(sysfs_remove_mount_point); diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 7c2867b..6c95628 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -90,7 +90,7 @@ static ssize_t sysfs_kf_bin_read(struct kernfs_open_file *of, char *buf, return 0; if (size) { - if (pos > size) + if (pos >= size) return 0; if (pos + count > size) count = size - pos; diff --git a/fs/sysfs/group.c b/fs/sysfs/group.c index b400c04..39a0199 100644 --- a/fs/sysfs/group.c +++ b/fs/sysfs/group.c @@ -135,7 +135,7 @@ static int internal_create_group(struct kobject *kobj, int update, * This function creates a group for the first time. It will explicitly * warn and error if any of the attribute files being created already exist. * - * Returns 0 on success or error. + * Returns 0 on success or error code on failure. */ int sysfs_create_group(struct kobject *kobj, const struct attribute_group *grp) @@ -155,7 +155,7 @@ EXPORT_SYMBOL_GPL(sysfs_create_group); * It will explicitly warn and error if any of the attribute files being * created already exist. * - * Returns 0 on success or error code from sysfs_create_group on error. + * Returns 0 on success or error code from sysfs_create_group on failure. */ int sysfs_create_groups(struct kobject *kobj, const struct attribute_group **groups) @@ -193,7 +193,7 @@ EXPORT_SYMBOL_GPL(sysfs_create_groups); * The primary use for this function is to call it after making a change * that affects group visibility. * - * Returns 0 on success or error. + * Returns 0 on success or error code on failure. */ int sysfs_update_group(struct kobject *kobj, const struct attribute_group *grp) diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c index 8a49486..1c6ac6f 100644 --- a/fs/sysfs/mount.c +++ b/fs/sysfs/mount.c @@ -31,9 +31,6 @@ static struct dentry *sysfs_mount(struct file_system_type *fs_type, bool new_sb; if (!(flags & MS_KERNMOUNT)) { - if (!capable(CAP_SYS_ADMIN) && !fs_fully_visible(fs_type)) - return ERR_PTR(-EPERM); - if (!kobj_ns_current_may_mount(KOBJ_NS_TYPE_NET)) return ERR_PTR(-EPERM); } @@ -58,7 +55,7 @@ static struct file_system_type sysfs_fs_type = { .name = "sysfs", .mount = sysfs_mount, .kill_sb = sysfs_kill_sb, - .fs_flags = FS_USERNS_MOUNT, + .fs_flags = FS_USERNS_VISIBLE | FS_USERNS_MOUNT, }; int __init sysfs_init(void) diff --git a/fs/tracefs/inode.c b/fs/tracefs/inode.c index 6e8a140..cbc8d5d 100644 --- a/fs/tracefs/inode.c +++ b/fs/tracefs/inode.c @@ -626,14 +626,12 @@ bool tracefs_initialized(void) return tracefs_registered; } -static struct kobject *trace_kobj; - static int __init tracefs_init(void) { int retval; - trace_kobj = kobject_create_and_add("tracing", kernel_kobj); - if (!trace_kobj) + retval = sysfs_create_mount_point(kernel_kobj, "tracing"); + if (retval) return -EINVAL; retval = register_filesystem(&trace_fs_type); diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 20f5dbd..9547a278 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -2246,7 +2246,9 @@ static int __init ubifs_init(void) if (!ubifs_inode_slab) return -ENOMEM; - register_shrinker(&ubifs_shrinker_info); + err = register_shrinker(&ubifs_shrinker_info); + if (err) + goto out_slab; err = ubifs_compressors_init(); if (err) @@ -2270,6 +2272,7 @@ out_compr: ubifs_compressors_exit(); out_shrinker: unregister_shrinker(&ubifs_shrinker_info); +out_slab: kmem_cache_destroy(ubifs_inode_slab); return err; } diff --git a/fs/udf/dir.c b/fs/udf/dir.c index 541a12b..541d9c6 100644 --- a/fs/udf/dir.c +++ b/fs/udf/dir.c @@ -168,7 +168,7 @@ static int udf_readdir(struct file *file, struct dir_context *ctx) } flen = udf_get_filename(sb, nameptr, lfi, fname, UDF_NAME_LEN); - if (!flen) + if (flen < 0) continue; tloc = lelb_to_cpu(cfi.icb.extLocation); diff --git a/fs/udf/file.c b/fs/udf/file.c index 7a95b8f..bddf3d0 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -152,8 +152,6 @@ out: mutex_unlock(&inode->i_mutex); if (retval > 0) { - ssize_t err; - mark_inode_dirty(inode); err = generic_write_sync(file, iocb->ki_pos - retval, retval); if (err < 0) diff --git a/fs/udf/namei.c b/fs/udf/namei.c index 5c03f0d..c97b5a8 100644 --- a/fs/udf/namei.c +++ b/fs/udf/namei.c @@ -138,6 +138,25 @@ int udf_write_fi(struct inode *inode, struct fileIdentDesc *cfi, return 0; } +/** + * udf_find_entry - find entry in given directory. + * + * @dir: directory inode to search in + * @child: qstr of the name + * @fibh: buffer head / inode with file identifier descriptor we found + * @cfi: found file identifier descriptor with given name + * + * This function searches in the directory @dir for a file name @child. When + * found, @fibh points to the buffer head(s) (bh is NULL for in ICB + * directories) containing the file identifier descriptor (FID). In that case + * the function returns pointer to the FID in the buffer or inode - but note + * that FID may be split among two buffers (blocks) so accessing it via that + * pointer isn't easily possible. This pointer can be used only as an iterator + * for other directory manipulation functions. For inspection of the FID @cfi + * can be used - the found FID is copied there. + * + * Returns pointer to FID, NULL when nothing found, or error code. + */ static struct fileIdentDesc *udf_find_entry(struct inode *dir, const struct qstr *child, struct udf_fileident_bh *fibh, @@ -167,8 +186,11 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir, fibh->soffset = fibh->eoffset = f_pos & (sb->s_blocksize - 1); if (dinfo->i_alloc_type != ICBTAG_FLAG_AD_IN_ICB) { if (inode_bmap(dir, f_pos >> sb->s_blocksize_bits, &epos, - &eloc, &elen, &offset) != (EXT_RECORDED_ALLOCATED >> 30)) + &eloc, &elen, &offset) != (EXT_RECORDED_ALLOCATED >> 30)) { + fi = ERR_PTR(-EIO); goto out_err; + } + block = udf_get_lb_pblock(sb, &eloc, offset); if ((++offset << sb->s_blocksize_bits) < elen) { if (dinfo->i_alloc_type == ICBTAG_FLAG_AD_SHORT) @@ -179,19 +201,25 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir, offset = 0; fibh->sbh = fibh->ebh = udf_tread(sb, block); - if (!fibh->sbh) + if (!fibh->sbh) { + fi = ERR_PTR(-EIO); goto out_err; + } } fname = kmalloc(UDF_NAME_LEN, GFP_NOFS); - if (!fname) + if (!fname) { + fi = ERR_PTR(-ENOMEM); goto out_err; + } while (f_pos < size) { fi = udf_fileident_read(dir, &f_pos, fibh, cfi, &epos, &eloc, &elen, &offset); - if (!fi) + if (!fi) { + fi = ERR_PTR(-EIO); goto out_err; + } liu = le16_to_cpu(cfi->lengthOfImpUse); lfi = cfi->lengthFileIdent; @@ -234,12 +262,17 @@ static struct fileIdentDesc *udf_find_entry(struct inode *dir, continue; flen = udf_get_filename(sb, nameptr, lfi, fname, UDF_NAME_LEN); - if (flen && udf_match(flen, fname, child->len, child->name)) + if (flen < 0) { + fi = ERR_PTR(flen); + goto out_err; + } + + if (udf_match(flen, fname, child->len, child->name)) goto out_ok; } -out_err: fi = NULL; +out_err: if (fibh->sbh != fibh->ebh) brelse(fibh->ebh); brelse(fibh->sbh); @@ -256,6 +289,7 @@ static struct dentry *udf_lookup(struct inode *dir, struct dentry *dentry, struct inode *inode = NULL; struct fileIdentDesc cfi; struct udf_fileident_bh fibh; + struct fileIdentDesc *fi; if (dentry->d_name.len > UDF_NAME_LEN - 2) return ERR_PTR(-ENAMETOOLONG); @@ -275,7 +309,11 @@ static struct dentry *udf_lookup(struct inode *dir, struct dentry *dentry, } else #endif /* UDF_RECOVERY */ - if (udf_find_entry(dir, &dentry->d_name, &fibh, &cfi)) { + fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi); + if (IS_ERR(fi)) + return ERR_CAST(fi); + + if (fi) { struct kernel_lb_addr loc; if (fibh.sbh != fibh.ebh) @@ -774,8 +812,11 @@ static int udf_rmdir(struct inode *dir, struct dentry *dentry) retval = -ENOENT; fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi); - if (!fi) + if (IS_ERR_OR_NULL(fi)) { + if (fi) + retval = PTR_ERR(fi); goto out; + } retval = -EIO; tloc = lelb_to_cpu(cfi.icb.extLocation); @@ -817,8 +858,12 @@ static int udf_unlink(struct inode *dir, struct dentry *dentry) retval = -ENOENT; fi = udf_find_entry(dir, &dentry->d_name, &fibh, &cfi); - if (!fi) + + if (IS_ERR_OR_NULL(fi)) { + if (fi) + retval = PTR_ERR(fi); goto out; + } retval = -EIO; tloc = lelb_to_cpu(cfi.icb.extLocation); @@ -1049,24 +1094,30 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry, struct udf_inode_info *old_iinfo = UDF_I(old_inode); ofi = udf_find_entry(old_dir, &old_dentry->d_name, &ofibh, &ocfi); - if (ofi) { - if (ofibh.sbh != ofibh.ebh) - brelse(ofibh.ebh); - brelse(ofibh.sbh); + if (IS_ERR(ofi)) { + retval = PTR_ERR(ofi); + goto end_rename; } + + if (ofibh.sbh != ofibh.ebh) + brelse(ofibh.ebh); + + brelse(ofibh.sbh); tloc = lelb_to_cpu(ocfi.icb.extLocation); if (!ofi || udf_get_lb_pblock(old_dir->i_sb, &tloc, 0) != old_inode->i_ino) goto end_rename; nfi = udf_find_entry(new_dir, &new_dentry->d_name, &nfibh, &ncfi); - if (nfi) { - if (!new_inode) { - if (nfibh.sbh != nfibh.ebh) - brelse(nfibh.ebh); - brelse(nfibh.sbh); - nfi = NULL; - } + if (IS_ERR(nfi)) { + retval = PTR_ERR(nfi); + goto end_rename; + } + if (nfi && !new_inode) { + if (nfibh.sbh != nfibh.ebh) + brelse(nfibh.ebh); + brelse(nfibh.sbh); + nfi = NULL; } if (S_ISDIR(old_inode->i_mode)) { int offset = udf_ext0_offset(old_inode); @@ -1221,7 +1272,7 @@ static struct dentry *udf_nfs_get_inode(struct super_block *sb, u32 block, static struct dentry *udf_fh_to_dentry(struct super_block *sb, struct fid *fid, int fh_len, int fh_type) { - if ((fh_len != 3 && fh_len != 5) || + if (fh_len < 3 || (fh_type != FILEID_UDF_WITH_PARENT && fh_type != FILEID_UDF_WITHOUT_PARENT)) return NULL; @@ -1233,7 +1284,7 @@ static struct dentry *udf_fh_to_dentry(struct super_block *sb, static struct dentry *udf_fh_to_parent(struct super_block *sb, struct fid *fid, int fh_len, int fh_type) { - if (fh_len != 5 || fh_type != FILEID_UDF_WITH_PARENT) + if (fh_len < 5 || fh_type != FILEID_UDF_WITH_PARENT) return NULL; return udf_nfs_get_inode(sb, fid->udf.parent_block, diff --git a/fs/udf/super.c b/fs/udf/super.c index 6299f34..b96f190 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -927,17 +927,23 @@ static int udf_load_pvoldesc(struct super_block *sb, sector_t block) #endif } - if (!udf_build_ustr(instr, pvoldesc->volIdent, 32)) - if (udf_CS0toUTF8(outstr, instr)) { - strncpy(UDF_SB(sb)->s_volume_ident, outstr->u_name, - outstr->u_len > 31 ? 31 : outstr->u_len); - udf_debug("volIdent[] = '%s'\n", - UDF_SB(sb)->s_volume_ident); - } + if (!udf_build_ustr(instr, pvoldesc->volIdent, 32)) { + ret = udf_CS0toUTF8(outstr, instr); + if (ret < 0) + goto out_bh; + + strncpy(UDF_SB(sb)->s_volume_ident, outstr->u_name, + outstr->u_len > 31 ? 31 : outstr->u_len); + udf_debug("volIdent[] = '%s'\n", UDF_SB(sb)->s_volume_ident); + } - if (!udf_build_ustr(instr, pvoldesc->volSetIdent, 128)) - if (udf_CS0toUTF8(outstr, instr)) - udf_debug("volSetIdent[] = '%s'\n", outstr->u_name); + if (!udf_build_ustr(instr, pvoldesc->volSetIdent, 128)) { + ret = udf_CS0toUTF8(outstr, instr); + if (ret < 0) + goto out_bh; + + udf_debug("volSetIdent[] = '%s'\n", outstr->u_name); + } ret = 0; out_bh: diff --git a/fs/udf/symlink.c b/fs/udf/symlink.c index 8dfbc40..862535b 100644 --- a/fs/udf/symlink.c +++ b/fs/udf/symlink.c @@ -82,6 +82,9 @@ static int udf_pc_to_char(struct super_block *sb, unsigned char *from, comp_len = udf_get_filename(sb, pc->componentIdent, pc->lengthComponentIdent, p, tolen); + if (comp_len < 0) + return comp_len; + p += comp_len; tolen -= comp_len; if (tolen == 0) diff --git a/fs/udf/unicode.c b/fs/udf/unicode.c index b84fee3..ab478e6 100644 --- a/fs/udf/unicode.c +++ b/fs/udf/unicode.c @@ -68,21 +68,16 @@ int udf_build_ustr(struct ustr *dest, dstring *ptr, int size) /* * udf_build_ustr_exact */ -static int udf_build_ustr_exact(struct ustr *dest, dstring *ptr, int exactsize) +static void udf_build_ustr_exact(struct ustr *dest, dstring *ptr, int exactsize) { - if ((!dest) || (!ptr) || (!exactsize)) - return -1; - memset(dest, 0, sizeof(struct ustr)); dest->u_cmpID = ptr[0]; dest->u_len = exactsize - 1; memcpy(dest->u_name, ptr + 1, exactsize - 1); - - return 0; } /* - * udf_ocu_to_utf8 + * udf_CS0toUTF8 * * PURPOSE * Convert OSTA Compressed Unicode to the UTF-8 equivalent. @@ -94,7 +89,7 @@ static int udf_build_ustr_exact(struct ustr *dest, dstring *ptr, int exactsize) * both of type "struct ustr *" * * POST-CONDITIONS - * <return> Zero on success. + * <return> >= 0 on success. * * HISTORY * November 12, 1997 - Andrew E. Mileski @@ -117,7 +112,7 @@ int udf_CS0toUTF8(struct ustr *utf_o, const struct ustr *ocu_i) memset(utf_o, 0, sizeof(struct ustr)); pr_err("unknown compression code (%d) stri=%s\n", cmp_id, ocu_i->u_name); - return 0; + return -EINVAL; } ocu = ocu_i->u_name; @@ -154,7 +149,7 @@ int udf_CS0toUTF8(struct ustr *utf_o, const struct ustr *ocu_i) /* * - * udf_utf8_to_ocu + * udf_UTF8toCS0 * * PURPOSE * Convert UTF-8 to the OSTA Compressed Unicode equivalent. @@ -270,7 +265,7 @@ static int udf_CS0toNLS(struct nls_table *nls, struct ustr *utf_o, memset(utf_o, 0, sizeof(struct ustr)); pr_err("unknown compression code (%d) stri=%s\n", cmp_id, ocu_i->u_name); - return 0; + return -EINVAL; } ocu = ocu_i->u_name; @@ -338,43 +333,51 @@ int udf_get_filename(struct super_block *sb, uint8_t *sname, int slen, uint8_t *dname, int dlen) { struct ustr *filename, *unifilename; - int len = 0; + int ret; + + if (!slen) + return -EIO; filename = kmalloc(sizeof(struct ustr), GFP_NOFS); if (!filename) - return 0; + return -ENOMEM; unifilename = kmalloc(sizeof(struct ustr), GFP_NOFS); - if (!unifilename) + if (!unifilename) { + ret = -ENOMEM; goto out1; + } - if (udf_build_ustr_exact(unifilename, sname, slen)) - goto out2; - + udf_build_ustr_exact(unifilename, sname, slen); if (UDF_QUERY_FLAG(sb, UDF_FLAG_UTF8)) { - if (!udf_CS0toUTF8(filename, unifilename)) { + ret = udf_CS0toUTF8(filename, unifilename); + if (ret < 0) { udf_debug("Failed in udf_get_filename: sname = %s\n", sname); goto out2; } } else if (UDF_QUERY_FLAG(sb, UDF_FLAG_NLS_MAP)) { - if (!udf_CS0toNLS(UDF_SB(sb)->s_nls_map, filename, - unifilename)) { + ret = udf_CS0toNLS(UDF_SB(sb)->s_nls_map, filename, + unifilename); + if (ret < 0) { udf_debug("Failed in udf_get_filename: sname = %s\n", sname); goto out2; } } else - goto out2; + BUG(); - len = udf_translate_to_linux(dname, dlen, + ret = udf_translate_to_linux(dname, dlen, filename->u_name, filename->u_len, unifilename->u_name, unifilename->u_len); + /* Zero length filename isn't valid... */ + if (ret == 0) + ret = -EINVAL; out2: kfree(unifilename); out1: kfree(filename); - return len; + return ret; } int udf_put_filename(struct super_block *sb, const uint8_t *sname, diff --git a/fs/ufs/super.c b/fs/ufs/super.c index dc33f94..250579a 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -80,6 +80,7 @@ #include <linux/stat.h> #include <linux/string.h> #include <linux/blkdev.h> +#include <linux/backing-dev.h> #include <linux/init.h> #include <linux/parser.h> #include <linux/buffer_head.h> @@ -298,18 +298,18 @@ vfs_removexattr(struct dentry *dentry, const char *name) mutex_lock(&inode->i_mutex); error = security_inode_removexattr(dentry, name); - if (error) { - mutex_unlock(&inode->i_mutex); - return error; - } + if (error) + goto out; error = inode->i_op->removexattr(dentry, name); - mutex_unlock(&inode->i_mutex); if (!error) { fsnotify_xattr(dentry); evm_inode_post_removexattr(dentry, name); } + +out: + mutex_unlock(&inode->i_mutex); return error; } EXPORT_SYMBOL_GPL(vfs_removexattr); diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index 516162b..f9e9ffe 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -149,13 +149,27 @@ xfs_alloc_compute_aligned( { xfs_agblock_t bno; xfs_extlen_t len; + xfs_extlen_t diff; /* Trim busy sections out of found extent */ xfs_extent_busy_trim(args, foundbno, foundlen, &bno, &len); + /* + * If we have a largish extent that happens to start before min_agbno, + * see if we can shift it into range... + */ + if (bno < args->min_agbno && bno + len > args->min_agbno) { + diff = args->min_agbno - bno; + if (len > diff) { + bno += diff; + len -= diff; + } + } + if (args->alignment > 1 && len >= args->minlen) { xfs_agblock_t aligned_bno = roundup(bno, args->alignment); - xfs_extlen_t diff = aligned_bno - bno; + + diff = aligned_bno - bno; *resbno = aligned_bno; *reslen = diff >= len ? 0 : len - diff; @@ -795,9 +809,13 @@ xfs_alloc_find_best_extent( * The good extent is closer than this one. */ if (!dir) { + if (*sbnoa > args->max_agbno) + goto out_use_good; if (*sbnoa >= args->agbno + gdiff) goto out_use_good; } else { + if (*sbnoa < args->min_agbno) + goto out_use_good; if (*sbnoa <= args->agbno - gdiff) goto out_use_good; } @@ -884,6 +902,17 @@ xfs_alloc_ag_vextent_near( dofirst = prandom_u32() & 1; #endif + /* handle unitialized agbno range so caller doesn't have to */ + if (!args->min_agbno && !args->max_agbno) + args->max_agbno = args->mp->m_sb.sb_agblocks - 1; + ASSERT(args->min_agbno <= args->max_agbno); + + /* clamp agbno to the range if it's outside */ + if (args->agbno < args->min_agbno) + args->agbno = args->min_agbno; + if (args->agbno > args->max_agbno) + args->agbno = args->max_agbno; + restart: bno_cur_lt = NULL; bno_cur_gt = NULL; @@ -976,6 +1005,8 @@ restart: <bnoa, <lena); if (ltlena < args->minlen) continue; + if (ltbnoa < args->min_agbno || ltbnoa > args->max_agbno) + continue; args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen); xfs_alloc_fix_len(args); ASSERT(args->len >= args->minlen); @@ -1096,11 +1127,11 @@ restart: XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0); xfs_alloc_compute_aligned(args, ltbno, ltlen, <bnoa, <lena); - if (ltlena >= args->minlen) + if (ltlena >= args->minlen && ltbnoa >= args->min_agbno) break; if ((error = xfs_btree_decrement(bno_cur_lt, 0, &i))) goto error0; - if (!i) { + if (!i || ltbnoa < args->min_agbno) { xfs_btree_del_cursor(bno_cur_lt, XFS_BTREE_NOERROR); bno_cur_lt = NULL; @@ -1112,11 +1143,11 @@ restart: XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0); xfs_alloc_compute_aligned(args, gtbno, gtlen, >bnoa, >lena); - if (gtlena >= args->minlen) + if (gtlena >= args->minlen && gtbnoa <= args->max_agbno) break; if ((error = xfs_btree_increment(bno_cur_gt, 0, &i))) goto error0; - if (!i) { + if (!i || gtbnoa > args->max_agbno) { xfs_btree_del_cursor(bno_cur_gt, XFS_BTREE_NOERROR); bno_cur_gt = NULL; @@ -1216,6 +1247,7 @@ restart: ASSERT(ltnew >= ltbno); ASSERT(ltnew + rlen <= ltbnoa + ltlena); ASSERT(ltnew + rlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length)); + ASSERT(ltnew >= args->min_agbno && ltnew <= args->max_agbno); args->agbno = ltnew; if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur_lt, ltbno, ltlen, @@ -1825,11 +1857,11 @@ xfs_alloc_compute_maxlevels( xfs_extlen_t xfs_alloc_longest_free_extent( struct xfs_mount *mp, - struct xfs_perag *pag) + struct xfs_perag *pag, + xfs_extlen_t need) { - xfs_extlen_t need, delta = 0; + xfs_extlen_t delta = 0; - need = XFS_MIN_FREELIST_PAG(pag, mp); if (need > pag->pagf_flcount) delta = need - pag->pagf_flcount; @@ -1838,131 +1870,150 @@ xfs_alloc_longest_free_extent( return pag->pagf_flcount > 0 || pag->pagf_longest > 0; } +unsigned int +xfs_alloc_min_freelist( + struct xfs_mount *mp, + struct xfs_perag *pag) +{ + unsigned int min_free; + + /* space needed by-bno freespace btree */ + min_free = min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_BNOi] + 1, + mp->m_ag_maxlevels); + /* space needed by-size freespace btree */ + min_free += min_t(unsigned int, pag->pagf_levels[XFS_BTNUM_CNTi] + 1, + mp->m_ag_maxlevels); + + return min_free; +} + +/* + * Check if the operation we are fixing up the freelist for should go ahead or + * not. If we are freeing blocks, we always allow it, otherwise the allocation + * is dependent on whether the size and shape of free space available will + * permit the requested allocation to take place. + */ +static bool +xfs_alloc_space_available( + struct xfs_alloc_arg *args, + xfs_extlen_t min_free, + int flags) +{ + struct xfs_perag *pag = args->pag; + xfs_extlen_t longest; + int available; + + if (flags & XFS_ALLOC_FLAG_FREEING) + return true; + + /* do we have enough contiguous free space for the allocation? */ + longest = xfs_alloc_longest_free_extent(args->mp, pag, min_free); + if ((args->minlen + args->alignment + args->minalignslop - 1) > longest) + return false; + + /* do have enough free space remaining for the allocation? */ + available = (int)(pag->pagf_freeblks + pag->pagf_flcount - + min_free - args->total); + if (available < (int)args->minleft) + return false; + + return true; +} + /* * Decide whether to use this allocation group for this allocation. * If so, fix up the btree freelist's size. */ STATIC int /* error */ xfs_alloc_fix_freelist( - xfs_alloc_arg_t *args, /* allocation argument structure */ - int flags) /* XFS_ALLOC_FLAG_... */ + struct xfs_alloc_arg *args, /* allocation argument structure */ + int flags) /* XFS_ALLOC_FLAG_... */ { - xfs_buf_t *agbp; /* agf buffer pointer */ - xfs_agf_t *agf; /* a.g. freespace structure pointer */ - xfs_buf_t *agflbp;/* agfl buffer pointer */ - xfs_agblock_t bno; /* freelist block */ - xfs_extlen_t delta; /* new blocks needed in freelist */ - int error; /* error result code */ - xfs_extlen_t longest;/* longest extent in allocation group */ - xfs_mount_t *mp; /* file system mount point structure */ - xfs_extlen_t need; /* total blocks needed in freelist */ - xfs_perag_t *pag; /* per-ag information structure */ - xfs_alloc_arg_t targs; /* local allocation arguments */ - xfs_trans_t *tp; /* transaction pointer */ - - mp = args->mp; + struct xfs_mount *mp = args->mp; + struct xfs_perag *pag = args->pag; + struct xfs_trans *tp = args->tp; + struct xfs_buf *agbp = NULL; + struct xfs_buf *agflbp = NULL; + struct xfs_alloc_arg targs; /* local allocation arguments */ + xfs_agblock_t bno; /* freelist block */ + xfs_extlen_t need; /* total blocks needed in freelist */ + int error; - pag = args->pag; - tp = args->tp; if (!pag->pagf_init) { - if ((error = xfs_alloc_read_agf(mp, tp, args->agno, flags, - &agbp))) - return error; + error = xfs_alloc_read_agf(mp, tp, args->agno, flags, &agbp); + if (error) + goto out_no_agbp; if (!pag->pagf_init) { ASSERT(flags & XFS_ALLOC_FLAG_TRYLOCK); ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING)); - args->agbp = NULL; - return 0; + goto out_agbp_relse; } - } else - agbp = NULL; + } /* - * If this is a metadata preferred pag and we are user data - * then try somewhere else if we are not being asked to - * try harder at this point + * If this is a metadata preferred pag and we are user data then try + * somewhere else if we are not being asked to try harder at this + * point */ if (pag->pagf_metadata && args->userdata && (flags & XFS_ALLOC_FLAG_TRYLOCK)) { ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING)); - args->agbp = NULL; - return 0; + goto out_agbp_relse; } - if (!