diff options
Diffstat (limited to 'fs')
159 files changed, 4118 insertions, 2715 deletions
diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 9ee5343..3662f1d 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -1127,7 +1127,7 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr) } /* Write all dirty data */ - if (S_ISREG(dentry->d_inode->i_mode)) + if (d_is_reg(dentry)) filemap_write_and_wait(dentry->d_inode->i_mapping); retval = p9_client_wstat(fid, &wstat); @@ -1296,7 +1296,7 @@ SYSCALL_DEFINE2(io_setup, unsigned, nr_events, aio_context_t __user *, ctxp) ret = -EINVAL; if (unlikely(ctx || nr_events == 0)) { - pr_debug("EINVAL: io_setup: ctx %lu nr_events %u\n", + pr_debug("EINVAL: ctx %lu nr_events %u\n", ctx, nr_events); goto out; } @@ -1344,7 +1344,7 @@ SYSCALL_DEFINE1(io_destroy, aio_context_t, ctx) return ret; } - pr_debug("EINVAL: io_destroy: invalid context id\n"); + pr_debug("EINVAL: invalid context id\n"); return -EINVAL; } @@ -1528,7 +1528,7 @@ static int io_submit_one(struct kioctx *ctx, struct iocb __user *user_iocb, (iocb->aio_nbytes != (size_t)iocb->aio_nbytes) || ((ssize_t)iocb->aio_nbytes < 0) )) { - pr_debug("EINVAL: io_submit: overflow check\n"); + pr_debug("EINVAL: overflow check\n"); return -EINVAL; } diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index aaf96cb..ac7d921 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -95,7 +95,7 @@ static int check_dev_ioctl_version(int cmd, struct autofs_dev_ioctl *param) */ static struct autofs_dev_ioctl *copy_dev_ioctl(struct autofs_dev_ioctl __user *in) { - struct autofs_dev_ioctl tmp; + struct autofs_dev_ioctl tmp, *res; if (copy_from_user(&tmp, in, sizeof(tmp))) return ERR_PTR(-EFAULT); @@ -106,7 +106,11 @@ static struct autofs_dev_ioctl *copy_dev_ioctl(struct autofs_dev_ioctl __user *i if (tmp.size > (PATH_MAX + sizeof(tmp))) return ERR_PTR(-ENAMETOOLONG); - return memdup_user(in, tmp.size); + res = memdup_user(in, tmp.size); + if (!IS_ERR(res)) + res->size = tmp.size; + + return res; } static inline void free_dev_ioctl(struct autofs_dev_ioctl *param) diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c index bfdbaba..11dd118 100644 --- a/fs/autofs4/expire.c +++ b/fs/autofs4/expire.c @@ -374,7 +374,7 @@ static struct dentry *should_expire(struct dentry *dentry, return NULL; } - if (dentry->d_inode && S_ISLNK(dentry->d_inode->i_mode)) { + if (dentry->d_inode && d_is_symlink(dentry)) { DPRINTK("checking symlink %p %pd", dentry, dentry); /* * A symlink can't be "busy" in the usual sense so diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index dbb5b72..7e44fdd 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -108,7 +108,7 @@ static int autofs4_dir_open(struct inode *inode, struct file *file) struct dentry *dentry = file->f_path.dentry; struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); - DPRINTK("file=%p dentry=%p %pD", file, dentry, dentry); + DPRINTK("file=%p dentry=%p %pd", file, dentry, dentry); if (autofs4_oz_mode(sbi)) goto out; @@ -371,7 +371,7 @@ static struct vfsmount *autofs4_d_automount(struct path *path) * having d_mountpoint() true, so there's no need to call back * to the daemon. */ - if (dentry->d_inode && S_ISLNK(dentry->d_inode->i_mode)) { + if (dentry->d_inode && d_is_symlink(dentry)) { spin_unlock(&sbi->fs_lock); goto done; } @@ -485,7 +485,7 @@ static int autofs4_d_manage(struct dentry *dentry, bool rcu_walk) * an incorrect ELOOP error return. */ if ((!d_mountpoint(dentry) && !simple_empty(dentry)) || - (dentry->d_inode && S_ISLNK(dentry->d_inode->i_mode))) + (dentry->d_inode && d_is_symlink(dentry))) status = -EISDIR; } spin_unlock(&sbi->fs_lock); diff --git a/fs/bad_inode.c b/fs/bad_inode.c index afd2b44..861b1e1 100644 --- a/fs/bad_inode.c +++ b/fs/bad_inode.c @@ -15,161 +15,14 @@ #include <linux/namei.h> #include <linux/poll.h> - -static loff_t bad_file_llseek(struct file *file, loff_t offset, int whence) -{ - return -EIO; -} - -static ssize_t bad_file_read(struct file *filp, char __user *buf, - size_t size, loff_t *ppos) -{ - return -EIO; -} - -static ssize_t bad_file_write(struct file *filp, const char __user *buf, - size_t siz, loff_t *ppos) -{ - return -EIO; -} - -static ssize_t bad_file_aio_read(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) -{ - return -EIO; -} - -static ssize_t bad_file_aio_write(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) -{ - return -EIO; -} - -static int bad_file_readdir(struct file *file, struct dir_context *ctx) -{ - return -EIO; -} - -static unsigned int bad_file_poll(struct file *filp, poll_table *wait) -{ - return POLLERR; -} - -static long bad_file_unlocked_ioctl(struct file *file, unsigned cmd, - unsigned long arg) -{ - return -EIO; -} - -static long bad_file_compat_ioctl(struct file *file, unsigned int cmd, - unsigned long arg) -{ - return -EIO; -} - -static int bad_file_mmap(struct file *file, struct vm_area_struct *vma) -{ - return -EIO; -} - static int bad_file_open(struct inode *inode, struct file *filp) { return -EIO; } -static int bad_file_flush(struct file *file, fl_owner_t id) -{ - return -EIO; -} - -static int bad_file_release(struct inode *inode, struct file *filp) -{ - return -EIO; -} - -static int bad_file_fsync(struct file *file, loff_t start, loff_t end, - int datasync) -{ - return -EIO; -} - -static int bad_file_aio_fsync(struct kiocb *iocb, int datasync) -{ - return -EIO; -} - -static int bad_file_fasync(int fd, struct file *filp, int on) -{ - return -EIO; -} - -static int bad_file_lock(struct file *file, int cmd, struct file_lock *fl) -{ - return -EIO; -} - -static ssize_t bad_file_sendpage(struct file *file, struct page *page, - int off, size_t len, loff_t *pos, int more) -{ - return -EIO; -} - -static unsigned long bad_file_get_unmapped_area(struct file *file, - unsigned long addr, unsigned long len, - unsigned long pgoff, unsigned long flags) -{ - return -EIO; -} - -static int bad_file_check_flags(int flags) -{ - return -EIO; -} - -static int bad_file_flock(struct file *filp, int cmd, struct file_lock *fl) -{ - return -EIO; -} - -static ssize_t bad_file_splice_write(struct pipe_inode_info *pipe, - struct file *out, loff_t *ppos, size_t len, - unsigned int flags) -{ - return -EIO; -} - -static ssize_t bad_file_splice_read(struct file *in, loff_t *ppos, - struct pipe_inode_info *pipe, size_t len, - unsigned int flags) -{ - return -EIO; -} - static const struct file_operations bad_file_ops = { - .llseek = bad_file_llseek, - .read = bad_file_read, - .write = bad_file_write, - .aio_read = bad_file_aio_read, - .aio_write = bad_file_aio_write, - .iterate = bad_file_readdir, - .poll = bad_file_poll, - .unlocked_ioctl = bad_file_unlocked_ioctl, - .compat_ioctl = bad_file_compat_ioctl, - .mmap = bad_file_mmap, .open = bad_file_open, - .flush = bad_file_flush, - .release = bad_file_release, - .fsync = bad_file_fsync, - .aio_fsync = bad_file_aio_fsync, - .fasync = bad_file_fasync, - .lock = bad_file_lock, - .sendpage = bad_file_sendpage, - .get_unmapped_area = bad_file_get_unmapped_area, - .check_flags = bad_file_check_flags, - .flock = bad_file_flock, - .splice_write = bad_file_splice_write, - .splice_read = bad_file_splice_read, }; static int bad_inode_create (struct inode *dir, struct dentry *dentry, diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index 02b1691..995986b 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -645,11 +645,12 @@ out: static unsigned long randomize_stack_top(unsigned long stack_top) { - unsigned int random_variable = 0; + unsigned long random_variable = 0; if ((current->flags & PF_RANDOMIZE) && !(current->personality & ADDR_NO_RANDOMIZE)) { - random_variable = get_random_int() & STACK_RND_MASK; + random_variable = (unsigned long) get_random_int(); + random_variable &= STACK_RND_MASK; random_variable <<= PAGE_SHIFT; } #ifdef CONFIG_STACK_GROWSUP diff --git a/fs/btrfs/backref.c b/fs/btrfs/backref.c index 8729cf6..f55721f 100644 --- a/fs/btrfs/backref.c +++ b/fs/btrfs/backref.c @@ -1246,25 +1246,6 @@ int btrfs_check_shared(struct btrfs_trans_handle *trans, return ret; } -/* - * this makes the path point to (inum INODE_ITEM ioff) - */ -int inode_item_info(u64 inum, u64 ioff, struct btrfs_root *fs_root, - struct btrfs_path *path) -{ - struct btrfs_key key; - return btrfs_find_item(fs_root, path, inum, ioff, - BTRFS_INODE_ITEM_KEY, &key); -} - -static int inode_ref_info(u64 inum, u64 ioff, struct btrfs_root *fs_root, - struct btrfs_path *path, - struct btrfs_key *found_key) -{ - return btrfs_find_item(fs_root, path, inum, ioff, - BTRFS_INODE_REF_KEY, found_key); -} - int btrfs_find_one_extref(struct btrfs_root *root, u64 inode_objectid, u64 start_off, struct btrfs_path *path, struct btrfs_inode_extref **ret_extref, @@ -1374,7 +1355,8 @@ char *btrfs_ref_to_path(struct btrfs_root *fs_root, struct btrfs_path *path, btrfs_tree_read_unlock_blocking(eb); free_extent_buffer(eb); } - ret = inode_ref_info(parent, 0, fs_root, path, &found_key); + ret = btrfs_find_item(fs_root, path, parent, 0, + BTRFS_INODE_REF_KEY, &found_key); if (ret > 0) ret = -ENOENT; if (ret) @@ -1727,8 +1709,10 @@ static int iterate_inode_refs(u64 inum, struct btrfs_root *fs_root, struct btrfs_key found_key; while (!ret) { - ret = inode_ref_info(inum, parent ? parent+1 : 0, fs_root, path, - &found_key); + ret = btrfs_find_item(fs_root, path, inum, + parent ? parent + 1 : 0, BTRFS_INODE_REF_KEY, + &found_key); + if (ret < 0) break; if (ret) { diff --git a/fs/btrfs/backref.h b/fs/btrfs/backref.h index 2a1ac6b..9c41fba 100644 --- a/fs/btrfs/backref.h +++ b/fs/btrfs/backref.h @@ -32,9 +32,6 @@ struct inode_fs_paths { typedef int (iterate_extent_inodes_t)(u64 inum, u64 offset, u64 root, void *ctx); -int inode_item_info(u64 inum, u64 ioff, struct btrfs_root *fs_root, - struct btrfs_path *path); - int extent_from_logical(struct btrfs_fs_info *fs_info, u64 logical, struct btrfs_path *path, struct btrfs_key *found_key, u64 *flags); diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index 4aadadc..de5e4f2 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -185,6 +185,9 @@ struct btrfs_inode { struct btrfs_delayed_node *delayed_node; + /* File creation time. */ + struct timespec i_otime; + struct inode vfs_inode; }; diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 14a72ed..6d67f32 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -213,11 +213,19 @@ static struct extent_buffer *btrfs_read_lock_root_node(struct btrfs_root *root) */ static void add_root_to_dirty_list(struct btrfs_root *root) { + if (test_bit(BTRFS_ROOT_DIRTY, &root->state) || + !test_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state)) + return; + spin_lock(&root->fs_info->trans_lock); - if (test_bit(BTRFS_ROOT_TRACK_DIRTY, &root->state) && - list_empty(&root->dirty_list)) { - list_add(&root->dirty_list, - &root->fs_info->dirty_cowonly_roots); + if (!test_and_set_bit(BTRFS_ROOT_DIRTY, &root->state)) { + /* Want the extent tree to be the last on the list */ + if (root->objectid == BTRFS_EXTENT_TREE_OBJECTID) + list_move_tail(&root->dirty_list, + &root->fs_info->dirty_cowonly_roots); + else + list_move(&root->dirty_list, + &root->fs_info->dirty_cowonly_roots); } spin_unlock(&root->fs_info->trans_lock); } @@ -1363,8 +1371,7 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct btrfs_path *path, if (tm->op == MOD_LOG_KEY_REMOVE_WHILE_FREEING) { BUG_ON(tm->slot != 0); - eb_rewin = alloc_dummy_extent_buffer(eb->start, - fs_info->tree_root->nodesize); + eb_rewin = alloc_dummy_extent_buffer(fs_info, eb->start); if (!eb_rewin) { btrfs_tree_read_unlock_blocking(eb); free_extent_buffer(eb); @@ -1444,7 +1451,7 @@ get_old_root(struct btrfs_root *root, u64 time_seq) } else if (old_root) { btrfs_tree_read_unlock(eb_root); free_extent_buffer(eb_root); - eb = alloc_dummy_extent_buffer(logical, root->nodesize); + eb = alloc_dummy_extent_buffer(root->fs_info, logical); } else { btrfs_set_lock_blocking_rw(eb_root, BTRFS_READ_LOCK); eb = btrfs_clone_extent_buffer(eb_root); @@ -1638,14 +1645,14 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans, parent_nritems = btrfs_header_nritems(parent); blocksize = root->nodesize; - end_slot = parent_nritems; + end_slot = parent_nritems - 1; - if (parent_nritems == 1) + if (parent_nritems <= 1) return 0; btrfs_set_lock_blocking(parent); - for (i = start_slot; i < end_slot; i++) { + for (i = start_slot; i <= end_slot; i++) { int close = 1; btrfs_node_key(parent, &disk_key, i); @@ -1662,7 +1669,7 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans, other = btrfs_node_blockptr(parent, i - 1); close = close_blocks(blocknr, other, blocksize); } - if (!close && i < end_slot - 2) { + if (!close && i < end_slot) { other = btrfs_node_blockptr(parent, i + 1); close = close_blocks(blocknr, other, blocksize); } @@ -2282,7 +2289,7 @@ static void reada_for_search(struct btrfs_root *root, if ((search <= target && target - search <= 65536) || (search > target && search - target <= 65536)) { gen = btrfs_node_ptr_generation(node, nr); - readahead_tree_block(root, search, blocksize); + readahead_tree_block(root, search); nread += blocksize; } nscan++; @@ -2301,7 +2308,6 @@ static noinline void reada_for_balance(struct btrfs_root *root, u64 gen; u64 block1 = 0; u64 block2 = 0; - int blocksize; parent = path->nodes[level + 1]; if (!parent) @@ -2309,7 +2315,6 @@ static noinline void reada_for_balance(struct btrfs_root *root, nritems = btrfs_header_nritems(parent); slot = path->slots[level + 1]; - blocksize = root->nodesize; if (slot > 0) { block1 = btrfs_node_blockptr(parent, slot - 1); @@ -2334,9 +2339,9 @@ static noinline void reada_for_balance(struct btrfs_root *root, } if (block1) - readahead_tree_block(root, block1, blocksize); + readahead_tree_block(root, block1); if (block2) - readahead_tree_block(root, block2, blocksize); + readahead_tree_block(root, block2); } @@ -2609,32 +2614,24 @@ static int key_search(struct extent_buffer *b, struct btrfs_key *key, return 0; } -int btrfs_find_item(struct btrfs_root *fs_root, struct btrfs_path *found_path, +int btrfs_find_item(struct btrfs_root *fs_root, struct btrfs_path *path, u64 iobjectid, u64 ioff, u8 key_type, struct btrfs_key *found_key) { int ret; struct btrfs_key key; struct extent_buffer *eb; - struct btrfs_path *path; + + ASSERT(path); + ASSERT(found_key); key.type = key_type; key.objectid = iobjectid; key.offset = ioff; - if (found_path == NULL) { - path = btrfs_alloc_path(); - if (!path) - return -ENOMEM; - } else - path = found_path; - ret = btrfs_search_slot(NULL, fs_root, &key, path, 0, 0); - if ((ret < 0) || (found_key == NULL)) { - if (path != found_path) - btrfs_free_path(path); + if (ret < 0) return ret; - } eb = path->nodes[0]; if (ret && path->slots[0] >= btrfs_header_nritems(eb)) { @@ -3383,7 +3380,7 @@ static noinline int insert_new_root(struct btrfs_trans_handle *trans, add_root_to_dirty_list(root); extent_buffer_get(c); path->nodes[level] = c; - path->locks[level] = BTRFS_WRITE_LOCK; + path->locks[level] = BTRFS_WRITE_LOCK_BLOCKING; path->slots[level] = 0; return 0; } @@ -4356,13 +4353,15 @@ static noinline int setup_leaf_for_split(struct btrfs_trans_handle *trans, path->search_for_split = 1; ret = btrfs_search_slot(trans, root, &key, path, 0, 1); path->search_for_split = 0; + if (ret > 0) + ret = -EAGAIN; if (ret < 0) goto err; ret = -EAGAIN; leaf = path->nodes[0]; - /* if our item isn't there or got smaller, return now */ - if (ret > 0 || item_size != btrfs_item_size_nr(leaf, path->slots[0])) + /* if our item isn't there, return now */ + if (item_size != btrfs_item_size_nr(leaf, path->slots[0])) goto err; /* the leaf has changed, it now has room. return now */ diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 0b18070..f9c89ca 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -198,6 +198,8 @@ static int btrfs_csum_sizes[] = { 4, 0 }; #define BTRFS_DIRTY_METADATA_THRESH (32 * 1024 * 1024) +#define BTRFS_MAX_EXTENT_SIZE (128 * 1024 * 1024) + /* * The key defines the order in the tree, and so it also defines (optimal) * block layout. @@ -1020,6 +1022,9 @@ enum btrfs_raid_types { BTRFS_BLOCK_GROUP_RAID6 | \ BTRFS_BLOCK_GROUP_DUP | \ BTRFS_BLOCK_GROUP_RAID10) +#define BTRFS_BLOCK_GROUP_RAID56_MASK (BTRFS_BLOCK_GROUP_RAID5 | \ + BTRFS_BLOCK_GROUP_RAID6) + /* * We need a bit for restriper to be able to tell when chunks of type * SINGLE are available. This "extended" profile format is used in @@ -1239,7 +1244,6 @@ enum btrfs_disk_cache_state { BTRFS_DC_ERROR = 1, BTRFS_DC_CLEAR = 2, BTRFS_DC_SETUP = 3, - BTRFS_DC_NEED_WRITE = 4, }; struct btrfs_caching_control { @@ -1277,7 +1281,6 @@ struct btrfs_block_group_cache { unsigned long full_stripe_len; unsigned int ro:1; - unsigned int dirty:1; unsigned int iref:1; unsigned int has_caching_ctl:1; unsigned int removed:1; @@ -1315,6 +1318,9 @@ struct btrfs_block_group_cache { struct list_head ro_list; atomic_t trimming; + + /* For dirty block groups */ + struct list_head dirty_list; }; /* delayed seq elem */ @@ -1741,6 +1747,7 @@ struct btrfs_fs_info { spinlock_t unused_bgs_lock; struct list_head unused_bgs; + struct mutex unused_bg_unpin_mutex; /* For btrfs to record security options */ struct security_mnt_opts security_opts; @@ -1776,6 +1783,7 @@ struct btrfs_subvolume_writers { #define BTRFS_ROOT_DEFRAG_RUNNING 6 #define BTRFS_ROOT_FORCE_COW 7 #define BTRFS_ROOT_MULTI_LOG_TASKS 8 +#define BTRFS_ROOT_DIRTY 9 /* * in ram representation of the tree. extent_root is used for all allocations @@ -1794,8 +1802,6 @@ struct btrfs_root { struct btrfs_fs_info *fs_info; struct extent_io_tree dirty_log_pages; - struct kobject root_kobj; - struct completion kobj_unregister; struct mutex objectid_mutex; spinlock_t accounting_lock; @@ -2465,31 +2471,6 @@ BTRFS_SETGET_STACK_FUNCS(stack_inode_gid, struct btrfs_inode_item, gid, 32); BTRFS_SETGET_STACK_FUNCS(stack_inode_mode, struct btrfs_inode_item, mode, 32); BTRFS_SETGET_STACK_FUNCS(stack_inode_rdev, struct btrfs_inode_item, rdev, 64); BTRFS_SETGET_STACK_FUNCS(stack_inode_flags, struct btrfs_inode_item, flags, 64); - -static inline struct btrfs_timespec * -btrfs_inode_atime(struct btrfs_inode_item *inode_item) -{ - unsigned long ptr = (unsigned long)inode_item; - ptr += offsetof(struct btrfs_inode_item, atime); - return (struct btrfs_timespec *)ptr; -} - -static inline struct btrfs_timespec * -btrfs_inode_mtime(struct btrfs_inode_item *inode_item) -{ - unsigned long ptr = (unsigned long)inode_item; - ptr += offsetof(struct btrfs_inode_item, mtime); - return (struct btrfs_timespec *)ptr; -} - -static inline struct btrfs_timespec * -btrfs_inode_ctime(struct btrfs_inode_item *inode_item) -{ - unsigned long ptr = (unsigned long)inode_item; - ptr += offsetof(struct btrfs_inode_item, ctime); - return (struct btrfs_timespec *)ptr; -} - BTRFS_SETGET_FUNCS(timespec_sec, struct btrfs_timespec, sec, 64); BTRFS_SETGET_FUNCS(timespec_nsec, struct btrfs_timespec, nsec, 32); BTRFS_SETGET_STACK_FUNCS(stack_timespec_sec, struct btrfs_timespec, sec, 64); @@ -3406,6 +3387,8 @@ int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, struct btrfs_root *root); +int btrfs_setup_space_cache(struct