(flags & XFS_ALLOC_FLAG_FREEING)) { - /* - * If it looks like there isn't a long enough extent, or enough - * total blocks, reject it. - */ - need = XFS_MIN_FREELIST_PAG(pag, mp); - longest = xfs_alloc_longest_free_extent(mp, pag); - if ((args->minlen + args->alignment + args->minalignslop - 1) > - longest || - ((int)(pag->pagf_freeblks + pag->pagf_flcount - - need - args->total) < (int)args->minleft)) { - if (agbp) - xfs_trans_brelse(tp, agbp); - args->agbp = NULL; - return 0; - } - } + need = xfs_alloc_min_freelist(mp, pag); + if (!xfs_alloc_space_available(args, need, flags)) + goto out_agbp_relse; /* * Get the a.g. freespace buffer. * Can fail if we're not blocking on locks, and it's held. */ - if (agbp == NULL) { - if ((error = xfs_alloc_read_agf(mp, tp, args->agno, flags, - &agbp))) - return error; - if (agbp == NULL) { + if (!agbp) { + error = xfs_alloc_read_agf(mp, tp, args->agno, flags, &agbp); + if (error) + goto out_no_agbp; + if (!agbp) { ASSERT(flags & XFS_ALLOC_FLAG_TRYLOCK); ASSERT(!(flags & XFS_ALLOC_FLAG_FREEING)); - args->agbp = NULL; - return 0; - } - } - /* - * Figure out how many blocks we should have in the freelist. - */ - agf = XFS_BUF_TO_AGF(agbp); - need = XFS_MIN_FREELIST(agf, mp); - /* - * If there isn't enough total or single-extent, reject it. - */ - if (!(flags & XFS_ALLOC_FLAG_FREEING)) { - delta = need > be32_to_cpu(agf->agf_flcount) ? - (need - be32_to_cpu(agf->agf_flcount)) : 0; - longest = be32_to_cpu(agf->agf_longest); - longest = (longest > delta) ? (longest - delta) : - (be32_to_cpu(agf->agf_flcount) > 0 || longest > 0); - if ((args->minlen + args->alignment + args->minalignslop - 1) > - longest || - ((int)(be32_to_cpu(agf->agf_freeblks) + - be32_to_cpu(agf->agf_flcount) - need - args->total) < - (int)args->minleft)) { - xfs_trans_brelse(tp, agbp); - args->agbp = NULL; - return 0; + goto out_no_agbp; } } + + /* If there isn't enough total space or single-extent, reject it. */ + need = xfs_alloc_min_freelist(mp, pag); + if (!xfs_alloc_space_available(args, need, flags)) + goto out_agbp_relse; + /* * Make the freelist shorter if it's too long. + * + * Note that from this point onwards, we will always release the agf and + * agfl buffers on error. This handles the case where we error out and + * the buffers are clean or may not have been joined to the transaction + * and hence need to be released manually. If they have been joined to + * the transaction, then xfs_trans_brelse() will handle them + * appropriately based on the recursion count and dirty state of the + * buffer. + * + * XXX (dgc): When we have lots of free space, does this buy us + * anything other than extra overhead when we need to put more blocks + * back on the free list? Maybe we should only do this when space is + * getting low or the AGFL is more than half full? */ - while (be32_to_cpu(agf->agf_flcount) > need) { - xfs_buf_t *bp; + while (pag->pagf_flcount > need) { + struct xfs_buf *bp; error = xfs_alloc_get_freelist(tp, agbp, &bno, 0); if (error) - return error; - if ((error = xfs_free_ag_extent(tp, agbp, args->agno, bno, 1, 1))) - return error; + goto out_agbp_relse; + error = xfs_free_ag_extent(tp, agbp, args->agno, bno, 1, 1); + if (error) + goto out_agbp_relse; bp = xfs_btree_get_bufs(mp, tp, args->agno, bno, 0); xfs_trans_binval(tp, bp); } - /* - * Initialize the args structure. - */ + memset(&targs, 0, sizeof(targs)); targs.tp = tp; targs.mp = mp; @@ -1971,21 +2022,20 @@ xfs_alloc_fix_freelist( targs.alignment = targs.minlen = targs.prod = targs.isfl = 1; targs.type = XFS_ALLOCTYPE_THIS_AG; targs.pag = pag; - if ((error = xfs_alloc_read_agfl(mp, tp, targs.agno, &agflbp))) - return error; - /* - * Make the freelist longer if it's too short. - */ - while (be32_to_cpu(agf->agf_flcount) < need) { + error = xfs_alloc_read_agfl(mp, tp, targs.agno, &agflbp); + if (error) + goto out_agbp_relse; + + /* Make the freelist longer if it's too short. */ + while (pag->pagf_flcount < need) { targs.agbno = 0; - targs.maxlen = need - be32_to_cpu(agf->agf_flcount); - /* - * Allocate as many blocks as possible at once. - */ - if ((error = xfs_alloc_ag_vextent(&targs))) { - xfs_trans_brelse(tp, agflbp); - return error; - } + targs.maxlen = need - pag->pagf_flcount; + + /* Allocate as many blocks as possible at once. */ + error = xfs_alloc_ag_vextent(&targs); + if (error) + goto out_agflbp_relse; + /* * Stop if we run out. Won't happen if callers are obeying * the restrictions correctly. Can happen for free calls @@ -1994,9 +2044,7 @@ xfs_alloc_fix_freelist( if (targs.agbno == NULLAGBLOCK) { if (flags & XFS_ALLOC_FLAG_FREEING) break; - xfs_trans_brelse(tp, agflbp); - args->agbp = NULL; - return 0; + goto out_agflbp_relse; } /* * Put each allocated block on the list. @@ -2005,12 +2053,21 @@ xfs_alloc_fix_freelist( error = xfs_alloc_put_freelist(tp, agbp, agflbp, bno, 0); if (error) - return error; + goto out_agflbp_relse; } } xfs_trans_brelse(tp, agflbp); args->agbp = agbp; return 0; + +out_agflbp_relse: + xfs_trans_brelse(tp, agflbp); +out_agbp_relse: + if (agbp) + xfs_trans_brelse(tp, agbp); +out_no_agbp: + args->agbp = NULL; + return error; } /* diff --git a/fs/xfs/libxfs/xfs_alloc.h b/fs/xfs/libxfs/xfs_alloc.h index d1b4b6a..ca1c816 100644 --- a/fs/xfs/libxfs/xfs_alloc.h +++ b/fs/xfs/libxfs/xfs_alloc.h @@ -112,6 +112,8 @@ typedef struct xfs_alloc_arg { xfs_extlen_t total; /* total blocks needed in xaction */ xfs_extlen_t alignment; /* align answer to multiple of this */ xfs_extlen_t minalignslop; /* slop for minlen+alignment calcs */ + xfs_agblock_t min_agbno; /* set an agbno range for NEAR allocs */ + xfs_agblock_t max_agbno; /* ... */ xfs_extlen_t len; /* output: actual size of extent */ xfs_alloctype_t type; /* allocation type XFS_ALLOCTYPE_... */ xfs_alloctype_t otype; /* original allocation type */ @@ -128,11 +130,9 @@ typedef struct xfs_alloc_arg { #define XFS_ALLOC_USERDATA 1 /* allocation is for user data*/ #define XFS_ALLOC_INITIAL_USER_DATA 2 /* special case start of file */ -/* - * Find the length of the longest extent in an AG. - */ -xfs_extlen_t -xfs_alloc_longest_free_extent(struct xfs_mount *mp, +xfs_extlen_t xfs_alloc_longest_free_extent(struct xfs_mount *mp, + struct xfs_perag *pag, xfs_extlen_t need); +unsigned int xfs_alloc_min_freelist(struct xfs_mount *mp, struct xfs_perag *pag); /* diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c index 0a472fb..3349c9a 100644 --- a/fs/xfs/libxfs/xfs_attr.c +++ b/fs/xfs/libxfs/xfs_attr.c @@ -266,7 +266,7 @@ xfs_attr_set( tres.tr_logflags = XFS_TRANS_PERM_LOG_RES; error = xfs_trans_reserve(args.trans, &tres, args.total, 0); if (error) { - xfs_trans_cancel(args.trans, 0); + xfs_trans_cancel(args.trans); return error; } xfs_ilock(dp, XFS_ILOCK_EXCL); @@ -276,7 +276,7 @@ xfs_attr_set( XFS_QMOPT_RES_REGBLKS); if (error) { xfs_iunlock(dp, XFS_ILOCK_EXCL); - xfs_trans_cancel(args.trans, XFS_TRANS_RELEASE_LOG_RES); + xfs_trans_cancel(args.trans); return error; } @@ -320,8 +320,7 @@ xfs_attr_set( xfs_trans_ichgtime(args.trans, dp, XFS_ICHGTIME_CHG); } - err2 = xfs_trans_commit(args.trans, - XFS_TRANS_RELEASE_LOG_RES); + err2 = xfs_trans_commit(args.trans); xfs_iunlock(dp, XFS_ILOCK_EXCL); return error ? error : err2; @@ -383,16 +382,14 @@ xfs_attr_set( * Commit the last in the sequence of transactions. */ xfs_trans_log_inode(args.trans, dp, XFS_ILOG_CORE); - error = xfs_trans_commit(args.trans, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(args.trans); xfs_iunlock(dp, XFS_ILOCK_EXCL); return error; out: - if (args.trans) { - xfs_trans_cancel(args.trans, - XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT); - } + if (args.trans) + xfs_trans_cancel(args.trans); xfs_iunlock(dp, XFS_ILOCK_EXCL); return error; } @@ -462,7 +459,7 @@ xfs_attr_remove( error = xfs_trans_reserve(args.trans, &M_RES(mp)->tr_attrrm, XFS_ATTRRM_SPACE_RES(mp), 0); if (error) { - xfs_trans_cancel(args.trans, 0); + xfs_trans_cancel(args.trans); return error; } @@ -501,16 +498,14 @@ xfs_attr_remove( * Commit the last in the sequence of transactions. */ xfs_trans_log_inode(args.trans, dp, XFS_ILOG_CORE); - error = xfs_trans_commit(args.trans, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(args.trans); xfs_iunlock(dp, XFS_ILOCK_EXCL); return error; out: - if (args.trans) { - xfs_trans_cancel(args.trans, - XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT); - } + if (args.trans) + xfs_trans_cancel(args.trans); xfs_iunlock(dp, XFS_ILOCK_EXCL); return error; } diff --git a/fs/xfs/libxfs/xfs_attr_leaf.c b/fs/xfs/libxfs/xfs_attr_leaf.c index 04e79d5..e9d401c 100644 --- a/fs/xfs/libxfs/xfs_attr_leaf.c +++ b/fs/xfs/libxfs/xfs_attr_leaf.c @@ -574,8 +574,8 @@ xfs_attr_shortform_add(xfs_da_args_t *args, int forkoff) * After the last attribute is removed revert to original inode format, * making all literal area available to the data fork once more. */ -STATIC void -xfs_attr_fork_reset( +void +xfs_attr_fork_remove( struct xfs_inode *ip, struct xfs_trans *tp) { @@ -641,7 +641,7 @@ xfs_attr_shortform_remove(xfs_da_args_t *args) (mp->m_flags & XFS_MOUNT_ATTR2) && (dp->i_d.di_format != XFS_DINODE_FMT_BTREE) && !(args->op_flags & XFS_DA_OP_ADDNAME)) { - xfs_attr_fork_reset(dp, args->trans); + xfs_attr_fork_remove(dp, args->trans); } else { xfs_idata_realloc(dp, -size, XFS_ATTR_FORK); dp->i_d.di_forkoff = xfs_attr_shortform_bytesfit(dp, totsize); @@ -905,7 +905,7 @@ xfs_attr3_leaf_to_shortform( if (forkoff == -1) { ASSERT(dp->i_mount->m_flags & XFS_MOUNT_ATTR2); ASSERT(dp->i_d.di_format != XFS_DINODE_FMT_BTREE); - xfs_attr_fork_reset(dp, args->trans); + xfs_attr_fork_remove(dp, args->trans); goto out; } diff --git a/fs/xfs/libxfs/xfs_attr_leaf.h b/fs/xfs/libxfs/xfs_attr_leaf.h index 025c4b8..882c8d3 100644 --- a/fs/xfs/libxfs/xfs_attr_leaf.h +++ b/fs/xfs/libxfs/xfs_attr_leaf.h @@ -53,7 +53,7 @@ int xfs_attr_shortform_remove(struct xfs_da_args *args); int xfs_attr_shortform_list(struct xfs_attr_list_context *context); int xfs_attr_shortform_allfit(struct xfs_buf *bp, struct xfs_inode *dp); int xfs_attr_shortform_bytesfit(xfs_inode_t *dp, int bytes); - +void xfs_attr_fork_remove(struct xfs_inode *ip, struct xfs_trans *tp); /* * Internal routines when attribute fork size == XFS_LBSIZE(mp). diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index aeffeaa..63e05b6 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -1112,7 +1112,6 @@ xfs_bmap_add_attrfork( int committed; /* xaction was committed */ int logflags; /* logging flags */ int error; /* error return value */ - int cancel_flags = 0; ASSERT(XFS_IFORK_Q(ip) == 0); @@ -1124,17 +1123,15 @@ xfs_bmap_add_attrfork( tp->t_flags |= XFS_TRANS_RESERVE; error = xfs_trans_reserve(tp, &M_RES(mp)->tr_addafork, blks, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } - cancel_flags = XFS_TRANS_RELEASE_LOG_RES; xfs_ilock(ip, XFS_ILOCK_EXCL); error = xfs_trans_reserve_quota_nblks(tp, ip, blks, 0, rsvd ? XFS_QMOPT_RES_REGBLKS | XFS_QMOPT_FORCE_RES : XFS_QMOPT_RES_REGBLKS); if (error) goto trans_cancel; - cancel_flags |= XFS_TRANS_ABORT; if (XFS_IFORK_Q(ip)) goto trans_cancel; if (ip->i_d.di_aformat != XFS_DINODE_FMT_EXTENTS) { @@ -1218,14 +1215,14 @@ xfs_bmap_add_attrfork( error = xfs_bmap_finish(&tp, &flist, &committed); if (error) goto bmap_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; bmap_cancel: xfs_bmap_cancel(&flist); trans_cancel: - xfs_trans_cancel(tp, cancel_flags); + xfs_trans_cancel(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; } @@ -3224,12 +3221,24 @@ xfs_bmap_extsize_align( align_alen += temp; align_off -= temp; } + + /* Same adjustment for the end of the requested area. */ + temp = (align_alen % extsz); + if (temp) + align_alen += extsz - temp; + /* - * Same adjustment for the end of the requested area. + * For large extent hint sizes, the aligned extent might be larger than + * MAXEXTLEN. In that case, reduce the size by an extsz so that it pulls + * the length back under MAXEXTLEN. The outer allocation loops handle + * short allocation just fine, so it is safe to do this. We only want to + * do it when we are forced to, though, because it means more allocation + * operations are required. */ - if ((temp = (align_alen % extsz))) { - align_alen += extsz - temp; - } + while (align_alen > MAXEXTLEN) + align_alen -= extsz; + ASSERT(align_alen <= MAXEXTLEN); + /* * If the previous block overlaps with this proposed allocation * then move the start forward without adjusting the length. @@ -3318,7 +3327,9 @@ xfs_bmap_extsize_align( return -EINVAL; } else { ASSERT(orig_off >= align_off); - ASSERT(orig_end <= align_off + align_alen); + /* see MAXEXTLEN handling above */ + ASSERT(orig_end <= align_off + align_alen || + align_alen + extsz > MAXEXTLEN); } #ifdef DEBUG @@ -3507,7 +3518,8 @@ xfs_bmap_longest_free_extent( } } - longest = xfs_alloc_longest_free_extent(mp, pag); + longest = xfs_alloc_longest_free_extent(mp, pag, + xfs_alloc_min_freelist(mp, pag)); if (*blen < longest) *blen = longest; @@ -4099,13 +4111,6 @@ xfs_bmapi_reserve_delalloc( /* Figure out the extent size, adjust alen */ extsz = xfs_get_extsz_hint(ip); if (extsz) { - /* - * Make sure we don't exceed a single extent length when we - * align the extent by reducing length we are going to - * allocate by the maximum amount extent size aligment may - * require. - */ - alen = XFS_FILBLKS_MIN(len, MAXEXTLEN - (2 * extsz - 1)); error = xfs_bmap_extsize_align(mp, got, prev, extsz, rt, eof, 1, 0, &aoff, &alen); ASSERT(!error); @@ -4417,7 +4422,15 @@ xfs_bmapi_convert_unwritten( error = xfs_bmap_add_extent_unwritten_real(bma->tp, bma->ip, &bma->idx, &bma->cur, mval, bma->firstblock, bma->flist, &tmp_logflags); - bma->logflags |= tmp_logflags; + /* + * Log the inode core unconditionally in the unwritten extent conversion + * path because the conversion might not have done so (e.g., if the + * extent count hasn't changed). We need to make sure the inode is dirty + * in the transaction for the sake of fsync(), even if nothing has + * changed, because fsync() will not force the log for this transaction + * unless it sees the inode pinned. + */ + bma->logflags |= tmp_logflags | XFS_ILOG_CORE; if (error) return error; @@ -5911,7 +5924,7 @@ xfs_bmap_split_extent( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write, XFS_DIOSTRAT_SPACE_RES(mp, 0), 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -5929,10 +5942,9 @@ xfs_bmap_split_extent( if (error) goto out; - return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); - + return xfs_trans_commit(tp); out: - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); return error; } diff --git a/fs/xfs/libxfs/xfs_format.h b/fs/xfs/libxfs/xfs_format.h index 4daaa66..a0ae572 100644 --- a/fs/xfs/libxfs/xfs_format.h +++ b/fs/xfs/libxfs/xfs_format.h @@ -170,7 +170,7 @@ typedef struct xfs_sb { __uint32_t sb_features_log_incompat; __uint32_t sb_crc; /* superblock crc */ - __uint32_t sb_pad; + xfs_extlen_t sb_spino_align; /* sparse inode chunk alignment */ xfs_ino_t sb_pquotino; /* project quota inode */ xfs_lsn_t sb_lsn; /* last write sequence */ @@ -256,7 +256,7 @@ typedef struct xfs_dsb { __be32 sb_features_log_incompat; __le32 sb_crc; /* superblock crc */ - __be32 sb_pad; + __be32 sb_spino_align; /* sparse inode chunk alignment */ __be64 sb_pquotino; /* project quota inode */ __be64 sb_lsn; /* last write sequence */ @@ -457,8 +457,10 @@ xfs_sb_has_ro_compat_feature( } #define XFS_SB_FEAT_INCOMPAT_FTYPE (1 << 0) /* filetype in dirent */ +#define XFS_SB_FEAT_INCOMPAT_SPINODES (1 << 1) /* sparse inode chunks */ #define XFS_SB_FEAT_INCOMPAT_ALL \ - (XFS_SB_FEAT_INCOMPAT_FTYPE) + (XFS_SB_FEAT_INCOMPAT_FTYPE| \ + XFS_SB_FEAT_INCOMPAT_SPINODES) #define XFS_SB_FEAT_INCOMPAT_UNKNOWN ~XFS_SB_FEAT_INCOMPAT_ALL static inline bool @@ -506,6 +508,12 @@ static inline int xfs_sb_version_hasfinobt(xfs_sb_t *sbp) (sbp->sb_features_ro_compat & XFS_SB_FEAT_RO_COMPAT_FINOBT); } +static inline bool xfs_sb_version_hassparseinodes(struct xfs_sb *sbp) +{ + return XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5 && + xfs_sb_has_incompat_feature(sbp, XFS_SB_FEAT_INCOMPAT_SPINODES); +} + /* * end of superblock version macros */ @@ -758,19 +766,6 @@ typedef struct xfs_agfl { #define XFS_AGFL_CRC_OFF offsetof(struct xfs_agfl, agfl_crc) - -#define XFS_AG_MAXLEVELS(mp) ((mp)->m_ag_maxlevels) -#define XFS_MIN_FREELIST_RAW(bl,cl,mp) \ - (MIN(bl + 1, XFS_AG_MAXLEVELS(mp)) + MIN(cl + 1, XFS_AG_MAXLEVELS(mp))) -#define XFS_MIN_FREELIST(a,mp) \ - (XFS_MIN_FREELIST_RAW( \ - be32_to_cpu((a)->agf_levels[XFS_BTNUM_BNOi]), \ - be32_to_cpu((a)->agf_levels[XFS_BTNUM_CNTi]), mp)) -#define XFS_MIN_FREELIST_PAG(pag,mp) \ - (XFS_MIN_FREELIST_RAW( \ - (unsigned int)(pag)->pagf_levels[XFS_BTNUM_BNOi], \ - (unsigned int)(pag)->pagf_levels[XFS_BTNUM_CNTi], mp)) - #define XFS_AGB_TO_FSB(mp,agno,agbno) \ (((xfs_fsblock_t)(agno) << (mp)->m_sb.sb_agblklog) | (agbno)) #define XFS_FSB_TO_AGNO(mp,fsbno) \ @@ -1216,26 +1211,54 @@ typedef __uint64_t xfs_inofree_t; #define XFS_INOBT_ALL_FREE ((xfs_inofree_t)-1) #define XFS_INOBT_MASK(i) ((xfs_inofree_t)1 << (i)) +#define XFS_INOBT_HOLEMASK_FULL 0 /* holemask for full chunk */ +#define XFS_INOBT_HOLEMASK_BITS (NBBY * sizeof(__uint16_t)) +#define XFS_INODES_PER_HOLEMASK_BIT \ + (XFS_INODES_PER_CHUNK / (NBBY * sizeof(__uint16_t))) + static inline xfs_inofree_t xfs_inobt_maskn(int i, int n) { return ((n >= XFS_INODES_PER_CHUNK ? 0 : XFS_INOBT_MASK(n)) - 1) << i; } /* - * Data record structure + * The on-disk inode record structure has two formats. The original "full" + * format uses a 4-byte freecount. The "sparse" format uses a 1-byte freecount + * and replaces the 3 high-order freecount bytes wth the holemask and inode + * count. + * + * The holemask of the sparse record format allows an inode chunk to have holes + * that refer to blocks not owned by the inode record. This facilitates inode + * allocation in the event of severe free space fragmentation. */ typedef struct xfs_inobt_rec { __be32 ir_startino; /* starting inode number */ - __be32 ir_freecount; /* count of free inodes (set bits) */ + union { + struct { + __be32 ir_freecount; /* count of free inodes */ + } f; + struct { + __be16 ir_holemask;/* hole mask for sparse chunks */ + __u8 ir_count; /* total inode count */ + __u8 ir_freecount; /* count of free inodes */ + } sp; + } ir_u; __be64 ir_free; /* free inode mask */ } xfs_inobt_rec_t; typedef struct xfs_inobt_rec_incore { xfs_agino_t ir_startino; /* starting inode number */ - __int32_t ir_freecount; /* count of free inodes (set bits) */ + __uint16_t ir_holemask; /* hole mask for sparse chunks */ + __uint8_t ir_count; /* total inode count */ + __uint8_t ir_freecount; /* count of free inodes (set bits) */ xfs_inofree_t ir_free; /* free inode mask */ } xfs_inobt_rec_incore_t; +static inline bool xfs_inobt_issparse(uint16_t holemask) +{ + /* non-zero holemask represents a sparse rec. */ + return holemask; +} /* * Key structure @@ -1453,8 +1476,8 @@ struct xfs_acl { sizeof(struct xfs_acl_entry) * XFS_ACL_MAX_ENTRIES((mp))) /* On-disk XFS extended attribute names */ -#define SGI_ACL_FILE (unsigned char *)"SGI_ACL_FILE" -#define SGI_ACL_DEFAULT (unsigned char *)"SGI_ACL_DEFAULT" +#define SGI_ACL_FILE "SGI_ACL_FILE" +#define SGI_ACL_DEFAULT "SGI_ACL_DEFAULT" #define SGI_ACL_FILE_SIZE (sizeof(SGI_ACL_FILE)-1) #define SGI_ACL_DEFAULT_SIZE (sizeof(SGI_ACL_DEFAULT)-1) diff --git a/fs/xfs/libxfs/xfs_fs.h b/fs/xfs/libxfs/xfs_fs.h index 18dc721..89689c6 100644 --- a/fs/xfs/libxfs/xfs_fs.h +++ b/fs/xfs/libxfs/xfs_fs.h @@ -239,6 +239,7 @@ typedef struct xfs_fsop_resblks { #define XFS_FSOP_GEOM_FLAGS_V5SB 0x8000 /* version 5 superblock */ #define XFS_FSOP_GEOM_FLAGS_FTYPE 0x10000 /* inode directory types */ #define XFS_FSOP_GEOM_FLAGS_FINOBT 0x20000 /* free inode btree */ +#define XFS_FSOP_GEOM_FLAGS_SPINODES 0x40000 /* sparse inode chunks */ /* * Minimum and maximum sizes need for growth checks. diff --git a/fs/xfs/libxfs/xfs_ialloc.c b/fs/xfs/libxfs/xfs_ialloc.c index 07349a1..66efc70 100644 --- a/fs/xfs/libxfs/xfs_ialloc.c +++ b/fs/xfs/libxfs/xfs_ialloc.c @@ -65,6 +65,8 @@ xfs_inobt_lookup( int *stat) /* success/failure */ { cur->bc_rec.i.ir_startino = ino; + cur->bc_rec.i.ir_holemask = 0; + cur->bc_rec.i.ir_count = 0; cur->bc_rec.i.ir_freecount = 0; cur->bc_rec.i.ir_free = 0; return xfs_btree_lookup(cur, dir, stat); @@ -82,7 +84,14 @@ xfs_inobt_update( union xfs_btree_rec rec; rec.inobt.ir_startino = cpu_to_be32(irec->ir_startino); - rec.inobt.ir_freecount = cpu_to_be32(irec->ir_freecount); + if (xfs_sb_version_hassparseinodes(&cur->bc_mp->m_sb)) { + rec.inobt.ir_u.sp.ir_holemask = cpu_to_be16(irec->ir_holemask); + rec.inobt.ir_u.sp.ir_count = irec->ir_count; + rec.inobt.ir_u.sp.ir_freecount = irec->ir_freecount; + } else { + /* ir_holemask/ir_count not supported on-disk */ + rec.inobt.ir_u.f.ir_freecount = cpu_to_be32(irec->ir_freecount); + } rec.inobt.ir_free = cpu_to_be64(irec->ir_free); return xfs_btree_update(cur, &rec); } @@ -100,12 +109,27 @@ xfs_inobt_get_rec( int error; error = xfs_btree_get_rec(cur, &rec, stat); - if (!error && *stat == 1) { - irec->ir_startino = be32_to_cpu(rec->inobt.ir_startino); - irec->ir_freecount = be32_to_cpu(rec->inobt.ir_freecount); - irec->ir_free = be64_to_cpu(rec->inobt.ir_free); + if (error || *stat == 0) + return error; + + irec->ir_startino = be32_to_cpu(rec->inobt.ir_startino); + if (xfs_sb_version_hassparseinodes(&cur->bc_mp->m_sb)) { + irec->ir_holemask = be16_to_cpu(rec->inobt.ir_u.sp.ir_holemask); + irec->ir_count = rec->inobt.ir_u.sp.ir_count; + irec->ir_freecount = rec->inobt.ir_u.sp.ir_freecount; + } else { + /* + * ir_holemask/ir_count not supported on-disk. Fill in hardcoded + * values for full inode chunks. + */ + irec->ir_holemask = XFS_INOBT_HOLEMASK_FULL; + irec->ir_count = XFS_INODES_PER_CHUNK; + irec->ir_freecount = + be32_to_cpu(rec->inobt.ir_u.f.ir_freecount); } - return error; + irec->ir_free = be64_to_cpu(rec->inobt.ir_free); + + return 0; } /* @@ -114,10 +138,14 @@ xfs_inobt_get_rec( STATIC int xfs_inobt_insert_rec( struct xfs_btree_cur *cur, + __uint16_t holemask, + __uint8_t count, __int32_t freecount, xfs_inofree_t free, int *stat) { + cur->bc_rec.i.ir_holemask = holemask; + cur->bc_rec.i.ir_count = count; cur->bc_rec.i.ir_freecount = freecount; cur->bc_rec.i.ir_free = free; return xfs_btree_insert(cur, stat); @@ -154,7 +182,9 @@ xfs_inobt_insert( } ASSERT(i == 0); - error = xfs_inobt_insert_rec(cur, XFS_INODES_PER_CHUNK, + error = xfs_inobt_insert_rec(cur, XFS_INOBT_HOLEMASK_FULL, + XFS_INODES_PER_CHUNK, + XFS_INODES_PER_CHUNK, XFS_INOBT_ALL_FREE, &i); if (error) { xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); @@ -220,6 +250,7 @@ xfs_ialloc_inode_init( struct xfs_mount *mp, struct xfs_trans *tp, struct list_head *buffer_list, + int icount, xfs_agnumber_t agno, xfs_agblock_t agbno, xfs_agblock_t length, @@ -275,7 +306,7 @@ xfs_ialloc_inode_init( * they track in the AIL as if they were physically logged. */ if (tp) - xfs_icreate_log(tp, agno, agbno, mp->m_ialloc_inos, + xfs_icreate_log(tp, agno, agbno, icount, mp->m_sb.sb_inodesize, length, gen); } else version = 2; @@ -347,6 +378,214 @@ xfs_ialloc_inode_init( } /* + * Align startino and allocmask for a recently allocated sparse chunk such that + * they are fit for insertion (or merge) into the on-disk inode btrees. + * + * Background: + * + * When enabled, sparse inode support increases the inode alignment from cluster + * size to inode chunk size. This means that the minimum range between two + * non-adjacent inode records in the inobt is large enough for a full inode + * record. This allows for cluster sized, cluster aligned block allocation + * without need to worry about whether the resulting inode record overlaps with + * another record in the tree. Without this basic rule, we would have to deal + * with the consequences of overlap by potentially undoing recent allocations in + * the inode allocation codepath. + * + * Because of this alignment rule (which is enforced on mount), there are two + * inobt possibilities for newly allocated sparse chunks. One is that the + * aligned inode record for the chunk covers a range of inodes not already + * covered in the inobt (i.e., it is safe to insert a new sparse record). The + * other is that a record already exists at the aligned startino that considers + * the newly allocated range as sparse. In the latter case, record content is + * merged in hope that sparse inode chunks fill to full chunks over time. + */ +STATIC void +xfs_align_sparse_ino( + struct xfs_mount *mp, + xfs_agino_t *startino, + uint16_t *allocmask) +{ + xfs_agblock_t agbno; + xfs_agblock_t mod; + int offset; + + agbno = XFS_AGINO_TO_AGBNO(mp, *startino); + mod = agbno % mp->m_sb.sb_inoalignmt; + if (!mod) + return; + + /* calculate the inode offset and align startino */ + offset = mod << mp->m_sb.sb_inopblog; + *startino -= offset; + + /* + * Since startino has been aligned down, left shift allocmask such that + * it continues to represent the same physical inodes relative to the + * new startino. + */ + *allocmask <<= offset / XFS_INODES_PER_HOLEMASK_BIT; +} + +/* + * Determine whether the source inode record can merge into the target. Both + * records must be sparse, the inode ranges must match and there must be no + * allocation overlap between the records. + */ +STATIC bool +__xfs_inobt_can_merge( + struct xfs_inobt_rec_incore *trec, /* tgt record */ + struct xfs_inobt_rec_incore *srec) /* src record */ +{ + uint64_t talloc; + uint64_t salloc; + + /* records must cover the same inode range */ + if (trec->ir_startino != srec->ir_startino) + return false; + + /* both records must be sparse */ + if (!xfs_inobt_issparse(trec->ir_holemask) || + !xfs_inobt_issparse(srec->ir_holemask)) + return false; + + /* both records must track some inodes */ + if (!trec->ir_count || !srec->ir_count) + return false; + + /* can't exceed capacity of a full record */ + if (trec->ir_count + srec->ir_count > XFS_INODES_PER_CHUNK) + return false; + + /* verify there is no allocation overlap */ + talloc = xfs_inobt_irec_to_allocmask(trec); + salloc = xfs_inobt_irec_to_allocmask(srec); + if (talloc & salloc) + return false; + + return true; +} + +/* + * Merge the source inode record into the target. The caller must call + * __xfs_inobt_can_merge() to ensure the merge is valid. + */ +STATIC void +__xfs_inobt_rec_merge( + struct xfs_inobt_rec_incore *trec, /* target */ + struct xfs_inobt_rec_incore *srec) /* src */ +{ + ASSERT(trec->ir_startino == srec->ir_startino); + + /* combine the counts */ + trec->ir_count += srec->ir_count; + trec->ir_freecount += srec->ir_freecount; + + /* + * Merge the holemask and free mask. For both fields, 0 bits refer to + * allocated inodes. We combine the allocated ranges with bitwise AND. + */ + trec->ir_holemask &= srec->ir_holemask; + trec->ir_free &= srec->ir_free; +} + +/* + * Insert a new sparse inode chunk into the associated inode btree. The inode + * record for the sparse chunk is pre-aligned to a startino that should match + * any pre-existing sparse inode record in the tree. This allows sparse chunks + * to fill over time. + * + * This function supports two modes of handling preexisting records depending on + * the merge flag. If merge is true, the provided record is merged with the + * existing record and updated in place. The merged record is returned in nrec. + * If merge is false, an existing record is replaced with the provided record. + * If no preexisting record exists, the provided record is always inserted. + * + * It is considered corruption if a merge is requested and not possible. Given + * the sparse inode alignment constraints, this should never happen. + */ +STATIC int +xfs_inobt_insert_sprec( + struct xfs_mount *mp, + struct xfs_trans *tp, + struct xfs_buf *agbp, + int btnum, + struct xfs_inobt_rec_incore *nrec, /* in/out: new/merged rec. */ + bool merge) /* merge or replace */ +{ + struct xfs_btree_cur *cur; + struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); + xfs_agnumber_t agno = be32_to_cpu(agi->agi_seqno); + int error; + int i; + struct xfs_inobt_rec_incore rec; + + cur = xfs_inobt_init_cursor(mp, tp, agbp, agno, btnum); + + /* the new record is pre-aligned so we know where to look */ + error = xfs_inobt_lookup(cur, nrec->ir_startino, XFS_LOOKUP_EQ, &i); + if (error) + goto error; + /* if nothing there, insert a new record and return */ + if (i == 0) { + error = xfs_inobt_insert_rec(cur, nrec->ir_holemask, + nrec->ir_count, nrec->ir_freecount, + nrec->ir_free, &i); + if (error) + goto error; + XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error); + + goto out; + } + + /* + * A record exists at this startino. Merge or replace the record + * depending on what we've been asked to do. + */ + if (merge) { + error = xfs_inobt_get_rec(cur, &rec, &i); + if (error) + goto error; + XFS_WANT_CORRUPTED_GOTO(mp, i == 1, error); + XFS_WANT_CORRUPTED_GOTO(mp, + rec.ir_startino == nrec->ir_startino, + error); + + /* + * This should never fail. If we have coexisting records that + * cannot merge, something is seriously wrong. + */ + XFS_WANT_CORRUPTED_GOTO(mp, __xfs_inobt_can_merge(nrec, &rec), + error); + + trace_xfs_irec_merge_pre(mp, agno, rec.ir_startino, + rec.ir_holemask, nrec->ir_startino, + nrec->ir_holemask); + + /* merge to nrec to output the updated record */ + __xfs_inobt_rec_merge(nrec, &rec); + + trace_xfs_irec_merge_post(mp, agno, nrec->ir_startino, + nrec->ir_holemask); + + error = xfs_inobt_rec_check_count(mp, nrec); + if (error) + goto error; + } + + error = xfs_inobt_update(cur, nrec); + if (error) + goto error; + +out: + xfs_btree_del_cursor(cur, XFS_BTREE_NOERROR); + return 0; +error: + xfs_btree_del_cursor(cur, XFS_BTREE_ERROR); + return error; +} + +/* * Allocate new inodes in the allocation group specified by agbp. * Return 0 for success, else error code. */ @@ -364,11 +603,22 @@ xfs_ialloc_ag_alloc( xfs_agino_t newlen; /* new number of inodes */ int isaligned = 0; /* inode allocation at stripe unit */ /* boundary */ + uint16_t allocmask = (uint16_t) -1; /* init. to full chunk */ + struct xfs_inobt_rec_incore rec; struct xfs_perag *pag; + int do_sparse = 0; memset(&args, 0, sizeof(args)); args.tp = tp; args.mp = tp->t_mountp; + args.fsbno = NULLFSBLOCK; + +#ifdef DEBUG + /* randomly do sparse inode allocations */ + if (xfs_sb_version_hassparseinodes(&tp->t_mountp->m_sb) && + args.mp->m_ialloc_min_blks < args.mp->m_ialloc_blks) + do_sparse = prandom_u32() & 1; +#endif /* * Locking will ensure that we don't have two callers in here @@ -376,7 +626,7 @@ xfs_ialloc_ag_alloc( */ newlen = args.mp->m_ialloc_inos; if (args.mp->m_maxicount && - percpu_counter_read(&args.mp->m_icount) + newlen > + percpu_counter_read_positive(&args.mp->m_icount) + newlen > args.mp->m_maxicount) return -ENOSPC; args.minlen = args.maxlen = args.mp->m_ialloc_blks; @@ -390,6 +640,8 @@ xfs_ialloc_ag_alloc( agno = be32_to_cpu(agi->agi_seqno); args.agbno = XFS_AGINO_TO_AGBNO(args.mp, newino) + args.mp->m_ialloc_blks; + if (do_sparse) + goto sparse_alloc; if (likely(newino != NULLAGINO && (args.agbno < be32_to_cpu(agi->agi_length)))) { args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno); @@ -428,8 +680,7 @@ xfs_ialloc_ag_alloc( * subsequent requests. */ args.minalignslop = 0; - } else - args.fsbno = NULLFSBLOCK; + } if (unlikely(args.fsbno == NULLFSBLOCK)) { /* @@ -480,6 +731,47 @@ xfs_ialloc_ag_alloc( return error; } + /* + * Finally, try a sparse allocation if the filesystem supports it and + * the sparse allocation length is smaller than a full chunk. + */ + if (xfs_sb_version_hassparseinodes(&args.mp->m_sb) && + args.mp->m_ialloc_min_blks < args.mp->m_ialloc_blks && + args.fsbno == NULLFSBLOCK) { +sparse_alloc: + args.type = XFS_ALLOCTYPE_NEAR_BNO; + args.agbno = be32_to_cpu(agi->agi_root); + args.fsbno = XFS_AGB_TO_FSB(args.mp, agno, args.agbno); + args.alignment = args.mp->m_sb.sb_spino_align; + args.prod = 1; + + args.minlen = args.mp->m_ialloc_min_blks; + args.maxlen = args.minlen; + + /* + * The inode record will be aligned to full chunk size. We must + * prevent sparse allocation from AG boundaries that result in + * invalid inode records, such as records that start at agbno 0 + * or extend beyond the AG. + * + * Set min agbno to the first aligned, non-zero agbno and max to + * the last aligned agbno that is at least one full chunk from + * the end of the AG. + */ + args.min_agbno = args.mp->m_sb.sb_inoalignmt; + args.max_agbno = round_down(args.mp->m_sb.sb_agblocks, + args.mp->m_sb.sb_inoalignmt) - + args.mp->m_ialloc_blks; + + error = xfs_alloc_vextent(&args); + if (error) + return error; + + newlen = args.len << args.mp->m_sb.sb_inopblog; + ASSERT(newlen <= XFS_INODES_PER_CHUNK); + allocmask = (1 << (newlen / XFS_INODES_PER_HOLEMASK_BIT)) - 1; + } + if (args.fsbno == NULLFSBLOCK) { *alloc = 0; return 0; @@ -495,8 +787,8 @@ xfs_ialloc_ag_alloc( * rather than a linear progression to prevent the next generation * number from being easily guessable. */ - error = xfs_ialloc_inode_init(args.mp, tp, NULL, agno, args.agbno, - args.len, prandom_u32()); + error = xfs_ialloc_inode_init(args.mp, tp, NULL, newlen, agno, + args.agbno, args.len, prandom_u32()); if (error) return error; @@ -504,6 +796,73 @@ xfs_ialloc_ag_alloc( * Convert the results. */ newino = XFS_OFFBNO_TO_AGINO(args.mp, args.agbno, 0); + + if (xfs_inobt_issparse(~allocmask)) { + /* + * We've allocated a sparse chunk. Align the startino and mask. + */ + xfs_align_sparse_ino(args.mp, &newino, &allocmask); + + rec.ir_startino = newino; + rec.ir_holemask = ~allocmask; + rec.ir_count = newlen; + rec.ir_freecount = newlen; + rec.ir_free = XFS_INOBT_ALL_FREE; + + /* + * Insert the sparse record into the inobt and allow for a merge + * if necessary. If a merge does occur, rec is updated to the + * merged record. + */ + error = xfs_inobt_insert_sprec(args.mp, tp, agbp, XFS_BTNUM_INO, + &rec, true); + if (error == -EFSCORRUPTED) { + xfs_alert(args.mp, + "invalid sparse inode record: ino 0x%llx holemask 0x%x count %u", + XFS_AGINO_TO_INO(args.mp, agno, + rec.ir_startino), + rec.ir_holemask, rec.ir_count); + xfs_force_shutdown(args.mp, SHUTDOWN_CORRUPT_INCORE); + } + if (error) + return error; + + /* + * We can't merge the part we've just allocated as for the inobt + * due to finobt semantics. The original record may or may not + * exist independent of whether physical inodes exist in this + * sparse chunk. + * + * We must update the finobt record based on the inobt record. + * rec contains the fully merged and up to date inobt record + * from the previous call. Set merge false to replace any + * existing record with this one. + */ + if (xfs_sb_version_hasfinobt(&args.mp->m_sb)) { + error = xfs_inobt_insert_sprec(args.mp, tp, agbp, + XFS_BTNUM_FINO, &rec, + false); + if (error) + return error; + } + } else { + /* full chunk - insert new records to both btrees */ + error = xfs_inobt_insert(args.mp, tp, agbp, newino, newlen, + XFS_BTNUM_INO); + if (error) + return error; + + if (xfs_sb_version_hasfinobt(&args.mp->m_sb)) { + error = xfs_inobt_insert(args.mp, tp, agbp, newino, + newlen, XFS_BTNUM_FINO); + if (error) + return error; + } + } + + /* + * Update AGI counts and newino. + */ be32_add_cpu(&agi->agi_count, newlen); be32_add_cpu(&agi->agi_freecount, newlen); pag = xfs_perag_get(args.mp, agno); @@ -512,20 +871,6 @@ xfs_ialloc_ag_alloc( agi->agi_newino = cpu_to_be32(newino); /* - * Insert records describing the new inode chunk into the btrees. - */ - error = xfs_inobt_insert(args.mp, tp, agbp, newino, newlen, - XFS_BTNUM_INO); - if (error) - return error; - - if (xfs_sb_version_hasfinobt(&args.mp->m_sb)) { - error = xfs_inobt_insert(args.mp, tp, agbp, newino, newlen, - XFS_BTNUM_FINO); - if (error) - return error; - } - /* * Log allocation group header fields */ xfs_ialloc_log_agi(tp, agbp, @@ -645,7 +990,7 @@ xfs_ialloc_ag_select( * if we fail allocation due to alignment issues then it is most * likely a real ENOSPC condition. */ - ineed = mp->m_ialloc_blks; + ineed = mp->m_ialloc_min_blks; if (flags && ineed > 1) ineed += xfs_ialloc_cluster_alignment(mp); longest = pag->pagf_longest; @@ -732,6 +1077,27 @@ xfs_ialloc_get_rec( } /* + * Return the offset of the first free inode in the record. If the inode chunk + * is sparsely allocated, we convert the record holemask to inode granularity + * and mask off the unallocated regions from the inode free mask. + */ +STATIC int +xfs_inobt_first_free_inode( + struct xfs_inobt_rec_incore *rec) +{ + xfs_inofree_t realfree; + + /* if there are no holes, return the first available offset */ + if (!xfs_inobt_issparse(rec->ir_holemask)) + return xfs_lowbit64(rec->ir_free); + + realfree = xfs_inobt_irec_to_allocmask(rec); + realfree &= rec->ir_free; + + return xfs_lowbit64(realfree); +} + +/* * Allocate an inode using the inobt-only algorithm. */ STATIC int @@ -961,7 +1327,7 @@ newino: } alloc_inode: - offset = xfs_lowbit64(rec.ir_free); + offset = xfs_inobt_first_free_inode(&rec); ASSERT(offset >= 0); ASSERT(offset < XFS_INODES_PER_CHUNK); ASSERT((XFS_AGINO_TO_OFFSET(mp, rec.ir_startino) % @@ -1210,7 +1576,7 @@ xfs_dialloc_ag( if (error) goto error_cur; - offset = xfs_lowbit64(rec.ir_free); + offset = xfs_inobt_first_free_inode(&rec); ASSERT(offset >= 0); ASSERT(offset < XFS_INODES_PER_CHUNK); ASSERT((XFS_AGINO_TO_OFFSET(mp, rec.ir_startino) % @@ -1339,10 +1705,13 @@ xfs_dialloc( * If we have already hit the ceiling of inode blocks then clear * okalloc so we scan all available agi structures for a free * inode. + * + * Read rough value of mp->m_icount by percpu_counter_read_positive, + * which will sacrifice the preciseness but improve the performance. */ if (mp->m_maxicount && - percpu_counter_read(&mp->m_icount) + mp->m_ialloc_inos > - mp->m_maxicount) { + percpu_counter_read_positive(&mp->m_icount) + mp->m_ialloc_inos + > mp->m_maxicount) { noroom = 1; okalloc = 0; } @@ -1436,6 +1805,83 @@ out_error: return error; } +/* + * Free the blocks of an inode chunk. We must consider that the inode chunk + * might be sparse and only free the regions that are allocated as part of the + * chunk. + */ +STATIC void +xfs_difree_inode_chunk( + struct xfs_mount *mp, + xfs_agnumber_t agno, + struct xfs_inobt_rec_incore *rec, + struct xfs_bmap_free *flist) +{ + xfs_agblock_t sagbno = XFS_AGINO_TO_AGBNO(mp, rec->ir_startino); + int startidx, endidx; + int nextbit; + xfs_agblock_t agbno; + int contigblk; + DECLARE_BITMAP(holemask, XFS_INOBT_HOLEMASK_BITS); + + if (!xfs_inobt_issparse(rec->ir_holemask)) { + /* not sparse, calculate extent info directly */ + xfs_bmap_add_free(XFS_AGB_TO_FSB(mp, agno, + XFS_AGINO_TO_AGBNO(mp, rec->ir_startino)), + mp->m_ialloc_blks, flist, mp); + return; + } + + /* holemask is only 16-bits (fits in an unsigned long) */ + ASSERT(sizeof(rec->ir_holemask) <= sizeof(holemask[0])); + holemask[0] = rec->ir_holemask; + + /* + * Find contiguous ranges of zeroes (i.e., allocated regions) in the + * holemask and convert the start/end index of each range to an extent. + * We start with the start and end index both pointing at the first 0 in + * the mask. + */ + startidx = endidx = find_first_zero_bit(holemask, + XFS_INOBT_HOLEMASK_BITS); + nextbit = startidx + 1; + while (startidx < XFS_INOBT_HOLEMASK_BITS) { + nextbit = find_next_zero_bit(holemask, XFS_INOBT_HOLEMASK_BITS, + nextbit); + /* + * If the next zero bit is contiguous, update the end index of + * the current range and continue. + */ + if (nextbit != XFS_INOBT_HOLEMASK_BITS && + nextbit == endidx + 1) { + endidx = nextbit; + goto next; + } + + /* + * nextbit is not contiguous with the current end index. Convert + * the current start/end to an extent and add it to the free + * list. + */ + agbno = sagbno + (startidx * XFS_INODES_PER_HOLEMASK_BIT) / + mp->m_sb.sb_inopblock; + contigblk = ((endidx - startidx + 1) * + XFS_INODES_PER_HOLEMASK_BIT) / + mp->m_sb.sb_inopblock; + + ASSERT(agbno % mp->m_sb.sb_spino_align == 0); + ASSERT(contigblk % mp->m_sb.sb_spino_align == 0); + xfs_bmap_add_free(XFS_AGB_TO_FSB(mp, agno, agbno), contigblk, + flist, mp); + + /* reset range to current bit and carry on... */ + startidx = endidx = nextbit; + +next: + nextbit++; + } +} + STATIC int xfs_difree_inobt( struct xfs_mount *mp, @@ -1443,8 +1889,7 @@ xfs_difree_inobt( struct xfs_buf *agbp, xfs_agino_t agino, struct xfs_bmap_free *flist, - int *deleted, - xfs_ino_t *first_ino, + struct xfs_icluster *xic, struct xfs_inobt_rec_incore *orec) { struct xfs_agi *agi = XFS_BUF_TO_AGI(agbp); @@ -1498,20 +1943,23 @@ xfs_difree_inobt( rec.ir_freecount++; /* - * When an inode cluster is free, it becomes eligible for removal + * When an inode chunk is free, it becomes eligible for removal. Don't + * remove the chunk if the block size is large enough for multiple inode + * chunks (that might not be free). */ if (!(mp->m_flags & XFS_MOUNT_IKEEP) && - (rec.ir_freecount == mp->m_ialloc_inos)) { - - *deleted = 1; - *first_ino = XFS_AGINO_TO_INO(mp, agno, rec.ir_startino); + rec.ir_free == XFS_INOBT_ALL_FREE && + mp->m_sb.sb_inopblock <= XFS_INODES_PER_CHUNK) { + xic->deleted = 1; + xic->first_ino = XFS_AGINO_TO_INO(mp, agno, rec.ir_startino); + xic->alloc = xfs_inobt_irec_to_allocmask(&rec); /* * Remove the inode cluster from the AGI B+Tree, adjust the * AGI and Superblock inode counts, and mark the disk space * to be freed when the transaction is committed. */ - ilen = mp->m_ialloc_inos; + ilen = rec.ir_freecount; be32_add_cpu(&agi->agi_count, -ilen); be32_add_cpu(&agi->agi_freecount, -(ilen - 1)); xfs_ialloc_log_agi(tp, agbp, XFS_AGI_COUNT | XFS_AGI_FREECOUNT); @@ -1527,11 +1975,9 @@ xfs_difree_inobt( goto error0; } - xfs_bmap_add_free(XFS_AGB_TO_FSB(mp, agno, - XFS_AGINO_TO_AGBNO(mp, rec.ir_startino)), - mp->m_ialloc_blks, flist, mp); + xfs_difree_inode_chunk(mp, agno, &rec, flist); } else { - *deleted = 0; + xic->deleted = 0; error = xfs_inobt_update(cur, &rec); if (error) { @@ -1596,7 +2042,9 @@ xfs_difree_finobt( */ XFS_WANT_CORRUPTED_GOTO(mp, ibtrec->ir_freecount == 1, error); - error = xfs_inobt_insert_rec(cur, ibtrec->ir_freecount, + error = xfs_inobt_insert_rec(cur, ibtrec->ir_holemask, + ibtrec->ir_count, + ibtrec->ir_freecount, ibtrec->ir_free, &i); if (error) goto error; @@ -1631,8 +2079,13 @@ xfs_difree_finobt( * free inode. Hence, if all of the inodes are free and we aren't * keeping inode chunks permanently on disk, remove the record. * Otherwise, update the record with the new information. + * + * Note that we currently can't free chunks when the block size is large + * enough for multiple chunks. Leave the finobt record to remain in sync + * with the inobt. */ - if (rec.ir_freecount == mp->m_ialloc_inos && + if (rec.ir_free == XFS_INOBT_ALL_FREE && + mp->m_sb.sb_inopblock <= XFS_INODES_PER_CHUNK && !