btrfs_trans_handle *trans, + struct btrfs_root *root); int btrfs_extent_readonly(struct btrfs_root *root, u64 bytenr); int btrfs_free_block_groups(struct btrfs_fs_info *info); int btrfs_read_block_groups(struct btrfs_root *root); @@ -3928,6 +3911,9 @@ int btrfs_prealloc_file_range_trans(struct inode *inode, loff_t actual_len, u64 *alloc_hint); int btrfs_inode_check_errors(struct inode *inode); extern const struct dentry_operations btrfs_dentry_operations; +#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS +void btrfs_test_inode_set_ops(struct inode *inode); +#endif /* ioctl.c */ long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg); diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c index de4e70f..82f0c7c 100644 --- a/fs/btrfs/delayed-inode.c +++ b/fs/btrfs/delayed-inode.c @@ -1755,27 +1755,31 @@ static void fill_stack_inode_item(struct btrfs_trans_handle *trans, btrfs_set_stack_inode_flags(inode_item, BTRFS_I(inode)->flags); btrfs_set_stack_inode_block_group(inode_item, 0); - btrfs_set_stack_timespec_sec(btrfs_inode_atime(inode_item), + btrfs_set_stack_timespec_sec(&inode_item->atime, inode->i_atime.tv_sec); - btrfs_set_stack_timespec_nsec(btrfs_inode_atime(inode_item), + btrfs_set_stack_timespec_nsec(&inode_item->atime, inode->i_atime.tv_nsec); - btrfs_set_stack_timespec_sec(btrfs_inode_mtime(inode_item), + btrfs_set_stack_timespec_sec(&inode_item->mtime, inode->i_mtime.tv_sec); - btrfs_set_stack_timespec_nsec(btrfs_inode_mtime(inode_item), + btrfs_set_stack_timespec_nsec(&inode_item->mtime, inode->i_mtime.tv_nsec); - btrfs_set_stack_timespec_sec(btrfs_inode_ctime(inode_item), + btrfs_set_stack_timespec_sec(&inode_item->ctime, inode->i_ctime.tv_sec); - btrfs_set_stack_timespec_nsec(btrfs_inode_ctime(inode_item), + btrfs_set_stack_timespec_nsec(&inode_item->ctime, inode->i_ctime.tv_nsec); + + btrfs_set_stack_timespec_sec(&inode_item->otime, + BTRFS_I(inode)->i_otime.tv_sec); + btrfs_set_stack_timespec_nsec(&inode_item->otime, + BTRFS_I(inode)->i_otime.tv_nsec); } int btrfs_fill_inode(struct inode *inode, u32 *rdev) { struct btrfs_delayed_node *delayed_node; struct btrfs_inode_item *inode_item; - struct btrfs_timespec *tspec; delayed_node = btrfs_get_delayed_node(inode); if (!delayed_node) @@ -1802,17 +1806,19 @@ int btrfs_fill_inode(struct inode *inode, u32 *rdev) *rdev = btrfs_stack_inode_rdev(inode_item); BTRFS_I(inode)->flags = btrfs_stack_inode_flags(inode_item); - tspec = btrfs_inode_atime(inode_item); - inode->i_atime.tv_sec = btrfs_stack_timespec_sec(tspec); - inode->i_atime.tv_nsec = btrfs_stack_timespec_nsec(tspec); + inode->i_atime.tv_sec = btrfs_stack_timespec_sec(&inode_item->atime); + inode->i_atime.tv_nsec = btrfs_stack_timespec_nsec(&inode_item->atime); + + inode->i_mtime.tv_sec = btrfs_stack_timespec_sec(&inode_item->mtime); + inode->i_mtime.tv_nsec = btrfs_stack_timespec_nsec(&inode_item->mtime); - tspec = btrfs_inode_mtime(inode_item); - inode->i_mtime.tv_sec = btrfs_stack_timespec_sec(tspec); - inode->i_mtime.tv_nsec = btrfs_stack_timespec_nsec(tspec); + inode->i_ctime.tv_sec = btrfs_stack_timespec_sec(&inode_item->ctime); + inode->i_ctime.tv_nsec = btrfs_stack_timespec_nsec(&inode_item->ctime); - tspec = btrfs_inode_ctime(inode_item); - inode->i_ctime.tv_sec = btrfs_stack_timespec_sec(tspec); - inode->i_ctime.tv_nsec = btrfs_stack_timespec_nsec(tspec); + BTRFS_I(inode)->i_otime.tv_sec = + btrfs_stack_timespec_sec(&inode_item->otime); + BTRFS_I(inode)->i_otime.tv_nsec = + btrfs_stack_timespec_nsec(&inode_item->otime); inode->i_generation = BTRFS_I(inode)->generation; BTRFS_I(inode)->index_cnt = (u64)-1; diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c index ca6a3a3..5ec03d9 100644 --- a/fs/btrfs/dev-replace.c +++ b/fs/btrfs/dev-replace.c @@ -440,18 +440,9 @@ leave: */ static void btrfs_rm_dev_replace_blocked(struct btrfs_fs_info *fs_info) { - s64 writers; - DEFINE_WAIT(wait); - set_bit(BTRFS_FS_STATE_DEV_REPLACING, &fs_info->fs_state); - do { - prepare_to_wait(&fs_info->replace_wait, &wait, - TASK_UNINTERRUPTIBLE); - writers = percpu_counter_sum(&fs_info->bio_counter); - if (writers) - schedule(); - finish_wait(&fs_info->replace_wait, &wait); - } while (writers); + wait_event(fs_info->replace_wait, !percpu_counter_sum( + &fs_info->bio_counter)); } /* @@ -932,15 +923,15 @@ void btrfs_bio_counter_sub(struct btrfs_fs_info *fs_info, s64 amount) void btrfs_bio_counter_inc_blocked(struct btrfs_fs_info *fs_info) { - DEFINE_WAIT(wait); -again: - percpu_counter_inc(&fs_info->bio_counter); - if (test_bit(BTRFS_FS_STATE_DEV_REPLACING, &fs_info->fs_state)) { + while (1) { + percpu_counter_inc(&fs_info->bio_counter); + if (likely(!test_bit(BTRFS_FS_STATE_DEV_REPLACING, + &fs_info->fs_state))) + break; + btrfs_bio_counter_dec(fs_info); wait_event(fs_info->replace_wait, !test_bit(BTRFS_FS_STATE_DEV_REPLACING, &fs_info->fs_state)); - goto again; } - } diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 1afb182..639f266 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -318,7 +318,7 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf, memcpy(&found, result, csum_size); read_extent_buffer(buf, &val, 0, csum_size); - printk_ratelimited(KERN_INFO + printk_ratelimited(KERN_WARNING "BTRFS: %s checksum verify failed on %llu wanted %X found %X " "level %d\n", root->fs_info->sb->s_id, buf->start, @@ -367,7 +367,8 @@ static int verify_parent_transid(struct extent_io_tree *io_tree, ret = 0; goto out; } - printk_ratelimited(KERN_INFO "BTRFS (device %s): parent transid verify failed on %llu wanted %llu found %llu\n", + printk_ratelimited(KERN_ERR + "BTRFS (device %s): parent transid verify failed on %llu wanted %llu found %llu\n", eb->fs_info->sb->s_id, eb->start, parent_transid, btrfs_header_generation(eb)); ret = 1; @@ -633,21 +634,21 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio, found_start = btrfs_header_bytenr(eb); if (found_start != eb->start) { - printk_ratelimited(KERN_INFO "BTRFS (device %s): bad tree block start " + printk_ratelimited(KERN_ERR "BTRFS (device %s): bad tree block start " "%llu %llu\n", eb->fs_info->sb->s_id, found_start, eb->start); ret = -EIO; goto err; } if (check_tree_block_fsid(root, eb)) { - printk_ratelimited(KERN_INFO "BTRFS (device %s): bad fsid on block %llu\n", + printk_ratelimited(KERN_ERR "BTRFS (device %s): bad fsid on block %llu\n", eb->fs_info->sb->s_id, eb->start); ret = -EIO; goto err; } found_level = btrfs_header_level(eb); if (found_level >= BTRFS_MAX_LEVEL) { - btrfs_info(root->fs_info, "bad tree block level %d", + btrfs_err(root->fs_info, "bad tree block level %d", (int)btrfs_header_level(eb)); ret = -EIO; goto err; @@ -1073,12 +1074,12 @@ static const struct address_space_operations btree_aops = { .set_page_dirty = btree_set_page_dirty, }; -void readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize) +void readahead_tree_block(struct btrfs_root *root, u64 bytenr) { struct extent_buffer *buf = NULL; struct inode *btree_inode = root->fs_info->btree_inode; - buf = btrfs_find_create_tree_block(root, bytenr, blocksize); + buf = btrfs_find_create_tree_block(root, bytenr); if (!buf) return; read_extent_buffer_pages(&BTRFS_I(btree_inode)->io_tree, @@ -1086,7 +1087,7 @@ void readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize) free_extent_buffer(buf); } -int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize, +int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, int mirror_num, struct extent_buffer **eb) { struct extent_buffer *buf = NULL; @@ -1094,7 +1095,7 @@ int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize, struct extent_io_tree *io_tree = &BTRFS_I(btree_inode)->io_tree; int ret; - buf = btrfs_find_create_tree_block(root, bytenr, blocksize); + buf = btrfs_find_create_tree_block(root, bytenr); if (!buf) return 0; @@ -1125,12 +1126,11 @@ struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root, } struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root, - u64 bytenr, u32 blocksize) + u64 bytenr) { if (btrfs_test_is_dummy_root(root)) - return alloc_test_extent_buffer(root->fs_info, bytenr, - blocksize); - return alloc_extent_buffer(root->fs_info, bytenr, blocksize); + return alloc_test_extent_buffer(root->fs_info, bytenr); + return alloc_extent_buffer(root->fs_info, bytenr); } @@ -1152,7 +1152,7 @@ struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr, struct extent_buffer *buf = NULL; int ret; - buf = btrfs_find_create_tree_block(root, bytenr, root->nodesize); + buf = btrfs_find_create_tree_block(root, bytenr); if (!buf) return NULL; @@ -1275,12 +1275,10 @@ static void __setup_root(u32 nodesize, u32 sectorsize, u32 stripesize, memset(&root->root_key, 0, sizeof(root->root_key)); memset(&root->root_item, 0, sizeof(root->root_item)); memset(&root->defrag_progress, 0, sizeof(root->defrag_progress)); - memset(&root->root_kobj, 0, sizeof(root->root_kobj)); if (fs_info) root->defrag_trans_start = fs_info->generation; else root->defrag_trans_start = 0; - init_completion(&root->kobj_unregister); root->root_key.objectid = objectid; root->anon_dev = 0; @@ -1630,6 +1628,8 @@ struct btrfs_root *btrfs_get_fs_root(struct btrfs_fs_info *fs_info, bool check_ref) { struct btrfs_root *root; + struct btrfs_path *path; + struct btrfs_key key; int ret; if (location->objectid == BTRFS_ROOT_TREE_OBJECTID) @@ -1669,8 +1669,17 @@ again: if (ret) goto fail; - ret = btrfs_find_item(fs_info->tree_root, NULL, BTRFS_ORPHAN_OBJECTID, - location->objectid, BTRFS_ORPHAN_ITEM_KEY, NULL); + path = btrfs_alloc_path(); + if (!path) { + ret = -ENOMEM; + goto fail; + } + key.objectid = BTRFS_ORPHAN_OBJECTID; + key.type = BTRFS_ORPHAN_ITEM_KEY; + key.offset = location->objectid; + + ret = btrfs_search_slot(NULL, fs_info->tree_root, &key, path, 0, 0); + btrfs_free_path(path); if (ret < 0) goto fail; if (ret == 0) @@ -2232,6 +2241,7 @@ int open_ctree(struct super_block *sb, spin_lock_init(&fs_info->qgroup_op_lock); spin_lock_init(&fs_info->buffer_lock); spin_lock_init(&fs_info->unused_bgs_lock); + mutex_init(&fs_info->unused_bg_unpin_mutex); rwlock_init(&fs_info->tree_mod_log_lock); mutex_init(&fs_info->reloc_mutex); mutex_init(&fs_info->delalloc_root_mutex); @@ -2496,7 +2506,7 @@ int open_ctree(struct super_block *sb, features |= BTRFS_FEATURE_INCOMPAT_COMPRESS_LZO; if (features & BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA) - printk(KERN_ERR "BTRFS: has skinny extents\n"); + printk(KERN_INFO "BTRFS: has skinny extents\n"); /* * flag our filesystem as having big metadata blocks if @@ -2520,7 +2530,7 @@ int open_ctree(struct super_block *sb, */ if ((features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) && (sectorsize != nodesize)) { - printk(KERN_WARNING "BTRFS: unequal leaf/node/sector sizes " + printk(KERN_ERR "BTRFS: unequal leaf/node/sector sizes " "are not allowed for mixed block groups on %s\n", sb->s_id); goto fail_alloc; @@ -2628,12 +2638,12 @@ int open_ctree(struct super_block *sb, sb->s_blocksize_bits = blksize_bits(sectorsize); if (btrfs_super_magic(disk_super) != BTRFS_MAGIC) { - printk(KERN_INFO "BTRFS: valid FS not found on %s\n", sb->s_id); + printk(KERN_ERR "BTRFS: valid FS not found on %s\n", sb->s_id); goto fail_sb_buffer; } if (sectorsize != PAGE_SIZE) { - printk(KERN_WARNING "BTRFS: Incompatible sector size(%lu) " + printk(KERN_ERR "BTRFS: incompatible sector size (%lu) " "found on %s\n", (unsigned long)sectorsize, sb->s_id); goto fail_sb_buffer; } @@ -2642,7 +2652,7 @@ int open_ctree(struct super_block *sb, ret = btrfs_read_sys_array(tree_root); mutex_unlock(&fs_info->chunk_mutex); if (ret) { - printk(KERN_WARNING "BTRFS: failed to read the system " + printk(KERN_ERR "BTRFS: failed to read the system " "array on %s\n", sb->s_id); goto fail_sb_buffer; } @@ -2657,7 +2667,7 @@ int open_ctree(struct super_block *sb, generation); if (!chunk_root->node || !test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) { - printk(KERN_WARNING "BTRFS: failed to read chunk root on %s\n", + printk(KERN_ERR "BTRFS: failed to read chunk root on %s\n", sb->s_id); goto fail_tree_roots; } @@ -2669,7 +2679,7 @@ int open_ctree(struct super_block *sb, ret = btrfs_read_chunk_tree(chunk_root); if (ret) { - printk(KERN_WARNING "BTRFS: failed to read chunk tree on %s\n", + printk(KERN_ERR "BTRFS: failed to read chunk tree on %s\n", sb->s_id); goto fail_tree_roots; } @@ -2681,7 +2691,7 @@ int open_ctree(struct super_block *sb, btrfs_close_extra_devices(fs_info, fs_devices, 0); if (!fs_devices->latest_bdev) { - printk(KERN_CRIT "BTRFS: failed to read devices on %s\n", + printk(KERN_ERR "BTRFS: failed to read devices on %s\n", sb->s_id); goto fail_tree_roots; } @@ -2765,7 +2775,7 @@ retry_root_backup: ret = btrfs_recover_balance(fs_info); if (ret) { - printk(KERN_WARNING "BTRFS: failed to recover balance\n"); + printk(KERN_ERR "BTRFS: failed to recover balance\n"); goto fail_block_groups; } @@ -3860,6 +3870,21 @@ static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info, printk(KERN_WARNING "BTRFS: log_root block unaligned: %llu\n", btrfs_super_log_root(sb)); + /* + * Check the lower bound, the alignment and other constraints are + * checked later. + */ + if (btrfs_super_nodesize(sb) < 4096) { + printk(KERN_ERR "BTRFS: nodesize too small: %u < 4096\n", + btrfs_super_nodesize(sb)); + ret = -EINVAL; + } + if (btrfs_super_sectorsize(sb) < 4096) { + printk(KERN_ERR "BTRFS: sectorsize too small: %u < 4096\n", + btrfs_super_sectorsize(sb)); + ret = -EINVAL; + } + if (memcmp(fs_info->fsid, sb->dev_item.fsid, BTRFS_UUID_SIZE) != 0) { printk(KERN_ERR "BTRFS: dev_item UUID does not match fsid: %pU != %pU\n", fs_info->fsid, sb->dev_item.fsid); @@ -3873,6 +3898,10 @@ static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info, if (btrfs_super_num_devices(sb) > (1UL << 31)) printk(KERN_WARNING "BTRFS: suspicious number of devices: %llu\n", btrfs_super_num_devices(sb)); + if (btrfs_super_num_devices(sb) == 0) { + printk(KERN_ERR "BTRFS: number of devices is 0\n"); + ret = -EINVAL; + } if (btrfs_super_bytenr(sb) != BTRFS_SUPER_INFO_OFFSET) { printk(KERN_ERR "BTRFS: super offset mismatch %llu != %u\n", @@ -3881,6 +3910,25 @@ static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info, } /* + * Obvious sys_chunk_array corruptions, it must hold at least one key + * and one chunk + */ + if (btrfs_super_sys_array_size(sb) > BTRFS_SYSTEM_CHUNK_ARRAY_SIZE) { + printk(KERN_ERR "BTRFS: system chunk array too big %u > %u\n", + btrfs_super_sys_array_size(sb), + BTRFS_SYSTEM_CHUNK_ARRAY_SIZE); + ret = -EINVAL; + } + if (btrfs_super_sys_array_size(sb) < sizeof(struct btrfs_disk_key) + + sizeof(struct btrfs_chunk)) { + printk(KERN_ERR "BTRFS: system chunk array too small %u < %zu\n", + btrfs_super_sys_array_size(sb), + sizeof(struct btrfs_disk_key) + + sizeof(struct btrfs_chunk)); + ret = -EINVAL; + } + + /* * The generation is a global counter, we'll trust it more than the others * but it's still possible that it's the one that's wrong. */ diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 4146518..27d44c0 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -46,11 +46,11 @@ struct btrfs_fs_devices; struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr, u64 parent_transid); -void readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize); -int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize, +void readahead_tree_block(struct btrfs_root *root, u64 bytenr); +int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, int mirror_num, struct extent_buffer **eb); struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root, - u64 bytenr, u32 blocksize); + u64 bytenr); void clean_tree_block(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf); int open_ctree(struct super_block *sb, diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index a684086..8b353ad 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -74,8 +74,9 @@ enum { RESERVE_ALLOC_NO_ACCOUNT = 2, }; -static int update_block_group(struct btrfs_root *root, - u64 bytenr, u64 num_bytes, int alloc); +static int update_block_group(struct btrfs_trans_handle *trans, + struct btrfs_root *root, u64 bytenr, + 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, @@ -1925,7 +1926,7 @@ int btrfs_discard_extent(struct btrfs_root *root, u64 bytenr, */ ret = 0; } - kfree(bbio); + btrfs_put_bbio(bbio); } if (actual_bytes) @@ -2768,7 +2769,6 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head *head; int ret; int run_all = count == (unsigned long)-1; - int run_most = 0; /* We'll clean this up in btrfs_cleanup_transaction */ if (trans->aborted) @@ -2778,10 +2778,8 @@ int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, root = root->fs_info->tree_root; delayed_refs = &trans->transaction->delayed_refs; - if (count == 0) { + if (count == 0) count = atomic_read(&delayed_refs->num_entries) * 2; - run_most = 1; - } again: #ifdef SCRAMBLE_DELAYED_REFS @@ -3210,6 +3208,8 @@ static int cache_save_setup(struct btrfs_block_group_cache *block_group, return 0; } + if (trans->aborted) + return 0; again: inode = lookup_free_space_inode(root, block_group, path); if (IS_ERR(inode) && PTR_ERR(inode) != -ENOENT) { @@ -3245,6 +3245,20 @@ again: */ BTRFS_I(inode)->generation = 0; ret = btrfs_update_inode(trans, root, inode); + if (ret) { + /* + * So theoretically we could recover from this, simply set the + * super cache generation to 0 so we know to invalidate the + * cache, but then we'd have to keep track of the block groups + * that fail this way so we know we _have_ to reset this cache + * before the next commit or risk reading stale cache. So to + * limit our exposure to horrible edge cases lets just abort the + * transaction, this only happens in really bad situations + * anyway. + */ + btrfs_abort_transaction(trans, root, ret); + goto out_put; + } WARN_ON(ret); if (i_size_read(inode) > 0) { @@ -3311,124 +3325,72 @@ out: return ret; } -int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, - struct btrfs_root *root) +int btrfs_setup_space_cache(struct btrfs_trans_handle *trans, + struct btrfs_root *root) { - struct btrfs_block_group_cache *cache; - int err = 0; + struct btrfs_block_group_cache *cache, *tmp; + struct btrfs_transaction *cur_trans = trans->transaction; struct btrfs_path *path; - u64 last = 0; + + if (list_empty(&cur_trans->dirty_bgs) || + !btrfs_test_opt(root, SPACE_CACHE)) + return 0; path = btrfs_alloc_path(); if (!path) return -ENOMEM; -again: - while (1) { - cache = btrfs_lookup_first_block_group(root->fs_info, last); - while (cache) { - if (cache->disk_cache_state == BTRFS_DC_CLEAR) - break; - cache = next_block_group(root, cache); - } - if (!cache) { - if (last == 0) - break; - last = 0; - continue; - } - err = cache_save_setup(cache, trans, path); - last = cache->key.objectid + cache->key.offset; - btrfs_put_block_group(cache); + /* Could add new block groups, use _safe just in case */ + list_for_each_entry_safe(cache, tmp, &cur_trans->dirty_bgs, + dirty_list) { + if (cache->disk_cache_state == BTRFS_DC_CLEAR) + cache_save_setup(cache, trans, path); } - while (1) { - if (last == 0) { - err = btrfs_run_delayed_refs(trans, root, - (unsigned long)-1); - if (err) /* File system offline */ - goto out; - } - - cache = btrfs_lookup_first_block_group(root->fs_info, last); - while (cache) { - if (cache->disk_cache_state == BTRFS_DC_CLEAR) { - btrfs_put_block_group(cache); - goto again; - } - - if (cache->dirty) - break; - cache = next_block_group(root, cache); - } - if (!cache) { - if (last == 0) - break; - last = 0; - continue; - } - - if (cache->disk_cache_state == BTRFS_DC_SETUP) - cache->disk_cache_state = BTRFS_DC_NEED_WRITE; - cache->dirty = 0; - last = cache->key.objectid + cache->key.offset; - - err = write_one_cache_group(trans, root, path, cache); - btrfs_put_block_group(cache); - if (err) /* File system offline */ - goto out; - } + btrfs_free_path(path); + return 0; +} - while (1) { - /* - * I don't think this is needed since we're just marking our - * preallocated extent as written, but just in case it can't - * hurt. - */ - if (last == 0) { - err = btrfs_run_delayed_refs(trans, root, - (unsigned long)-1); - if (err) /* File system offline */ - goto out; - } +int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, + struct btrfs_root *root) +{ + struct btrfs_block_group_cache *cache; + struct btrfs_transaction *cur_trans = trans->transaction; + int ret = 0; + struct btrfs_path *path; - cache = btrfs_lookup_first_block_group(root->fs_info, last); - while (cache) { - /* - * Really this shouldn't happen, but it could if we - * couldn't write the entire preallocated extent and - * splitting the extent resulted in a new block. - */ - if (cache->dirty) { - btrfs_put_block_group(cache); - goto again; - } - if (cache->disk_cache_state == BTRFS_DC_NEED_WRITE) - break; - cache = next_block_group(root, cache); - } - if (!cache) { - if (last == 0) - break; - last = 0; - continue; - } + if (list_empty(&cur_trans->dirty_bgs)) + return 0; - err = btrfs_write_out_cache(root, trans, cache, path); + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; - /* - * If we didn't have an error then the cache state is still - * NEED_WRITE, so we can set it to WRITTEN. - */ - if (!err && cache->disk_cache_state == BTRFS_DC_NEED_WRITE) - cache->disk_cache_state = BTRFS_DC_WRITTEN; - last = cache->key.objectid + cache->key.offset; + /* + * We don't need the lock here since we are protected by the transaction + * commit. We want to do the cache_save_setup first and then run the + * delayed refs to make sure we have the best chance at doing this all + * in one shot. + */ + while (!list_empty(&cur_trans->dirty_bgs)) { + cache = list_first_entry(&cur_trans->dirty_bgs, + struct btrfs_block_group_cache, + dirty_list); + list_del_init(&cache->dirty_list); + if (cache->disk_cache_state == BTRFS_DC_CLEAR) + cache_save_setup(cache, trans, path); + if (!ret) + ret = btrfs_run_delayed_refs(trans, root, + (unsigned long) -1); + if (!ret && cache->disk_cache_state == BTRFS_DC_SETUP) + btrfs_write_out_cache(root, trans, cache, path); + if (!ret) + ret = write_one_cache_group(trans, root, path, cache); btrfs_put_block_group(cache); } -out: btrfs_free_path(path); - return err; + return ret; } int btrfs_extent_readonly(struct btrfs_root *root, u64 bytenr) @@ -5043,19 +5005,25 @@ void btrfs_subvolume_release_metadata(struct btrfs_root *root, /** * drop_outstanding_extent - drop an outstanding extent * @inode: the inode we're dropping the extent for + * @num_bytes: the number of bytes we're relaseing. * * This is called when we are freeing up an outstanding extent, either called * after an error or after an extent is written. This will return the number of * reserved extents that need to be freed. This must be called with * BTRFS_I(inode)->lock held. */ -static unsigned drop_outstanding_extent(struct inode *inode) +static unsigned drop_outstanding_extent(struct inode *inode, u64 num_bytes) { unsigned drop_inode_space = 0; unsigned dropped_extents = 0; + unsigned num_extents = 0; - BUG_ON(!BTRFS_I(inode)->outstanding_extents); - BTRFS_I(inode)->outstanding_extents--; + num_extents = (unsigned)div64_u64(num_bytes + + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + ASSERT(num_extents); + ASSERT(BTRFS_I(inode)->outstanding_extents >= num_extents); + BTRFS_I(inode)->outstanding_extents -= num_extents; if (BTRFS_I(inode)->outstanding_extents == 0 && test_and_clear_bit(BTRFS_INODE_DELALLOC_META_RESERVED, @@ -5168,7 +5136,11 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) num_bytes = ALIGN(num_bytes, root->sectorsize); spin_lock(&BTRFS_I(inode)->lock); - BTRFS_I(inode)->outstanding_extents++; + nr_extents = (unsigned)div64_u64(num_bytes + + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + BTRFS_I(inode)->outstanding_extents += nr_extents; + nr_extents = 0; if (BTRFS_I(inode)->outstanding_extents > BTRFS_I(inode)->reserved_extents) @@ -5226,7 +5198,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) out_fail: spin_lock(&BTRFS_I(inode)->lock); - dropped = drop_outstanding_extent(inode); + dropped = drop_outstanding_extent(inode, num_bytes); /* * If the inodes csum_bytes is the same as the original * csum_bytes then we know we haven't raced with any free()ers @@ -5305,7 +5277,7 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) num_bytes = ALIGN(num_bytes, root->sectorsize); spin_lock(&BTRFS_I(inode)->lock); - dropped = drop_outstanding_extent(inode); + dropped = drop_outstanding_extent(inode, num_bytes); if (num_bytes) to_free = calc_csum_metadata_size(inode, num_bytes, 0); @@ -5313,6 +5285,9 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) if (dropped > 0) to_free += btrfs_calc_trans_metadata_size(root, dropped); + if (btrfs_test_is_dummy_root(root)) + return; + trace_btrfs_space_reservation(root->fs_info, "delalloc", btrfs_ino(inode), to_free, 0); if (root->fs_info->quota_enabled) { @@ -5375,8 +5350,9 @@ void btrfs_delalloc_release_space(struct inode *inode, u64 num_bytes) btrfs_free_reserved_data_space(inode, num_bytes); } -static int update_block_group(struct btrfs_root *root, - u64 bytenr, u64 num_bytes, int alloc) +static int update_block_group(struct btrfs_trans_handle *trans, + struct btrfs_root *root, u64 bytenr, + u64 num_bytes, int alloc) { struct btrfs_block_group_cache *cache = NULL; struct btrfs_fs_info *info = root->fs_info; @@ -5414,6 +5390,14 @@ static int update_block_group(struct btrfs_root *root, if (!alloc && cache->cached == BTRFS_CACHE_NO) cache_block_group(cache, 1); + spin_lock(&trans->transaction->dirty_bgs_lock); + if (list_empty(&cache->dirty_list)) { + list_add_tail(&cache->dirty_list, + &trans->transaction->dirty_bgs); + btrfs_get_block_group(cache); + } + spin_unlock(&trans->transaction->dirty_bgs_lock); + byte_in_group = bytenr - cache->key.objectid; WARN_ON(byte_in_group > cache->key.offset); @@ -5424,7 +5408,6 @@ static int update_block_group(struct btrfs_root *root, cache->disk_cache_state < BTRFS_DC_CLEAR) cache->disk_cache_state = BTRFS_DC_CLEAR; - cache->dirty = 1; old_val = btrfs_block_group_used(&cache->item); num_bytes = min(total, cache->key.offset - byte_in_group); if (alloc) { @@ -5807,10 +5790,13 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, unpin = &fs_info->freed_extents[0]; while (1) { + mutex_lock(&fs_info->unused_bg_unpin_mutex); ret = find_first_extent_bit(unpin, 0, &start, &end, EXTENT_DIRTY, NULL); - if (ret) + if (ret) { + mutex_unlock(&fs_info->unused_bg_unpin_mutex); break; + } if (btrfs_test_opt(root, DISCARD)) ret = btrfs_discard_extent(root, start, @@ -5818,6 +5804,7 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans, clear_extent_dirty(unpin, start, end, GFP_NOFS); unpin_extent_range(root, start, end, true); + mutex_unlock(&fs_info->unused_bg_unpin_mutex); cond_resched(); } @@ -6103,7 +6090,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, } } - ret = update_block_group(root, bytenr, num_bytes, 0); + ret = update_block_group(trans, root, bytenr, num_bytes, 0); if (ret) { btrfs_abort_transaction(trans, extent_root, ret); goto out; @@ -6205,7 +6192,6 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, struct extent_buffer *buf, u64 parent, int last_ref) { - struct btrfs_block_group_cache *cache = NULL; int pin = 1; int ret; @@ -6221,17 +6207,20 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, if (!last_ref) return; - cache = btrfs_lookup_block_group(root->fs_info, buf->start); - if (btrfs_header_generation(buf) == trans->transid) { + struct btrfs_block_group_cache *cache; + if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) { ret = check_ref_cleanup(trans, root, buf->start); if (!ret) goto out; } + cache = btrfs_lookup_block_group(root->fs_info, buf->start); + if (btrfs_header_flag(buf, BTRFS_HEADER_FLAG_WRITTEN)) { pin_down_extent(root, cache, buf->start, buf->len, 1); + btrfs_put_block_group(cache); goto out; } @@ -6239,6 +6228,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, btrfs_add_free_space(cache, buf->start, buf->len); btrfs_update_reserved_bytes(cache, buf->len, RESERVE_FREE, 0); + btrfs_put_block_group(cache); trace_btrfs_reserved_extent_free(root, buf->start, buf->len); pin = 0; } @@ -6253,7 +6243,6 @@ out: * anymore. */ clear_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags); - btrfs_put_block_group(cache); } /* Can return -ENOMEM */ @@ -7063,7 +7052,7 @@ static int alloc_reserved_file_extent(struct btrfs_trans_handle *trans, if (ret) return ret; - ret = update_block_group(root, ins->objectid, ins->offset, 1); + 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", ins->objectid, ins->offset); @@ -7152,7 +7141,8 @@ static int alloc_reserved_tree_block(struct btrfs_trans_handle *trans, return ret; } - ret = update_block_group(root, ins->objectid, root->nodesize, 1); + ret = update_block_group(trans, root, ins->objectid, root->nodesize, + 1); if (ret) { /* -ENOENT, logic error */ btrfs_err(fs_info, "update block group failed for %llu %llu", ins->objectid, ins->offset); @@ -7217,11 +7207,11 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans, static struct extent_buffer * btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root, - u64 bytenr, u32 blocksize, int level) + u64 bytenr, int level) { struct extent_buffer *buf; - buf = btrfs_find_create_tree_block(root, bytenr, blocksize); + buf = btrfs_find_create_tree_block(root, bytenr); if (!buf) return ERR_PTR(-ENOMEM); btrfs_set_header_generation(buf, trans->transid); @@ -7340,7 +7330,7 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans, if (btrfs_test_is_dummy_root(root)) { buf = btrfs_init_new_buffer(trans, root, root->alloc_bytenr, - blocksize, level); + level); if (!IS_ERR(buf)) root->alloc_bytenr += blocksize; return buf; @@ -7357,8 +7347,7 @@ struct extent_buffer *btrfs_alloc_tree_block(struct btrfs_trans_handle *trans, return ERR_PTR(ret); } - buf = btrfs_init_new_buffer(trans, root, ins.objectid, - blocksize, level); + buf = btrfs_init_new_buffer(trans, root, ins.objectid, level); BUG_ON(IS_ERR(buf)); /* -ENOMEM */ if (root_objectid == BTRFS_TREE_RELOC_OBJECTID) { @@ -7487,7 +7476,7 @@ static noinline void reada_walk_down(struct btrfs_trans_handle *trans, continue; } reada: - readahead_tree_block(root, bytenr, blocksize); + readahead_tree_block(root, bytenr); nread++; } wc->reada_slot = slot; @@ -7828,7 +7817,7 @@ static noinline int do_walk_down(struct btrfs_trans_handle *trans, next = btrfs_find_tree_block(root, bytenr); if (!next) { - next = btrfs_find_create_tree_block(root, bytenr, blocksize); + next = btrfs_find_create_tree_block(root, bytenr); if (!next) return -ENOMEM; btrfs_set_buffer_lockdep_class(root->root_key.objectid, next, @@ -8548,14 +8537,6 @@ int btrfs_set_block_group_ro(struct btrfs_root *root, if (IS_ERR(trans)) return PTR_ERR(trans); - 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); - if (ret < 0) - goto out; - } - ret = set_block_group_ro(cache, 0); if (!ret) goto out; @@ -8566,6 +8547,11 @@ int btrfs_set_block_group_ro(struct btrfs_root *root, goto out; ret = set_block_group_ro(cache, 0); out: + if (cache->flags & BTRFS_BLOCK_GROUP_SYSTEM) { + alloc_flags = update_block_group_flags(root, cache->flags); + check_system_chunk(trans, root, alloc_flags); + } + btrfs_end_transaction(trans, root); return ret; } @@ -9005,6 +8991,7 @@ btrfs_create_block_group_cache(struct btrfs_root *root, u64 start, u64 size) INIT_LIST_HEAD(&cache->cluster_list); INIT_LIST_HEAD(&cache->bg_list); INIT_LIST_HEAD(&cache->ro_list); + INIT_LIST_HEAD(&cache->dirty_list); btrfs_init_free_space_ctl(cache); atomic_set(&cache->trimming, 0); @@ -9068,9 +9055,8 @@ int btrfs_read_block_groups(struct btrfs_root *root) * b) Setting 'dirty flag' makes sure that we flush * the new space cache info onto disk. */ - cache->disk_cache_state = BTRFS_DC_CLEAR; if (btrfs_test_opt(root, SPACE_CACHE)) - cache->dirty = 1; + cache->disk_cache_state = BTRFS_DC_CLEAR; } read_extent_buffer(leaf, &cache->item, @@ -9460,6 +9446,13 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, } } + spin_lock(&trans->transaction->dirty_bgs_lock); + if (!list_empty(&block_group->dirty_list)) { + list_del_init(&block_group->dirty_list); + btrfs_put_block_group(block_group); + } + spin_unlock(&trans->transaction->dirty_bgs_lock); + btrfs_remove_free_space_cache(block_group); spin_lock(&block_group->space_info->lock); @@ -9611,7 +9604,8 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info) * Want to do this before we do anything else so we can recover * properly if we fail to join the transaction. */ - trans = btrfs_join_transaction(root); + /* 1 for btrfs_orphan_reserve_metadata() */ + trans = btrfs_start_transaction(root, 1); if (IS_ERR(trans)) { btrfs_set_block_group_rw(root, block_group); ret = PTR_ERR(trans); @@ -9624,18 +9618,33 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info) */ start = block_group->key.objectid; end = start + block_group->key.offset - 1; + /* + * Hold the unused_bg_unpin_mutex lock to avoid racing with + * btrfs_finish_extent_commit(). If we are at transaction N, + * another task might be running finish_extent_commit() for the + * previous transaction N - 1, and have seen a range belonging + * to the block group in freed_extents[] before we were able to + * clear the whole block group range from freed_extents[]. This + * means that task can lookup for the block group after we + * unpinned it from freed_extents[] and removed it, leading to + * a BUG_ON() at btrfs_unpin_extent_range(). + */ + mutex_lock(&fs_info->unused_bg_unpin_mutex); ret = clear_extent_bits(&fs_info->freed_extents[0], start, end, EXTENT_DIRTY, GFP_NOFS); if (ret) { + mutex_unlock(&fs_info->unused_bg_unpin_mutex); btrfs_set_block_group_rw(root, block_group); goto end_trans; } ret = clear_extent_bits(&fs_info->freed_extents[1], start, end, EXTENT_DIRTY, GFP_NOFS); if (ret) { + mutex_unlock(&fs_info->unused_bg_unpin_mutex); btrfs_set_block_group_rw(root, block_group); goto end_trans; } + mutex_unlock(&fs_info->unused_bg_unpin_mutex); /* Reset pinned so btrfs_put_block_group doesn't complain */ block_group->pinned = 0; diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index c73df6a..d688cfe 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -64,7 +64,7 @@ void btrfs_leak_debug_check(void) while (!list_empty(&states)) { state = list_entry(states.next, struct extent_state, leak_list); - pr_err("BTRFS: state leak: start %llu end %llu state %lu in tree %d refs %d\n", + pr_err("BTRFS: state leak: start %llu end %llu state %u in tree %d refs %d\n", state->start, state->end, state->state, extent_state_in_tree(state), atomic_read(&state->refs)); @@ -396,21 +396,21 @@ static void merge_state(struct extent_io_tree *tree, } static void set_state_cb(struct extent_io_tree *tree, - struct extent_state *state, unsigned long *bits) + struct extent_state *state, unsigned *bits) { if (tree->ops && tree->ops->set_bit_hook) tree->ops->set_bit_hook(tree->mapping->host, state, bits); } static void clear_state_cb(struct extent_io_tree *tree, - struct extent_state *state, unsigned long *bits) + struct extent_state *state, unsigned *bits) { if (tree->ops && tree->ops->clear_bit_hook) tree->ops->clear_bit_hook(tree->mapping->host, state, bits); } static void set_state_bits(struct extent_io_tree *tree, - struct extent_state *state, unsigned long *bits); + struct extent_state *state, unsigned *bits); /* * insert an extent_state struct into the tree. 'bits' are set on the @@ -426,7 +426,7 @@ static int insert_state(struct extent_io_tree *tree, struct extent_state *state, u64 start, u64 end, struct rb_node ***p, struct rb_node **parent, - unsigned long *bits) + unsigned *bits) { struct rb_node *node; @@ -511,10 +511,10 @@ static struct extent_state *next_state(struct extent_state *state) */ static struct extent_state *clear_state_bit(struct extent_io_tree *tree, struct extent_state *state, - unsigned long *bits, int wake) + unsigned *bits, int wake) { struct extent_state *next; - unsigned long bits_to_clear = *bits & ~EXTENT_CTLBITS; + unsigned bits_to_clear = *bits & ~EXTENT_CTLBITS; if ((bits_to_clear & EXTENT_DIRTY) && (state->state & EXTENT_DIRTY)) { u64 range = state->end - state->start + 1; @@ -570,7 +570,7 @@ static void extent_io_tree_panic(struct extent_io_tree *tree, int err) * This takes the tree lock, and returns 0 on success and < 0 on error. */ int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, int wake, int delete, + unsigned bits, int wake, int delete, struct extent_state **cached_state, gfp_t mask) { @@ -789,9 +789,9 @@ out: static void set_state_bits(struct extent_io_tree *tree, struct extent_state *state, - unsigned long *bits) + unsigned *bits) { - unsigned long bits_to_set = *bits & ~EXTENT_CTLBITS; + unsigned bits_to_set = *bits & ~EXTENT_CTLBITS; set_state_cb(tree, state, bits); if ((bits_to_set & EXTENT_DIRTY) && !(state->state & EXTENT_DIRTY)) { @@ -803,7 +803,7 @@ static void set_state_bits(struct extent_io_tree *tree, static void cache_state_if_flags(struct extent_state *state, struct extent_state **cached_ptr, - const u64 flags) + unsigned flags) { if (cached_ptr && !(*cached_ptr)) { if (!flags || (state->state & flags)) { @@ -833,7 +833,7 @@ static void cache_state(struct extent_state *state, static int __must_check __set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, unsigned long exclusive_bits, + unsigned bits, unsigned exclusive_bits, u64 *failed_start, struct extent_state **cached_state, gfp_t mask) { @@ -1034,7 +1034,7 @@ search_again: } int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, u64 * failed_start, + unsigned bits, u64 * failed_start, struct extent_state **cached_state, gfp_t mask) { return __set_extent_bit(tree, start, end, bits, 0, failed_start, @@ -1060,7 +1060,7 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, * boundary bits like LOCK. */ int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, unsigned long clear_bits, + unsigned bits, unsigned clear_bits, struct extent_state **cached_state, gfp_t mask) { struct extent_state *state; @@ -1268,14 +1268,14 @@ int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, } int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, gfp_t mask) + unsigned bits, gfp_t mask) { return set_extent_bit(tree, start, end, bits, NULL, NULL, mask); } int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, gfp_t mask) + unsigned bits, gfp_t mask) { return clear_extent_bit(tree, start, end, bits, 0, 0, NULL, mask); } @@ -1330,10 +1330,11 @@ int clear_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end, * us if waiting is desired. */ int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, struct extent_state **cached_state) + unsigned bits, struct extent_state **cached_state) { int err; u64 failed_start; + while (1) { err = __set_extent_bit(tree, start, end, EXTENT_LOCKED | bits, EXTENT_LOCKED, &failed_start, @@ -1440,7 +1441,7 @@ static int set_range_writeback(struct extent_io_tree *tree, u64 start, u64 end) */ static struct extent_state * find_first_extent_bit_state(struct extent_io_tree *tree, - u64 start, unsigned long bits) + u64 start, unsigned bits) { struct rb_node *node; struct extent_state *state; @@ -1474,7 +1475,7 @@ out: * If nothing was found, 1 is returned. If found something, return 0. */ int find_first_extent_bit(struct extent_io_tree *tree, u64 start, - u64 *start_ret, u64 *end_ret, unsigned long bits, + u64 *start_ret, u64 *end_ret, unsigned bits, struct extent_state **cached_state) { struct extent_state *state; @@ -1753,7 +1754,7 @@ out_failed: int extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end, struct page *locked_page, - unsigned long clear_bits, + unsigned clear_bits, unsigned long page_ops) { struct extent_io_tree *tree = &BTRFS_I(inode)->io_tree; @@ -1810,7 +1811,7 @@ int extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end, */ u64 count_range_bits(struct extent_io_tree *tree, u64 *start, u64 search_end, u64 max_bytes, - unsigned long bits, int contig) + unsigned bits, int contig) { struct rb_node *node; struct extent_state *state; @@ -1928,7 +1929,7 @@ out: * range is found set. */ int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, int filled, struct extent_state *cached) + unsigned bits, int filled, struct extent_state *cached) { struct extent_state *state = NULL; struct rb_node *node; @@ -2057,7 +2058,7 @@ int repair_io_failure(struct inode *inode, u64 start, u64 length, u64 logical, sector = bbio->stripes[mirror_num-1].physical >> 9; bio->bi_iter.bi_sector = sector; dev = bbio->stripes[mirror_num-1].dev; - kfree(bbio); + btrfs_put_bbio(bbio); if (!dev || !dev->bdev || !dev->writeable) { bio_put(bio); return -EIO; @@ -2816,8 +2817,10 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree, bio_add_page(bio, page, page_size, offset) < page_size) { ret = submit_one_bio(rw, bio, mirror_num, prev_bio_flags); - if (ret < 0) + if (ret < 0) { + *bio_ret = NULL; return ret; + } bio = NULL; } else { return 0; @@ -3239,7 +3242,7 @@ static noinline_for_stack int writepage_delalloc(struct inode *inode, page, &delalloc_start, &delalloc_end, - 128 * 1024 * 1024); + BTRFS_MAX_EXTENT_SIZE); if (nr_delalloc == 0) { delalloc_start = delalloc_end + 1; continue; @@ -4598,11 +4601,11 @@ static inline void btrfs_release_extent_buffer(struct extent_buffer *eb) static struct extent_buffer * __alloc_extent_buffer(struct btrfs_fs_info *fs_info, u64 start, - unsigned long len, gfp_t mask) + unsigned long len) { struct extent_buffer *eb = NULL; - eb = kmem_cache_zalloc(extent_buffer_cache, mask); + eb = kmem_cache_zalloc(extent_buffer_cache, GFP_NOFS); if (eb == NULL) return NULL; eb->start = start; @@ -4643,7 +4646,7 @@ struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src) struct extent_buffer *new; unsigned long num_pages = num_extent_pages(src->start, src->len); - new = __alloc_extent_buffer(NULL, src->start, src->len, GFP_NOFS); + new = __alloc_extent_buffer(src->fs_info, src->start, src->len); if (new == NULL) return NULL; @@ -4666,13 +4669,26 @@ struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src) return new; } -struct extent_buffer *alloc_dummy_extent_buffer(u64 start, unsigned long len) +struct extent_buffer *alloc_dummy_extent_buffer(struct btrfs_fs_info *fs_info, + u64 start) { struct extent_buffer *eb; - unsigned long num_pages = num_extent_pages(0, len); + unsigned long len; + unsigned long num_pages; unsigned long i; - eb = __alloc_extent_buffer(NULL, start, len, GFP_NOFS); + if (!fs_info) { + /* + * Called only from tests that don't always have a fs_info + * available, but we know that nodesize is 4096 + */ + len = 4096; + } else { + len = fs_info->tree_root->nodesize; + } + num_pages = num_extent_pages(0, len); + + eb = __alloc_extent_buffer(fs_info, start, len); if (!eb) return NULL; @@ -4762,7 +4778,7 @@ struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info, #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info, - u64 start, unsigned long len) + u64 start) { struct extent_buffer *eb, *exists = NULL; int ret; @@ -4770,7 +4786,7 @@ struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info, eb = find_extent_buffer(fs_info, start); if (eb) return eb; - eb = alloc_dummy_extent_buffer(start, len); + eb = alloc_dummy_extent_buffer(fs_info, start); if (!eb) return NULL; eb->fs_info = fs_info; @@ -4808,8 +4824,9 @@ free_eb: #endif struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info, - u64 start, unsigned long len) + u64 start) { + unsigned long len = fs_info->tree_root->nodesize; unsigned long num_pages = num_extent_pages(start, len); unsigned long i; unsigned long index = start >> PAGE_CACHE_SHIFT; @@ -4824,7 +4841,7 @@ struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info, if (eb) return eb; - eb = __alloc_extent_buffer(fs_info, start, len, GFP_NOFS); + eb = __alloc_extent_buffer(fs_info, start, len); if (!eb) return NULL; @@ -4951,6 +4968,12 @@ static int release_extent_buffer(struct extent_buffer *eb) /* Should be safe to release our pages at this point */ btrfs_release_extent_buffer_page(eb); +#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS + if (unlikely(test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags))) { + __free_extent_buffer(eb); + return 1; + } +#endif call_rcu(&eb->rcu_head, btrfs_release_extent_buffer_rcu); return 1; } diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index ece9ce8..695b0cc 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -4,22 +4,22 @@ #include <linux/rbtree.h> /* bits for the extent state */ -#define EXTENT_DIRTY 1 -#define EXTENT_WRITEBACK (1 << 1) -#define EXTENT_UPTODATE (1 << 2) -#define EXTENT_LOCKED (1 << 3) -#define EXTENT_NEW (1 << 4) -#define EXTENT_DELALLOC (1 << 5) -#define EXTENT_DEFRAG (1 << 6) -#define EXTENT_BOUNDARY (1 << 9) -#define EXTENT_NODATASUM (1 << 10) -#define EXTENT_DO_ACCOUNTING (1 << 11) -#define EXTENT_FIRST_DELALLOC (1 << 12) -#define EXTENT_NEED_WAIT (1 << 13) -#define EXTENT_DAMAGED (1 << 14) -#define EXTENT_NORESERVE (1 << 15) -#define EXTENT_IOBITS (EXTENT_LOCKED | EXTENT_WRITEBACK) -#define EXTENT_CTLBITS (EXTENT_DO_ACCOUNTING | EXTENT_FIRST_DELALLOC) +#define EXTENT_DIRTY (1U << 0) +#define EXTENT_WRITEBACK (1U << 1) +#define EXTENT_UPTODATE (1U << 2) +#define EXTENT_LOCKED (1U << 3) +#define EXTENT_NEW (1U << 4) +#define EXTENT_DELALLOC (1U << 5) +#define EXTENT_DEFRAG (1U << 6) +#define EXTENT_BOUNDARY (1U << 9) +#define EXTENT_NODATASUM (1U << 10) +#define EXTENT_DO_ACCOUNTING (1U << 11) +#define EXTENT_FIRST_DELALLOC (1U << 12) +#define EXTENT_NEED_WAIT (1U << 13) +#define EXTENT_DAMAGED (1U << 14) +#define EXTENT_NORESERVE (1U << 15) +#define EXTENT_IOBITS (EXTENT_LOCKED | EXTENT_WRITEBACK) +#define EXTENT_CTLBITS (EXTENT_DO_ACCOUNTING | EXTENT_FIRST_DELALLOC) /* * flags for bio submission. The high bits indicate the compression @@ -81,9 +81,9 @@ struct extent_io_ops { int (*writepage_end_io_hook)(struct page *page, u64 start, u64 end, struct extent_state *state, int uptodate); void (*set_bit_hook)(struct inode *inode, struct extent_state *state, - unsigned long *bits); + unsigned *bits); void (*clear_bit_hook)(struct inode *inode, struct extent_state *state, - unsigned long *bits); + unsigned *bits); void (*merge_extent_hook)(struct inode *inode, struct extent_state *new, struct extent_state *other); @@ -108,7 +108,7 @@ struct extent_state { /* ADD NEW ELEMENTS AFTER THIS */ wait_queue_head_t wq; atomic_t refs; - unsigned long state; + unsigned state; /* for use by the FS */ u64 private; @@ -188,7 +188,7 @@ int try_release_extent_mapping(struct extent_map_tree *map, int try_release_extent_buffer(struct page *page); int lock_extent(struct extent_io_tree *tree, u64 start, u64 end); int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, struct extent_state **cached); + unsigned bits, struct extent_state **cached); int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end); int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end, struct extent_state **cached, gfp_t mask); @@ -202,21 +202,21 @@ void extent_io_exit(void); u64 count_range_bits(struct extent_io_tree *tree, u64 *start, u64 search_end, - u64 max_bytes, unsigned long bits, int contig); + u64 max_bytes, unsigned bits, int contig); void free_extent_state(struct extent_state *state); int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, int filled, + unsigned bits, int filled, struct extent_state *cached_state); int clear_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, gfp_t mask); + unsigned bits, gfp_t mask); int clear_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, int wake, int delete, + unsigned bits, int wake, int delete, struct extent_state **cached, gfp_t mask); int set_extent_bits(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, gfp_t mask); + unsigned bits, gfp_t mask); int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, u64 *failed_start, + unsigned bits, u64 *failed_start, struct extent_state **cached_state, gfp_t mask); int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end, struct extent_state **cached_state, gfp_t mask); @@ -229,14 +229,14 @@ int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask); int convert_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, - unsigned long bits, unsigned long clear_bits, + unsigned bits, unsigned clear_bits, struct extent_state **cached_state, gfp_t mask); int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end, struct extent_state **cached_state, gfp_t mask); int set_extent_defrag(struct extent_io_tree *tree, u64 start, u64 end, struct extent_state **cached_state, gfp_t mask); int find_first_extent_bit(struct extent_io_tree *tree, u64 start, - u64 *start_ret, u64 *end_ret, unsigned long bits, + u64 *start_ret, u64 *end_ret, unsigned bits, struct extent_state **cached_state); int extent_invalidatepage(struct extent_io_tree *tree, struct page *page, unsigned long offset); @@ -262,8 +262,9 @@ int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private); void set_page_extent_mapped(struct page *page); struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info, - u64 start, unsigned long len); -struct extent_buffer *alloc_dummy_extent_buffer(u64 start, unsigned long len); + u64 start); +struct extent_buffer *alloc_dummy_extent_buffer(struct btrfs_fs_info *fs_info, + u64 start); struct extent_buffer *btrfs_clone_extent_buffer(struct extent_buffer *src); struct extent_buffer *find_extent_buffer(struct btrfs_fs_info *fs_info, u64 start); @@ -322,7 +323,7 @@ int extent_range_clear_dirty_for_io(struct inode *inode, u64 start, u64 end); int extent_range_redirty_for_io(struct inode *inode, u64 start, u64 end); int extent_clear_unlock_delalloc(struct inode *inode, u64 start, u64 end, struct page *locked_page, - unsigned long bits_to_clear, + unsigned bits_to_clear, unsigned long page_ops); struct bio * btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs, @@ -377,5 +378,5 @@ noinline u64 find_lock_delalloc_range(struct inode *inode, u64 *end, u64 max_bytes); #endif struct extent_buffer *alloc_test_extent_buffer(struct btrfs_fs_info *fs_info, - u64 start, unsigned long len); + u64 start); #endif diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 69c9508..aee18f8 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1811,22 +1811,10 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb, mutex_unlock(&inode->i_mutex); /* - * we want to make sure fsync finds this change - * but we haven't joined a transaction running right now. - * - * Later on, someone is sure to update the inode and get the - * real transid recorded. - * - * We set last_trans now to the fs_info generation + 1, - * this will either be one more than the running transaction - * or the generation used for the next transaction if there isn't - * one running right now. - * * We also have to set last_sub_trans to the current log transid, * otherwise subsequent syncs to a file that's been synced in this * transaction will appear to have already occured. */ - BTRFS_I(inode)->last_trans = root->fs_info->generation + 1; BTRFS_I(inode)->last_sub_trans = root->log_transid; if (num_written > 0) { err = generic_write_sync(file, pos, num_written); @@ -1959,25 +1947,37 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) atomic_inc(&root->log_batch); /* - * check the transaction that last modified this inode - * and see if its already been committed - */ - if (!BTRFS_I(inode)->last_trans) { - mutex_unlock(&inode->i_mutex); - goto out; - } - - /* - * if the last transaction that changed this file was before - * the current transaction, we can bail out now without any - * syncing + * If the last transaction that changed this file was before the current + * transaction and we have the full sync flag set in our inode, we can + * bail out now without any syncing. + * + * Note that we can't bail out if the full sync flag isn't set. This is + * because when the full sync flag is set we start all ordered extents + * and wait for them to fully complete - when they complete they update + * the inode's last_trans field through: + * + * btrfs_finish_ordered_io() -> + * btrfs_update_inode_fallback() -> + * btrfs_update_inode() -> + * btrfs_set_inode_last_trans() + * + * So we are sure that last_trans is up to date and can do this check to + * bail out safely. For the fast path, when the full sync flag is not + * set in our inode, we can not do it because we start only our ordered + * extents and don't wait for them to complete (that is when + * btrfs_finish_ordered_io runs), so here at this point their last_trans + * value might be less than or equals to fs_info->last_trans_committed, + * and setting a speculative last_trans for an inode when a buffered + * write is made (such as fs_info->generation + 1 for example) would not + * be reliable since after setting the value and before fsync is called + * any number of transactions can start and commit (transaction kthread + * commits the current transaction periodically), and a transaction + * commit does not start nor waits for ordered extents to complete. */ smp_mb(); if (btrfs_inode_in_log(inode, root->fs_info->generation) || - BTRFS_I(inode)->last_trans <= - root->fs_info->last_trans_committed) { - BTRFS_I(inode)->last_trans = 0; - + (full_sync && BTRFS_I(inode)->last_trans <= + root->fs_info->last_trans_committed)) { /* * We'v had everything committed since the last time we were * modified so clear this flag in case it was set for whatever @@ -2275,6 +2275,8 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len) bool same_page; bool no_holes = btrfs_fs_incompat(root->fs_info, NO_HOLES); u64 ino_size; + bool truncated_page = false; + bool updated_inode = false; ret = btrfs_wait_ordered_range(inode, offset, len); if (ret) @@ -2306,13 +2308,18 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len) * entire page. */ if (same_page && len < PAGE_CACHE_SIZE) { - if (offset < ino_size) + if (offset < ino_size) { + truncated_page = true; ret = btrfs_truncate_page(inode, offset, len, 0); + } else { + ret = 0; + } goto out_only_mutex; } /* zero back part of the first page */ if (offset < ino_size) { + truncated_page = true; ret = btrfs_truncate_page(inode, offset, 0, 0); if (ret) { mutex_unlock(&inode->i_mutex); @@ -2348,6 +2355,7 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len) if (!ret) { /* zero the front end of the last page */ if (tail_start + tail_len < ino_size) { + truncated_page = true; ret = btrfs_truncate_page(inode, tail_start + tail_len, 0, 1); if (ret) @@ -2357,8 +2365,8 @@ static int btrfs_punch_hole(struct inode *inode, loff_t offset, loff_t len) } if (lockend < lockstart) { - mutex_unlock(&inode->i_mutex); - return 0; + ret = 0; + goto out_only_mutex; } while (1) { @@ -2506,6 +2514,7 @@ out_trans: trans->block_rsv = &root->fs_info->trans_block_rsv; ret = btrfs_update_inode(trans, root, inode); + updated_inode = true; btrfs_end_transaction(trans, root); btrfs_btree_balance_dirty(root); out_free: @@ -2515,6 +2524,22 @@ out: unlock_extent_cached(&BTRFS_I(inode)->io_tree, lockstart, lockend, &cached_state, GFP_NOFS); out_only_mutex: + if (!updated_inode && truncated_page && !ret && !err) { + /* + * If we only end up zeroing part of a page, we still need to + * update the inode item, so that all the time fields are + * updated as well as the necessary btrfs inode in memory fields + * for detecting, at fsync time, if the inode isn't yet in the + * log tree or it's there but not up to date. + */ + trans = btrfs_start_transaction(root, 1); + if (IS_ERR(trans)) { + err = PTR_ERR(trans); + } else { + err = btrfs_update_inode(trans, root, inode); + ret = btrfs_end_transaction(trans, root); + } + } mutex_unlock(&inode->i_mutex); if (ret && !err) err = ret; diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index d6c03f7..a719785 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -651,15 +651,13 @@ static int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, struct io_ctl io_ctl; struct btrfs_key key; struct btrfs_free_space *e, *n; - struct list_head bitmaps; + LIST_HEAD(bitmaps); u64 num_entries; u64 num_bitmaps; u64 generation; u8 type; int ret = 0; - INIT_LIST_HEAD(&bitmaps); - /* Nothing in the space cache, goodbye */ if (!i_size_read(inode)) return 0; @@ -1243,6 +1241,7 @@ int btrfs_write_out_cache(struct btrfs_root *root, struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct inode *inode; int ret = 0; + enum btrfs_disk_cache_state dcs = BTRFS_DC_WRITTEN; root = root->fs_info->tree_root; @@ -1266,9 +1265,7 @@ int btrfs_write_out_cache(struct btrfs_root *root, ret = __btrfs_write_out_cache(root, inode, ctl, block_group, trans, path, block_group->key.objectid); if (ret) { - spin_lock(&block_group->lock); - block_group->disk_cache_state = BTRFS_DC_ERROR; - spin_unlock(&block_group->lock); + dcs = BTRFS_DC_ERROR; ret = 0; #ifdef DEBUG btrfs_err(root->fs_info, @@ -1277,6 +1274,9 @@ int btrfs_write_out_cache(struct btrfs_root *root, #endif } + spin_lock(&block_group->lock); + block_group->disk_cache_state = dcs; + spin_unlock(&block_group->lock); iput(inode); return ret; } @@ -2903,7 +2903,6 @@ int btrfs_find_space_cluster(struct btrfs_root *root, trace_btrfs_find_cluster(block_group, offset, bytes, empty_size, min_bytes); - INIT_LIST_HEAD(&bitmaps); ret = setup_cluster_no_bitmap(block_group, cluster, &bitmaps, offset, bytes + empty_size, cont1_bytes, min_bytes); diff --git a/fs/btrfs/inode-item.c b/fs/btrfs/inode-item.c index 8ffa478..265e03c 100644 --- a/fs/btrfs/inode-item.c +++ b/fs/btrfs/inode-item.c @@ -344,6 +344,7 @@ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans, return -ENOMEM; path->leave_spinning = 1; + path->skip_release_on_error = 1; ret = btrfs_insert_empty_item(trans, root, path, &key, ins_len); if (ret == -EEXIST) { @@ -362,8 +363,12 @@ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans, ptr = (unsigned long)(ref + 1); ret = 0; } else if (ret < 0) { - if (ret == -EOVERFLOW) - ret = -EMLINK; + if (ret == -EOVERFLOW) { + if (find_name_in_backref(path, name, name_len, &ref)) + ret = -EEXIST; + else + ret = -EMLINK; + } goto out; } else { ref = btrfs_item_ptr(path->nodes[0], path->slots[0], diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index b214ab1..686331f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -108,6 +108,13 @@ static struct extent_map *create_pinned_em(struct inode *inode, u64 start, static int btrfs_dirty_inode(struct inode *inode); +#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS +void btrfs_test_inode_set_ops(struct inode *inode) +{ + BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops; +} +#endif + static int btrfs_init_inode_security(struct btrfs_trans_handle *trans, struct inode *inode, struct inode *dir, const struct qstr *qstr) @@ -1530,10 +1537,32 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page, static void btrfs_split_extent_hook(struct inode *inode, struct extent_state *orig, u64 split) { + u64 size; + /* not delalloc, ignore it */ if (!