(mp->m_flags & XFS_MOUNT_IKEEP)) { error = xfs_btree_delete(cur, &i); if (error) @@ -1668,8 +2121,7 @@ xfs_difree( struct xfs_trans *tp, /* transaction pointer */ xfs_ino_t inode, /* inode to be freed */ struct xfs_bmap_free *flist, /* extents to free */ - int *deleted,/* set if inode cluster was deleted */ - xfs_ino_t *first_ino)/* first inode in deleted cluster */ + struct xfs_icluster *xic) /* cluster info if deleted */ { /* REFERENCED */ xfs_agblock_t agbno; /* block number containing inode */ @@ -1720,8 +2172,7 @@ xfs_difree( /* * Fix up the inode allocation btree. */ - error = xfs_difree_inobt(mp, tp, agbp, agino, flist, deleted, first_ino, - &rec); + error = xfs_difree_inobt(mp, tp, agbp, agino, flist, xic, &rec); if (error) goto error0; diff --git a/fs/xfs/libxfs/xfs_ialloc.h b/fs/xfs/libxfs/xfs_ialloc.h index 100007d..6e450df 100644 --- a/fs/xfs/libxfs/xfs_ialloc.h +++ b/fs/xfs/libxfs/xfs_ialloc.h @@ -28,6 +28,13 @@ struct xfs_btree_cur; /* Move inodes in clusters of this size */ #define XFS_INODE_BIG_CLUSTER_SIZE 8192 +struct xfs_icluster { + bool deleted; /* record is deleted */ + xfs_ino_t first_ino; /* first inode number */ + uint64_t alloc; /* inode phys. allocation bitmap for + * sparse chunks */ +}; + /* Calculate and return the number of filesystem blocks per inode cluster */ static inline int xfs_icluster_size_fsb( @@ -44,8 +51,7 @@ xfs_icluster_size_fsb( static inline struct xfs_dinode * xfs_make_iptr(struct xfs_mount *mp, struct xfs_buf *b, int o) { - return (struct xfs_dinode *) - (xfs_buf_offset(b, o << (mp)->m_sb.sb_inodelog)); + return xfs_buf_offset(b, o << (mp)->m_sb.sb_inodelog); } /* @@ -90,8 +96,7 @@ xfs_difree( struct xfs_trans *tp, /* transaction pointer */ xfs_ino_t inode, /* inode to be freed */ struct xfs_bmap_free *flist, /* extents to free */ - int *deleted, /* set if inode cluster was deleted */ - xfs_ino_t *first_ino); /* first inode in deleted cluster */ + struct xfs_icluster *ifree); /* cluster info if deleted */ /* * Return the location of the inode in imap, for mapping it into a buffer. @@ -156,7 +161,7 @@ int xfs_inobt_get_rec(struct xfs_btree_cur *cur, * Inode chunk initialisation routine */ int xfs_ialloc_inode_init(struct xfs_mount *mp, struct xfs_trans *tp, - struct list_head *buffer_list, + struct list_head *buffer_list, int icount, xfs_agnumber_t agno, xfs_agblock_t agbno, xfs_agblock_t length, unsigned int gen); diff --git a/fs/xfs/libxfs/xfs_ialloc_btree.c b/fs/xfs/libxfs/xfs_ialloc_btree.c index 964c465..674ad8f 100644 --- a/fs/xfs/libxfs/xfs_ialloc_btree.c +++ b/fs/xfs/libxfs/xfs_ialloc_btree.c @@ -167,7 +167,16 @@ xfs_inobt_init_rec_from_cur( union xfs_btree_rec *rec) { rec->inobt.ir_startino = cpu_to_be32(cur->bc_rec.i.ir_startino); - rec->inobt.ir_freecount = cpu_to_be32(cur->bc_rec.i.ir_freecount); + if (xfs_sb_version_hassparseinodes(&cur->bc_mp->m_sb)) { + rec->inobt.ir_u.sp.ir_holemask = + cpu_to_be16(cur->bc_rec.i.ir_holemask); + rec->inobt.ir_u.sp.ir_count = cur->bc_rec.i.ir_count; + rec->inobt.ir_u.sp.ir_freecount = cur->bc_rec.i.ir_freecount; + } else { + /* ir_holemask/ir_count not supported on-disk */ + rec->inobt.ir_u.f.ir_freecount = + cpu_to_be32(cur->bc_rec.i.ir_freecount); + } rec->inobt.ir_free = cpu_to_be64(cur->bc_rec.i.ir_free); } @@ -418,3 +427,85 @@ xfs_inobt_maxrecs( return blocklen / sizeof(xfs_inobt_rec_t); return blocklen / (sizeof(xfs_inobt_key_t) + sizeof(xfs_inobt_ptr_t)); } + +/* + * Convert the inode record holemask to an inode allocation bitmap. The inode + * allocation bitmap is inode granularity and specifies whether an inode is + * physically allocated on disk (not whether the inode is considered allocated + * or free by the fs). + * + * A bit value of 1 means the inode is allocated, a value of 0 means it is free. + */ +uint64_t +xfs_inobt_irec_to_allocmask( + struct xfs_inobt_rec_incore *rec) +{ + uint64_t bitmap = 0; + uint64_t inodespbit; + int nextbit; + uint allocbitmap; + + /* + * The holemask has 16-bits for a 64 inode record. Therefore each + * holemask bit represents multiple inodes. Create a mask of bits to set + * in the allocmask for each holemask bit. + */ + inodespbit = (1 << XFS_INODES_PER_HOLEMASK_BIT) - 1; + + /* + * Allocated inodes are represented by 0 bits in holemask. Invert the 0 + * bits to 1 and convert to a uint so we can use xfs_next_bit(). Mask + * anything beyond the 16 holemask bits since this casts to a larger + * type. + */ + allocbitmap = ~rec->ir_holemask & ((1 << XFS_INOBT_HOLEMASK_BITS) - 1); + + /* + * allocbitmap is the inverted holemask so every set bit represents + * allocated inodes. To expand from 16-bit holemask granularity to + * 64-bit (e.g., bit-per-inode), set inodespbit bits in the target + * bitmap for every holemask bit. + */ + nextbit = xfs_next_bit(&allocbitmap, 1, 0); + while (nextbit != -1) { + ASSERT(nextbit < (sizeof(rec->ir_holemask) * NBBY)); + + bitmap |= (inodespbit << + (nextbit * XFS_INODES_PER_HOLEMASK_BIT)); + + nextbit = xfs_next_bit(&allocbitmap, 1, nextbit + 1); + } + + return bitmap; +} + +#if defined(DEBUG) || defined(XFS_WARN) +/* + * Verify that an in-core inode record has a valid inode count. + */ +int +xfs_inobt_rec_check_count( + struct xfs_mount *mp, + struct xfs_inobt_rec_incore *rec) +{ + int inocount = 0; + int nextbit = 0; + uint64_t allocbmap; + int wordsz; + + wordsz = sizeof(allocbmap) / sizeof(unsigned int); + allocbmap = xfs_inobt_irec_to_allocmask(rec); + + nextbit = xfs_next_bit((uint *) &allocbmap, wordsz, nextbit); + while (nextbit != -1) { + inocount++; + nextbit = xfs_next_bit((uint *) &allocbmap, wordsz, + nextbit + 1); + } + + if (inocount != rec->ir_count) + return -EFSCORRUPTED; + + return 0; +} +#endif /* DEBUG */ diff --git a/fs/xfs/libxfs/xfs_ialloc_btree.h b/fs/xfs/libxfs/xfs_ialloc_btree.h index d7ebea7..bd88453 100644 --- a/fs/xfs/libxfs/xfs_ialloc_btree.h +++ b/fs/xfs/libxfs/xfs_ialloc_btree.h @@ -62,4 +62,14 @@ extern struct xfs_btree_cur *xfs_inobt_init_cursor(struct xfs_mount *, xfs_btnum_t); extern int xfs_inobt_maxrecs(struct xfs_mount *, int, int); +/* ir_holemask to inode allocation bitmap conversion */ +uint64_t xfs_inobt_irec_to_allocmask(struct xfs_inobt_rec_incore *); + +#if defined(DEBUG) || defined(XFS_WARN) +int xfs_inobt_rec_check_count(struct xfs_mount *, + struct xfs_inobt_rec_incore *); +#else +#define xfs_inobt_rec_check_count(mp, rec) 0 +#endif /* DEBUG */ + #endif /* __XFS_IALLOC_BTREE_H__ */ diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c index 002b6b3..6526e76 100644 --- a/fs/xfs/libxfs/xfs_inode_buf.c +++ b/fs/xfs/libxfs/xfs_inode_buf.c @@ -46,8 +46,7 @@ xfs_inobp_check( j = mp->m_inode_cluster_size >> mp->m_sb.sb_inodelog; for (i = 0; i < j; i++) { - dip = (xfs_dinode_t *)xfs_buf_offset(bp, - i * mp->m_sb.sb_inodesize); + dip = xfs_buf_offset(bp, i * mp->m_sb.sb_inodesize); if (!dip->di_next_unlinked) { xfs_alert(mp, "Detected bogus zero next_unlinked field in inode %d buffer 0x%llx.", @@ -86,8 +85,7 @@ xfs_inode_buf_verify( int di_ok; xfs_dinode_t *dip; - dip = (struct xfs_dinode *)xfs_buf_offset(bp, - (i << mp->m_sb.sb_inodelog)); + dip = xfs_buf_offset(bp, (i << mp->m_sb.sb_inodelog)); di_ok = dip->di_magic == cpu_to_be16(XFS_DINODE_MAGIC) && XFS_DINODE_GOOD_VERSION(dip->di_version); if (unlikely(XFS_TEST_ERROR(!di_ok, mp, @@ -186,7 +184,7 @@ xfs_imap_to_bp( } *bpp = bp; - *dipp = (struct xfs_dinode *)xfs_buf_offset(bp, imap->im_boffset); + *dipp = xfs_buf_offset(bp, imap->im_boffset); return 0; } diff --git a/fs/xfs/libxfs/xfs_sb.c b/fs/xfs/libxfs/xfs_sb.c index dc4bfc5..df9851c 100644 --- a/fs/xfs/libxfs/xfs_sb.c +++ b/fs/xfs/libxfs/xfs_sb.c @@ -174,6 +174,27 @@ xfs_mount_validate_sb( return -EFSCORRUPTED; } + /* + * Full inode chunks must be aligned to inode chunk size when + * sparse inodes are enabled to support the sparse chunk + * allocation algorithm and prevent overlapping inode records. + */ + if (xfs_sb_version_hassparseinodes(sbp)) { + uint32_t align; + + xfs_alert(mp, + "EXPERIMENTAL sparse inode feature enabled. Use at your own risk!"); + + align = XFS_INODES_PER_CHUNK * sbp->sb_inodesize + >> sbp->sb_blocklog; + if (sbp->sb_inoalignmt != align) { + xfs_warn(mp, +"Inode block alignment (%u) must match chunk size (%u) for sparse inodes.", + sbp->sb_inoalignmt, align); + return -EINVAL; + } + } + if (unlikely( sbp->sb_logstart == 0 && mp->m_logdev_targp == mp->m_ddev_targp)) { xfs_warn(mp, @@ -374,7 +395,7 @@ __xfs_sb_from_disk( be32_to_cpu(from->sb_features_log_incompat); /* crc is only used on disk, not in memory; just init to 0 here. */ to->sb_crc = 0; - to->sb_pad = 0; + to->sb_spino_align = be32_to_cpu(from->sb_spino_align); to->sb_pquotino = be64_to_cpu(from->sb_pquotino); to->sb_lsn = be64_to_cpu(from->sb_lsn); /* Convert on-disk flags to in-memory flags? */ @@ -516,7 +537,7 @@ xfs_sb_to_disk( cpu_to_be32(from->sb_features_incompat); to->sb_features_log_incompat = cpu_to_be32(from->sb_features_log_incompat); - to->sb_pad = 0; + to->sb_spino_align = cpu_to_be32(from->sb_spino_align); to->sb_lsn = cpu_to_be64(from->sb_lsn); } } @@ -689,6 +710,11 @@ xfs_sb_mount_common( mp->m_ialloc_inos = (int)MAX((__uint16_t)XFS_INODES_PER_CHUNK, sbp->sb_inopblock); mp->m_ialloc_blks = mp->m_ialloc_inos >> sbp->sb_inopblog; + + if (sbp->sb_spino_align) + mp->m_ialloc_min_blks = sbp->sb_spino_align; + else + mp->m_ialloc_min_blks = mp->m_ialloc_blks; } /* @@ -792,12 +818,12 @@ xfs_sync_sb( tp = _xfs_trans_alloc(mp, XFS_TRANS_SB_CHANGE, KM_SLEEP); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_sb, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } xfs_log_sb(tp); if (wait) xfs_trans_set_sync(tp); - return xfs_trans_commit(tp, 0); + return xfs_trans_commit(tp); } diff --git a/fs/xfs/libxfs/xfs_shared.h b/fs/xfs/libxfs/xfs_shared.h index 8dda4b3..5be5297 100644 --- a/fs/xfs/libxfs/xfs_shared.h +++ b/fs/xfs/libxfs/xfs_shared.h @@ -182,12 +182,6 @@ int xfs_log_calc_minimum_size(struct xfs_mount *); #define XFS_TRANS_FREEZE_PROT 0x40 /* Transaction has elevated writer count in superblock */ /* - * Values for call flags parameter. - */ -#define XFS_TRANS_RELEASE_LOG_RES 0x4 -#define XFS_TRANS_ABORT 0x8 - -/* * Field values for xfs_trans_mod_sb. */ #define XFS_TRANS_SB_ICOUNT 0x00000001 diff --git a/fs/xfs/libxfs/xfs_trans_resv.h b/fs/xfs/libxfs/xfs_trans_resv.h index 2d5bdfc..7978150 100644 --- a/fs/xfs/libxfs/xfs_trans_resv.h +++ b/fs/xfs/libxfs/xfs_trans_resv.h @@ -73,9 +73,9 @@ struct xfs_trans_resv { * 2 trees * (2 blocks/level * max depth - 1) * block size */ #define XFS_ALLOCFREE_LOG_RES(mp,nx) \ - ((nx) * (2 * XFS_FSB_TO_B((mp), 2 * XFS_AG_MAXLEVELS(mp) - 1))) + ((nx) * (2 * XFS_FSB_TO_B((mp), 2 * (mp)->m_ag_maxlevels - 1))) #define XFS_ALLOCFREE_LOG_COUNT(mp,nx) \ - ((nx) * (2 * (2 * XFS_AG_MAXLEVELS(mp) - 1))) + ((nx) * (2 * (2 * (mp)->m_ag_maxlevels - 1))) /* * Per-directory log reservation for any directory change. diff --git a/fs/xfs/libxfs/xfs_trans_space.h b/fs/xfs/libxfs/xfs_trans_space.h index bf9c457..41e0428 100644 --- a/fs/xfs/libxfs/xfs_trans_space.h +++ b/fs/xfs/libxfs/xfs_trans_space.h @@ -67,7 +67,7 @@ #define XFS_DIOSTRAT_SPACE_RES(mp, v) \ (XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK) + (v)) #define XFS_GROWFS_SPACE_RES(mp) \ - (2 * XFS_AG_MAXLEVELS(mp)) + (2 * (mp)->m_ag_maxlevels) #define XFS_GROWFSRT_SPACE_RES(mp,b) \ ((b) + XFS_EXTENTADD_SPACE_RES(mp, XFS_DATA_FORK)) #define XFS_LINK_SPACE_RES(mp,nl) \ diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index a56960d..3859f5e 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -109,7 +109,7 @@ xfs_setfilesize_trans_alloc( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_fsyncts, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -145,7 +145,7 @@ xfs_setfilesize( isize = xfs_new_eof(ip, offset + size); if (!isize) { xfs_iunlock(ip, XFS_ILOCK_EXCL); - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return 0; } @@ -155,7 +155,7 @@ xfs_setfilesize( xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - return xfs_trans_commit(tp, 0); + return xfs_trans_commit(tp); } STATIC int @@ -356,7 +356,6 @@ xfs_end_bio( { xfs_ioend_t *ioend = bio->bi_private; - ASSERT(atomic_read(&bio->bi_cnt) >= 1); ioend->io_error = test_bit(BIO_UPTODATE, &bio->bi_flags) ? 0 : error; /* Toss bio and pass work off to an xfsdatad thread */ @@ -1349,7 +1348,7 @@ __xfs_get_blocks( sector_t iblock, struct buffer_head *bh_result, int create, - int direct) + bool direct) { struct xfs_inode *ip = XFS_I(inode); struct xfs_mount *mp = ip->i_mount; @@ -1414,6 +1413,7 @@ __xfs_get_blocks( if (error) return error; new = 1; + } else { /* * Delalloc reservations do not require a transaction, @@ -1508,49 +1508,29 @@ xfs_get_blocks( struct buffer_head *bh_result, int create) { - return __xfs_get_blocks(inode, iblock, bh_result, create, 0); + return __xfs_get_blocks(inode, iblock, bh_result, create, false); } -STATIC int +int xfs_get_blocks_direct( struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) { - return __xfs_get_blocks(inode, iblock, bh_result, create, 1); + return __xfs_get_blocks(inode, iblock, bh_result, create, true); } -/* - * Complete a direct I/O write request. - * - * The ioend structure is passed from __xfs_get_blocks() to tell us what to do. - * If no ioend exists (i.e. @private == NULL) then the write IO is an overwrite - * wholly within the EOF and so there is nothing for us to do. Note that in this - * case the completion can be called in interrupt context, whereas if we have an - * ioend we will always be called in task context (i.e. from a workqueue). - */ -STATIC void -xfs_end_io_direct_write( - struct kiocb *iocb, +static void +__xfs_end_io_direct_write( + struct inode *inode, + struct xfs_ioend *ioend, loff_t offset, - ssize_t size, - void *private) + ssize_t size) { - struct inode *inode = file_inode(iocb->ki_filp); - struct xfs_inode *ip = XFS_I(inode); - struct xfs_mount *mp = ip->i_mount; - struct xfs_ioend *ioend = private; - - trace_xfs_gbmap_direct_endio(ip, offset, size, - ioend ? ioend->io_type : 0, NULL); + struct xfs_mount *mp = XFS_I(inode)->i_mount; - if (!ioend) { - ASSERT(offset + size <= i_size_read(inode)); - return; - } - - if (XFS_FORCED_SHUTDOWN(mp)) + if (XFS_FORCED_SHUTDOWN(mp) || ioend->io_error) goto out_end_io; /* @@ -1587,10 +1567,10 @@ xfs_end_io_direct_write( * here can result in EOF moving backwards and Bad Things Happen when * that occurs. */ - spin_lock(&ip->i_flags_lock); + spin_lock(&XFS_I(inode)->i_flags_lock); if (offset + size > i_size_read(inode)) i_size_write(inode, offset + size); - spin_unlock(&ip->i_flags_lock); + spin_unlock(&XFS_I(inode)->i_flags_lock); /* * If we are doing an append IO that needs to update the EOF on disk, @@ -1607,6 +1587,98 @@ out_end_io: return; } +/* + * Complete a direct I/O write request. + * + * The ioend structure is passed from __xfs_get_blocks() to tell us what to do. + * If no ioend exists (i.e. @private == NULL) then the write IO is an overwrite + * wholly within the EOF and so there is nothing for us to do. Note that in this + * case the completion can be called in interrupt context, whereas if we have an + * ioend we will always be called in task context (i.e. from a workqueue). + */ +STATIC void +xfs_end_io_direct_write( + struct kiocb *iocb, + loff_t offset, + ssize_t size, + void *private) +{ + struct inode *inode = file_inode(iocb->ki_filp); + struct xfs_ioend *ioend = private; + + trace_xfs_gbmap_direct_endio(XFS_I(inode), offset, size, + ioend ? ioend->io_type : 0, NULL); + + if (!ioend) { + ASSERT(offset + size <= i_size_read(inode)); + return; + } + + __xfs_end_io_direct_write(inode, ioend, offset, size); +} + +/* + * For DAX we need a mapping buffer callback for unwritten extent conversion + * when page faults allocate blocks and then zero them. Note that in this + * case the mapping indicated by the ioend may extend beyond EOF. We most + * definitely do not want to extend EOF here, so we trim back the ioend size to + * EOF. + */ +#ifdef CONFIG_FS_DAX +void +xfs_end_io_dax_write( + struct buffer_head *bh, + int uptodate) +{ + struct xfs_ioend *ioend = bh->b_private; + struct inode *inode = ioend->io_inode; + ssize_t size = ioend->io_size; + + ASSERT(IS_DAX(ioend->io_inode)); + + /* if there was an error zeroing, then don't convert it */ + if (!uptodate) + ioend->io_error = -EIO; + + /* + * Trim update to EOF, so we don't extend EOF during unwritten extent + * conversion of partial EOF blocks. + */ + spin_lock(&XFS_I(inode)->i_flags_lock); + if (ioend->io_offset + size > i_size_read(inode)) + size = i_size_read(inode) - ioend->io_offset; + spin_unlock(&XFS_I(inode)->i_flags_lock); + + __xfs_end_io_direct_write(inode, ioend, ioend->io_offset, size); + +} +#else +void xfs_end_io_dax_write(struct buffer_head *bh, int uptodate) { } +#endif + +static inline ssize_t +xfs_vm_do_dio( + struct inode *inode, + struct kiocb *iocb, + struct iov_iter *iter, + loff_t offset, + void (*endio)(struct kiocb *iocb, + loff_t offset, + ssize_t size, + void *private), + int flags) +{ + struct block_device *bdev; + + if (IS_DAX(inode)) + return dax_do_io(iocb, inode, iter, offset, + xfs_get_blocks_direct, endio, 0); + + bdev = xfs_find_bdev_for_inode(inode); + return __blockdev_direct_IO(iocb, inode, bdev, iter, offset, + xfs_get_blocks_direct, endio, NULL, flags); +} + STATIC ssize_t xfs_vm_direct_IO( struct kiocb *iocb, @@ -1614,16 +1686,11 @@ xfs_vm_direct_IO( loff_t offset) { struct inode *inode = iocb->ki_filp->f_mapping->host; - struct block_device *bdev = xfs_find_bdev_for_inode(inode); - if (iov_iter_rw(iter) == WRITE) { - return __blockdev_direct_IO(iocb, inode, bdev, iter, offset, - xfs_get_blocks_direct, - xfs_end_io_direct_write, NULL, - DIO_ASYNC_EXTEND); - } - return __blockdev_direct_IO(iocb, inode, bdev, iter, offset, - xfs_get_blocks_direct, NULL, NULL, 0); + if (iov_iter_rw(iter) == WRITE) + return xfs_vm_do_dio(inode, iocb, iter, offset, + xfs_end_io_direct_write, DIO_ASYNC_EXTEND); + return xfs_vm_do_dio(inode, iocb, iter, offset, NULL, 0); } /* @@ -1874,6 +1941,7 @@ xfs_vm_set_page_dirty( loff_t end_offset; loff_t offset; int newly_dirty; + struct mem_cgroup *memcg; if (unlikely(!mapping)) return !TestSetPageDirty(page); @@ -1893,6 +1961,11 @@ xfs_vm_set_page_dirty( offset += 1 << inode->i_blkbits; } while (bh != head); } + /* + * Use mem_group_begin_page_stat() to keep PageDirty synchronized with + * per-memcg dirty page counters. + */ + memcg = mem_cgroup_begin_page_stat(page); newly_dirty = !TestSetPageDirty(page); spin_unlock(&mapping->private_lock); @@ -1903,13 +1976,15 @@ xfs_vm_set_page_dirty( spin_lock_irqsave(&mapping->tree_lock, flags); if (page->mapping) { /* Race with truncate? */ WARN_ON_ONCE(!PageUptodate(page)); - account_page_dirtied(page, mapping); + account_page_dirtied(page, mapping, memcg); radix_tree_tag_set(&mapping->page_tree, page_index(page), PAGECACHE_TAG_DIRTY); } spin_unlock_irqrestore(&mapping->tree_lock, flags); - __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); } + mem_cgroup_end_page_stat(memcg); + if (newly_dirty) + __mark_inode_dirty(mapping->host, I_DIRTY_PAGES); return newly_dirty; } diff --git a/fs/xfs/xfs_aops.h b/fs/xfs/xfs_aops.h index ac644e0..86afd1a 100644 --- a/fs/xfs/xfs_aops.h +++ b/fs/xfs/xfs_aops.h @@ -53,7 +53,12 @@ typedef struct xfs_ioend { } xfs_ioend_t; extern const struct address_space_operations xfs_address_space_operations; -extern int xfs_get_blocks(struct inode *, sector_t, struct buffer_head *, int); + +int xfs_get_blocks(struct inode *inode, sector_t offset, + struct buffer_head *map_bh, int create); +int xfs_get_blocks_direct(struct inode *inode, sector_t offset, + struct buffer_head *map_bh, int create); +void xfs_end_io_dax_write(struct buffer_head *bh, int uptodate); extern void xfs_count_page_state(struct page *, int *, int *); diff --git a/fs/xfs/xfs_attr_inactive.c b/fs/xfs/xfs_attr_inactive.c index f9c1c64..2bb959a 100644 --- a/fs/xfs/xfs_attr_inactive.c +++ b/fs/xfs/xfs_attr_inactive.c @@ -380,23 +380,30 @@ xfs_attr3_root_inactive( return error; } +/* + * xfs_attr_inactive kills all traces of an attribute fork on an inode. It + * removes both the on-disk and in-memory inode fork. Note that this also has to + * handle the condition of inodes without attributes but with an attribute fork + * configured, so we can't use xfs_inode_hasattr() here. + * + * The in-memory attribute fork is removed even on error. + */ int -xfs_attr_inactive(xfs_inode_t *dp) +xfs_attr_inactive( + struct xfs_inode *dp) { - xfs_trans_t *trans; - xfs_mount_t *mp; - int error; + struct xfs_trans *trans; + struct xfs_mount *mp; + int lock_mode = XFS_ILOCK_SHARED; + int error = 0; mp = dp->i_mount; ASSERT(! XFS_NOT_DQATTACHED(mp, dp)); - xfs_ilock(dp, XFS_ILOCK_SHARED); - if (!xfs_inode_hasattr(dp) || - dp->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) { - xfs_iunlock(dp, XFS_ILOCK_SHARED); - return 0; - } - xfs_iunlock(dp, XFS_ILOCK_SHARED); + xfs_ilock(dp, lock_mode); + if (!XFS_IFORK_Q(dp)) + goto out_destroy_fork; + xfs_iunlock(dp, lock_mode); /* * Start our first transaction of the day. @@ -408,13 +415,17 @@ xfs_attr_inactive(xfs_inode_t *dp) * the inode in every transaction to let it float upward through * the log. */ + lock_mode = 0; trans = xfs_trans_alloc(mp, XFS_TRANS_ATTRINVAL); error = xfs_trans_reserve(trans, &M_RES(mp)->tr_attrinval, 0, 0); - if (error) { - xfs_trans_cancel(trans, 0); - return error; - } - xfs_ilock(dp, XFS_ILOCK_EXCL); + if (error) + goto out_cancel; + + lock_mode = XFS_ILOCK_EXCL; + xfs_ilock(dp, lock_mode); + + if (!XFS_IFORK_Q(dp)) + goto out_cancel; /* * No need to make quota reservations here. We expect to release some @@ -423,28 +434,36 @@ xfs_attr_inactive(xfs_inode_t *dp) xfs_trans_ijoin(trans, dp, 0); /* - * Decide on what work routines to call based on the inode size. + * Invalidate and truncate the attribute fork extents. Make sure the + * fork actually has attributes as otherwise the invalidation has no + * blocks to read and returns an error. In this case, just do the fork + * removal below. */ - if (!xfs_inode_hasattr(dp) || - dp->i_d.di_aformat == XFS_DINODE_FMT_LOCAL) { - error = 0; - goto out; - } - error = xfs_attr3_root_inactive(&trans, dp); - if (error) - goto out; + if (xfs_inode_hasattr(dp) && + dp->i_d.di_aformat != XFS_DINODE_FMT_LOCAL) { + error = xfs_attr3_root_inactive(&trans, dp); + if (error) + goto out_cancel; - error = xfs_itruncate_extents(&trans, dp, XFS_ATTR_FORK, 0); - if (error) - goto out; + error = xfs_itruncate_extents(&trans, dp, XFS_ATTR_FORK, 0); + if (error) + goto out_cancel; + } - error = xfs_trans_commit(trans, XFS_TRANS_RELEASE_LOG_RES); - xfs_iunlock(dp, XFS_ILOCK_EXCL); + /* Reset the attribute fork - this also destroys the in-core fork */ + xfs_attr_fork_remove(dp, trans); + error = xfs_trans_commit(trans); + xfs_iunlock(dp, lock_mode); return error; -out: - xfs_trans_cancel(trans, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT); - xfs_iunlock(dp, XFS_ILOCK_EXCL); +out_cancel: + xfs_trans_cancel(trans); +out_destroy_fork: + /* kill the in-core attr fork before we drop the inode lock */ + if (dp->i_afp) + xfs_idestroy_fork(dp, XFS_ATTR_FORK); + if (lock_mode) + xfs_iunlock(dp, lock_mode); return error; } diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c index a52bbd3..0f34886c 100644 --- a/fs/xfs/xfs_bmap_util.c +++ b/fs/xfs/xfs_bmap_util.c @@ -75,28 +75,20 @@ xfs_bmap_finish( xfs_efi_log_item_t *efi; /* extent free intention */ int error; /* error return value */ xfs_bmap_free_item_t *free; /* free extent item */ - struct xfs_trans_res tres; /* new log reservation */ xfs_mount_t *mp; /* filesystem mount structure */ xfs_bmap_free_item_t *next; /* next item on free list */ - xfs_trans_t *ntp; /* new transaction pointer */ ASSERT((*tp)->t_flags & XFS_TRANS_PERM_LOG_RES); if (flist->xbf_count == 0) { *committed = 0; return 0; } - ntp = *tp; - efi = xfs_trans_get_efi(ntp, flist->xbf_count); + efi = xfs_trans_get_efi(*tp, flist->xbf_count); for (free = flist->xbf_first; free; free = free->xbfi_next) - xfs_trans_log_efi_extent(ntp, efi, free->xbfi_startblock, + xfs_trans_log_efi_extent(*tp, efi, free->xbfi_startblock, free->xbfi_blockcount); - tres.tr_logres = ntp->t_log_res; - tres.tr_logcount = ntp->t_log_count; - tres.tr_logflags = XFS_TRANS_PERM_LOG_RES; - ntp = xfs_trans_dup(*tp); - error = xfs_trans_commit(*tp, 0); - *tp = ntp; + error = xfs_trans_roll(tp, NULL); *committed = 1; /* * We have a new transaction, so we should return committed=1, @@ -105,19 +97,10 @@ xfs_bmap_finish( if (error) return error; - /* - * transaction commit worked ok so we can drop the extra ticket - * reference that we gained in xfs_trans_dup() - */ - xfs_log_ticket_put(ntp->t_ticket); - - error = xfs_trans_reserve(ntp, &tres, 0, 0); - if (error) - return error; - efd = xfs_trans_get_efd(ntp, efi, flist->xbf_count); + efd = xfs_trans_get_efd(*tp, efi, flist->xbf_count); for (free = flist->xbf_first; free != NULL; free = next) { next = free->xbfi_next; - if ((error = xfs_free_extent(ntp, free->xbfi_startblock, + if ((error = xfs_free_extent(*tp, free->xbfi_startblock, free->xbfi_blockcount))) { /* * The bmap free list will be cleaned up at a @@ -127,7 +110,7 @@ xfs_bmap_finish( * happens, since this transaction may not be * dirty yet. */ - mp = ntp->t_mountp; + mp = (*tp)->t_mountp; if (!XFS_FORCED_SHUTDOWN(mp)) xfs_force_shutdown(mp, (error == -EFSCORRUPTED) ? @@ -135,7 +118,7 @@ xfs_bmap_finish( SHUTDOWN_META_IO_ERROR); return error; } - xfs_trans_log_efd_extent(ntp, efd, free->xbfi_startblock, + xfs_trans_log_efd_extent(*tp, efd, free->xbfi_startblock, free->xbfi_blockcount); xfs_bmap_del_free(flist, NULL, free); } @@ -878,7 +861,7 @@ xfs_free_eofblocks( if (need_iolock) { if (!xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL)) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return -EAGAIN; } } @@ -886,7 +869,7 @@ xfs_free_eofblocks( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0); if (error) { ASSERT(XFS_FORCED_SHUTDOWN(mp)); - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); if (need_iolock) xfs_iunlock(ip, XFS_IOLOCK_EXCL); return error; @@ -908,12 +891,9 @@ xfs_free_eofblocks( * If we get an error at this point we simply don't * bother truncating the file. */ - xfs_trans_cancel(tp, - (XFS_TRANS_RELEASE_LOG_RES | - XFS_TRANS_ABORT)); + xfs_trans_cancel(tp); } else { - error = xfs_trans_commit(tp, - XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (!error) xfs_inode_clear_eofblocks_tag(ip); } @@ -1026,7 +1006,7 @@ xfs_alloc_file_space( * Free the transaction structure. */ ASSERT(error == -ENOSPC || XFS_FORCED_SHUTDOWN(mp)); - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); break; } xfs_ilock(ip, XFS_ILOCK_EXCL); @@ -1053,7 +1033,7 @@ xfs_alloc_file_space( goto error0; } - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); if (error) { break; @@ -1077,7 +1057,7 @@ error0: /* Cancel bmap, unlock inode, unreserve quota blocks, cancel trans */ xfs_trans_unreserve_quota_nblks(tp, ip, (long)qblocks, 0, quota_flag); error1: /* Just cancel transaction */ - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; } @@ -1133,14 +1113,29 @@ xfs_zero_remaining_bytes( break; ASSERT(imap.br_blockcount >= 1); ASSERT(imap.br_startoff == offset_fsb); + ASSERT(imap.br_startblock != DELAYSTARTBLOCK); + + if (imap.br_startblock == HOLESTARTBLOCK || + imap.br_state == XFS_EXT_UNWRITTEN) { + /* skip the entire extent */ + lastoffset = XFS_FSB_TO_B(mp, imap.br_startoff + + imap.br_blockcount) - 1; + continue; + } + lastoffset = XFS_FSB_TO_B(mp, imap.br_startoff + 1) - 1; if (lastoffset > endoff) lastoffset = endoff; - if (imap.br_startblock == HOLESTARTBLOCK) - continue; - ASSERT(imap.br_startblock != DELAYSTARTBLOCK); - if (imap.br_state == XFS_EXT_UNWRITTEN) + + /* DAX can just zero the backing device directly */ + if (IS_DAX(VFS_I(ip))) { + error = dax_zero_page_range(VFS_I(ip), offset, + lastoffset - offset + 1, + xfs_get_blocks_direct); + if (error) + return error; continue; + } error = xfs_buf_read_uncached(XFS_IS_REALTIME_INODE(ip) ? mp->m_rtdev_targp : mp->m_ddev_targp, @@ -1289,7 +1284,7 @@ xfs_free_file_space( * Free the transaction structure. */ ASSERT(error == -ENOSPC || XFS_FORCED_SHUTDOWN(mp)); - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); break; } xfs_ilock(ip, XFS_ILOCK_EXCL); @@ -1320,7 +1315,7 @@ xfs_free_file_space( goto error0; } - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); } @@ -1330,7 +1325,7 @@ xfs_free_file_space( error0: xfs_bmap_cancel(&free_list); error1: - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); goto out; } @@ -1462,7 +1457,7 @@ xfs_shift_file_space( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write, XFS_DIOSTRAT_SPACE_RES(mp, 0), 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); break; } @@ -1492,13 +1487,13 @@ xfs_shift_file_space( if (error) goto out; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); } return error; out: - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); return error; } @@ -1718,7 +1713,7 @@ xfs_swap_extents( tp = xfs_trans_alloc(mp, XFS_TRANS_SWAPEXT); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); goto out_unlock; } @@ -1901,7 +1896,7 @@ xfs_swap_extents( if (mp->m_flags & XFS_MOUNT_WSYNC) xfs_trans_set_sync(tp); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); trace_xfs_swap_extent_after(ip, 0); trace_xfs_swap_extent_after(tip, 1); @@ -1915,6 +1910,6 @@ out_unlock: goto out; out_trans_cancel: - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); goto out; } diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 1790b00..a4b7d92 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -1419,9 +1419,9 @@ xfs_buf_submit_wait( return error; } -xfs_caddr_t +void * xfs_buf_offset( - xfs_buf_t *bp, + struct xfs_buf *bp, size_t offset) { struct page *page; @@ -1431,7 +1431,7 @@ xfs_buf_offset( offset += bp->b_offset; page = bp->b_pages[offset >> PAGE_SHIFT]; - return (xfs_caddr_t)page_address(page) + (offset & (PAGE_SIZE-1)); + return page_address(page) + (offset & (PAGE_SIZE-1)); } /* diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h index 75ff5d5..331c1cc 100644 --- a/fs/xfs/xfs_buf.h +++ b/fs/xfs/xfs_buf.h @@ -299,7 +299,7 @@ extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, void *, xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO) /* Buffer Utility Routines */ -extern xfs_caddr_t xfs_buf_offset(xfs_buf_t *, size_t); +extern void *xfs_buf_offset(struct xfs_buf *, size_t); /* Delayed Write Buffer Routines */ extern bool xfs_buf_delwri_queue(struct xfs_buf *, struct list_head *); diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c index 02c01bb..4143dc7 100644 --- a/fs/xfs/xfs_dquot.c +++ b/fs/xfs/xfs_dquot.c @@ -568,8 +568,6 @@ xfs_qm_dqread( struct xfs_buf *bp; struct xfs_trans *tp = NULL; int error; - int cancelflags = 0; - dqp = kmem_zone_zalloc(xfs_qm_dqzone, KM_SLEEP); @@ -617,7 +615,6 @@ xfs_qm_dqread( XFS_QM_DQALLOC_SPACE_RES(mp), 0); if (error) goto error1; - cancelflags = XFS_TRANS_RELEASE_LOG_RES; } /* @@ -632,7 +629,6 @@ xfs_qm_dqread( * allocate (ENOENT). */ trace_xfs_dqread_fail(dqp); - cancelflags |= XFS_TRANS_ABORT; goto error1; } @@ -670,7 +666,7 @@ xfs_qm_dqread( xfs_trans_brelse(tp, bp); if (tp) { - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto error0; } @@ -680,7 +676,7 @@ xfs_qm_dqread( error1: if (tp) - xfs_trans_cancel(tp, cancelflags); + xfs_trans_cancel(tp); error0: xfs_qm_dqdestroy(dqp); *O_dqpp = NULL; diff --git a/fs/xfs/xfs_error.c b/fs/xfs/xfs_error.c index 338e50b..74d0e59 100644 --- a/fs/xfs/xfs_error.c +++ b/fs/xfs/xfs_error.c @@ -127,7 +127,7 @@ xfs_error_report( struct xfs_mount *mp, const char *filename, int linenum, - inst_t *ra) + void *ra) { if (level <= xfs_error_level) { xfs_alert_tag(mp, XFS_PTAG_ERROR_REPORT, @@ -146,7 +146,7 @@ xfs_corruption_error( void *p, const char *filename, int linenum, - inst_t *ra) + void *ra) { if (level <= xfs_error_level) xfs_hex_dump(p, 64); diff --git a/fs/xfs/xfs_error.h b/fs/xfs/xfs_error.h index c0394ed..4ed3042 100644 --- a/fs/xfs/xfs_error.h +++ b/fs/xfs/xfs_error.h @@ -21,10 +21,10 @@ struct xfs_mount; extern void xfs_error_report(const char *tag, int level, struct xfs_mount *mp, - const char *filename, int linenum, inst_t *ra); + const char *filename, int linenum, void *ra); extern void xfs_corruption_error(const char *tag, int level, struct xfs_mount *mp, void *p, const char *filename, - int linenum, inst_t *ra); + int linenum, void *ra); extern void xfs_verifier_error(struct xfs_buf *bp); #define XFS_ERROR_REPORT(e, lvl, mp) \ diff --git a/fs/xfs/xfs_extfree_item.c b/fs/xfs/xfs_extfree_item.c index cb7fe64..adc8f8f 100644 --- a/fs/xfs/xfs_extfree_item.c +++ b/fs/xfs/xfs_extfree_item.c @@ -239,7 +239,7 @@ xfs_efi_init( xfs_log_item_init(mp, &efip->efi_item, XFS_LI_EFI, &xfs_efi_item_ops); efip->efi_format.efi_nextents = nextents; - efip->efi_format.efi_id = (__psint_t)(void*)efip; + efip->efi_format.efi_id = (uintptr_t)(void *)efip; atomic_set(&efip->efi_next_extent, 0); atomic_set(&efip->efi_refcount, 2); diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index 71c2c712..f0e8249 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -41,6 +41,7 @@ #include <linux/dcache.h> #include <linux/falloc.h> #include <linux/pagevec.h> +#include <linux/backing-dev.h> static const struct vm_operations_struct xfs_file_vm_ops; @@ -79,14 +80,15 @@ xfs_rw_ilock_demote( } /* - * xfs_iozero + * xfs_iozero clears the specified range supplied via the page cache (except in + * the DAX case). Writes through the page cache will allocate blocks over holes, + * though the callers usually map the holes first and avoid them. If a block is + * not completely zeroed, then it will be read from disk before being partially + * zeroed. * - * xfs_iozero clears the specified range of buffer supplied, - * and marks all the affected blocks as valid and modified. If - * an affected block is not allocated, it will be allocated. If - * an affected block is not completely overwritten, and is not - * valid before the operation, it will be read from disk before - * being partially zeroed. + * In the DAX case, we can just directly write to the underlying pages. This + * will not allocate blocks, but will avoid holes and unwritten extents and so + * not do unnecessary work. */ int xfs_iozero( @@ -96,7 +98,8 @@ xfs_iozero( { struct page *page; struct address_space *mapping; - int status; + int status = 0; + mapping = VFS_I(ip)->i_mapping; do { @@ -108,23 +111,30 @@ xfs_iozero( if (bytes > count) bytes = count; - status = pagecache_write_begin(NULL, mapping, pos, bytes, - AOP_FLAG_UNINTERRUPTIBLE, - &page, &fsdata); - if (status) - break; + if (IS_DAX(VFS_I(ip))) { + status = dax_zero_page_range(VFS_I(ip), pos, bytes, + xfs_get_blocks_direct); + if (status) + break; + } else { + status = pagecache_write_begin(NULL, mapping, pos, bytes, + AOP_FLAG_UNINTERRUPTIBLE, + &page, &fsdata); + if (status) + break; - zero_user(page, offset, bytes); + zero_user(page, offset, bytes); - status = pagecache_write_end(NULL, mapping, pos, bytes, bytes, - page, fsdata); - WARN_ON(status <= 0); /* can't return less than zero! */ + status = pagecache_write_end(NULL, mapping, pos, bytes, + bytes, page, fsdata); + WARN_ON(status <= 0); /* can't return less than zero! */ + status = 0; + } pos += bytes; count -= bytes; - status = 0; } while (count); - return (-status); + return status; } int @@ -138,7 +148,7 @@ xfs_update_prealloc_flags( tp = xfs_trans_alloc(ip->i_mount, XFS_TRANS_WRITEID); error = xfs_trans_reserve(tp, &M_RES(ip->i_mount)->tr_writeid, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -160,7 +170,7 @@ xfs_update_prealloc_flags( xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); if (flags & XFS_PREALLOC_SYNC) xfs_trans_set_sync(tp); - return xfs_trans_commit(tp, 0); + return xfs_trans_commit(tp); } /* @@ -284,7 +294,7 @@ xfs_file_read_iter( if (file->f_mode & FMODE_NOCMTIME) ioflags |= XFS_IO_INVIS; - if (unlikely(ioflags & XFS_IO_ISDIRECT)) { + if ((ioflags & XFS_IO_ISDIRECT) && !IS_DAX(inode)) { xfs_buftarg_t *target = XFS_IS_REALTIME_INODE(ip) ? mp->m_rtdev_targp : mp->m_ddev_targp; @@ -378,7 +388,11 @@ xfs_file_splice_read( trace_xfs_file_splice_read(ip, count, *ppos, ioflags); - ret = generic_file_splice_read(infilp, ppos, pipe, count, flags); + /* for dax, we need to avoid the page cache */ + if (IS_DAX(VFS_I(ip))) + ret = default_file_splice_read(infilp, ppos, pipe, count, flags); + else + ret = generic_file_splice_read(infilp, ppos, pipe, count, flags); if (ret > 0) XFS_STATS_ADD(xs_read_bytes, ret); @@ -681,7 +695,7 @@ xfs_file_dio_aio_write( mp->m_rtdev_targp : mp->m_ddev_targp; /* DIO must be aligned to device logical sector size */ - if ((pos | count) & target->bt_logical_sectormask) + if (!IS_DAX(inode) && ((pos | count) & target->bt_logical_sectormask)) return -EINVAL; /* "unaligned" here means not aligned to a filesystem block */ @@ -767,8 +781,11 @@ xfs_file_dio_aio_write( out: xfs_rw_iunlock(ip, iolock); - /* No fallback to buffered IO on errors for XFS. */ - ASSERT(ret < 0 || ret == count); + /* + * No fallback to buffered IO on errors for XFS. DAX can result in + * partial writes, but direct IO will either complete fully or fail. + */ + ASSERT(ret < 0 || ret == count || IS_DAX(VFS_I(ip))); return ret; } @@ -851,7 +868,7 @@ xfs_file_write_iter( if (XFS_FORCED_SHUTDOWN(ip->i_mount)) return -EIO; - if (unlikely(iocb->ki_flags & IOCB_DIRECT)) + if ((iocb->ki_flags & IOCB_DIRECT) || IS_DAX(inode)) ret = xfs_file_dio_aio_write(iocb, from); else ret = xfs_file_buffered_aio_write(iocb, from); @@ -1072,17 +1089,6 @@ xfs_file_readdir( return xfs_readdir(ip, ctx, bufsize); } -STATIC int -xfs_file_mmap( - struct file *filp, - struct vm_area_struct *vma) -{ - vma->vm_ops = &xfs_file_vm_ops; - - file_accessed(filp); - return 0; -} - /* * This type is designed to indicate the type of offset we would like * to search from page cache for xfs_seek_hole_data(). @@ -1463,48 +1469,83 @@ xfs_file_llseek( * ordering of: * * mmap_sem (MM) - * i_mmap_lock (XFS - truncate serialisation) - * page_lock (MM) - * i_lock (XFS - extent map serialisation) + * sb_start_pagefault(vfs, freeze) + * i_mmap_lock (XFS - truncate serialisation) + * page_lock (MM) + * i_lock (XFS - extent map serialisation) + */ + +/* + * mmap()d file has taken write protection fault and is being made writable. We + * can set the page state up correctly for a writable page, which means we can + * do correct delalloc accounting (ENOSPC checking!) and unwritten extent + * mapping. */ STATIC int -xfs_filemap_fault( +xfs_filemap_page_mkwrite( struct vm_area_struct *vma, struct vm_fault *vmf) { - struct xfs_inode *ip = XFS_I(vma->vm_file->f_mapping->host); - int error; + struct inode *inode = file_inode(vma->vm_file); + int ret; - trace_xfs_filemap_fault(ip); + trace_xfs_filemap_page_mkwrite(XFS_I(inode)); - xfs_ilock(ip, XFS_MMAPLOCK_SHARED); - error = filemap_fault(vma, vmf); - xfs_iunlock(ip, XFS_MMAPLOCK_SHARED); + sb_start_pagefault(inode->i_sb); + file_update_time(vma->vm_file); + xfs_ilock(XFS_I(inode), XFS_MMAPLOCK_SHARED); - return error; + if (IS_DAX(inode)) { + ret = __dax_mkwrite(vma, vmf, xfs_get_blocks_direct, + xfs_end_io_dax_write); + } else { + ret = __block_page_mkwrite(vma, vmf, xfs_get_blocks); + ret = block_page_mkwrite_return(ret); + } + + xfs_iunlock(XFS_I(inode), XFS_MMAPLOCK_SHARED); + sb_end_pagefault(inode->i_sb); + + return ret; } -/* - * mmap()d file has taken write protection fault and is being made writable. We - * can set the page state up correctly for a writable page, which means we can - * do correct delalloc accounting (ENOSPC checking!) and unwritten extent - * mapping. - */ STATIC int -xfs_filemap_page_mkwrite( +xfs_filemap_fault( struct vm_area_struct *vma, struct vm_fault *vmf) { - struct xfs_inode *ip = XFS_I(vma->vm_file->f_mapping->host); - int error; + struct xfs_inode *ip = XFS_I(file_inode(vma->vm_file)); + int ret; + + trace_xfs_filemap_fault(ip); - trace_xfs_filemap_page_mkwrite(ip); + /* DAX can shortcut the normal fault path on write faults! */ + if ((vmf->flags & FAULT_FLAG_WRITE) && IS_DAX(VFS_I(ip))) + return xfs_filemap_page_mkwrite(vma, vmf); xfs_ilock(ip, XFS_MMAPLOCK_SHARED); - error = block_page_mkwrite(vma, vmf, xfs_get_blocks); + ret = filemap_fault(vma, vmf); xfs_iunlock(ip, XFS_MMAPLOCK_SHARED); - return error; + return ret; +} + +static const struct vm_operations_struct xfs_file_vm_ops = { + .fault = xfs_filemap_fault, + .map_pages = filemap_map_pages, + .page_mkwrite = xfs_filemap_page_mkwrite, +}; + +STATIC int +xfs_file_mmap( + struct file *filp, + struct vm_area_struct *vma) +{ + file_accessed(filp); + vma->vm_ops = &xfs_file_vm_ops; + if (IS_DAX(file_inode(filp))) + vma->vm_flags |= VM_MIXEDMAP; + return 0; } const struct file_operations xfs_file_operations = { @@ -1535,9 +1576,3 @@ const struct file_operations xfs_dir_file_operations = { #endif .fsync = xfs_dir_fsync, }; - -static const struct vm_operations_struct xfs_file_vm_ops = { - .fault = xfs_filemap_fault, - .map_pages = filemap_map_pages, - .page_mkwrite = xfs_filemap_page_mkwrite, -}; diff --git a/fs/xfs/xfs_filestream.c b/fs/xfs/xfs_filestream.c index da82f1c..c4c130f 100644 --- a/fs/xfs/xfs_filestream.c +++ b/fs/xfs/xfs_filestream.c @@ -196,7 +196,8 @@ xfs_filestream_pick_ag( goto next_ag; } - longest = xfs_alloc_longest_free_extent(mp, pag); + longest = xfs_alloc_longest_free_extent(mp, pag, + xfs_alloc_min_freelist(mp, pag)); if (((minlen && longest >= minlen) || (!minlen && pag->pagf_freeblks >= minfree)) && (!pag->pagf_metadata || !(flags & XFS_PICK_USERDATA) || diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index cb7e8a2..9b3438a 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -101,7 +101,9 @@ xfs_fs_geometry( (xfs_sb_version_hasftype(&mp->m_sb) ? XFS_FSOP_GEOM_FLAGS_FTYPE : 0) | (xfs_sb_version_hasfinobt(&mp->m_sb) ? - XFS_FSOP_GEOM_FLAGS_FINOBT : 0); + XFS_FSOP_GEOM_FLAGS_FINOBT : 0) | + (xfs_sb_version_hassparseinodes(&mp->m_sb) ? + XFS_FSOP_GEOM_FLAGS_SPINODES : 0); geo->logsectsize = xfs_sb_version_hassector(&mp->m_sb) ? mp->m_sb.sb_logsectsize : BBSIZE; geo->rtsectsize = mp->m_sb.sb_blocksize; @@ -201,7 +203,7 @@ xfs_growfs_data_private( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_growdata, XFS_GROWFS_SPACE_RES(mp), 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -489,7 +491,7 @@ xfs_growfs_data_private( if (dpct) xfs_trans_mod_sb(tp, XFS_TRANS_SB_IMAXPCT, dpct); xfs_trans_set_sync(tp); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); if (error) return error; @@ -557,7 +559,7 @@ xfs_growfs_data_private( return saved_error ? saved_error : error; error0: - xfs_trans_cancel(tp, XFS_TRANS_ABORT); + xfs_trans_cancel(tp); return error; } diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index d6ebc85..3da9f4d 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -905,7 +905,6 @@ xfs_dir_ialloc( { xfs_trans_t *tp; - xfs_trans_t *ntp; xfs_inode_t *ip; xfs_buf_t *ialloc_context = NULL; int code; @@ -954,8 +953,6 @@ xfs_dir_ialloc( * to succeed the second time. */ if (ialloc_context) { - struct xfs_trans_res tres; - /* * Normally, xfs_trans_commit releases all the locks. * We call bhold to hang on to the ialloc_context across @@ -964,12 +961,6 @@ xfs_dir_ialloc( * allocation group. */ xfs_trans_bhold(tp, ialloc_context); - /* - * Save the log reservation so we can use - * them in the next transaction. - */ - tres.tr_logres = xfs_trans_get_log_res(tp); - tres.tr_logcount = xfs_trans_get_log_count(tp); /* * We want the quota changes to be associated with the next @@ -985,35 +976,9 @@ xfs_dir_ialloc( tp->t_flags &= ~(XFS_TRANS_DQ_DIRTY); } - ntp = xfs_trans_dup(tp); - code = xfs_trans_commit(tp, 0); - tp = ntp; - if (committed != NULL) { + code = xfs_trans_roll(&tp, 0); + if (committed != NULL) *committed = 1; - } - /* - * If we get an error during the commit processing, - * release the buffer that is still held and return - * to the caller. - */ - if (code) { - xfs_buf_relse(ialloc_context); - if (dqinfo) { - tp->t_dqinfo = dqinfo; - xfs_trans_free_dqinfo(tp); - } - *tpp = ntp; - *ipp = NULL; - return code; - } - - /* - * transaction commit worked ok so we can drop the extra ticket - * reference that we gained in xfs_trans_dup() - */ - xfs_log_ticket_put(tp->t_ticket); - tres.tr_logflags = XFS_TRANS_PERM_LOG_RES; - code = xfs_trans_reserve(tp, &tres, 0, 0); /* * Re-attach the quota info that we detached from prev trx. @@ -1025,7 +990,7 @@ xfs_dir_ialloc( if (code) { xfs_buf_relse(ialloc_context); - *tpp = ntp; + *tpp = tp; *ipp = NULL; return code; } @@ -1127,7 +1092,6 @@ xfs_create( xfs_bmap_free_t free_list; xfs_fsblock_t first_block; bool unlock_dp_on_error = false; - uint cancel_flags; int committed; prid_t prid; struct xfs_dquot *udqp = NULL; @@ -1164,8 +1128,6 @@ xfs_create( tp = xfs_trans_alloc(mp, XFS_TRANS_CREATE); } - cancel_flags = XFS_TRANS_RELEASE_LOG_RES; - /* * Initially assume that the file does not exist and * reserve the resources for that case. If that is not @@ -1183,10 +1145,9 @@ xfs_create( resblks = 0; error = xfs_trans_reserve(tp, tres, 0, 0); } - if (error) { - cancel_flags = 0; + if (error) goto out_trans_cancel; - } + xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT); unlock_dp_on_error = true; @@ -1217,7 +1178,7 @@ xfs_create( if (error) { if (error == -ENOSPC) goto out_trans_cancel; - goto out_trans_abort; + goto out_trans_cancel; } /* @@ -1235,7 +1196,7 @@ xfs_create( resblks - XFS_IALLOC_SPACE_RES(mp) : 0); if (error) { ASSERT(error != -ENOSPC); - goto out_trans_abort; + goto out_trans_cancel; } xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE); @@ -1269,7 +1230,7 @@ xfs_create( if (error) goto out_bmap_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto out_release_inode; @@ -1282,10 +1243,8 @@ xfs_create( out_bmap_cancel: xfs_bmap_cancel(&free_list); - out_trans_abort: - cancel_flags |= XFS_TRANS_ABORT; out_trans_cancel: - xfs_trans_cancel(tp, cancel_flags); + xfs_trans_cancel(tp); out_release_inode: /* * Wait until after the current transaction is aborted to finish the @@ -1317,7 +1276,6 @@ xfs_create_tmpfile( struct xfs_inode *ip = NULL; struct xfs_trans *tp = NULL; int error; - uint cancel_flags = XFS_TRANS_RELEASE_LOG_RES; prid_t prid; struct xfs_dquot *udqp = NULL; struct xfs_dquot *gdqp = NULL; @@ -1350,10 +1308,8 @@ xfs_create_tmpfile( resblks = 0; error = xfs_trans_reserve(tp, tres, 0, 0); } - if (error) { - cancel_flags = 0; + if (error) goto out_trans_cancel; - } error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp, pdqp, resblks, 1, 0); @@ -1365,7 +1321,7 @@ xfs_create_tmpfile( if (error) { if (error == -ENOSPC) goto out_trans_cancel; - goto out_trans_abort; + goto out_trans_cancel; } if (mp->m_flags & XFS_MOUNT_WSYNC) @@ -1381,9 +1337,9 @@ xfs_create_tmpfile( ip->i_d.di_nlink--; error = xfs_iunlink(tp, ip); if (error) - goto out_trans_abort; + goto out_trans_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto out_release_inode; @@ -1394,10 +1350,8 @@ xfs_create_tmpfile( *ipp = ip; return 0; - out_trans_abort: - cancel_flags |= XFS_TRANS_ABORT; out_trans_cancel: - xfs_trans_cancel(tp, cancel_flags); + xfs_trans_cancel(tp); out_release_inode: /* * Wait until after the current transaction is aborted to finish the @@ -1427,7 +1381,6 @@ xfs_link( int error; xfs_bmap_free_t free_list; xfs_fsblock_t first_block; - int cancel_flags; int committed; int resblks; @@ -1447,17 +1400,14 @@ xfs_link( goto std_return; tp = xfs_trans_alloc(mp, XFS_TRANS_LINK); - cancel_flags = XFS_TRANS_RELEASE_LOG_RES; resblks = XFS_LINK_SPACE_RES(mp, target_name->len); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_link, resblks, 0); if (error == -ENOSPC) { resblks = 0; error = xfs_trans_reserve(tp, &M_RES(mp)->tr_link, 0, 0); } - if (error) { - cancel_flags = 0; + if (error) goto error_return; - } xfs_lock_two_inodes(sip, tdp, XFS_ILOCK_EXCL); @@ -1486,19 +1436,19 @@ xfs_link( if (sip->i_d.di_nlink == 0) { error = xfs_iunlink_remove(tp, sip); if (error) - goto abort_return; + goto error_return; } error = xfs_dir_createname(tp, tdp, target_name, sip->i_ino, &first_block, &free_list, resblks); if (error) - goto abort_return; + goto error_return; xfs_trans_ichgtime(tp, tdp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); xfs_trans_log_inode(tp, tdp, XFS_ILOG_CORE); error = xfs_bumplink(tp, sip); if (error) - goto abort_return; + goto error_return; /* * If this is a synchronous mount, make sure that the @@ -1512,15 +1462,13 @@ xfs_link( error = xfs_bmap_finish (&tp, &free_list, &committed); if (error) { xfs_bmap_cancel(&free_list); - goto abort_return; + goto error_return; } - return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + return xfs_trans_commit(tp); - abort_return: - cancel_flags |= XFS_TRANS_ABORT; error_return: - xfs_trans_cancel(tp, cancel_flags); + xfs_trans_cancel(tp); std_return: return error; } @@ -1555,7 +1503,6 @@ xfs_itruncate_extents( { struct xfs_mount *mp = ip->i_mount; struct xfs_trans *tp = *tpp; - struct xfs_trans *ntp; xfs_bmap_free_t free_list; xfs_fsblock_t first_block; xfs_fileoff_t first_unmap_block; @@ -1613,29 +1560,7 @@ xfs_itruncate_extents( if (error) goto out_bmap_cancel; - if (committed) { - /* - * Mark the inode dirty so it will be logged and - * moved forward in the log as part of every commit. - */ - xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - } - - ntp = xfs_trans_dup(tp); - error = xfs_trans_commit(tp, 0); - tp = ntp; - - xfs_trans_ijoin(tp, ip, 0); - - if (error) - goto out; - - /* - * Transaction commit worked ok so we can drop the extra ticket - * reference that we gained in xfs_trans_dup() - */ - xfs_log_ticket_put(tp->t_ticket); - error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0); + error = xfs_trans_roll(&tp, ip); if (error) goto out; } @@ -1756,7 +1681,7 @@ xfs_inactive_truncate( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0); if (error) { ASSERT(XFS_FORCED_SHUTDOWN(mp)); - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -1777,7 +1702,7 @@ xfs_inactive_truncate( ASSERT(ip->i_d.di_nextents == 0); - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto error_unlock; @@ -1785,7 +1710,7 @@ xfs_inactive_truncate( return 0; error_trans_cancel: - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); error_unlock: xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; @@ -1835,7 +1760,7 @@ xfs_inactive_ifree( } else { ASSERT(XFS_FORCED_SHUTDOWN(mp)); } - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES); + xfs_trans_cancel(tp); return error; } @@ -1855,7 +1780,7 @@ xfs_inactive_ifree( __func__, error); xfs_force_shutdown(mp, SHUTDOWN_META_IO_ERROR); } - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT); + xfs_trans_cancel(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; } @@ -1874,7 +1799,7 @@ xfs_inactive_ifree( if (error) xfs_notice(mp, "%s: xfs_bmap_finish returned error %d", __func__, error); - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) xfs_notice(mp, "%s: xfs_trans_commit returned error %d", __func__, error); @@ -1946,21 +1871,17 @@ xfs_inactive( /* * If there are attributes associated with the file then blow them away * now. The code calls a routine that recursively deconstructs the - * attribute fork. We need to just commit the current transaction - * because we can't use it for xfs_attr_inactive(). + * attribute fork. If also blows away the in-core attribute fork. */ - if (ip->i_d.di_anextents > 0) { - ASSERT(ip->i_d.di_forkoff != 0); - + if (XFS_IFORK_Q(ip)) { error = xfs_attr_inactive(ip); if (error) return; } - if (ip->i_afp) - xfs_idestroy_fork(ip, XFS_ATTR_FORK); - + ASSERT(!ip->i_afp); ASSERT(ip->i_d.di_anextents == 0); + ASSERT(ip->i_d.di_forkoff == 0); /* * Free the inode. @@ -2239,28 +2160,42 @@ xfs_iunlink_remove( */ STATIC int xfs_ifree_cluster( - xfs_inode_t *free_ip, - xfs_trans_t *tp, - xfs_ino_t inum) + xfs_inode_t *free_ip, + xfs_trans_t *tp, + struct xfs_icluster *xic) { xfs_mount_t *mp = free_ip->i_mount; int blks_per_cluster; int inodes_per_cluster; int nbufs; int i, j; + int ioffset; xfs_daddr_t blkno; xfs_buf_t *bp; xfs_inode_t *ip; xfs_inode_log_item_t *iip; xfs_log_item_t *lip; struct xfs_perag *pag; + xfs_ino_t inum; + inum = xic->first_ino; pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, inum)); blks_per_cluster = xfs_icluster_size_fsb(mp); inodes_per_cluster = blks_per_cluster << mp->m_sb.sb_inopblog; nbufs = mp->m_ialloc_blks / blks_per_cluster; for (j = 0; j < nbufs; j++, inum += inodes_per_cluster) { + /* + * The allocation bitmap tells us which inodes of the chunk were + * physically allocated. Skip the cluster if an inode falls into + * a sparse region. + */ + ioffset = inum - xic->first_ino; + if ((xic->alloc & XFS_INOBT_MASK(ioffset)) == 0) { + ASSERT(do_mod(ioffset, inodes_per_cluster) == 0); + continue; + } + blkno = XFS_AGB_TO_DADDR(mp, XFS_INO_TO_AGNO(mp, inum), XFS_INO_TO_AGBNO(mp, inum)); @@ -2418,8 +2353,7 @@ xfs_ifree( xfs_bmap_free_t *flist) { int error; - int delete; - xfs_ino_t first_ino; + struct xfs_icluster xic = { 0 }; ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); ASSERT(ip->i_d.di_nlink == 0); @@ -2435,7 +2369,7 @@ xfs_ifree( if (error) return error; - error = xfs_difree(tp, ip->i_ino, flist, &delete, &first_ino); + error = xfs_difree(tp, ip->i_ino, flist, &xic); if (error) return error; @@ -2452,8 +2386,8 @@ xfs_ifree( ip->i_d.di_gen++; xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - if (delete) - error = xfs_ifree_cluster(ip, tp, first_ino); + if (xic.deleted) + error = xfs_ifree_cluster(ip, tp, &xic); return error; } @@ -2540,7 +2474,6 @@ xfs_remove( int error = 0; xfs_bmap_free_t free_list; xfs_fsblock_t first_block; - int cancel_flags; int committed; uint resblks; @@ -2561,7 +2494,6 @@ xfs_remove( tp = xfs_trans_alloc(mp, XFS_TRANS_RMDIR); else tp = xfs_trans_alloc(mp, XFS_TRANS_REMOVE); - cancel_flags = XFS_TRANS_RELEASE_LOG_RES; /* * We try to get the real space reservation first, @@ -2580,7 +2512,6 @@ xfs_remove( } if (error) { ASSERT(error != -ENOSPC); - cancel_flags = 0; goto out_trans_cancel; } @@ -2592,7 +2523,6 @@ xfs_remove( /* * If we're removing a directory perform some additional validation. */ - cancel_flags |= XFS_TRANS_ABORT; if (is_dir) { ASSERT(ip->i_d.di_nlink >= 2); if (ip->i_d.di_nlink != 2) { @@ -2648,7 +2578,7 @@ xfs_remove( if (error) goto out_bmap_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto std_return; @@ -2660,7 +2590,7 @@ xfs_remove( out_bmap_cancel: xfs_bmap_cancel(&free_list); out_trans_cancel: - xfs_trans_cancel(tp, cancel_flags); + xfs_trans_cancel(tp); std_return: return error; } @@ -2734,11 +2664,11 @@ xfs_finish_rename( error = xfs_bmap_finish(&tp, free_list, &committed); if (error) { xfs_bmap_cancel(free_list); - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT); + xfs_trans_cancel(tp); return error; } - return xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + return xfs_trans_commit(tp); } /* @@ -2859,7 +2789,7 @@ xfs_cross_rename( out_trans_abort: xfs_bmap_cancel(free_list); - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES|XFS_TRANS_ABORT); + xfs_trans_cancel(tp); return error; } @@ -2883,7 +2813,13 @@ xfs_rename_alloc_whiteout( if (error) return error; - /* Satisfy xfs_bumplink that this is a real tmpfile */ + /* + * Prepare the tmpfile inode as if it were created through the VFS. + * Otherwise, the link increment paths will complain about nlink 0->1. + * Drop the link count as done by d_tmpfile(), complete the inode setup + * and flag it as linkable. + */ + drop_nlink(VFS_I(tmpfile)); xfs_finish_inode_setup(tmpfile); VFS_I(tmpfile)->i_state |= I_LINKABLE; @@ -2913,7 +2849,6 @@ xfs_rename( int num_inodes = __XFS_SORT_INODES; bool new_parent = (src_dp != target_dp); bool src_is_directory = S_ISDIR(src_ip->i_d.di_mode); - int cancel_flags = 0; int spaceres; int error; @@ -2949,7 +2884,6 @@ xfs_rename( } if (error) goto out_trans_cancel; - cancel_flags = XFS_TRANS_RELEASE_LOG_RES; /* * Attach the dquots to the inodes @@ -3020,10 +2954,8 @@ xfs_rename( error = xfs_dir_createname(tp, target_dp, target_name, src_ip->i_ino, &first_block, &free_list, spaceres); - if (error == -ENOSPC) - goto out_bmap_cancel; if (error) - goto out_trans_abort; + goto out_bmap_cancel; xfs_trans_ichgtime(tp, target_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); @@ -3031,7 +2963,7 @@ xfs_rename( if (new_parent && src_is_directory) { error = xfs_bumplink(tp, target_dp); if (error) - goto out_trans_abort; + goto out_bmap_cancel; } } else { /* target_ip != NULL */ /* @@ -3063,7 +2995,7 @@ xfs_rename( src_ip->i_ino, &first_block, &free_list, spaceres); if (error) - goto out_trans_abort; + goto out_bmap_cancel; xfs_trans_ichgtime(tp, target_dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); @@ -3074,7 +3006,7 @@ xfs_rename( */ error = xfs_droplink(tp, target_ip); if (error) - goto out_trans_abort; + goto out_bmap_cancel; if (src_is_directory) { /* @@ -3082,7 +3014,7 @@ xfs_rename( */ error = xfs_droplink(tp, target_ip); if (error) - goto out_trans_abort; + goto out_bmap_cancel; } } /* target_ip != NULL */ @@ -3099,7 +3031,7 @@ xfs_rename( &first_block, &free_list, spaceres); ASSERT(error != -EEXIST); if (error) - goto out_trans_abort; + goto out_bmap_cancel; } /* @@ -3125,7 +3057,7 @@ xfs_rename( */ error = xfs_droplink(tp, src_dp); if (error) - goto out_trans_abort; + goto out_bmap_cancel; } /* @@ -3140,7 +3072,7 @@ xfs_rename( error = xfs_dir_removename(tp, src_dp, src_name, src_ip->i_ino, &first_block, &free_list, spaceres); if (error) - goto out_trans_abort; + goto out_bmap_cancel; /* * For whiteouts, we need to bump the link count on the whiteout inode. @@ -3151,13 +3083,13 @@ xfs_rename( * intermediate state on disk. */ if (wip) { - ASSERT(wip->i_d.di_nlink == 0); + ASSERT(VFS_I(wip)->i_nlink == 0 && wip->i_d.di_nlink == 0); error = xfs_bumplink(tp, wip); if (error) - goto out_trans_abort; + goto out_bmap_cancel; error = xfs_iunlink_remove(tp, wip); if (error) - goto out_trans_abort; + goto out_bmap_cancel; xfs_trans_log_inode(tp, wip, XFS_ILOG_CORE); /* @@ -3178,12 +3110,10 @@ xfs_rename( IRELE(wip); return error; -out_trans_abort: - cancel_flags |= XFS_TRANS_ABORT; out_bmap_cancel: xfs_bmap_cancel(&free_list); out_trans_cancel: - xfs_trans_cancel(tp, cancel_flags); + xfs_trans_cancel(tp); if (wip) IRELE(wip); return error; @@ -3462,7 +3392,7 @@ xfs_iflush_int( ASSERT(ip->i_d.di_version > 1); /* set *dip = inode's place in the buffer */ - dip = (xfs_dinode_t *)xfs_buf_offset(bp, ip->i_imap.im_boffset); + dip = xfs_buf_offset(bp, ip->i_imap.im_boffset); if (XFS_TEST_ERROR(dip->di_magic != cpu_to_be16(XFS_DINODE_MAGIC), mp, XFS_ERRTAG_IFLUSH_1, XFS_RANDOM_IFLUSH_1)) { diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c index 87f67c6..ea7d85a 100644 --- a/fs/xfs/xfs_ioctl.c +++ b/fs/xfs/xfs_ioctl.c @@ -336,7 +336,7 @@ xfs_set_dmattrs( tp = xfs_trans_alloc(mp, XFS_TRANS_SET_DMATTRS); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } xfs_ilock(ip, XFS_ILOCK_EXCL); @@ -346,7 +346,7 @@ xfs_set_dmattrs( ip->i_d.di_dmstate = state; xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); return error; } @@ -1076,7 +1076,7 @@ xfs_ioctl_setattr_get_trans( return tp; out_cancel: - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return ERR_PTR(error); } @@ -1253,7 +1253,7 @@ xfs_ioctl_setattr( else ip->i_d.di_extsize = 0; - code = xfs_trans_commit(tp, 0); + code = xfs_trans_commit(tp); /* * Release any dquot(s) the inode had kept before chown. @@ -1265,7 +1265,7 @@ xfs_ioctl_setattr( return code; error_trans_cancel: - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); error_free_dquots: xfs_qm_dqrele(udqp); xfs_qm_dqrele(pdqp); @@ -1338,11 +1338,11 @@ xfs_ioc_setxflags( error = xfs_ioctl_setattr_xflags(tp, ip, &fa); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); goto out_drop_write; } - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); out_drop_write: mnt_drop_write_file(filp); return error; diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 38e633b..1f86033 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -183,7 +183,7 @@ xfs_iomap_write_direct( * Check for running out of space, note: need lock to return */ if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -213,7 +213,7 @@ xfs_iomap_write_direct( error = xfs_bmap_finish(&tp, &free_list, &committed); if (error) goto out_bmap_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto out_unlock; @@ -236,7 +236,7 @@ out_bmap_cancel: xfs_bmap_cancel(&free_list); xfs_trans_unreserve_quota_nblks(tp, ip, (long)qblocks, 0, quota_flag); out_trans_cancel: - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); goto out_unlock; } @@ -690,7 +690,7 @@ xfs_iomap_write_allocate( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write, nres, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } xfs_ilock(ip, XFS_ILOCK_EXCL); @@ -760,7 +760,7 @@ xfs_iomap_write_allocate( if (error) goto trans_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto error0; @@ -791,7 +791,7 @@ xfs_iomap_write_allocate( trans_cancel: xfs_bmap_cancel(&free_list); - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); error0: xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; @@ -853,7 +853,7 @@ xfs_iomap_write_unwritten( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_write, resblks, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -890,7 +890,7 @@ xfs_iomap_write_unwritten( if (error) goto error_on_bmapi_transaction; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); if (error) return error; @@ -914,7 +914,7 @@ xfs_iomap_write_unwritten( error_on_bmapi_transaction: xfs_bmap_cancel(&free_list); - xfs_trans_cancel(tp, (XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT)); + xfs_trans_cancel(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; } diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index 7f51f39..766b23f 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -699,7 +699,7 @@ xfs_setattr_nonsize( if (mp->m_flags & XFS_MOUNT_WSYNC) xfs_trans_set_sync(tp); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); @@ -730,7 +730,7 @@ xfs_setattr_nonsize( return 0; out_trans_cancel: - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); xfs_iunlock(ip, XFS_ILOCK_EXCL); out_dqrele: xfs_qm_dqrele(udqp); @@ -752,7 +752,6 @@ xfs_setattr_size( struct xfs_trans *tp; int error; uint lock_flags = 0; - uint commit_flags = 0; bool did_zeroing = false; trace_xfs_setattr(ip); @@ -848,7 +847,11 @@ xfs_setattr_size( * to hope that the caller sees ENOMEM and retries the truncate * operation. */ - error = block_truncate_page(inode->i_mapping, newsize, xfs_get_blocks); + if (IS_DAX(inode)) + error = dax_truncate_page(inode, newsize, xfs_get_blocks_direct); + else + error = block_truncate_page(inode->i_mapping, newsize, + xfs_get_blocks); if (error) return error; truncate_setsize(inode, newsize); @@ -858,7 +861,6 @@ xfs_setattr_size( if (error) goto out_trans_cancel; - commit_flags = XFS_TRANS_RELEASE_LOG_RES; lock_flags |= XFS_ILOCK_EXCL; xfs_ilock(ip, XFS_ILOCK_EXCL); xfs_trans_ijoin(tp, ip, 0); @@ -898,7 +900,7 @@ xfs_setattr_size( if (newsize <= oldsize) { error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, newsize); if (error) - goto out_trans_abort; + goto out_trans_cancel; /* * Truncated "down", so we're removing references to old data @@ -925,16 +927,14 @@ xfs_setattr_size( if (mp->m_flags & XFS_MOUNT_WSYNC) xfs_trans_set_sync(tp); - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); out_unlock: if (lock_flags) xfs_iunlock(ip, lock_flags); return error; -out_trans_abort: - commit_flags |= XFS_TRANS_ABORT; out_trans_cancel: - xfs_trans_cancel(tp, commit_flags); + xfs_trans_cancel(tp); goto out_unlock; } @@ -981,7 +981,7 @@ xfs_vn_update_time( tp = xfs_trans_alloc(mp, XFS_TRANS_FSYNC_TS); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_fsyncts, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -1003,7 +1003,7 @@ xfs_vn_update_time( } xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); xfs_trans_log_inode(tp, ip, XFS_ILOG_TIMESTAMP); - return xfs_trans_commit(tp, 0); + return xfs_trans_commit(tp); } #define XFS_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) @@ -1188,22 +1188,22 @@ xfs_diflags_to_iflags( struct inode *inode, struct xfs_inode *ip) { - if (ip->i_d.di_flags & XFS_DIFLAG_IMMUTABLE) + uint16_t flags = ip->i_d.di_flags; + + inode->i_flags &= ~(S_IMMUTABLE | S_APPEND | S_SYNC | + S_NOATIME | S_DAX); + + if (flags & XFS_DIFLAG_IMMUTABLE) inode->i_flags |= S_IMMUTABLE; - else - inode->i_flags &= ~S_IMMUTABLE; - if (ip->i_d.di_flags & XFS_DIFLAG_APPEND) + if (flags & XFS_DIFLAG_APPEND) inode->i_flags |= S_APPEND; - else - inode->i_flags &= ~S_APPEND; - if (ip->i_d.di_flags & XFS_DIFLAG_SYNC) + if (flags & XFS_DIFLAG_SYNC) inode->i_flags |= S_SYNC; - else - inode->i_flags &= ~S_SYNC; - if (ip->i_d.di_flags & XFS_DIFLAG_NOATIME) + if (flags & XFS_DIFLAG_NOATIME) inode->i_flags |= S_NOATIME; - else - inode->i_flags &= ~S_NOATIME; + /* XXX: Also needs an on-disk per inode flag! */ + if (ip->i_mount->m_flags & XFS_MOUNT_DAX) + inode->i_flags |= S_DAX; } /* diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index 8042989..f41b0c3 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -252,7 +252,7 @@ xfs_bulkstat_grab_ichunk( } irec->ir_free |= xfs_inobt_maskn(0, idx); - *icount = XFS_INODES_PER_CHUNK - irec->ir_freecount; + *icount = irec->ir_count - irec->ir_freecount; } return 0; @@ -415,6 +415,8 @@ xfs_bulkstat( goto del_cursor; if (icount) { irbp->ir_startino = r.ir_startino; + irbp->ir_holemask = r.ir_holemask; + irbp->ir_count = r.ir_count; irbp->ir_freecount = r.ir_freecount; irbp->ir_free = r.ir_free; irbp++; @@ -447,13 +449,15 @@ xfs_bulkstat( * If this chunk has any allocated inodes, save it. * Also start read-ahead now for this chunk. */ - if (r.ir_freecount < XFS_INODES_PER_CHUNK) { + if (r.ir_freecount < r.ir_count) { xfs_bulkstat_ichunk_ra(mp, agno, &r); irbp->ir_startino = r.ir_startino; + irbp->ir_holemask = r.ir_holemask; + irbp->ir_count = r.ir_count; irbp->ir_freecount = r.ir_freecount; irbp->ir_free = r.ir_free; irbp++; - icount += XFS_INODES_PER_CHUNK - r.ir_freecount; + icount += r.ir_count - r.ir_freecount; } error = xfs_btree_increment(cur, 0, &stat); if (error || stat == 0) { @@ -599,8 +603,7 @@ xfs_inumbers( agino = r.ir_startino + XFS_INODES_PER_CHUNK - 1; buffer[bufidx].xi_startino = XFS_AGINO_TO_INO(mp, agno, r.ir_startino); - buffer[bufidx].xi_alloccount = - XFS_INODES_PER_CHUNK - r.ir_freecount; + buffer[bufidx].xi_alloccount = r.ir_count - r.ir_freecount; buffer[bufidx].xi_allocmask = ~r.ir_free; if (++bufidx == bcount) { long written; diff --git a/fs/xfs/xfs_linux.h b/fs/xfs/xfs_linux.h index 7c7842c..85f883d 100644 --- a/fs/xfs/xfs_linux.h +++ b/fs/xfs/xfs_linux.h @@ -32,26 +32,12 @@ typedef unsigned int __uint32_t; typedef signed long long int __int64_t; typedef unsigned long long int __uint64_t; -typedef __uint32_t inst_t; /* an instruction */ - typedef __s64 xfs_off_t; /* <file offset> type */ typedef unsigned long long xfs_ino_t; /* <inode> type */ typedef __s64 xfs_daddr_t; /* <disk address> type */ -typedef char * xfs_caddr_t; /* <core address> type */ typedef __u32 xfs_dev_t; typedef __u32 xfs_nlink_t; -/* __psint_t is the same size as a pointer */ -#if (BITS_PER_LONG == 32) -typedef __int32_t __psint_t; -typedef __uint32_t __psunsigned_t; -#elif (BITS_PER_LONG == 64) -typedef __int64_t __psint_t; -typedef __uint64_t __psunsigned_t; -#else -#error BITS_PER_LONG must be 32 or 64 -#endif - #include "xfs_types.h" #include "kmem.h" diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index bcc7cfa..08d4fe4 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -109,7 +109,7 @@ xlog_ungrant_log_space( STATIC void xlog_verify_dest_ptr( struct xlog *log, - char *ptr); + void *ptr); STATIC void xlog_verify_grant_tail( struct xlog *log); @@ -513,7 +513,7 @@ xfs_log_done( struct xfs_mount *mp, struct xlog_ticket *ticket, struct xlog_in_core **iclog, - uint flags) + bool regrant) { struct xlog *log = mp->m_log; xfs_lsn_t lsn = 0; @@ -526,14 +526,11 @@ xfs_log_done( (((ticket->t_flags & XLOG_TIC_INITED) == 0) && (xlog_commit_record(log, ticket, iclog, &lsn)))) { lsn = (xfs_lsn_t) -1; - if (ticket->t_flags & XLOG_TIC_PERM_RESERV) { - flags |= XFS_LOG_REL_PERM_RESERV; - } + regrant = false; } - if ((ticket->t_flags & XLOG_TIC_PERM_RESERV) == 0 || - (flags & XFS_LOG_REL_PERM_RESERV)) { + if (!regrant) { trace_xfs_log_done_nonperm(log, ticket); /* @@ -541,7 +538,6 @@ xfs_log_done( * request has been made to release a permanent reservation. */ xlog_ungrant_log_space(log, ticket); - xfs_log_ticket_put(ticket); } else { trace_xfs_log_done_perm(log, ticket); @@ -553,6 +549,7 @@ xfs_log_done( ticket->t_flags |= XLOG_TIC_INITED; } + xfs_log_ticket_put(ticket); return lsn; } @@ -1447,7 +1444,7 @@ xlog_alloc_log( iclog->ic_bp = bp; iclog->ic_data = bp->b_addr; #ifdef DEBUG - log->l_iclog_bak[i] = (xfs_caddr_t)&(iclog->ic_header); + log->l_iclog_bak[i] = &iclog->ic_header; #endif head = &iclog->ic_header; memset(head, 0, sizeof(xlog_rec_header_t)); @@ -1602,7 +1599,7 @@ xlog_pack_data( int i, j, k; int size = iclog->ic_offset + roundoff; __be32 cycle_lsn; - xfs_caddr_t dp; + char *dp; cycle_lsn = CYCLE_LSN_DISK(iclog->ic_header.h_lsn); @@ -3664,7 +3661,7 @@ xlog_ticket_alloc( void xlog_verify_dest_ptr( struct xlog *log, - char *ptr) + void *ptr) { int i; int good_ptr = 0; @@ -3767,9 +3764,8 @@ xlog_verify_iclog( xlog_op_header_t *ophead; xlog_in_core_t *icptr; xlog_in_core_2_t *xhdr; - xfs_caddr_t ptr; - xfs_caddr_t base_ptr; - __psint_t field_offset; + void *base_ptr, *ptr, *p; + ptrdiff_t field_offset; __uint8_t clientid; int len, i, j, k, op_len; int idx; @@ -3788,9 +3784,9 @@ xlog_verify_iclog( if (iclog->ic_header.h_magicno != cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) xfs_emerg(log->l_mp, "%s: invalid magic num", __func__); - ptr = (xfs_caddr_t) &iclog->ic_header; - for (ptr += BBSIZE; ptr < ((xfs_caddr_t)&iclog->ic_header) + count; - ptr += BBSIZE) { + base_ptr = ptr = &iclog->ic_header; + p = &iclog->ic_header; + for (ptr += BBSIZE; ptr < base_ptr + count; ptr += BBSIZE) { if (*(__be32 *)ptr == cpu_to_be32(XLOG_HEADER_MAGIC_NUM)) xfs_emerg(log->l_mp, "%s: unexpected magic num", __func__); @@ -3798,20 +3794,19 @@ xlog_verify_iclog( /* check fields */ len = be32_to_cpu(iclog->ic_header.h_num_logops); - ptr = iclog->ic_datap; - base_ptr = ptr; - ophead = (xlog_op_header_t *)ptr; + base_ptr = ptr = iclog->ic_datap; + ophead = ptr; xhdr = iclog->ic_data; for (i = 0; i < len; i++) { - ophead = (xlog_op_header_t *)ptr; + ophead = ptr; /* clientid is only 1 byte */ - field_offset = (__psint_t) - ((xfs_caddr_t)&(ophead->oh_clientid) - base_ptr); + p = &ophead->oh_clientid; + field_offset = p - base_ptr; if (!syncing || (field_offset & 0x1ff)) { clientid = ophead->oh_clientid; } else { - idx = BTOBBT((xfs_caddr_t)&(ophead->oh_clientid) - iclog->ic_datap); + idx = BTOBBT((char *)&ophead->oh_clientid - iclog->ic_datap); if (idx >= (XLOG_HEADER_CYCLE_SIZE / BBSIZE)) { j = idx / (XLOG_HEADER_CYCLE_SIZE / BBSIZE); k = idx % (XLOG_HEADER_CYCLE_SIZE / BBSIZE); @@ -3829,13 +3824,13 @@ xlog_verify_iclog( (unsigned long)field_offset); /* check length */ - field_offset = (__psint_t) - ((xfs_caddr_t)&(ophead->oh_len) - base_ptr); + p = &ophead->oh_len; + field_offset = p - base_ptr; if (!syncing || (field_offset & 0x1ff)) { op_len = be32_to_cpu(ophead->oh_len); } else { - idx = BTOBBT((__psint_t)&ophead->oh_len - - (__psint_t)iclog->ic_datap); + idx = BTOBBT((uintptr_t)&ophead->oh_len - + (uintptr_t)iclog->ic_datap); if (idx >= (XLOG_HEADER_CYCLE_SIZE / BBSIZE)) { j = idx / (XLOG_HEADER_CYCLE_SIZE / BBSIZE); k = idx % (XLOG_HEADER_CYCLE_SIZE / BBSIZE); diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h index 84e0deb..fa27aae 100644 --- a/fs/xfs/xfs_log.h +++ b/fs/xfs/xfs_log.h @@ -111,15 +111,6 @@ static inline xfs_lsn_t _lsn_cmp(xfs_lsn_t lsn1, xfs_lsn_t lsn2) #define XFS_LSN_CMP(x,y) _lsn_cmp(x,y) /* - * Macros, structures, prototypes for interface to the log manager. - */ - -/* - * Flags to xfs_log_done() - */ -#define XFS_LOG_REL_PERM_RESERV 0x1 - -/* * Flags to xfs_log_force() * * XFS_LOG_SYNC: Synchronous force in-core log to disk @@ -138,7 +129,7 @@ struct xfs_log_callback; xfs_lsn_t xfs_log_done(struct xfs_mount *mp, struct xlog_ticket *ticket, struct xlog_in_core **iclog, - uint flags); + bool regrant); int _xfs_log_force(struct xfs_mount *mp, uint flags, int *log_forced); @@ -183,7 +174,7 @@ struct xlog_ticket *xfs_log_ticket_get(struct xlog_ticket *ticket); void xfs_log_ticket_put(struct xlog_ticket *ticket); void xfs_log_commit_cil(struct xfs_mount *mp, struct xfs_trans *tp, - xfs_lsn_t *commit_lsn, int flags); + xfs_lsn_t *commit_lsn, bool regrant); bool xfs_log_item_in_current_chkpt(struct xfs_log_item *lip); void xfs_log_work_queue(struct xfs_mount *mp); diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c index 45cc0ce..abc2ccb 100644 --- a/fs/xfs/xfs_log_cil.c +++ b/fs/xfs/xfs_log_cil.c @@ -624,7 +624,7 @@ restart: spin_unlock(&cil->xc_push_lock); /* xfs_log_done always frees the ticket on error. */ - commit_lsn = xfs_log_done(log->l_mp, tic, &commit_iclog, 0); + commit_lsn = xfs_log_done(log->l_mp, tic, &commit_iclog, false); if (commit_lsn == -1) goto out_abort; @@ -773,14 +773,10 @@ xfs_log_commit_cil( struct xfs_mount *mp, struct xfs_trans *tp, xfs_lsn_t *commit_lsn, - int flags) + bool regrant) { struct xlog *log = mp->m_log; struct xfs_cil *cil = log->l_cilp; - int log_flags = 0; - - if (flags & XFS_TRANS_RELEASE_LOG_RES) - log_flags = XFS_LOG_REL_PERM_RESERV; /* lock out background commit */ down_read(&cil->xc_ctx_lock); @@ -795,7 +791,7 @@ xfs_log_commit_cil( if (commit_lsn) *commit_lsn = tp->t_commit_lsn; - xfs_log_done(mp, tp->t_ticket, NULL, log_flags); + xfs_log_done(mp, tp->t_ticket, NULL, regrant); xfs_trans_unreserve_and_mod_sb(tp); /* @@ -809,7 +805,7 @@ xfs_log_commit_cil( * the log items. This affects (at least) processing of stale buffers, * inodes and EFIs. */ - xfs_trans_free_items(tp, tp->t_commit_lsn, 0); + xfs_trans_free_items(tp, tp->t_commit_lsn, false); xlog_cil_push_background(log); diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h index db7cbde..1c87c8a 100644 --- a/fs/xfs/xfs_log_priv.h +++ b/fs/xfs/xfs_log_priv.h @@ -409,7 +409,7 @@ struct xlog { /* The following field are used for debugging; need to hold icloglock */ #ifdef DEBUG - char *l_iclog_bak[XLOG_MAX_ICLOGS]; + void *l_iclog_bak[XLOG_MAX_ICLOGS]; #endif }; diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 4f5784f..01dd228 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -147,7 +147,7 @@ xlog_put_bp( * Return the address of the start of the given block number's data * in a log buffer. The buffer covers a log sector-aligned region. */ -STATIC xfs_caddr_t +STATIC char * xlog_align( struct xlog *log, xfs_daddr_t blk_no, @@ -203,7 +203,7 @@ xlog_bread( xfs_daddr_t blk_no, int nbblks, struct xfs_buf *bp, - xfs_caddr_t *offset) + char **offset) { int error; @@ -225,9 +225,9 @@ xlog_bread_offset( xfs_daddr_t blk_no, /* block to read from */ int nbblks, /* blocks to read */ struct xfs_buf *bp, - xfs_caddr_t offset) + char *offset) { - xfs_caddr_t orig_offset = bp->b_addr; + char *orig_offset = bp->b_addr; int orig_len = BBTOB(bp->b_length); int error, error2; @@ -396,7 +396,7 @@ xlog_find_cycle_start( xfs_daddr_t *last_blk, uint cycle) { - xfs_caddr_t offset; + char *offset; xfs_daddr_t mid_blk; xfs_daddr_t end_blk; uint mid_cycle; @@ -443,7 +443,7 @@ xlog_find_verify_cycle( uint cycle; xfs_buf_t *bp; xfs_daddr_t bufblks; - xfs_caddr_t buf = NULL; + char *buf = NULL; int error = 0; /* @@ -509,7 +509,7 @@ xlog_find_verify_log_record( { xfs_daddr_t i; xfs_buf_t *bp; - xfs_caddr_t offset = NULL; + char *offset = NULL; xlog_rec_header_t *head = NULL; int error = 0; int smallmem = 0; @@ -616,7 +616,7 @@ xlog_find_head( xfs_daddr_t *return_head_blk) { xfs_buf_t *bp; - xfs_caddr_t offset; + char *offset; xfs_daddr_t new_blk, first_blk, start_blk, last_blk, head_blk; int num_scan_bblks; uint first_half_cycle, last_half_cycle; @@ -891,7 +891,7 @@ xlog_find_tail( { xlog_rec_header_t *rhead; xlog_op_header_t *op_head; - xfs_caddr_t offset = NULL; + char *offset = NULL; xfs_buf_t *bp; int error, i, found; xfs_daddr_t umount_data_blk; @@ -1099,7 +1099,7 @@ xlog_find_zeroed( xfs_daddr_t *blk_no) { xfs_buf_t *bp; - xfs_caddr_t offset; + char *offset; uint first_cycle, last_cycle; xfs_daddr_t new_blk, last_blk, start_blk; xfs_daddr_t num_scan_bblks; @@ -1199,7 +1199,7 @@ bp_err: STATIC void xlog_add_record( struct xlog *log, - xfs_caddr_t buf, + char *buf, int cycle, int block, int tail_cycle, @@ -1227,7 +1227,7 @@ xlog_write_log_records( int tail_cycle, int tail_block) { - xfs_caddr_t offset; + char *offset; xfs_buf_t *bp; int balign, ealign; int sectbb = log->l_sectBBsize; @@ -1789,8 +1789,7 @@ xlog_recover_do_inode_buffer( return -EFSCORRUPTED; } - buffer_nextp = (xfs_agino_t *)xfs_buf_offset(bp, - next_unlinked_offset); + buffer_nextp = xfs_buf_offset(bp, next_unlinked_offset); *buffer_nextp = *logged_nextp; /* @@ -1798,7 +1797,7 @@ xlog_recover_do_inode_buffer( * have to leave the inode in a consistent state for whoever * reads it next.... */ - xfs_dinode_calc_crc(mp, (struct xfs_dinode *) + xfs_dinode_calc_crc(mp, xfs_buf_offset(bp, i * mp->m_sb.sb_inodesize)); } @@ -2503,8 +2502,8 @@ xlog_recover_inode_pass2( xfs_buf_t *bp; xfs_dinode_t *dip; int len; - xfs_caddr_t src; - xfs_caddr_t dest; + char *src; + char *dest; int error; int attr_index; uint fields; @@ -2546,7 +2545,7 @@ xlog_recover_inode_pass2( goto out_release; } ASSERT(in_f->ilf_fields & XFS_ILOG_CORE); - dip = (xfs_dinode_t *)xfs_buf_offset(bp, in_f->ilf_boffset); + dip = xfs_buf_offset(bp, in_f->ilf_boffset); /* * Make sure the place we're flushing out to really looks @@ -2885,7 +2884,7 @@ xlog_recover_dquot_pass2( return error; ASSERT(bp); - ddq = (xfs_disk_dquot_t *)xfs_buf_offset(bp, dq_f->qlf_boffset); + ddq = xfs_buf_offset(bp, dq_f->qlf_boffset); /* * If the dquot has an LSN in it, recover the dquot only if it's less @@ -3068,12 +3067,22 @@ xlog_recover_do_icreate_pass2( return -EINVAL; } - /* existing allocation is fixed value */ - ASSERT(count == mp->m_ialloc_inos); - ASSERT(length == mp->m_ialloc_blks); - if (count != mp->m_ialloc_inos || - length != mp->m_ialloc_blks) { - xfs_warn(log->l_mp, "xlog_recover_do_icreate_trans: bad count 2"); + /* + * The inode chunk is either full or sparse and we only support + * m_ialloc_min_blks sized sparse allocations at this time. + */ + if (length != mp->m_ialloc_blks && + length != mp->m_ialloc_min_blks) { + xfs_warn(log->l_mp, + "%s: unsupported chunk length", __FUNCTION__); + return -EINVAL; + } + + /* verify inode count is consistent with extent length */ + if ((count >> mp->m_sb.sb_inopblog) != length) { + xfs_warn(log->l_mp, + "%s: inconsistent inode count and chunk length", + __FUNCTION__); return -EINVAL; } @@ -3091,8 +3100,8 @@ xlog_recover_do_icreate_pass2( XFS_AGB_TO_DADDR(mp, agno, agbno), length, 0)) return 0; - xfs_ialloc_inode_init(mp, NULL, buffer_list, agno, agbno, length, - be32_to_cpu(icl->icl_gen)); + xfs_ialloc_inode_init(mp, NULL, buffer_list, count, agno, agbno, length, + be32_to_cpu(icl->icl_gen)); return 0; } @@ -3364,17 +3373,17 @@ STATIC int xlog_recover_add_to_cont_trans( struct xlog *log, struct xlog_recover *trans, - xfs_caddr_t dp, + char *dp, int len) { xlog_recover_item_t *item; - xfs_caddr_t ptr, old_ptr; + char *ptr, *old_ptr; int old_len; if (list_empty(&trans->r_itemq)) { /* finish copying rest of trans header */ xlog_recover_add_item(&trans->r_itemq); - ptr = (xfs_caddr_t) &trans->r_theader + + ptr = (char *)&trans->r_theader + sizeof(xfs_trans_header_t) - len; memcpy(ptr, dp, len); return 0; @@ -3410,12 +3419,12 @@ STATIC int xlog_recover_add_to_trans( struct xlog *log, struct xlog_recover *trans, - xfs_caddr_t dp, + char *dp, int len) { xfs_inode_log_format_t *in_f; /* any will do */ xlog_recover_item_t *item; - xfs_caddr_t ptr; + char *ptr; if (!len) return 0; @@ -3504,7 +3513,7 @@ STATIC int xlog_recovery_process_trans( struct xlog *log, struct xlog_recover *trans, - xfs_caddr_t dp, + char *dp, unsigned int len, unsigned int flags, int pass) @@ -3611,8 +3620,8 @@ xlog_recover_process_ophdr( struct hlist_head rhash[], struct xlog_rec_header *rhead, struct xlog_op_header *ohead, - xfs_caddr_t dp, - xfs_caddr_t end, + char *dp, + char *end, int pass) { struct xlog_recover *trans; @@ -3661,11 +3670,11 @@ xlog_recover_process_data( struct xlog *log, struct hlist_head rhash[], struct xlog_rec_header *rhead, - xfs_caddr_t dp, + char *dp, int pass) { struct xlog_op_header *ohead; - xfs_caddr_t end; + char *end; int num_logops; int error; @@ -3751,11 +3760,11 @@ xlog_recover_process_efi( } set_bit(XFS_EFI_RECOVERED, &efip->efi_flags); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); return error; abort_error: - xfs_trans_cancel(tp, XFS_TRANS_ABORT); + xfs_trans_cancel(tp); return error; } @@ -3857,13 +3866,13 @@ xlog_recover_clear_agi_bucket( xfs_trans_log_buf(tp, agibp, offset, (offset + sizeof(xfs_agino_t) - 1)); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); if (error) goto out_error; return; out_abort: - xfs_trans_cancel(tp, XFS_TRANS_ABORT); + xfs_trans_cancel(tp); out_error: xfs_warn(mp, "%s: failed to clear agi %d. Continuing.", __func__, agno); return; @@ -4010,7 +4019,7 @@ xlog_recover_process_iunlinks( STATIC int xlog_unpack_data_crc( struct xlog_rec_header *rhead, - xfs_caddr_t dp, + char *dp, struct xlog *log) { __le32 crc; @@ -4040,7 +4049,7 @@ xlog_unpack_data_crc( STATIC int xlog_unpack_data( struct xlog_rec_header *rhead, - xfs_caddr_t dp, + char *dp, struct xlog *log) { int i, j, k; @@ -4122,7 +4131,7 @@ xlog_do_recovery_pass( { xlog_rec_header_t *rhead; xfs_daddr_t blk_no; - xfs_caddr_t offset; + char *offset; xfs_buf_t *hbp, *dbp; int error = 0, h_size; int bblks, split_bblks; diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index 2ce7ee3..461e791 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -725,6 +725,22 @@ xfs_mountfs( } /* + * If enabled, sparse inode chunk alignment is expected to match the + * cluster size. Full inode chunk alignment must match the chunk size, + * but that is checked on sb read verification... + */ + if (xfs_sb_version_hassparseinodes(&mp->m_sb) && + mp->m_sb.sb_spino_align != + XFS_B_TO_FSBT(mp, mp->m_inode_cluster_size)) { + xfs_warn(mp, + "Sparse inode block alignment (%u) must match cluster size (%llu).", + mp->m_sb.sb_spino_align, + XFS_B_TO_FSBT(mp, mp->m_inode_cluster_size)); + error = -EINVAL; + goto out_remove_uuid; + } + + /* * Set inode alignment fields */ xfs_set_inoalignment(mp); @@ -1084,14 +1100,18 @@ xfs_log_sbcount(xfs_mount_t *mp) return xfs_sync_sb(mp, true); } +/* + * Deltas for the inode count are +/-64, hence we use a large batch size + * of 128 so we don't need to take the counter lock on every update. + */ +#define XFS_ICOUNT_BATCH 128 int xfs_mod_icount( struct xfs_mount *mp, int64_t delta) { - /* deltas are +/-64, hence the large batch size of 128. */ - __percpu_counter_add(&mp->m_icount, delta, 128); - if (percpu_counter_compare(&mp->m_icount, 0) < 0) { + __percpu_counter_add(&mp->m_icount, delta, XFS_ICOUNT_BATCH); + if (__percpu_counter_compare(&mp->m_icount, 0, XFS_ICOUNT_BATCH) < 0) { ASSERT(0); percpu_counter_add(&mp->m_icount, -delta); return -EINVAL; @@ -1113,6 +1133,14 @@ xfs_mod_ifree( return 0; } +/* + * Deltas for the block count can vary from 1 to very large, but lock contention + * only occurs on frequent small block count updates such as in the delayed + * allocation path for buffered writes (page a time updates). Hence we set + * a large batch count (1024) to minimise global counter updates except when + * we get near to ENOSPC and we have to be very accurate with our updates. + */ +#define XFS_FDBLOCKS_BATCH 1024 int xfs_mod_fdblocks( struct xfs_mount *mp, @@ -1151,25 +1179,19 @@ xfs_mod_fdblocks( * Taking blocks away, need to be more accurate the closer we * are to zero. * - * batch size is set to a maximum of 1024 blocks - if we are - * allocating of freeing extents larger than this then we aren't - * going to be hammering the counter lock so a lock per update - * is not a problem. - * * If the counter has a value of less than 2 * max batch size, * then make everything serialise as we are real close to * ENOSPC. */ -#define __BATCH 1024 - if (percpu_counter_compare(&mp->m_fdblocks, 2 * __BATCH) < 0) + if (__percpu_counter_compare(&mp->m_fdblocks, 2 * XFS_FDBLOCKS_BATCH, + XFS_FDBLOCKS_BATCH) < 0) batch = 1; else - batch = __BATCH; -#undef __BATCH + batch = XFS_FDBLOCKS_BATCH; __percpu_counter_add(&mp->m_fdblocks, delta, batch); - if (percpu_counter_compare(&mp->m_fdblocks, - XFS_ALLOC_SET_ASIDE(mp)) >= 0) { + if (__percpu_counter_compare(&mp->m_fdblocks, XFS_ALLOC_SET_ASIDE(mp), + XFS_FDBLOCKS_BATCH) >= 0) { /* we had space! */ return 0; } diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 8c995a2..7999e91 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -101,6 +101,8 @@ typedef struct xfs_mount { __uint64_t m_flags; /* global mount flags */ int m_ialloc_inos; /* inodes in inode allocation */ int m_ialloc_blks; /* blocks in inode allocation */ + int m_ialloc_min_blks;/* min blocks in sparse inode + * allocation */ int m_inoalign_mask;/* mask sb_inoalignmt if used */ uint m_qflags; /* quota status flags */ struct xfs_trans_resv m_resv; /* precomputed res values */ @@ -179,6 +181,8 @@ typedef struct xfs_mount { allocator */ #define XFS_MOUNT_NOATTR2 (1ULL << 25) /* disable use of attr2 format */ +#define XFS_MOUNT_DAX (1ULL << 62) /* TEST ONLY! */ + /* * Default minimum read and write sizes. diff --git a/fs/xfs/xfs_pnfs.c b/fs/xfs/xfs_pnfs.c index 981a657..ab4a606 100644 --- a/fs/xfs/xfs_pnfs.c +++ b/fs/xfs/xfs_pnfs.c @@ -306,7 +306,7 @@ xfs_fs_commit_blocks( tp = xfs_trans_alloc(mp, XFS_TRANS_SETATTR_NOT_SIZE); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_ichange, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); goto out_drop_iolock; } @@ -321,7 +321,7 @@ xfs_fs_commit_blocks( } xfs_trans_set_sync(tp); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); out_drop_iolock: xfs_iunlock(ip, XFS_IOLOCK_EXCL); diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 5538468..eac9549 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -756,7 +756,7 @@ xfs_qm_qino_alloc( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_create, XFS_QM_QINOCREATE_SPACE_RES(mp), 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -764,8 +764,7 @@ xfs_qm_qino_alloc( error = xfs_dir_ialloc(&tp, NULL, S_IFREG, 1, 0, 0, 1, ip, &committed); if (error) { - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | - XFS_TRANS_ABORT); + xfs_trans_cancel(tp); return error; } } @@ -796,7 +795,7 @@ xfs_qm_qino_alloc( spin_unlock(&mp->m_sb_lock); xfs_log_sb(tp); - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) { ASSERT(XFS_FORCED_SHUTDOWN(mp)); xfs_alert(mp, "%s failed (error %d)!", __func__, error); diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 9a25c92..3640c6e 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -239,7 +239,7 @@ xfs_qm_scall_trunc_qfile( tp = xfs_trans_alloc(mp, XFS_TRANS_TRUNCATE_FILE); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); xfs_iunlock(ip, XFS_IOLOCK_EXCL); goto out_put; } @@ -252,15 +252,14 @@ xfs_qm_scall_trunc_qfile( error = xfs_itruncate_extents(&tp, ip, XFS_DATA_FORK, 0); if (error) { - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | - XFS_TRANS_ABORT); + xfs_trans_cancel(tp); goto out_unlock; } ASSERT(ip->i_d.di_nextents == 0); xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG); - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); out_unlock: xfs_iunlock(ip, XFS_ILOCK_EXCL | XFS_IOLOCK_EXCL); @@ -437,7 +436,7 @@ xfs_qm_scall_setqlim( tp = xfs_trans_alloc(mp, XFS_TRANS_QM_SETQLIM); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_setqlim, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); goto out_rele; } @@ -548,7 +547,7 @@ xfs_qm_scall_setqlim( dqp->dq_flags |= XFS_DQ_DIRTY; xfs_trans_log_dquot(tp, dqp); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); out_rele: xfs_qm_dqrele(dqp); @@ -571,7 +570,7 @@ xfs_qm_log_quotaoff_end( error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_equotaoff, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -585,8 +584,7 @@ xfs_qm_log_quotaoff_end( * We don't care about quotoff's performance. */ xfs_trans_set_sync(tp); - error = xfs_trans_commit(tp, 0); - return error; + return xfs_trans_commit(tp); } @@ -605,7 +603,7 @@ xfs_qm_log_quotaoff( tp = xfs_trans_alloc(mp, XFS_TRANS_QM_QUOTAOFF); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_qm_quotaoff, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); goto out; } @@ -624,7 +622,7 @@ xfs_qm_log_quotaoff( * We don't care about quotoff's performance. */ xfs_trans_set_sync(tp); - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); if (error) goto out; diff --git a/fs/xfs/xfs_quota.h b/fs/xfs/xfs_quota.h index 5376dd4..ce6506a 100644 --- a/fs/xfs/xfs_quota.h +++ b/fs/xfs/xfs_quota.h @@ -55,7 +55,6 @@ struct xfs_trans; typedef struct xfs_dqtrx { struct xfs_dquot *qt_dquot; /* the dquot this refers to */ ulong qt_blk_res; /* blks reserved on a dquot */ - ulong qt_blk_res_used; /* blks used from the reservation */ ulong qt_ino_res; /* inode reserved on a dquot */ ulong qt_ino_res_used; /* inodes used from the reservation */ long qt_bcount_delta; /* dquot blk count changes */ diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c index f2079b6..f4e8c06 100644 --- a/fs/xfs/xfs_rtalloc.c +++ b/fs/xfs/xfs_rtalloc.c @@ -780,7 +780,6 @@ xfs_growfs_rt_alloc( * Allocate space to the file, as necessary. */ while (oblocks < nblocks) { - int cancelflags = 0; xfs_trans_t *tp; tp = xfs_trans_alloc(mp, XFS_TRANS_GROWFSRT_ALLOC); @@ -792,7 +791,6 @@ xfs_growfs_rt_alloc( resblks, 0); if (error) goto error_cancel; - cancelflags = XFS_TRANS_RELEASE_LOG_RES; /* * Lock the inode. */ @@ -804,7 +802,6 @@ xfs_growfs_rt_alloc( * Allocate blocks to the bitmap file. */ nmap = 1; - cancelflags |= XFS_TRANS_ABORT; error = xfs_bmapi_write(tp, ip, oblocks, nblocks - oblocks, XFS_BMAPI_METADATA, &firstblock, resblks, &map, &nmap, &flist); @@ -818,14 +815,13 @@ xfs_growfs_rt_alloc( error = xfs_bmap_finish(&tp, &flist, &committed); if (error) goto error_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto error; /* * Now we need to clear the allocated blocks. * Do this one block per transaction, to keep it simple. */ - cancelflags = 0; for (bno = map.br_startoff, fsbno = map.br_startblock; bno < map.br_startoff + map.br_blockcount; bno++, fsbno++) { @@ -851,7 +847,7 @@ xfs_growfs_rt_alloc( if (bp == NULL) { error = -EIO; error_cancel: - xfs_trans_cancel(tp, cancelflags); + xfs_trans_cancel(tp); goto error; } memset(bp->b_addr, 0, mp->m_sb.sb_blocksize); @@ -859,7 +855,7 @@ error_cancel: /* * Commit the transaction. */ - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); if (error) goto error; } @@ -973,7 +969,6 @@ xfs_growfs_rt( bmbno < nrbmblocks; bmbno++) { xfs_trans_t *tp; - int cancelflags = 0; *nmp = *mp; nsbp = &nmp->m_sb; @@ -1015,7 +1010,6 @@ xfs_growfs_rt( mp->m_rbmip->i_d.di_size = nsbp->sb_rbmblocks * nsbp->sb_blocksize; xfs_trans_log_inode(tp, mp->m_rbmip, XFS_ILOG_CORE); - cancelflags |= XFS_TRANS_ABORT; /* * Get the summary inode into the transaction. */ @@ -1062,7 +1056,7 @@ xfs_growfs_rt( nsbp->sb_rextents - sbp->sb_rextents, &bp, &sumbno); if (error) { error_cancel: - xfs_trans_cancel(tp, cancelflags); + xfs_trans_cancel(tp); break; } /* @@ -1076,7 +1070,7 @@ error_cancel: mp->m_rsumlevels = nrsumlevels; mp->m_rsumsize = nrsumsize; - error = xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp); if (error) break; } diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 858e1e6..1fb16562 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -112,6 +112,8 @@ static struct xfs_kobj xfs_dbg_kobj; /* global debug sysfs attrs */ #define MNTOPT_DISCARD "discard" /* Discard unused blocks */ #define MNTOPT_NODISCARD "nodiscard" /* Do not discard unused blocks */ +#define MNTOPT_DAX "dax" /* Enable direct access to bdev pages */ + /* * Table driven mount option parser. * @@ -363,6 +365,10 @@ xfs_parseargs( mp->m_flags |= XFS_MOUNT_DISCARD; } else if (!strcmp(this_char, MNTOPT_NODISCARD)) { mp->m_flags &= ~XFS_MOUNT_DISCARD; +#ifdef CONFIG_FS_DAX + } else if (!strcmp(this_char, MNTOPT_DAX)) { + mp->m_flags |= XFS_MOUNT_DAX; +#endif } else { xfs_warn(mp, "unknown mount option [%s].", this_char); return -EINVAL; @@ -452,8 +458,8 @@ done: } struct proc_xfs_info { - int flag; - char *str; + uint64_t flag; + char *str; }; STATIC int @@ -474,6 +480,7 @@ xfs_showargs( { XFS_MOUNT_GRPID, "," MNTOPT_GRPID }, { XFS_MOUNT_DISCARD, "," MNTOPT_DISCARD }, { XFS_MOUNT_SMALL_INUMS, "," MNTOPT_32BITINODE }, + { XFS_MOUNT_DAX, "," MNTOPT_DAX }, { 0, NULL } }; static struct proc_xfs_info xfs_info_unset[] = { @@ -1507,6 +1514,20 @@ xfs_fs_fill_super( if (XFS_SB_VERSION_NUM(&mp->m_sb) == XFS_SB_VERSION_5) sb->s_flags |= MS_I_VERSION; + if (mp->m_flags & XFS_MOUNT_DAX) { + xfs_warn(mp, + "DAX enabled. Warning: EXPERIMENTAL, use at your own risk"); + if (sb->s_blocksize != PAGE_SIZE) { + xfs_alert(mp, + "Filesystem block size invalid for DAX Turning DAX off."); + mp->m_flags &= ~XFS_MOUNT_DAX; + } else if (!sb->s_bdev->bd_disk->fops->direct_access) { + xfs_alert(mp, + "Block device does not support DAX Turning DAX off."); + mp->m_flags &= ~XFS_MOUNT_DAX; + } + } + error = xfs_mountfs(mp); if (error) goto out_filestream_unmount; diff --git a/fs/xfs/xfs_symlink.c b/fs/xfs/xfs_symlink.c index 3df411e..4be27b0 100644 --- a/fs/xfs/xfs_symlink.c +++ b/fs/xfs/xfs_symlink.c @@ -104,7 +104,7 @@ xfs_readlink_bmap( cur_chunk += sizeof(struct xfs_dsymlink_hdr); } - memcpy(link + offset, bp->b_addr, byte_cnt); + memcpy(link + offset, cur_chunk, byte_cnt); pathlen -= byte_cnt; offset += byte_cnt; @@ -178,7 +178,6 @@ xfs_symlink( struct xfs_bmap_free free_list; xfs_fsblock_t first_block; bool unlock_dp_on_error = false; - uint cancel_flags; int committed; xfs_fileoff_t first_fsb; xfs_filblks_t fs_blocks; @@ -224,7 +223,6 @@ xfs_symlink( return error; tp = xfs_trans_alloc(mp, XFS_TRANS_SYMLINK); - cancel_flags = XFS_TRANS_RELEASE_LOG_RES; /* * The symlink will fit into the inode data fork? * There can't be any attributes so we get the whole variable part. @@ -239,10 +237,8 @@ xfs_symlink( resblks = 0; error = xfs_trans_reserve(tp, &M_RES(mp)->tr_symlink, 0, 0); } - if (error) { - cancel_flags = 0; + if (error) goto out_trans_cancel; - } xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT); unlock_dp_on_error = true; @@ -394,7 +390,7 @@ xfs_symlink( if (error) goto out_bmap_cancel; - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) goto out_release_inode; @@ -407,9 +403,8 @@ xfs_symlink( out_bmap_cancel: xfs_bmap_cancel(&free_list); - cancel_flags |= XFS_TRANS_ABORT; out_trans_cancel: - xfs_trans_cancel(tp, cancel_flags); + xfs_trans_cancel(tp); out_release_inode: /* * Wait until after the current transaction is aborted to finish the @@ -464,7 +459,7 @@ xfs_inactive_symlink_rmt( tp = xfs_trans_alloc(mp, XFS_TRANS_INACTIVE); error = xfs_trans_reserve(tp, &M_RES(mp)->tr_itruncate, 0, 0); if (error) { - xfs_trans_cancel(tp, 0); + xfs_trans_cancel(tp); return error; } @@ -533,7 +528,7 @@ xfs_inactive_symlink_rmt( /* * Commit the transaction containing extent freeing and EFDs. */ - error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + error = xfs_trans_commit(tp); if (error) { ASSERT(XFS_FORCED_SHUTDOWN(mp)); goto error_unlock; @@ -552,7 +547,7 @@ xfs_inactive_symlink_rmt( error_bmap_cancel: xfs_bmap_cancel(&free_list); error_trans_cancel: - xfs_trans_cancel(tp, XFS_TRANS_RELEASE_LOG_RES | XFS_TRANS_ABORT); + xfs_trans_cancel(tp); error_unlock: xfs_iunlock(ip, XFS_ILOCK_EXCL); return error; diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 615781b..8d916d3 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -738,6 +738,53 @@ TRACE_EVENT(xfs_iomap_prealloc_size, __entry->blocks, __entry->shift, __entry->writeio_blocks) ) +TRACE_EVENT(xfs_irec_merge_pre, + TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, xfs_agino_t agino, + uint16_t holemask, xfs_agino_t nagino, uint16_t nholemask), + TP_ARGS(mp, agno, agino, holemask, nagino, nholemask), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_agnumber_t, agno) + __field(xfs_agino_t, agino) + __field(uint16_t, holemask) + __field(xfs_agino_t, nagino) + __field(uint16_t, nholemask) + ), + TP_fast_assign( + __entry->dev = mp->m_super->s_dev; + __entry->agno = agno; + __entry->agino = agino; + __entry->holemask = holemask; + __entry->nagino = nagino; + __entry->nholemask = holemask; + ), + TP_printk("dev %d:%d agno %d inobt (%u:0x%x) new (%u:0x%x)", + MAJOR(__entry->dev), MINOR(__entry->dev), __entry->agno, + __entry->agino, __entry->holemask, __entry->nagino, + __entry->nholemask) +) + +TRACE_EVENT(xfs_irec_merge_post, + TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, xfs_agino_t agino, + uint16_t holemask), + TP_ARGS(mp, agno, agino, holemask), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_agnumber_t, agno) + __field(xfs_agino_t, agino) + __field(uint16_t, holemask) + ), + TP_fast_assign( + __entry->dev = mp->m_super->s_dev; + __entry->agno = agno; + __entry->agino = agino; + __entry->holemask = holemask; + ), + TP_printk("dev %d:%d agno %d inobt (%u:0x%x)", MAJOR(__entry->dev), + MINOR(__entry->dev), __entry->agno, __entry->agino, + __entry->holemask) +) + #define DEFINE_IREF_EVENT(name) \ DEFINE_EVENT(xfs_iref_class, name, \ TP_PROTO(struct xfs_inode *ip, unsigned long caller_ip), \ diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 220ef2c..0582a27 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -113,7 +113,7 @@ xfs_trans_free( * blocks. Locks and log items, however, are no inherited. They must * be added to the new transaction explicitly. */ -xfs_trans_t * +STATIC xfs_trans_t * xfs_trans_dup( xfs_trans_t *tp) { @@ -251,14 +251,7 @@ xfs_trans_reserve( */ undo_log: if (resp->tr_logres > 0) { - int log_flags; - - if (resp->tr_logflags & XFS_TRANS_PERM_LOG_RES) { - log_flags = XFS_LOG_REL_PERM_RESERV; - } else { - log_flags = 0; - } - xfs_log_done(tp->t_mountp, tp->t_ticket, NULL, log_flags); + xfs_log_done(tp->t_mountp, tp->t_ticket, NULL, false); tp->t_ticket = NULL; tp->t_log_res = 0; tp->t_flags &= ~XFS_TRANS_PERM_LOG_RES; @@ -744,7 +737,7 @@ void xfs_trans_free_items( struct xfs_trans *tp, xfs_lsn_t commit_lsn, - int flags) + bool abort) { struct xfs_log_item_desc *lidp, *next; @@ -755,7 +748,7 @@ xfs_trans_free_items( if (commit_lsn != NULLCOMMITLSN) lip->li_ops->iop_committing(lip, commit_lsn); - if (flags & XFS_TRANS_ABORT) + if (abort) lip->li_flags |= XFS_LI_ABORTED; lip->li_ops->iop_unlock(lip); @@ -892,27 +885,17 @@ xfs_trans_committed_bulk( * have already been unlocked as if the commit had succeeded. * Do not reference the transaction structure after this call. */ -int -xfs_trans_commit( +static int +__xfs_trans_commit( struct xfs_trans *tp, - uint flags) + bool regrant) { struct xfs_mount *mp = tp->t_mountp; xfs_lsn_t commit_lsn = -1; int error = 0; - int log_flags = 0; int sync = tp->t_flags & XFS_TRANS_SYNC; /* - * Determine whether this commit is releasing a permanent - * log reservation or not. - */ - if (flags & XFS_TRANS_RELEASE_LOG_RES) { - ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES); - log_flags = XFS_LOG_REL_PERM_RESERV; - } - - /* * If there is nothing to be logged by the transaction, * then unlock all of the items associated with the * transaction and free the transaction structure. @@ -936,7 +919,7 @@ xfs_trans_commit( xfs_trans_apply_sb_deltas(tp); xfs_trans_apply_dquot_deltas(tp); - xfs_log_commit_cil(mp, tp, &commit_lsn, flags); + xfs_log_commit_cil(mp, tp, &commit_lsn, regrant); current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS); xfs_trans_free(tp); @@ -964,18 +947,25 @@ out_unreserve: */ xfs_trans_unreserve_and_mod_dquots(tp); if (tp->t_ticket) { - commit_lsn = xfs_log_done(mp, tp->t_ticket, NULL, log_flags); + commit_lsn = xfs_log_done(mp, tp->t_ticket, NULL, regrant); if (commit_lsn == -1 && !error) error = -EIO; } current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS); - xfs_trans_free_items(tp, NULLCOMMITLSN, error ? XFS_TRANS_ABORT : 0); + xfs_trans_free_items(tp, NULLCOMMITLSN, !!error); xfs_trans_free(tp); XFS_STATS_INC(xs_trans_empty); return error; } +int +xfs_trans_commit( + struct xfs_trans *tp) +{ + return __xfs_trans_commit(tp, false); +} + /* * Unlock all of the transaction's items and free the transaction. * The transaction must not have modified any of its items, because @@ -986,29 +976,22 @@ out_unreserve: */ void xfs_trans_cancel( - xfs_trans_t *tp, - int flags) + struct xfs_trans *tp) { - int log_flags; - xfs_mount_t *mp = tp->t_mountp; + struct xfs_mount *mp = tp->t_mountp; + bool dirty = (tp->t_flags & XFS_TRANS_DIRTY); /* - * See if the caller is being too lazy to figure out if - * the transaction really needs an abort. - */ - if ((flags & XFS_TRANS_ABORT) && !(tp->t_flags & XFS_TRANS_DIRTY)) - flags &= ~XFS_TRANS_ABORT; - /* * See if the caller is relying on us to shut down the * filesystem. This happens in paths where we detect * corruption and decide to give up. */ - if ((tp->t_flags & XFS_TRANS_DIRTY) && !XFS_FORCED_SHUTDOWN(mp)) { + if (dirty && !XFS_FORCED_SHUTDOWN(mp)) { XFS_ERROR_REPORT("xfs_trans_cancel", XFS_ERRLEVEL_LOW, mp); xfs_force_shutdown(mp, SHUTDOWN_CORRUPT_INCORE); } #ifdef DEBUG - if (!(flags & XFS_TRANS_ABORT) && !XFS_FORCED_SHUTDOWN(mp)) { + if (!dirty && !XFS_FORCED_SHUTDOWN(mp)) { struct xfs_log_item_desc *lidp; list_for_each_entry(lidp, &tp->t_items, lid_trans) @@ -1018,27 +1001,20 @@ xfs_trans_cancel( xfs_trans_unreserve_and_mod_sb(tp); xfs_trans_unreserve_and_mod_dquots(tp); - if (tp->t_ticket) { - if (flags & XFS_TRANS_RELEASE_LOG_RES) { - ASSERT(tp->t_flags & XFS_TRANS_PERM_LOG_RES); - log_flags = XFS_LOG_REL_PERM_RESERV; - } else { - log_flags = 0; - } - xfs_log_done(mp, tp->t_ticket, NULL, log_flags); - } + if (tp->t_ticket) + xfs_log_done(mp, tp->t_ticket, NULL, false); /* mark this thread as no longer being in a transaction */ current_restore_flags_nested(&tp->t_pflags, PF_FSTRANS); - xfs_trans_free_items(tp, NULLCOMMITLSN, flags); + xfs_trans_free_items(tp, NULLCOMMITLSN, dirty); xfs_trans_free(tp); } /* * Roll from one trans in the sequence of PERMANENT transactions to * the next: permanent transactions are only flushed out when - * committed with XFS_TRANS_RELEASE_LOG_RES, but we still want as soon + * committed with xfs_trans_commit(), but we still want as soon * as possible to let chunks of it go to the log. So we commit the * chunk we've been working on and get a new transaction to continue. */ @@ -1055,7 +1031,8 @@ xfs_trans_roll( * Ensure that the inode is always logged. */ trans = *tpp; - xfs_trans_log_inode(trans, dp, XFS_ILOG_CORE); + if (dp) + xfs_trans_log_inode(trans, dp, XFS_ILOG_CORE); /* * Copy the critical parameters from one trans to the next. @@ -1071,20 +1048,13 @@ xfs_trans_roll( * is in progress. The caller takes the responsibility to cancel * the duplicate transaction that gets returned. */ - error = xfs_trans_commit(trans, 0); + error = __xfs_trans_commit(trans, true); if (error) return error; trans = *tpp; /* - * transaction commit worked ok so we can drop the extra ticket - * reference that we gained in xfs_trans_dup() - */ - xfs_log_ticket_put(trans->t_ticket); - - - /* * Reserve space in the log for th next transaction. * This also pushes items in the "AIL", the list of logged items, * out to disk if they are taking up space at the tail of the log @@ -1100,6 +1070,7 @@ xfs_trans_roll( if (error) return error; - xfs_trans_ijoin(trans, dp, 0); + if (dp) + xfs_trans_ijoin(trans, dp, 0); return 0; } diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index b5bc1ab..3b21b4e 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -133,8 +133,6 @@ typedef struct xfs_trans { * XFS transaction mechanism exported interfaces that are * actually macros. */ -#define xfs_trans_get_log_res(tp) ((tp)->t_log_res) -#define xfs_trans_get_log_count(tp) ((tp)->t_log_count) #define xfs_trans_get_block_res(tp) ((tp)->t_blk_res) #define xfs_trans_set_sync(tp) ((tp)->t_flags |= XFS_TRANS_SYNC) @@ -153,7 +151,6 @@ typedef struct xfs_trans { */ xfs_trans_t *xfs_trans_alloc(struct xfs_mount *, uint); xfs_trans_t *_xfs_trans_alloc(struct xfs_mount *, uint, xfs_km_flags_t); -xfs_trans_t *xfs_trans_dup(xfs_trans_t *); int xfs_trans_reserve(struct xfs_trans *, struct xfs_trans_res *, uint, uint); void xfs_trans_mod_sb(xfs_trans_t *, uint, int64_t); @@ -228,9 +225,9 @@ void xfs_trans_log_efd_extent(xfs_trans_t *, struct xfs_efd_log_item *, xfs_fsblock_t, xfs_extlen_t); -int xfs_trans_commit(xfs_trans_t *, uint flags); +int xfs_trans_commit(struct xfs_trans *); int xfs_trans_roll(struct xfs_trans **, struct xfs_inode *); -void xfs_trans_cancel(xfs_trans_t *, int); +void xfs_trans_cancel(xfs_trans_t *); int xfs_trans_ail_init(struct xfs_mount *); void xfs_trans_ail_destroy(struct xfs_mount *); diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index 573aefb..1098cf4 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -159,7 +159,7 @@ xfs_trans_ail_cursor_next( { struct xfs_log_item *lip = cur->item; - if ((__psint_t)lip & 1) + if ((uintptr_t)lip & 1) lip = xfs_ail_min(ailp); if (lip) cur->item = xfs_ail_next(ailp, lip); @@ -196,7 +196,7 @@ xfs_trans_ail_cursor_clear( list_for_each_entry(cur, &ailp->xa_cursors, list) { if (cur->item == lip) cur->item = (struct xfs_log_item *) - ((__psint_t)cur->item | 1); + ((uintptr_t)cur->item | 1); } } @@ -287,7 +287,7 @@ xfs_ail_splice( * find the place in the AIL where the items belong. */ lip = cur ? cur->item : NULL; - if (!lip || (__psint_t) lip & 1) + if (!lip || (uintptr_t)lip & 1) lip = __xfs_trans_ail_cursor_last(ailp, lsn); /* diff --git a/fs/xfs/xfs_trans_dquot.c b/fs/xfs/xfs_trans_dquot.c index 76a16df..ce78534 100644 --- a/fs/xfs/xfs_trans_dquot.c +++ b/fs/xfs/xfs_trans_dquot.c @@ -90,8 +90,9 @@ xfs_trans_dup_dqinfo( xfs_trans_t *ntp) { xfs_dqtrx_t *oq, *nq; - int i,j; + int i, j; xfs_dqtrx_t *oqa, *nqa; + ulong blk_res_used; if (!otp->t_dqinfo) return; @@ -102,18 +103,23 @@ xfs_trans_dup_dqinfo( * Because the quota blk reservation is carried forward, * it is also necessary to carry forward the DQ_DIRTY flag. */ - if(otp->t_flags & XFS_TRANS_DQ_DIRTY) + if (otp->t_flags & XFS_TRANS_DQ_DIRTY) ntp->t_flags |= XFS_TRANS_DQ_DIRTY; for (j = 0; j < XFS_QM_TRANS_DQTYPES; j++) { oqa = otp->t_dqinfo->dqs[j]; nqa = ntp->t_dqinfo->dqs[j]; for (i = 0; i < XFS_QM_TRANS_MAXDQS; i++) { + blk_res_used = 0; + if (oqa[i].qt_dquot == NULL) break; oq = &oqa[i]; nq = &nqa[i]; + if (oq->qt_blk_res && oq->qt_bcount_delta > 0) + blk_res_used = oq->qt_bcount_delta; + nq->qt_dquot = oq->qt_dquot; nq->qt_bcount_delta = nq->qt_icount_delta = 0; nq->qt_rtbcount_delta = 0; @@ -121,8 +127,8 @@ xfs_trans_dup_dqinfo( /* * Transfer whatever is left of the reservations. */ - nq->qt_blk_res = oq->qt_blk_res - oq->qt_blk_res_used; - oq->qt_blk_res = oq->qt_blk_res_used; + nq->qt_blk_res = oq->qt_blk_res - blk_res_used; + oq->qt_blk_res = blk_res_used; nq->qt_rtblk_res = oq->qt_rtblk_res - oq->qt_rtblk_res_used; @@ -239,10 +245,6 @@ xfs_trans_mod_dquot( * disk blocks used. */ case XFS_TRANS_DQ_BCOUNT: - if (qtrx->qt_blk_res && delta > 0) { - qtrx->qt_blk_res_used += (ulong)delta; - ASSERT(qtrx->qt_blk_res >= qtrx->qt_blk_res_used); - } qtrx->qt_bcount_delta += delta; break; @@ -423,15 +425,19 @@ xfs_trans_apply_dquot_deltas( * reservation that a transaction structure knows of. */ if (qtrx->qt_blk_res != 0) { - if (qtrx->qt_blk_res != qtrx->qt_blk_res_used) { - if (qtrx->qt_blk_res > - qtrx->qt_blk_res_used) + ulong blk_res_used = 0; + + if (qtrx->qt_bcount_delta > 0) + blk_res_used = qtrx->qt_bcount_delta; + + if (qtrx->qt_blk_res != blk_res_used) { + if (qtrx->qt_blk_res > blk_res_used) dqp->q_res_bcount -= (xfs_qcnt_t) (qtrx->qt_blk_res - - qtrx->qt_blk_res_used); + blk_res_used); else dqp->q_res_bcount -= (xfs_qcnt_t) - (qtrx->qt_blk_res_used - + (blk_res_used - qtrx->qt_blk_res); } } else { diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index bd12818..1b73629 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -30,7 +30,7 @@ void xfs_trans_init(struct xfs_mount *); void xfs_trans_add_item(struct xfs_trans *, struct xfs_log_item *); void xfs_trans_del_item(struct xfs_log_item *); void xfs_trans_free_items(struct xfs_trans *tp, xfs_lsn_t commit_lsn, - int flags); + bool abort); void xfs_trans_unreserve_and_mod_sb(struct xfs_trans *tp); void xfs_trans_committed_bulk(struct xfs_ail *ailp, struct xfs_log_vec *lv, |