(orig->state & EXTENT_DELALLOC)) return; + size = orig->end - orig->start + 1; + if (size > BTRFS_MAX_EXTENT_SIZE) { + u64 num_extents; + u64 new_size; + + /* + * See the explanation in btrfs_merge_extent_hook, the same + * applies here, just in reverse. + */ + new_size = orig->end - split + 1; + num_extents = div64_u64(new_size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + new_size = split - orig->start; + num_extents += div64_u64(new_size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + if (div64_u64(size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE) >= num_extents) + return; + } + spin_lock(&BTRFS_I(inode)->lock); BTRFS_I(inode)->outstanding_extents++; spin_unlock(&BTRFS_I(inode)->lock); @@ -1549,10 +1578,55 @@ static void btrfs_merge_extent_hook(struct inode *inode, struct extent_state *new, struct extent_state *other) { + u64 new_size, old_size; + u64 num_extents; + /* not delalloc, ignore it */ if (!(other->state & EXTENT_DELALLOC)) return; + if (new->start > other->start) + new_size = new->end - other->start + 1; + else + new_size = other->end - new->start + 1; + + /* we're not bigger than the max, unreserve the space and go */ + if (new_size <= BTRFS_MAX_EXTENT_SIZE) { + spin_lock(&BTRFS_I(inode)->lock); + BTRFS_I(inode)->outstanding_extents--; + spin_unlock(&BTRFS_I(inode)->lock); + return; + } + + /* + * We have to add up either side to figure out how many extents were + * accounted for before we merged into one big extent. If the number of + * extents we accounted for is <= the amount we need for the new range + * then we can return, otherwise drop. Think of it like this + * + * [ 4k][MAX_SIZE] + * + * So we've grown the extent by a MAX_SIZE extent, this would mean we + * need 2 outstanding extents, on one side we have 1 and the other side + * we have 1 so they are == and we can return. But in this case + * + * [MAX_SIZE+4k][MAX_SIZE+4k] + * + * Each range on their own accounts for 2 extents, but merged together + * they are only 3 extents worth of accounting, so we need to drop in + * this case. + */ + old_size = other->end - other->start + 1; + num_extents = div64_u64(old_size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + old_size = new->end - new->start + 1; + num_extents += div64_u64(old_size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + + if (div64_u64(new_size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE) >= num_extents) + return; + spin_lock(&BTRFS_I(inode)->lock); BTRFS_I(inode)->outstanding_extents--; spin_unlock(&BTRFS_I(inode)->lock); @@ -1604,7 +1678,7 @@ static void btrfs_del_delalloc_inode(struct btrfs_root *root, * have pending delalloc work to be done. */ static void btrfs_set_bit_hook(struct inode *inode, - struct extent_state *state, unsigned long *bits) + struct extent_state *state, unsigned *bits) { if ((*bits & EXTENT_DEFRAG) && !(*bits & EXTENT_DELALLOC)) @@ -1627,6 +1701,10 @@ static void btrfs_set_bit_hook(struct inode *inode, spin_unlock(&BTRFS_I(inode)->lock); } + /* For sanity tests */ + if (btrfs_test_is_dummy_root(root)) + return; + __percpu_counter_add(&root->fs_info->delalloc_bytes, len, root->fs_info->delalloc_batch); spin_lock(&BTRFS_I(inode)->lock); @@ -1645,9 +1723,11 @@ static void btrfs_set_bit_hook(struct inode *inode, */ static void btrfs_clear_bit_hook(struct inode *inode, struct extent_state *state, - unsigned long *bits) + unsigned *bits) { u64 len = state->end + 1 - state->start; + u64 num_extents = div64_u64(len + BTRFS_MAX_EXTENT_SIZE -1, + BTRFS_MAX_EXTENT_SIZE); spin_lock(&BTRFS_I(inode)->lock); if ((state->state & EXTENT_DEFRAG) && (*bits & EXTENT_DEFRAG)) @@ -1667,7 +1747,7 @@ static void btrfs_clear_bit_hook(struct inode *inode, *bits &= ~EXTENT_FIRST_DELALLOC; } else if (!(*bits & EXTENT_DO_ACCOUNTING)) { spin_lock(&BTRFS_I(inode)->lock); - BTRFS_I(inode)->outstanding_extents--; + BTRFS_I(inode)->outstanding_extents -= num_extents; spin_unlock(&BTRFS_I(inode)->lock); } @@ -1680,6 +1760,10 @@ static void btrfs_clear_bit_hook(struct inode *inode, root != root->fs_info->tree_root) btrfs_delalloc_release_metadata(inode, len); + /* For sanity tests. */ + if (btrfs_test_is_dummy_root(root)) + return; + if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID && do_list && !(state->state & EXTENT_NORESERVE)) btrfs_free_reserved_data_space(inode, len); @@ -2945,7 +3029,7 @@ static int __readpage_endio_check(struct inode *inode, return 0; zeroit: if (__ratelimit(&_rs)) - btrfs_info(BTRFS_I(inode)->root->fs_info, + btrfs_warn(BTRFS_I(inode)->root->fs_info, "csum failed ino %llu off %llu csum %u expected csum %u", btrfs_ino(inode), start, csum, csum_expected); memset(kaddr + pgoff, 1, len); @@ -3407,7 +3491,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) out: if (ret) - btrfs_crit(root->fs_info, + btrfs_err(root->fs_info, "could not do orphan cleanup %d", ret); btrfs_free_path(path); return ret; @@ -3490,7 +3574,6 @@ static void btrfs_read_locked_inode(struct inode *inode) struct btrfs_path *path; struct extent_buffer *leaf; struct btrfs_inode_item *inode_item; - struct btrfs_timespec *tspec; struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_key location; unsigned long ptr; @@ -3527,17 +3610,19 @@ static void btrfs_read_locked_inode(struct inode *inode) i_gid_write(inode, btrfs_inode_gid(leaf, inode_item)); btrfs_i_size_write(inode, btrfs_inode_size(leaf, inode_item)); - tspec = btrfs_inode_atime(inode_item); - inode->i_atime.tv_sec = btrfs_timespec_sec(leaf, tspec); - inode->i_atime.tv_nsec = btrfs_timespec_nsec(leaf, tspec); + inode->i_atime.tv_sec = btrfs_timespec_sec(leaf, &inode_item->atime); + inode->i_atime.tv_nsec = btrfs_timespec_nsec(leaf, &inode_item->atime); - tspec = btrfs_inode_mtime(inode_item); - inode->i_mtime.tv_sec = btrfs_timespec_sec(leaf, tspec); - inode->i_mtime.tv_nsec = btrfs_timespec_nsec(leaf, tspec); + inode->i_mtime.tv_sec = btrfs_timespec_sec(leaf, &inode_item->mtime); + inode->i_mtime.tv_nsec = btrfs_timespec_nsec(leaf, &inode_item->mtime); - tspec = btrfs_inode_ctime(inode_item); - inode->i_ctime.tv_sec = btrfs_timespec_sec(leaf, tspec); - inode->i_ctime.tv_nsec = btrfs_timespec_nsec(leaf, tspec); + inode->i_ctime.tv_sec = btrfs_timespec_sec(leaf, &inode_item->ctime); + inode->i_ctime.tv_nsec = btrfs_timespec_nsec(leaf, &inode_item->ctime); + + BTRFS_I(inode)->i_otime.tv_sec = + btrfs_timespec_sec(leaf, &inode_item->otime); + BTRFS_I(inode)->i_otime.tv_nsec = + btrfs_timespec_nsec(leaf, &inode_item->otime); inode_set_bytes(inode, btrfs_inode_nbytes(leaf, inode_item)); BTRFS_I(inode)->generation = btrfs_inode_generation(leaf, inode_item); @@ -3656,21 +3741,26 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token); btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token); - btrfs_set_token_timespec_sec(leaf, btrfs_inode_atime(item), + btrfs_set_token_timespec_sec(leaf, &item->atime, inode->i_atime.tv_sec, &token); - btrfs_set_token_timespec_nsec(leaf, btrfs_inode_atime(item), + btrfs_set_token_timespec_nsec(leaf, &item->atime, inode->i_atime.tv_nsec, &token); - btrfs_set_token_timespec_sec(leaf, btrfs_inode_mtime(item), + btrfs_set_token_timespec_sec(leaf, &item->mtime, inode->i_mtime.tv_sec, &token); - btrfs_set_token_timespec_nsec(leaf, btrfs_inode_mtime(item), + btrfs_set_token_timespec_nsec(leaf, &item->mtime, inode->i_mtime.tv_nsec, &token); - btrfs_set_token_timespec_sec(leaf, btrfs_inode_ctime(item), + btrfs_set_token_timespec_sec(leaf, &item->ctime, inode->i_ctime.tv_sec, &token); - btrfs_set_token_timespec_nsec(leaf, btrfs_inode_ctime(item), + btrfs_set_token_timespec_nsec(leaf, &item->ctime, inode->i_ctime.tv_nsec, &token); + btrfs_set_token_timespec_sec(leaf, &item->otime, + BTRFS_I(inode)->i_otime.tv_sec, &token); + btrfs_set_token_timespec_nsec(leaf, &item->otime, + BTRFS_I(inode)->i_otime.tv_nsec, &token); + btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode), &token); btrfs_set_token_inode_generation(leaf, item, BTRFS_I(inode)->generation, @@ -5007,6 +5097,7 @@ static int fixup_tree_root_location(struct btrfs_root *root, struct btrfs_root *new_root; struct btrfs_root_ref *ref; struct extent_buffer *leaf; + struct btrfs_key key; int ret; int err = 0; @@ -5017,9 +5108,12 @@ static int fixup_tree_root_location(struct btrfs_root *root, } err = -ENOENT; - ret = btrfs_find_item(root->fs_info->tree_root, path, - BTRFS_I(dir)->root->root_key.objectid, - location->objectid, BTRFS_ROOT_REF_KEY, NULL); + key.objectid = BTRFS_I(dir)->root->root_key.objectid; + key.type = BTRFS_ROOT_REF_KEY; + key.offset = location->objectid; + + ret = btrfs_search_slot(NULL, root->fs_info->tree_root, &key, path, + 0, 0); if (ret) { if (ret < 0) err = ret; @@ -5258,7 +5352,10 @@ static struct inode *new_simple_dir(struct super_block *s, inode->i_op = &btrfs_dir_ro_inode_operations; inode->i_fop = &simple_dir_operations; inode->i_mode = S_IFDIR | S_IRUGO | S_IWUSR | S_IXUGO; - inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; + inode->i_mtime = CURRENT_TIME; + inode->i_atime = inode->i_mtime; + inode->i_ctime = inode->i_mtime; + BTRFS_I(inode)->i_otime = inode->i_mtime; return inode; } @@ -5826,7 +5923,12 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, inode_init_owner(inode, dir, mode); inode_set_bytes(inode, 0); - inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; + + inode->i_mtime = CURRENT_TIME; + inode->i_atime = inode->i_mtime; + inode->i_ctime = inode->i_mtime; + BTRFS_I(inode)->i_otime = inode->i_mtime; + inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0], struct btrfs_inode_item); memset_extent_buffer(path->nodes[0], 0, (unsigned long)inode_item, @@ -7134,17 +7236,28 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, u64 start = iblock << inode->i_blkbits; u64 lockstart, lockend; u64 len = bh_result->b_size; + u64 *outstanding_extents = NULL; int unlock_bits = EXTENT_LOCKED; int ret = 0; if (create) - unlock_bits |= EXTENT_DELALLOC | EXTENT_DIRTY; + unlock_bits |= EXTENT_DIRTY; else len = min_t(u64, len, root->sectorsize); lockstart = start; lockend = start + len - 1; + if (current->journal_info) { + /* + * Need to pull our outstanding extents and set journal_info to NULL so + * that anything that needs to check if there's a transction doesn't get + * confused. + */ + outstanding_extents = current->journal_info; + current->journal_info = NULL; + } + /* * If this errors out it's because we couldn't invalidate pagecache for * this range and we need to fallback to buffered. @@ -7205,7 +7318,6 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, ((BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW) && em->block_start != EXTENT_MAP_HOLE)) { int type; - int ret; u64 block_start, orig_start, orig_block_len, ram_bytes; if (test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) @@ -7269,14 +7381,21 @@ unlock: if (start + len > i_size_read(inode)) i_size_write(inode, start + len); - spin_lock(&BTRFS_I(inode)->lock); - BTRFS_I(inode)->outstanding_extents++; - spin_unlock(&BTRFS_I(inode)->lock); + /* + * If we have an outstanding_extents count still set then we're + * within our reservation, otherwise we need to adjust our inode + * counter appropriately. + */ + if (*outstanding_extents) { + (*outstanding_extents)--; + } else { + spin_lock(&BTRFS_I(inode)->lock); + BTRFS_I(inode)->outstanding_extents++; + spin_unlock(&BTRFS_I(inode)->lock); + } - ret = set_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, - lockstart + len - 1, EXTENT_DELALLOC, NULL, - &cached_state, GFP_NOFS); - BUG_ON(ret); + current->journal_info = outstanding_extents; + btrfs_free_reserved_data_space(inode, len); } /* @@ -7299,6 +7418,8 @@ unlock: unlock_err: clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend, unlock_bits, 1, 0, &cached_state, GFP_NOFS); + if (outstanding_extents) + current->journal_info = outstanding_extents; return ret; } @@ -7805,8 +7926,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, } /* async crcs make it difficult to collect full stripe writes. */ - if (btrfs_get_alloc_profile(root, 1) & - (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) + if (btrfs_get_alloc_profile(root, 1) & BTRFS_BLOCK_GROUP_RAID56_MASK) async_submit = 0; else async_submit = 1; @@ -7999,6 +8119,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; + u64 outstanding_extents = 0; size_t count = 0; int flags = 0; bool wakeup = true; @@ -8036,6 +8157,16 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, ret = btrfs_delalloc_reserve_space(inode, count); if (ret) goto out; + outstanding_extents = div64_u64(count + + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + + /* + * We need to know how many extents we reserved so that we can + * do the accounting properly if we go over the number we + * originally calculated. Abuse current->journal_info for this. + */ + current->journal_info = &outstanding_extents; } else if (test_bit(BTRFS_INODE_READDIO_NEED_LOCK, &BTRFS_I(inode)->runtime_flags)) { inode_dio_done(inode); @@ -8048,13 +8179,12 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, iter, offset, btrfs_get_blocks_direct, NULL, btrfs_submit_direct, flags); if (rw & WRITE) { + current->journal_info = NULL; if (ret < 0 && ret != -EIOCBQUEUED) btrfs_delalloc_release_space(inode, count); else if (ret >= 0 && (size_t)ret < count) btrfs_delalloc_release_space(inode, count - (size_t)ret); - else - btrfs_delalloc_release_metadata(inode, 0); } out: if (wakeup) @@ -8575,6 +8705,9 @@ struct inode *btrfs_alloc_inode(struct super_block *sb) ei->delayed_node = NULL; + ei->i_otime.tv_sec = 0; + ei->i_otime.tv_nsec = 0; + inode = &ei->vfs_inode; extent_map_tree_init(&ei->extent_tree); extent_io_tree_init(&ei->io_tree, &inode->i_data); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index d49fe8a..74609b9 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -776,11 +776,11 @@ static int btrfs_may_delete(struct inode *dir, struct dentry *victim, int isdir) IS_IMMUTABLE(victim->d_inode) || IS_SWAPFILE(victim->d_inode)) return -EPERM; if (isdir) { - if (!S_ISDIR(victim->d_inode->i_mode)) + if (!d_is_dir(victim)) return -ENOTDIR; if (IS_ROOT(victim)) return -EBUSY; - } else if (S_ISDIR(victim->d_inode->i_mode)) + } else if (d_is_dir(victim)) return -EISDIR; if (IS_DEADDIR(dir)) return -ENOENT; diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index 534544e..157cc54 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -452,9 +452,7 @@ void btrfs_get_logged_extents(struct inode *inode, continue; if (entry_end(ordered) <= start) break; - if (!list_empty(&ordered->log_list)) - continue; - if (test_bit(BTRFS_ORDERED_LOGGED, &ordered->flags)) + if (test_and_set_bit(BTRFS_ORDERED_LOGGED, &ordered->flags)) continue; list_add(&ordered->log_list, logged_list); atomic_inc(&ordered->refs); @@ -511,8 +509,7 @@ void btrfs_wait_logged_extents(struct btrfs_trans_handle *trans, wait_event(ordered->wait, test_bit(BTRFS_ORDERED_IO_DONE, &ordered->flags)); - if (!test_and_set_bit(BTRFS_ORDERED_LOGGED, &ordered->flags)) - list_add_tail(&ordered->trans_list, &trans->ordered); + list_add_tail(&ordered->trans_list, &trans->ordered); spin_lock_irq(&log->log_extents_lock[index]); } spin_unlock_irq(&log->log_extents_lock[index]); diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 48b60db..058c79e 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -1259,7 +1259,7 @@ static int comp_oper(struct btrfs_qgroup_operation *oper1, if (oper1->seq < oper2->seq) return -1; if (oper1->seq > oper2->seq) - return -1; + return 1; if (oper1->ref_root < oper2->ref_root) return -1; if (oper1->ref_root > oper2->ref_root) @@ -1431,9 +1431,8 @@ static int qgroup_excl_accounting(struct btrfs_fs_info *fs_info, qgroup = u64_to_ptr(unode->aux); qgroup->rfer += sign * oper->num_bytes; qgroup->rfer_cmpr += sign * oper->num_bytes; + WARN_ON(sign < 0 && qgroup->excl < oper->num_bytes); qgroup->excl += sign * oper->num_bytes; - if (sign < 0) - WARN_ON(qgroup->excl < oper->num_bytes); qgroup->excl_cmpr += sign * oper->num_bytes; qgroup_dirty(fs_info, qgroup); diff --git a/fs/btrfs/raid56.c b/fs/btrfs/raid56.c index 8ab2a17..5264858 100644 --- a/fs/btrfs/raid56.c +++ b/fs/btrfs/raid56.c @@ -58,15 +58,6 @@ */ #define RBIO_CACHE_READY_BIT 3 -/* - * bbio and raid_map is managed by the caller, so we shouldn't free - * them here. And besides that, all rbios with this flag should not - * be cached, because we need raid_map to check the rbios' stripe - * is the same or not, but it is very likely that the caller has - * free raid_map, so don't cache those rbios. - */ -#define RBIO_HOLD_BBIO_MAP_BIT 4 - #define RBIO_CACHE_SIZE 1024 enum btrfs_rbio_ops { @@ -79,13 +70,6 @@ struct btrfs_raid_bio { struct btrfs_fs_info *fs_info; struct btrfs_bio *bbio; - /* - * logical block numbers for the start of each stripe - * The last one or two are p/q. These are sorted, - * so raid_map[0] is the start of our full stripe - */ - u64 *raid_map; - /* while we're doing rmw on a stripe * we put it into a hash table so we can * lock the stripe and merge more rbios @@ -303,7 +287,7 @@ static void cache_rbio_pages(struct btrfs_raid_bio *rbio) */ static int rbio_bucket(struct btrfs_raid_bio *rbio) { - u64 num = rbio->raid_map[0]; + u64 num = rbio->bbio->raid_map[0]; /* * we shift down quite a bit. We're using byte @@ -606,8 +590,8 @@ static int rbio_can_merge(struct btrfs_raid_bio *last, test_bit(RBIO_CACHE_BIT, &cur->flags)) return 0; - if (last->raid_map[0] != - cur->raid_map[0]) + if (last->bbio->raid_map[0] != + cur->bbio->raid_map[0]) return 0; /* we can't merge with different operations */ @@ -689,7 +673,7 @@ static noinline int lock_stripe_add(struct btrfs_raid_bio *rbio) spin_lock_irqsave(&h->lock, flags); list_for_each_entry(cur, &h->hash_list, hash_list) { walk++; - if (cur->raid_map[0] == rbio->raid_map[0]) { + if (cur->bbio->raid_map[0] == rbio->bbio->raid_map[0]) { spin_lock(&cur->bio_list_lock); /* can we steal this cached rbio's pages? */ @@ -841,21 +825,6 @@ done_nolock: remove_rbio_from_cache(rbio); } -static inline void -__free_bbio_and_raid_map(struct btrfs_bio *bbio, u64 *raid_map, int need) -{ - if (need) { - kfree(raid_map); - kfree(bbio); - } -} - -static inline void free_bbio_and_raid_map(struct btrfs_raid_bio *rbio) -{ - __free_bbio_and_raid_map(rbio->bbio, rbio->raid_map, - !test_bit(RBIO_HOLD_BBIO_MAP_BIT, &rbio->flags)); -} - static void __free_raid_bio(struct btrfs_raid_bio *rbio) { int i; @@ -875,8 +844,7 @@ static void __free_raid_bio(struct btrfs_raid_bio *rbio) } } - free_bbio_and_raid_map(rbio); - + btrfs_put_bbio(rbio->bbio); kfree(rbio); } @@ -985,8 +953,7 @@ static unsigned long rbio_nr_pages(unsigned long stripe_len, int nr_stripes) * this does not allocate any pages for rbio->pages. */ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, - struct btrfs_bio *bbio, u64 *raid_map, - u64 stripe_len) + struct btrfs_bio *bbio, u64 stripe_len) { struct btrfs_raid_bio *rbio; int nr_data = 0; @@ -1007,7 +974,6 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, INIT_LIST_HEAD(&rbio->stripe_cache); INIT_LIST_HEAD(&rbio->hash_list); rbio->bbio = bbio; - rbio->raid_map = raid_map; rbio->fs_info = root->fs_info; rbio->stripe_len = stripe_len; rbio->nr_pages = num_pages; @@ -1028,10 +994,12 @@ static struct btrfs_raid_bio *alloc_rbio(struct btrfs_root *root, rbio->bio_pages = p + sizeof(struct page *) * num_pages; rbio->dbitmap = p + sizeof(struct page *) * num_pages * 2; - if (raid_map[real_stripes - 1] == RAID6_Q_STRIPE) + if (bbio->map_type & BTRFS_BLOCK_GROUP_RAID5) + nr_data = real_stripes - 1; + else if (bbio->map_type & BTRFS_BLOCK_GROUP_RAID6) nr_data = real_stripes - 2; else - nr_data = real_stripes - 1; + BUG(); rbio->nr_data = nr_data; return rbio; @@ -1182,7 +1150,7 @@ static void index_rbio_pages(struct btrfs_raid_bio *rbio) spin_lock_irq(&rbio->bio_list_lock); bio_list_for_each(bio, &rbio->bio_list) { start = (u64)bio->bi_iter.bi_sector << 9; - stripe_offset = start - rbio->raid_map[0]; + stripe_offset = start - rbio->bbio->raid_map[0]; page_index = stripe_offset >> PAGE_CACHE_SHIFT; for (i = 0; i < bio->bi_vcnt; i++) { @@ -1402,7 +1370,7 @@ static int find_logical_bio_stripe(struct btrfs_raid_bio *rbio, logical <<= 9; for (i = 0; i < rbio->nr_data; i++) { - stripe_start = rbio->raid_map[i]; + stripe_start = rbio->bbio->raid_map[i]; if (logical >= stripe_start && logical < stripe_start + rbio->stripe_len) { return i; @@ -1776,17 +1744,16 @@ static void btrfs_raid_unplug(struct blk_plug_cb *cb, bool from_schedule) * our main entry point for writes from the rest of the FS. */ int raid56_parity_write(struct btrfs_root *root, struct bio *bio, - struct btrfs_bio *bbio, u64 *raid_map, - u64 stripe_len) + struct btrfs_bio *bbio, u64 stripe_len) { struct btrfs_raid_bio *rbio; struct btrfs_plug_cb *plug = NULL; struct blk_plug_cb *cb; int ret; - rbio = alloc_rbio(root, bbio, raid_map, stripe_len); + rbio = alloc_rbio(root, bbio, stripe_len); if (IS_ERR(rbio)) { - __free_bbio_and_raid_map(bbio, raid_map, 1); + btrfs_put_bbio(bbio); return PTR_ERR(rbio); } bio_list_add(&rbio->bio_list, bio); @@ -1885,9 +1852,7 @@ static void __raid_recover_end_io(struct btrfs_raid_bio *rbio) } /* all raid6 handling here */ - if (rbio->raid_map[rbio->real_stripes - 1] == - RAID6_Q_STRIPE) { - + if (rbio->bbio->map_type & BTRFS_BLOCK_GROUP_RAID6) { /* * single failure, rebuild from parity raid5 * style @@ -1922,8 +1887,9 @@ static void __raid_recover_end_io(struct btrfs_raid_bio *rbio) * here due to a crc mismatch and we can't give them the * data they want */ - if (rbio->raid_map[failb] == RAID6_Q_STRIPE) { - if (rbio->raid_map[faila] == RAID5_P_STRIPE) { + if (rbio->bbio->raid_map[failb] == RAID6_Q_STRIPE) { + if (rbio->bbio->raid_map[faila] == + RAID5_P_STRIPE) { err = -EIO; goto cleanup; } @@ -1934,7 +1900,7 @@ static void __raid_recover_end_io(struct btrfs_raid_bio *rbio) goto pstripe; } - if (rbio->raid_map[failb] == RAID5_P_STRIPE) { + if (rbio->bbio->raid_map[failb] == RAID5_P_STRIPE) { raid6_datap_recov(rbio->real_stripes, PAGE_SIZE, faila, pointers); } else { @@ -2001,8 +1967,7 @@ cleanup: cleanup_io: if (rbio->operation == BTRFS_RBIO_READ_REBUILD) { - if (err == 0 && - !test_bit(RBIO_HOLD_BBIO_MAP_BIT, &rbio->flags)) + if (err == 0) cache_rbio_pages(rbio); else clear_bit(RBIO_CACHE_READY_BIT, &rbio->flags); @@ -2156,15 +2121,16 @@ cleanup: * of the drive. */ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio, - struct btrfs_bio *bbio, u64 *raid_map, - u64 stripe_len, int mirror_num, int generic_io) + struct btrfs_bio *bbio, u64 stripe_len, + int mirror_num, int generic_io) { struct btrfs_raid_bio *rbio; int ret; - rbio = alloc_rbio(root, bbio, raid_map, stripe_len); + rbio = alloc_rbio(root, bbio, stripe_len); if (IS_ERR(rbio)) { - __free_bbio_and_raid_map(bbio, raid_map, generic_io); + if (generic_io) + btrfs_put_bbio(bbio); return PTR_ERR(rbio); } @@ -2175,7 +2141,8 @@ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio, rbio->faila = find_logical_bio_stripe(rbio, bio); if (rbio->faila == -1) { BUG(); - __free_bbio_and_raid_map(bbio, raid_map, generic_io); + if (generic_io) + btrfs_put_bbio(bbio); kfree(rbio); return -EIO; } @@ -2184,7 +2151,7 @@ int raid56_parity_recover(struct btrfs_root *root, struct bio *bio, btrfs_bio_counter_inc_noblocked(root->fs_info); rbio->generic_bio_cnt = 1; } else { - set_bit(RBIO_HOLD_BBIO_MAP_BIT, &rbio->flags); + btrfs_get_bbio(bbio); } /* @@ -2240,14 +2207,14 @@ static void read_rebuild_work(struct btrfs_work *work) struct btrfs_raid_bio * raid56_parity_alloc_scrub_rbio(struct btrfs_root *root, struct bio *bio, - struct btrfs_bio *bbio, u64 *raid_map, - u64 stripe_len, struct btrfs_device *scrub_dev, + struct btrfs_bio *bbio, u64 stripe_len, + struct btrfs_device *scrub_dev, unsigned long *dbitmap, int stripe_nsectors) { struct btrfs_raid_bio *rbio; int i; - rbio = alloc_rbio(root, bbio, raid_map, stripe_len); + rbio = alloc_rbio(root, bbio, stripe_len); if (IS_ERR(rbio)) return NULL; bio_list_add(&rbio->bio_list, bio); @@ -2279,10 +2246,10 @@ void raid56_parity_add_scrub_pages(struct btrfs_raid_bio *rbio, int stripe_offset; int index; - ASSERT(logical >= rbio->raid_map[0]); - ASSERT(logical + PAGE_SIZE <= rbio->raid_map[0] + + ASSERT(logical >= rbio->bbio->raid_map[0]); + ASSERT(logical + PAGE_SIZE <= rbio->bbio->raid_map[0] + rbio->stripe_len * rbio->nr_data); - stripe_offset = (int)(logical - rbio->raid_map[0]); + stripe_offset = (int)(logical - rbio->bbio->raid_map[0]); index = stripe_offset >> PAGE_CACHE_SHIFT; rbio->bio_pages[index] = page; } diff --git a/fs/btrfs/raid56.h b/fs/btrfs/raid56.h index 31d4a15..2b5d797 100644 --- a/fs/btrfs/raid56.h +++ b/fs/btrfs/raid56.h @@ -43,16 +43,15 @@ struct btrfs_raid_bio; struct btrfs_device; int raid56_parity_recover(struct btrfs_root *root, struct bio *bio, - struct btrfs_bio *bbio, u64 *raid_map, - u64 stripe_len, int mirror_num, int generic_io); + struct btrfs_bio *bbio, u64 stripe_len, + int mirror_num, int generic_io); int raid56_parity_write(struct btrfs_root *root, struct bio *bio, - struct btrfs_bio *bbio, u64 *raid_map, - u64 stripe_len); + struct btrfs_bio *bbio, u64 stripe_len); struct btrfs_raid_bio * raid56_parity_alloc_scrub_rbio(struct btrfs_root *root, struct bio *bio, - struct btrfs_bio *bbio, u64 *raid_map, - u64 stripe_len, struct btrfs_device *scrub_dev, + struct btrfs_bio *bbio, u64 stripe_len, + struct btrfs_device *scrub_dev, unsigned long *dbitmap, int stripe_nsectors); void raid56_parity_add_scrub_pages(struct btrfs_raid_bio *rbio, struct page *page, u64 logical); diff --git a/fs/btrfs/reada.c b/fs/btrfs/reada.c index b63ae20..0e7beea 100644 --- a/fs/btrfs/reada.c +++ b/fs/btrfs/reada.c @@ -66,7 +66,6 @@ struct reada_extctl { struct reada_extent { u64 logical; struct btrfs_key top; - u32 blocksize; int err; struct list_head extctl; int refcnt; @@ -349,7 +348,6 @@ static struct reada_extent *reada_find_extent(struct btrfs_root *root, blocksize = root->nodesize; re->logical = logical; - re->blocksize = blocksize; re->top = *top; INIT_LIST_HEAD(&re->extctl); spin_lock_init(&re->lock); @@ -463,7 +461,7 @@ static struct reada_extent *reada_find_extent(struct btrfs_root *root, spin_unlock(&fs_info->reada_lock); btrfs_dev_replace_unlock(&fs_info->dev_replace); - kfree(bbio); + btrfs_put_bbio(bbio); return re; error: @@ -488,7 +486,7 @@ error: kref_put(&zone->refcnt, reada_zone_release); spin_unlock(&fs_info->reada_lock); } - kfree(bbio); + btrfs_put_bbio(bbio); kfree(re); return re_exist; } @@ -660,7 +658,6 @@ static int reada_start_machine_dev(struct btrfs_fs_info *fs_info, int mirror_num = 0; struct extent_buffer *eb = NULL; u64 logical; - u32 blocksize; int ret; int i; int need_kick = 0; @@ -694,7 +691,7 @@ static int reada_start_machine_dev(struct btrfs_fs_info *fs_info, spin_unlock(&fs_info->reada_lock); return 0; } - dev->reada_next = re->logical + re->blocksize; + dev->reada_next = re->logical + fs_info->tree_root->nodesize; re->refcnt++; spin_unlock(&fs_info->reada_lock); @@ -709,7 +706,6 @@ static int reada_start_machine_dev(struct btrfs_fs_info *fs_info, } } logical = re->logical; - blocksize = re->blocksize; spin_lock(&re->lock); if (re->scheduled_for == NULL) { @@ -724,8 +720,8 @@ static int reada_start_machine_dev(struct btrfs_fs_info *fs_info, return 0; atomic_inc(&dev->reada_in_flight); - ret = reada_tree_block_flagged(fs_info->extent_root, logical, blocksize, - mirror_num, &eb); + ret = reada_tree_block_flagged(fs_info->extent_root, logical, + mirror_num, &eb); if (ret) __readahead_hook(fs_info->extent_root, NULL, logical, ret); else if (eb) @@ -851,7 +847,7 @@ static void dump_devs(struct btrfs_fs_info *fs_info, int all) break; printk(KERN_DEBUG " re: logical %llu size %u empty %d for %lld", - re->logical, re->blocksize, + re->logical, fs_info->tree_root->nodesize, list_empty(&re->extctl), re->scheduled_for ? re->scheduled_for->devid : -1); @@ -886,7 +882,8 @@ static void dump_devs(struct btrfs_fs_info *fs_info, int all) } printk(KERN_DEBUG "re: logical %llu size %u list empty %d for %lld", - re->logical, re->blocksize, list_empty(&re->extctl), + re->logical, fs_info->tree_root->nodesize, + list_empty(&re->extctl), re->scheduled_for ? re->scheduled_for->devid : -1); for (i = 0; i < re->nzones; ++i) { printk(KERN_CONT " zone %llu-%llu devs", diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 74257d6..d830853 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -2855,9 +2855,10 @@ static void update_processed_blocks(struct reloc_control *rc, } } -static int tree_block_processed(u64 bytenr, u32 blocksize, - struct reloc_control *rc) +static int tree_block_processed(u64 bytenr, struct reloc_control *rc) { + u32 blocksize = rc->extent_root->nodesize; + if (test_range_bit(&rc->processed_blocks, bytenr, bytenr + blocksize - 1, EXTENT_DIRTY, 1, NULL)) return 1; @@ -2965,8 +2966,7 @@ int relocate_tree_blocks(struct btrfs_trans_handle *trans, while (rb_node) { block = rb_entry(rb_node, struct tree_block, rb_node); if (!block->key_ready) - readahead_tree_block(rc->extent_root, block->bytenr, - block->key.objectid); + readahead_tree_block(rc->extent_root, block->bytenr); rb_node = rb_next(rb_node); } @@ -3353,7 +3353,7 @@ static int __add_tree_block(struct reloc_control *rc, bool skinny = btrfs_fs_incompat(rc->extent_root->fs_info, SKINNY_METADATA); - if (tree_block_processed(bytenr, blocksize, rc)) + if (tree_block_processed(bytenr, rc)) return 0; if (tree_search(blocks, bytenr)) @@ -3611,7 +3611,7 @@ static int find_data_references(struct reloc_control *rc, if (added) goto next; - if (!tree_block_processed(leaf->start, leaf->len, rc)) { + if (!tree_block_processed(leaf->start, rc)) { block = kmalloc(sizeof(*block), GFP_NOFS); if (!block) { err = -ENOMEM; diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c index e427cb7e..ec57687 100644 --- a/fs/btrfs/scrub.c +++ b/fs/btrfs/scrub.c @@ -66,7 +66,6 @@ struct scrub_ctx; struct scrub_recover { atomic_t refs; struct btrfs_bio *bbio; - u64 *raid_map; u64 map_length; }; @@ -80,7 +79,7 @@ struct scrub_page { u64 logical; u64 physical; u64 physical_for_dev_replace; - atomic_t ref_count; + atomic_t refs; struct { unsigned int mirror_num:8; unsigned int have_csum:1; @@ -113,7 +112,7 @@ struct scrub_block { struct scrub_page *pagev[SCRUB_MAX_PAGES_PER_BLOCK]; int page_count; atomic_t outstanding_pages; - atomic_t ref_count; /* free mem on transition to zero */ + atomic_t refs; /* free mem on transition to zero */ struct scrub_ctx *sctx; struct scrub_parity *sparity; struct { @@ -142,7 +141,7 @@ struct scrub_parity { int stripe_len; - atomic_t ref_count; + atomic_t refs; struct list_head spages; @@ -194,6 +193,15 @@ struct scrub_ctx { */ struct btrfs_scrub_progress stat; spinlock_t stat_lock; + + /* + * Use a ref counter to avoid use-after-free issues. Scrub workers + * decrement bios_in_flight and workers_pending and then do a wakeup + * on the list_wait wait queue. We must ensure the main scrub task + * doesn't free the scrub context before or while the workers are + * doing the wakeup() call. + */ + atomic_t refs; }; struct scrub_fixup_nodatasum { @@ -236,10 +244,7 @@ static void scrub_pending_bio_dec(struct scrub_ctx *sctx); static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx); static void scrub_pending_trans_workers_dec(struct scrub_ctx *sctx); static int scrub_handle_errored_block(struct scrub_block *sblock_to_check); -static int scrub_setup_recheck_block(struct scrub_ctx *sctx, - struct btrfs_fs_info *fs_info, - struct scrub_block *original_sblock, - u64 length, u64 logical, +static int scrub_setup_recheck_block(struct scrub_block *original_sblock, struct scrub_block *sblocks_for_recheck); static void scrub_recheck_block(struct btrfs_fs_info *fs_info, struct scrub_block *sblock, int is_metadata, @@ -251,8 +256,7 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info, const u8 *csum, u64 generation, u16 csum_size); static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad, - struct scrub_block *sblock_good, - int force_write); + struct scrub_block *sblock_good); static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad, struct scrub_block *sblock_good, int page_num, int force_write); @@ -302,10 +306,12 @@ static int copy_nocow_pages(struct scrub_ctx *sctx, u64 logical, u64 len, static void copy_nocow_pages_worker(struct btrfs_work *work); static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info); static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info); +static void scrub_put_ctx(struct scrub_ctx *sctx); static void scrub_pending_bio_inc(struct scrub_ctx *sctx) { + atomic_inc(&sctx->refs); atomic_inc(&sctx->bios_in_flight); } @@ -313,6 +319,7 @@ static void scrub_pending_bio_dec(struct scrub_ctx *sctx) { atomic_dec(&sctx->bios_in_flight); wake_up(&sctx->list_wait); + scrub_put_ctx(sctx); } static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info) @@ -346,6 +353,7 @@ static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx) { struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; + atomic_inc(&sctx->refs); /* * increment scrubs_running to prevent cancel requests from * completing as long as a worker is running. we must also @@ -388,6 +396,7 @@ static void scrub_pending_trans_workers_dec(struct scrub_ctx *sctx) atomic_dec(&sctx->workers_pending); wake_up(&fs_info->scrub_pause_wait); wake_up(&sctx->list_wait); + scrub_put_ctx(sctx); } static void scrub_free_csums(struct scrub_ctx *sctx) @@ -433,6 +442,12 @@ static noinline_for_stack void scrub_free_ctx(struct scrub_ctx *sctx) kfree(sctx); } +static void scrub_put_ctx(struct scrub_ctx *sctx) +{ + if (atomic_dec_and_test(&sctx->refs)) + scrub_free_ctx(sctx); +} + static noinline_for_stack struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace) { @@ -457,6 +472,7 @@ struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace) sctx = kzalloc(sizeof(*sctx), GFP_NOFS); if (!sctx) goto nomem; + atomic_set(&sctx->refs, 1); sctx->is_dev_replace = is_dev_replace; sctx->pages_per_rd_bio = pages_per_rd_bio; sctx->curr = -1; @@ -520,6 +536,7 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root, struct inode_fs_paths *ipath = NULL; struct btrfs_root *local_root; struct btrfs_key root_key; + struct btrfs_key key; root_key.objectid = root; root_key.type = BTRFS_ROOT_ITEM_KEY; @@ -530,7 +547,14 @@ static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root, goto err; } - ret = inode_item_info(inum, 0, local_root, swarn->path); + /* + * this makes the path point to (inum INODE_ITEM ioff) + */ + key.objectid = inum; + key.type = BTRFS_INODE_ITEM_KEY; + key.offset = 0; + + ret = btrfs_search_slot(NULL, local_root, &key, swarn->path, 0, 0); if (ret) { btrfs_release_path(swarn->path); goto err; @@ -848,8 +872,7 @@ static inline void scrub_get_recover(struct scrub_recover *recover) static inline void scrub_put_recover(struct scrub_recover *recover) { if (atomic_dec_and_test(&recover->refs)) { - kfree(recover->bbio); - kfree(recover->raid_map); + btrfs_put_bbio(recover->bbio); kfree(recover); } } @@ -955,8 +978,7 @@ static int scrub_handle_errored_block(struct scrub_block *sblock_to_check) } /* setup the context, map the logical blocks and alloc the pages */ - ret = scrub_setup_recheck_block(sctx, fs_info, sblock_to_check, length, - logical, sblocks_for_recheck); + ret = scrub_setup_recheck_block(sblock_to_check, sblocks_for_recheck); if (ret) { spin_lock(&sctx->stat_lock); sctx->stat.read_errors++; @@ -1030,9 +1052,10 @@ static int scrub_handle_errored_block(struct scrub_block *sblock_to_check) if (!is_metadata && !have_csum) { struct scrub_fixup_nodatasum *fixup_nodatasum; -nodatasum_case: WARN_ON(sctx->is_dev_replace); +nodatasum_case: + /* * !is_metadata and !have_csum, this means that the data * might not be COW'ed, that it might be modified @@ -1091,76 +1114,20 @@ nodatasum_case: sblock_other->no_io_error_seen) { if (sctx->is_dev_replace) { scrub_write_block_to_dev_replace(sblock_other); + goto corrected_error; } else { - int force_write = is_metadata || have_csum; - ret = scrub_repair_block_from_good_copy( - sblock_bad, sblock_other, - force_write); + sblock_bad, sblock_other); + if (!ret) + goto corrected_error; } - if (0 == ret) - goto corrected_error; } } - /* - * for dev_replace, pick good pages and write to the target device. - */ - if (sctx->is_dev_replace) { - success = 1; - for (page_num = 0; page_num < sblock_bad->page_count; - page_num++) { - int sub_success; - - sub_success = 0; - for (mirror_index = 0; - mirror_index < BTRFS_MAX_MIRRORS && - sblocks_for_recheck[mirror_index].page_count > 0; - mirror_index++) { - struct scrub_block *sblock_other = - sblocks_for_recheck + mirror_index; - struct scrub_page *page_other = - sblock_other->pagev[page_num]; - - if (!page_other->io_error) { - ret = scrub_write_page_to_dev_replace( - sblock_other, page_num); - if (ret == 0) { - /* succeeded for this page */ - sub_success = 1; - break; - } else { - btrfs_dev_replace_stats_inc( - &sctx->dev_root-> - fs_info->dev_replace. - num_write_errors); - } - } - } - - if (!sub_success) { - /* - * did not find a mirror to fetch the page - * from. scrub_write_page_to_dev_replace() - * handles this case (page->io_error), by - * filling the block with zeros before - * submitting the write request - */ - success = 0; - ret = scrub_write_page_to_dev_replace( - sblock_bad, page_num); - if (ret) - btrfs_dev_replace_stats_inc( - &sctx->dev_root->fs_info-> - dev_replace.num_write_errors); - } - } - - goto out; - } + if (sblock_bad->no_io_error_seen && !sctx->is_dev_replace) + goto did_not_correct_error; /* - * for regular scrub, repair those pages that are errored. * In case of I/O errors in the area that is supposed to be * repaired, continue by picking good copies of those pages. * Select the good pages from mirrors to rewrite bad pages from @@ -1184,44 +1151,64 @@ nodatasum_case: * mirror, even if other 512 byte sectors in the same PAGE_SIZE * area are unreadable. */ - - /* can only fix I/O errors from here on */ - if (sblock_bad->no_io_error_seen) - goto did_not_correct_error; - success = 1; - for (page_num = 0; page_num < sblock_bad->page_count; page_num++) { + for (page_num = 0; page_num < sblock_bad->page_count; + page_num++) { struct scrub_page *page_bad = sblock_bad->pagev[page_num]; + struct scrub_block *sblock_other = NULL; - if (!page_bad->io_error) + /* skip no-io-error page in scrub */ + if (!page_bad->io_error && !sctx->is_dev_replace) continue; - for (mirror_index = 0; - mirror_index < BTRFS_MAX_MIRRORS && - sblocks_for_recheck[mirror_index].page_count > 0; - mirror_index++) { - struct scrub_block *sblock_other = sblocks_for_recheck + - mirror_index; - struct scrub_page *page_other = sblock_other->pagev[ - page_num]; - - if (!page_other->io_error) { - ret = scrub_repair_page_from_good_copy( - sblock_bad, sblock_other, page_num, 0); - if (0 == ret) { - page_bad->io_error = 0; - break; /* succeeded for this page */ + /* try to find no-io-error page in mirrors */ + if (page_bad->io_error) { + for (mirror_index = 0; + mirror_index < BTRFS_MAX_MIRRORS && + sblocks_for_recheck[mirror_index].page_count > 0; + mirror_index++) { + if (!sblocks_for_recheck[mirror_index]. + pagev[page_num]->io_error) { + sblock_other = sblocks_for_recheck + + mirror_index; + break; } } + if (!sblock_other) + success = 0; } - if (page_bad->io_error) { - /* did not find a mirror to copy the page from */ - success = 0; + if (sctx->is_dev_replace) { + /* + * did not find a mirror to fetch the page + * from. scrub_write_page_to_dev_replace() + * handles this case (page->io_error), by + * filling the block with zeros before + * submitting the write request + */ + if (!sblock_other) + sblock_other = sblock_bad; + + if (scrub_write_page_to_dev_replace(sblock_other, + page_num) != 0) { + btrfs_dev_replace_stats_inc( + &sctx->dev_root-> + fs_info->dev_replace. + num_write_errors); + success = 0; + } + } else if (sblock_other) { + ret = scrub_repair_page_from_good_copy(sblock_bad, + sblock_other, + page_num, 0); + if (0 == ret) + page_bad->io_error = 0; + else + success = 0; } } - if (success) { + if (success && !sctx->is_dev_replace) { if (is_metadata || have_csum) { /* * need to verify the checksum now that all @@ -1288,19 +1275,18 @@ out: return 0; } -static inline int scrub_nr_raid_mirrors(struct btrfs_bio *bbio, u64 *raid_map) +static inline int scrub_nr_raid_mirrors(struct btrfs_bio *bbio) { - if (raid_map) { - if (raid_map[bbio->num_stripes - 1] == RAID6_Q_STRIPE) - return 3; - else - return 2; - } else { + if (bbio->map_type & BTRFS_BLOCK_GROUP_RAID5) + return 2; + else if (bbio->map_type & BTRFS_BLOCK_GROUP_RAID6) + return 3; + else return (int)bbio->num_stripes; - } } -static inline void scrub_stripe_index_and_offset(u64 logical, u64 *raid_map, +static inline void scrub_stripe_index_and_offset(u64 logical, u64 map_type, + u64 *raid_map, u64 mapped_length, int nstripes, int mirror, int *stripe_index, @@ -1308,7 +1294,7 @@ static inline void scrub_stripe_index_and_offset(u64 logical, u64 *raid_map, { int i; - if (raid_map) { + if (map_type & BTRFS_BLOCK_GROUP_RAID56_MASK) { /* RAID5/6 */ for (i = 0; i < nstripes; i++) { if (raid_map[i] == RAID6_Q_STRIPE || @@ -1329,72 +1315,65 @@ static inline void scrub_stripe_index_and_offset(u64 logical, u64 *raid_map, } } -static int scrub_setup_recheck_block(struct scrub_ctx *sctx, - struct btrfs_fs_info *fs_info, - struct scrub_block *original_sblock, - u64 length, u64 logical, +static int scrub_setup_recheck_block(struct scrub_block *original_sblock, struct scrub_block *sblocks_for_recheck) { + struct scrub_ctx *sctx = original_sblock->sctx; + struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; + u64 length = original_sblock->page_count * PAGE_SIZE; + u64 logical = original_sblock->pagev[0]->logical; struct scrub_recover *recover; struct btrfs_bio *bbio; - u64 *raid_map; u64 sublen; u64 mapped_length; u64 stripe_offset; int stripe_index; - int page_index; + int page_index = 0; int mirror_index; int nmirrors; int ret; /* - * note: the two members ref_count and outstanding_pages + * note: the two members refs and outstanding_pages * are not used (and not set) in the blocks that are used for * the recheck procedure */ - page_index = 0; while (length > 0) { sublen = min_t(u64, length, PAGE_SIZE); mapped_length = sublen; bbio = NULL; - raid_map = NULL; /* * with a length of PAGE_SIZE, each returned stripe * represents one mirror */ ret = btrfs_map_sblock(fs_info, REQ_GET_READ_MIRRORS, logical, - &mapped_length, &bbio, 0, &raid_map); + &mapped_length, &bbio, 0, 1); if (ret || !bbio || mapped_length < sublen) { - kfree(bbio); - kfree(raid_map); + btrfs_put_bbio(bbio); return -EIO; } recover = kzalloc(sizeof(struct scrub_recover), GFP_NOFS); if (!recover) { - kfree(bbio); - kfree(raid_map); + btrfs_put_bbio(bbio); return -ENOMEM; } atomic_set(&recover->refs, 1); recover->bbio = bbio; - recover->raid_map = raid_map; recover->map_length = mapped_length; BUG_ON(page_index >= SCRUB_PAGES_PER_RD_BIO); - nmirrors = scrub_nr_raid_mirrors(bbio, raid_map); + nmirrors = min(scrub_nr_raid_mirrors(bbio), BTRFS_MAX_MIRRORS); + for (mirror_index = 0; mirror_index < nmirrors; mirror_index++) { struct scrub_block *sblock; struct scrub_page *page; - if (mirror_index >= BTRFS_MAX_MIRRORS) - continue; - sblock = sblocks_for_recheck + mirror_index; sblock->sctx = sctx; page = kzalloc(sizeof(*page), GFP_NOFS); @@ -1410,9 +1389,12 @@ leave_nomem: sblock->pagev[page_index] = page; page->logical = logical; - scrub_stripe_index_and_offset(logical, raid_map, + scrub_stripe_index_and_offset(logical, + bbio->map_type, + bbio->raid_map, mapped_length, - bbio->num_stripes, + bbio->num_stripes - + bbio->num_tgtdevs, mirror_index, &stripe_index, &stripe_offset); @@ -1458,7 +1440,8 @@ static void scrub_bio_wait_endio(struct bio *bio, int error) static inline int scrub_is_page_on_raid56(struct scrub_page *page) { - return page->recover && page->recover->raid_map; + return page->recover && + (page->recover->bbio->map_type & BTRFS_BLOCK_GROUP_RAID56_MASK); } static int scrub_submit_raid56_bio_wait(struct btrfs_fs_info *fs_info, @@ -1475,7 +1458,6 @@ static int scrub_submit_raid56_bio_wait(struct btrfs_fs_info *fs_info, bio->bi_end_io = scrub_bio_wait_endio; ret = raid56_parity_recover(fs_info->fs_root, bio, page->recover->bbio, - page->recover->raid_map, page->recover->map_length, page->mirror_num, 0); if (ret) @@ -1615,8 +1597,7 @@ static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info, } static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad, - struct scrub_block *sblock_good, - int force_write) + struct scrub_block *sblock_good) { int page_num; int ret = 0; @@ -1626,8 +1607,7 @@ static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad, ret_sub = scrub_repair_page_from_good_copy(sblock_bad, sblock_good, - page_num, - force_write); + page_num, 1); if (ret_sub) ret = ret_sub; } @@ -2067,12 +2047,12 @@ static int scrub_checksum_super(struct scrub_block *sblock) static void scrub_block_get(struct scrub_block *sblock) { - atomic_inc(&sblock->ref_count); + atomic_inc(&sblock->refs); } static void scrub_block_put(struct scrub_block *sblock) { - if (atomic_dec_and_test(&sblock->ref_count)) { + if (atomic_dec_and_test(&sblock->refs)) { int i; if (sblock->sparity) @@ -2086,12 +2066,12 @@ static void scrub_block_put(struct scrub_block *sblock) static void scrub_page_get(struct scrub_page *spage) { - atomic_inc(&spage->ref_count); + atomic_inc(&spage->refs); } static void scrub_page_put(struct scrub_page *spage) { - if (atomic_dec_and_test(&spage->ref_count)) { + if (atomic_dec_and_test(&spage->refs)) { if (spage->page) __free_page(spage->page); kfree(spage); @@ -2217,7 +2197,7 @@ static int scrub_pages(struct scrub_ctx *sctx, u64 logical, u64 len, /* one ref inside this function, plus one for each page added to * a bio later on */ - atomic_set(&sblock->ref_count, 1); + atomic_set(&sblock->refs, 1); sblock->sctx = sctx; sblock->no_io_error_seen = 1; @@ -2510,7 +2490,7 @@ static int scrub_pages_for_parity(struct scrub_parity *sparity, /* one ref inside this function, plus one for each page added to * a bio later on */ - atomic_set(&sblock->ref_count, 1); + atomic_set(&sblock->refs, 1); sblock->sctx = sctx; sblock->no_io_error_seen = 1; sblock->sparity = sparity; @@ -2705,7 +2685,6 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) struct btrfs_raid_bio *rbio; struct scrub_page *spage; struct btrfs_bio *bbio = NULL; - u64 *raid_map = NULL; u64 length; int ret; @@ -2716,8 +2695,8 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) length = sparity->logic_end - sparity->logic_start + 1; ret = btrfs_map_sblock(sctx->dev_root->fs_info, WRITE, sparity->logic_start, - &length, &bbio, 0, &raid_map); - if (ret || !bbio || !raid_map) + &length, &bbio, 0, 1); + if (ret || !bbio || !bbio->raid_map) goto bbio_out; bio = btrfs_io_bio_alloc(GFP_NOFS, 0); @@ -2729,8 +2708,7 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) bio->bi_end_io = scrub_parity_bio_endio; rbio = raid56_parity_alloc_scrub_rbio(sctx->dev_root, bio, bbio, - raid_map, length, - sparity->scrub_dev, + length, sparity->scrub_dev, sparity->dbitmap, sparity->nsectors); if (!rbio) @@ -2747,8 +2725,7 @@ static void scrub_parity_check_and_repair(struct scrub_parity *sparity) rbio_out: bio_put(bio); bbio_out: - kfree(bbio); - kfree(raid_map); + btrfs_put_bbio(bbio); bitmap_or(sparity->ebitmap, sparity->ebitmap, sparity->dbitmap, sparity->nsectors); spin_lock(&sctx->stat_lock); @@ -2765,12 +2742,12 @@ static inline int scrub_calc_parity_bitmap_len(int nsectors) static void scrub_parity_get(struct scrub_parity *sparity) { - atomic_inc(&sparity->ref_count); + atomic_inc(&sparity->refs); } static void scrub_parity_put(struct scrub_parity *sparity) { - if (!atomic_dec_and_test(&sparity->ref_count)) + if (!atomic_dec_and_test(&sparity->refs)) return; scrub_parity_check_and_repair(sparity); @@ -2820,7 +2797,7 @@ static noinline_for_stack int scrub_raid56_parity(struct scrub_ctx *sctx, sparity->scrub_dev = sdev; sparity->logic_start = logic_start; sparity->logic_end = logic_end; - atomic_set(&sparity->ref_count, 1); + atomic_set(&sparity->refs, 1); INIT_LIST_HEAD(&sparity->spages); sparity->dbitmap = sparity->bitmap; sparity->ebitmap = (void *)sparity->bitmap + bitmap_len; @@ -3037,8 +3014,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, } else if (map->type & BTRFS_BLOCK_GROUP_DUP) { increment = map->stripe_len; mirror_num = num % map->num_stripes + 1; - } else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) { + } else if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { get_raid56_logic_offset(physical, num, map, &offset, NULL); increment = map->stripe_len * nr_data_stripes(map); mirror_num = 1; @@ -3074,8 +3050,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, */ logical = base + offset; physical_end = physical + nstripes * map->stripe_len; - if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { get_raid56_logic_offset(physical_end, num, map, &logic_end, NULL); logic_end += base; @@ -3121,8 +3096,7 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, ret = 0; while (physical < physical_end) { /* for raid56, we skip parity stripe */ - if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { ret = get_raid56_logic_offset(physical, num, map, &logical, &stripe_logical); logical += base; @@ -3280,8 +3254,7 @@ again: scrub_free_csums(sctx); if (extent_logical + extent_len < key.objectid + bytes) { - if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { /* * loop until we find next data stripe * or we have finished all stripes. @@ -3775,7 +3748,7 @@ int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start, scrub_workers_put(fs_info); mutex_unlock(&fs_info->scrub_lock); - scrub_free_ctx(sctx); + scrub_put_ctx(sctx); return ret; } @@ -3881,14 +3854,14 @@ static void scrub_remap_extent(struct btrfs_fs_info *fs_info, &mapped_length, &bbio, 0); if (ret || !bbio || mapped_length < extent_len || !bbio->stripes[0].dev->bdev) { - kfree(bbio); + btrfs_put_bbio(bbio); return; } *extent_physical = bbio->stripes[0].physical; *extent_mirror_num = bbio->mirror_num; *extent_dev = bbio->stripes[0].dev; - kfree(bbio); + btrfs_put_bbio(bbio); } static int scrub_setup_wr_ctx(struct scrub_ctx *sctx, diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index 804432d..d6033f5 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -230,6 +230,7 @@ struct pending_dir_move { u64 parent_ino; u64 ino; u64 gen; + bool is_orphan; struct list_head update_refs; }; @@ -2471,12 +2472,9 @@ verbose_printk("btrfs: send_utimes %llu\n", ino); if (ret < 0) goto out; TLV_PUT_PATH(sctx, BTRFS_SEND_A_PATH, p); - TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_ATIME, eb, - btrfs_inode_atime(ii)); - TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_MTIME, eb, - btrfs_inode_mtime(ii)); - TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_CTIME, eb, - btrfs_inode_ctime(ii)); + TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_ATIME, eb, &ii->atime); + TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_MTIME, eb, &ii->mtime); + TLV_PUT_BTRFS_TIMESPEC(sctx, BTRFS_SEND_A_CTIME, eb, &ii->ctime); /* TODO Add otime support when the otime patches get into upstream */ ret = send_cmd(sctx); @@ -2987,7 +2985,8 @@ static int add_pending_dir_move(struct send_ctx *sctx, u64 ino_gen, u64 parent_ino, struct list_head *new_refs, - struct list_head *deleted_refs) + struct list_head *deleted_refs, + const bool is_orphan) { struct rb_node **p = &sctx->pending_dir_moves.rb_node; struct rb_node *parent = NULL; @@ -3002,6 +3001,7 @@ static int add_pending_dir_move(struct send_ctx *sctx, pm->parent_ino = parent_ino; pm->ino = ino; pm->gen = ino_gen; + pm->is_orphan = is_orphan; INIT_LIST_HEAD(&pm->list); INIT_LIST_HEAD(&pm->update_refs); RB_CLEAR_NODE(&pm->node); @@ -3134,16 +3134,20 @@ static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm) rmdir_ino = dm->rmdir_ino; free_waiting_dir_move(sctx, dm); - ret = get_first_ref(sctx->parent_root, pm->ino, - &parent_ino, &parent_gen, name); - if (ret < 0) - goto out; - - ret = get_cur_path(sctx, parent_ino, parent_gen, - from_path); - if (ret < 0) - goto out; - ret = fs_path_add_path(from_path, name); + if (pm->is_orphan) { + ret = gen_unique_name(sctx, pm->ino, + pm->gen, from_path); + } else { + ret = get_first_ref(sctx->parent_root, pm->ino, + &parent_ino, &parent_gen, name); + if (ret < 0) + goto out; + ret = get_cur_path(sctx, parent_ino, parent_gen, + from_path); + if (ret < 0) + goto out; + ret = fs_path_add_path(from_path, name); + } if (ret < 0) goto out; @@ -3153,7 +3157,8 @@ static int apply_dir_move(struct send_ctx *sctx, struct pending_dir_move *pm) LIST_HEAD(deleted_refs); ASSERT(ancestor > BTRFS_FIRST_FREE_OBJECTID); ret = add_pending_dir_move(sctx, pm->ino, pm->gen, ancestor, - &pm->update_refs, &deleted_refs); + &pm->update_refs, &deleted_refs, + pm->is_orphan); if (ret < 0) goto out; if (rmdir_ino) { @@ -3286,6 +3291,127 @@ out: return ret; } +/* + * We might need to delay a directory rename even when no ancestor directory + * (in the send root) with a higher inode number than ours (sctx->cur_ino) was + * renamed. This happens when we rename a directory to the old name (the name + * in the parent root) of some other unrelated directory that got its rename + * delayed due to some ancestor with higher number that got renamed. + * + * Example: + * + * Parent snapshot: + * . (ino 256) + * |---- a/ (ino 257) + * | |---- file (ino 260) + * | + * |---- b/ (ino 258) + * |---- c/ (ino 259) + * + * Send snapshot: + * . (ino 256) + * |---- a/ (ino 258) + * |---- x/ (ino 259) + * |---- y/ (ino 257) + * |----- file (ino 260) + * + * Here we can not rename 258 from 'b' to 'a' without the rename of inode 257 + * from 'a' to 'x/y' happening first, which in turn depends on the rename of + * inode 259 from 'c' to 'x'. So the order of rename commands the send stream + * must issue is: + * + * 1 - rename 259 from 'c' to 'x' + * 2 - rename 257 from 'a' to 'x/y' + * 3 - rename 258 from 'b' to 'a' + * + * Returns 1 if the rename of sctx->cur_ino needs to be delayed, 0 if it can + * be done right away and < 0 on error. + */ +static int wait_for_dest_dir_move(struct send_ctx *sctx, + struct recorded_ref *parent_ref, + const bool is_orphan) +{ + struct btrfs_path *path; + struct btrfs_key key; + struct btrfs_key di_key; + struct btrfs_dir_item *di; + u64 left_gen; + u64 right_gen; + int ret = 0; + + if (RB_EMPTY_ROOT(&sctx->waiting_dir_moves)) + return 0; + + path = alloc_path_for_send(); + if (!path) + return -ENOMEM; + + key.objectid = parent_ref->dir; + key.type = BTRFS_DIR_ITEM_KEY; + key.offset = btrfs_name_hash(parent_ref->name, parent_ref->name_len); + + ret = btrfs_search_slot(NULL, sctx->parent_root, &key, path, 0, 0); + if (ret < 0) { + goto out; + } else if (ret > 0) { + ret = 0; + goto out; + } + + di = btrfs_match_dir_item_name(sctx->parent_root, path, + parent_ref->name, parent_ref->name_len); + if (!di) { + ret = 0; + goto out; + } + /* + * di_key.objectid has the number of the inode that has a dentry in the + * parent directory with the same name that sctx->cur_ino is being + * renamed to. We need to check if that inode is in the send root as + * well and if it is currently marked as an inode with a pending rename, + * if it is, we need to delay the rename of sctx->cur_ino as well, so + * that it happens after that other inode is renamed. + */ + btrfs_dir_item_key_to_cpu(path->nodes[0], di, &di_key); + if (di_key.type != BTRFS_INODE_ITEM_KEY) { + ret = 0; + goto out; + } + + ret = get_inode_info(sctx->parent_root, di_key.objectid, NULL, + &left_gen, NULL, NULL, NULL, NULL); + if (ret < 0) + goto out; + ret = get_inode_info(sctx->send_root, di_key.objectid, NULL, + &right_gen, NULL, NULL, NULL, NULL); + if (ret < 0) { + if (ret == -ENOENT) + ret = 0; + goto out; + } + + /* Different inode, no need to delay the rename of sctx->cur_ino */ + if (right_gen != left_gen) { + ret = 0; + goto out; + } + + if (is_waiting_for_move(sctx, di_key.objectid)) { + ret = add_pending_dir_move(sctx, + sctx->cur_ino, + sctx->cur_inode_gen, + di_key.objectid, + &sctx->new_refs, + &sctx->deleted_refs, + is_orphan); + if (!ret) + ret = 1; + } +out: + btrfs_free_path(path); + return ret; +} + static int wait_for_parent_move(struct send_ctx *sctx, struct recorded_ref *parent_ref) { @@ -3352,7 +3478,8 @@ out: sctx->cur_inode_gen, ino, &sctx->new_refs, - &sctx->deleted_refs); + &sctx->deleted_refs, + false); if (!ret) ret = 1; } @@ -3375,6 +3502,7 @@ static int process_recorded_refs(struct send_ctx *sctx, int *pending_move) int did_overwrite = 0; int is_orphan = 0; u64 last_dir_ino_rm = 0; + bool can_rename = true; verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino); @@ -3493,12 +3621,22 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino); } } + if (S_ISDIR(sctx->cur_inode_mode) && sctx->parent_root) { + ret = wait_for_dest_dir_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 * it depending on the inode mode. */ - if (is_orphan) { + if (is_orphan && can_rename) { ret = send_rename(sctx, valid_path, cur->full_path); if (ret < 0) goto out; @@ -3506,7 +3644,7 @@ verbose_printk("btrfs: process_recorded_refs %llu\n", sctx->cur_ino); ret = fs_path_copy(valid_path, cur->full_path); if (ret < 0) goto out; - } else { + } else if (can_rename) { if (S_ISDIR(sctx->cur_inode_mode)) { /* * Dirs can't be linked, so move it. For moved diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 6f49b28..05fef19 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -1958,11 +1958,6 @@ static int btrfs_freeze(struct super_block *sb) return btrfs_commit_transaction(trans, root); } -static int btrfs_unfreeze(struct super_block *sb) -{ - return 0; -} - static int btrfs_show_devname(struct seq_file *m, struct dentry *root) { struct btrfs_fs_info *fs_info = btrfs_sb(root->d_sb); @@ -2011,7 +2006,6 @@ static const struct super_operations btrfs_super_ops = { .statfs = btrfs_statfs, .remount_fs = btrfs_remount, .freeze_fs = btrfs_freeze, - .unfreeze_fs = btrfs_unfreeze, }; static const struct file_operations btrfs_ctl_fops = { diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c index 92db3f6..94edb0a 100644 --- a/fs/btrfs/sysfs.c +++ b/fs/btrfs/sysfs.c @@ -733,10 +733,18 @@ int btrfs_init_sysfs(void) ret = btrfs_init_debugfs(); if (ret) - return ret; + goto out1; init_feature_attrs(); ret = sysfs_create_group(&btrfs_kset->kobj, &btrfs_feature_attr_group); + if (ret) + goto out2; + + return 0; +out2: + debugfs_remove_recursive(btrfs_debugfs_root_dentry); +out1: + kset_unregister(btrfs_kset); return ret; } diff --git a/fs/btrfs/tests/extent-buffer-tests.c b/fs/btrfs/tests/extent-buffer-tests.c index cc286ce..f51963a 100644 --- a/fs/btrfs/tests/extent-buffer-tests.c +++ b/fs/btrfs/tests/extent-buffer-tests.c @@ -53,7 +53,7 @@ static int test_btrfs_split_item(void) return -ENOMEM; } - path->nodes[0] = eb = alloc_dummy_extent_buffer(0, 4096); + path->nodes[0] = eb = alloc_dummy_extent_buffer(NULL, 4096); if (!eb) { test_msg("Could not allocate dummy buffer\n"); ret = -ENOMEM; diff --git a/fs/btrfs/tests/extent-io-tests.c b/fs/btrfs/tests/extent-io-tests.c index 7e99c2f..9e9f236 100644 --- a/fs/btrfs/tests/extent-io-tests.c +++ b/fs/btrfs/tests/extent-io-tests.c @@ -258,8 +258,7 @@ static int test_find_delalloc(void) } ret = 0; out_bits: - clear_extent_bits(&tmp, 0, total_dirty - 1, - (unsigned long)-1, GFP_NOFS); + clear_extent_bits(&tmp, 0, total_dirty - 1, (unsigned)-1, GFP_NOFS); out: if (locked_page) page_cache_release(locked_page); diff --git a/fs/btrfs/tests/inode-tests.c b/fs/btrfs/tests/inode-tests.c index 3ae0f5b..054fc0d 100644 --- a/fs/btrfs/tests/inode-tests.c +++ b/fs/btrfs/tests/inode-tests.c @@ -255,7 +255,7 @@ static noinline int test_btrfs_get_extent(void) goto out; } - root->node = alloc_dummy_extent_buffer(0, 4096); + root->node = alloc_dummy_extent_buffer(NULL, 4096); if (!root->node) { test_msg("Couldn't allocate dummy buffer\n"); goto out; @@ -843,7 +843,7 @@ static int test_hole_first(void) goto out; } - root->node = alloc_dummy_extent_buffer(0, 4096); + root->node = alloc_dummy_extent_buffer(NULL, 4096); if (!root->node) { test_msg("Couldn't allocate dummy buffer\n"); goto out; @@ -911,6 +911,197 @@ out: return ret; } +static int test_extent_accounting(void) +{ + struct inode *inode = NULL; + struct btrfs_root *root = NULL; + int ret = -ENOMEM; + + inode = btrfs_new_test_inode(); + if (!inode) { + test_msg("Couldn't allocate inode\n"); + return ret; + } + + root = btrfs_alloc_dummy_root(); + if (IS_ERR(root)) { + test_msg("Couldn't allocate root\n"); + goto out; + } + + root->fs_info = btrfs_alloc_dummy_fs_info(); + if (!root->fs_info) { + test_msg("Couldn't allocate dummy fs info\n"); + goto out; + } + + BTRFS_I(inode)->root = root; + btrfs_test_inode_set_ops(inode); + + /* [BTRFS_MAX_EXTENT_SIZE] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, 0, BTRFS_MAX_EXTENT_SIZE - 1, + NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 1) { + ret = -EINVAL; + test_msg("Miscount, wanted 1, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE][4k] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE, + BTRFS_MAX_EXTENT_SIZE + 4095, NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 2) { + ret = -EINVAL; + test_msg("Miscount, wanted 2, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE/2][4K HOLE][the rest] */ + ret = clear_extent_bit(&BTRFS_I(inode)->io_tree, + BTRFS_MAX_EXTENT_SIZE >> 1, + (BTRFS_MAX_EXTENT_SIZE >> 1) + 4095, + EXTENT_DELALLOC | EXTENT_DIRTY | + EXTENT_UPTODATE | EXTENT_DO_ACCOUNTING, 0, 0, + NULL, GFP_NOFS); + if (ret) { + test_msg("clear_extent_bit returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 2) { + ret = -EINVAL; + test_msg("Miscount, wanted 2, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE][4K] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE >> 1, + (BTRFS_MAX_EXTENT_SIZE >> 1) + 4095, + NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 2) { + ret = -EINVAL; + test_msg("Miscount, wanted 2, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* + * [BTRFS_MAX_EXTENT_SIZE+4K][4K HOLE][BTRFS_MAX_EXTENT_SIZE+4K] + * + * I'm artificially adding 2 to outstanding_extents because in the + * buffered IO case we'd add things up as we go, but I don't feel like + * doing that here, this isn't the interesting case we want to test. + */ + BTRFS_I(inode)->outstanding_extents += 2; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE + 8192, + (BTRFS_MAX_EXTENT_SIZE << 1) + 12287, + NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 4) { + ret = -EINVAL; + test_msg("Miscount, wanted 4, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE+4k][4k][BTRFS_MAX_EXTENT_SIZE+4k] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE+4096, + BTRFS_MAX_EXTENT_SIZE+8191, NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 3) { + ret = -EINVAL; + test_msg("Miscount, wanted 3, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE+4k][4K HOLE][BTRFS_MAX_EXTENT_SIZE+4k] */ + ret = clear_extent_bit(&BTRFS_I(inode)->io_tree, + BTRFS_MAX_EXTENT_SIZE+4096, + BTRFS_MAX_EXTENT_SIZE+8191, + EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0, + NULL, GFP_NOFS); + if (ret) { + test_msg("clear_extent_bit returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 4) { + ret = -EINVAL; + test_msg("Miscount, wanted 4, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* + * Refill the hole again just for good measure, because I thought it + * might fail and I'd rather satisfy my paranoia at this point. + */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE+4096, + BTRFS_MAX_EXTENT_SIZE+8191, NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 3) { + ret = -EINVAL; + test_msg("Miscount, wanted 3, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* Empty */ + ret = clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, (u64)-1, + EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0, + NULL, GFP_NOFS); + if (ret) { + test_msg("clear_extent_bit returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents) { + ret = -EINVAL; + test_msg("Miscount, wanted 0, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + ret = 0; +out: + if (ret) + clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, (u64)-1, + EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0, + NULL, GFP_NOFS); + iput(inode); + btrfs_free_dummy_root(root); + return ret; +} + int btrfs_test_inodes(void) { int ret; @@ -924,5 +1115,9 @@ int btrfs_test_inodes(void) if (ret) return ret; test_msg("Running hole first btrfs_get_extent test\n"); - return test_hole_first(); + ret = test_hole_first(); + if (ret) + return ret; + test_msg("Running outstanding_extents tests\n"); + return test_extent_accounting(); } diff --git a/fs/btrfs/tests/qgroup-tests.c b/fs/btrfs/tests/qgroup-tests.c index ec3dcb2..73f299e 100644 --- a/fs/btrfs/tests/qgroup-tests.c +++ b/fs/btrfs/tests/qgroup-tests.c @@ -404,12 +404,22 @@ int btrfs_test_qgroups(void) ret = -ENOMEM; goto out; } + /* We are using this root as our extent root */ + root->fs_info->extent_root = root; + + /* + * Some of the paths we test assume we have a filled out fs_info, so we + * just need to add the root in there so we don't panic. + */ + root->fs_info->tree_root = root; + root->fs_info->quota_root = root; + root->fs_info->quota_enabled = 1; /* * Can't use bytenr 0, some things freak out * *cough*backref walking code*cough* */ - root->node = alloc_test_extent_buffer(root->fs_info, 4096, 4096); + root->node = alloc_test_extent_buffer(root->fs_info, 4096); if (!root->node) { test_msg("Couldn't allocate dummy buffer\n"); ret = -ENOMEM; @@ -448,17 +458,6 @@ int btrfs_test_qgroups(void) goto out; } - /* We are using this root as our extent root */ - root->fs_info->extent_root = root; - - /* - * Some of the paths we test assume we have a filled out fs_info, so we - * just need to addt he root in there so we don't panic. - */ - root->fs_info->tree_root = root; - root->fs_info->quota_root = root; - root->fs_info->quota_enabled = 1; - test_msg("Running qgroup tests\n"); ret = test_no_shared_qgroup(root); if (ret) diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index e88b59d..8be4278 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -220,6 +220,7 @@ loop: * commit the transaction. */ atomic_set(&cur_trans->use_count, 2); + cur_trans->have_free_bgs = 0; cur_trans->start_time = get_seconds(); cur_trans->delayed_refs.href_root = RB_ROOT; @@ -248,6 +249,8 @@ loop: INIT_LIST_HEAD(&cur_trans->pending_chunks); INIT_LIST_HEAD(&cur_trans->switch_commits); INIT_LIST_HEAD(&cur_trans->pending_ordered); + INIT_LIST_HEAD(&cur_trans->dirty_bgs); + spin_lock_init(&cur_trans->dirty_bgs_lock); list_add_tail(&cur_trans->list, &fs_info->trans_list); extent_io_tree_init(&cur_trans->dirty_pages, fs_info->btree_inode->i_mapping); @@ -1022,7 +1025,6 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, struct btrfs_root *tree_root = root->fs_info->tree_root; old_root_used = btrfs_root_used(&root->root_item); - btrfs_write_dirty_block_groups(trans, root); while (1) { old_root_bytenr = btrfs_root_bytenr(&root->root_item); @@ -1038,9 +1040,6 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, return ret; old_root_used = btrfs_root_used(&root->root_item); - ret = btrfs_write_dirty_block_groups(trans, root); - if (ret) - return ret; } return 0; @@ -1057,14 +1056,11 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, struct btrfs_root *root) { struct btrfs_fs_info *fs_info = root->fs_info; + struct list_head *dirty_bgs = &trans->transaction->dirty_bgs; struct list_head *next; struct extent_buffer *eb; int ret; - ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); - if (ret) - return ret; - eb = btrfs_lock_root_node(fs_info->tree_root); ret = btrfs_cow_block(trans, fs_info->tree_root, eb, NULL, 0, &eb); @@ -1088,15 +1084,20 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, if (ret) return ret; + ret = btrfs_setup_space_cache(trans, root); + if (ret) + return ret; + /* run_qgroups might have added some more refs */ ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); if (ret) return ret; - +again: while (!list_empty(&fs_info->dirty_cowonly_roots)) { next = fs_info->dirty_cowonly_roots.next; list_del_init(next); root = list_entry(next, struct btrfs_root, dirty_list); + clear_bit(BTRFS_ROOT_DIRTY, &root->state); if (root != fs_info->extent_root) list_add_tail(&root->dirty_list, @@ -1104,8 +1105,23 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, ret = update_cowonly_root(trans, root); if (ret) return ret; + ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); + if (ret) + return ret; + } + + while (!list_empty(dirty_bgs)) { + ret = btrfs_write_dirty_block_groups(trans, root); + if (ret) + return ret; + ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); + if (ret) + return ret; } + if (!list_empty(&fs_info->dirty_cowonly_roots)) + goto again; + list_add_tail(&fs_info->extent_root->dirty_list, &trans->transaction->switch_commits); btrfs_after_dev_replace_commit(fs_info); @@ -1803,6 +1819,9 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, wait_for_commit(root, cur_trans); + if (unlikely(cur_trans->aborted)) + ret = cur_trans->aborted; + btrfs_put_transaction(cur_trans); return ret; @@ -1983,6 +2002,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, switch_commit_roots(cur_trans, root->fs_info); assert_qgroups_uptodate(trans); + ASSERT(list_empty(&cur_trans->dirty_bgs)); update_super_roots(root); btrfs_set_super_log_root(root->fs_info->super_copy, 0); @@ -2026,6 +2046,9 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, btrfs_finish_extent_commit(trans, root); + if (cur_trans->have_free_bgs) + btrfs_clear_space_info_full(root->fs_info); + root->fs_info->last_trans_committed = cur_trans->transid; /* * We needn't acquire the lock here because there is no other task diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 00ed29c..937050a 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -47,6 +47,11 @@ struct btrfs_transaction { atomic_t num_writers; atomic_t use_count; + /* + * true if there is free bgs operations in this transaction + */ + int have_free_bgs; + /* Be protected by fs_info->trans_lock when we want to change it. */ enum btrfs_trans_state state; struct list_head list; @@ -58,6 +63,8 @@ struct btrfs_transaction { struct list_head pending_chunks; struct list_head pending_ordered; struct list_head switch_commits; + struct list_head dirty_bgs; + spinlock_t dirty_bgs_lock; struct btrfs_delayed_ref_root delayed_refs; int aborted; }; diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 1a9585d..c5b8ba3 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -453,11 +453,13 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans, insert: btrfs_release_path(path); /* try to insert the key into the destination tree */ + path->skip_release_on_error = 1; ret = btrfs_insert_empty_item(trans, root, path, key, item_size); + path->skip_release_on_error = 0; /* make sure any existing item is the correct size */ - if (ret == -EEXIST) { + if (ret == -EEXIST || ret == -EOVERFLOW) { u32 found_size; found_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]); @@ -488,8 +490,20 @@ insert: src_item = (struct btrfs_inode_item *)src_ptr; dst_item = (struct btrfs_inode_item *)dst_ptr; - if (btrfs_inode_generation(eb, src_item) == 0) + if (btrfs_inode_generation(eb, src_item) == 0) { + struct extent_buffer *dst_eb = path->nodes[0]; + + if (S_ISREG(btrfs_inode_mode(eb, src_item)) && + S_ISREG(btrfs_inode_mode(dst_eb, dst_item))) { + struct btrfs_map_token token; + u64 ino_size = btrfs_inode_size(eb, src_item); + + btrfs_init_map_token(&token); + btrfs_set_token_inode_size(dst_eb, dst_item, + ino_size, &token); + } goto no_copy; + } if (overwrite_root && S_ISDIR(btrfs_inode_mode(eb, src_item)) && @@ -844,7 +858,7 @@ out: static noinline int backref_in_log(struct btrfs_root *log, struct btrfs_key *key, u64 ref_objectid, - char *name, int namelen) + const char *name, int namelen) { struct btrfs_path *path; struct btrfs_inode_ref *ref; @@ -998,7 +1012,7 @@ again: base = btrfs_item_ptr_offset(leaf, path->slots[0]); while (cur_offset < item_size) { - extref = (struct btrfs_inode_extref *)base + cur_offset; + extref = (struct btrfs_inode_extref *)(base + cur_offset); victim_name_len = btrfs_inode_extref_name_len(leaf, extref); @@ -1254,13 +1268,14 @@ out: } static int insert_orphan_item(struct btrfs_trans_handle *trans, - struct btrfs_root *root, u64 offset) + struct btrfs_root *root, u64 ino) { int ret; - ret = btrfs_find_item(root, NULL, BTRFS_ORPHAN_OBJECTID, - offset, BTRFS_ORPHAN_ITEM_KEY, NULL); - if (ret > 0) - ret = btrfs_insert_orphan_item(trans, root, offset); + + ret = btrfs_insert_orphan_item(trans, root, ino); + if (ret == -EEXIST) + ret = 0; + return ret; } @@ -1287,6 +1302,7 @@ static int count_inode_extrefs(struct btrfs_root *root, leaf = path->nodes[0]; item_size = btrfs_item_size_nr(leaf, path->slots[0]); ptr = btrfs_item_ptr_offset(leaf, path->slots[0]); + cur_offset = 0; while (cur_offset < item_size) { extref = (struct btrfs_inode_extref *) (ptr + cur_offset); @@ -1302,7 +1318,7 @@ static int count_inode_extrefs(struct btrfs_root *root, } btrfs_release_path(path); - if (ret < 0) + if (ret < 0 && ret != -ENOENT) return ret; return nlink; } @@ -1394,9 +1410,6 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, nlink = ret; ret = count_inode_extrefs(root, inode, path); - if (ret == -ENOENT) - ret = 0; - if (ret < 0) goto out; @@ -1557,6 +1570,30 @@ static noinline int insert_one_name(struct btrfs_trans_handle *trans, } /* + * Return true if an inode reference exists in the log for the given name, + * inode and parent inode. + */ +static bool name_in_log_ref(struct btrfs_root *log_root, + const char *name, const int name_len, + const u64 dirid, const u64 ino) +{ + struct btrfs_key search_key; + + search_key.objectid = ino; + search_key.type = BTRFS_INODE_REF_KEY; + search_key.offset = dirid; + if (backref_in_log(log_root, &search_key, dirid, name, name_len)) + return true; + + search_key.type = BTRFS_INODE_EXTREF_KEY; + search_key.offset = btrfs_extref_hash(dirid, name, name_len); + if (backref_in_log(log_root, &search_key, dirid, name, name_len)) + return true; + + return false; +} + +/* * take a single entry in a log directory item and replay it into * the subvolume. * @@ -1666,10 +1703,17 @@ out: return ret; insert: + if (name_in_log_ref(root->log_root, name, name_len, + key->objectid, log_key.objectid)) { + /* The dentry will be added later. */ + ret = 0; + update_size = false; + goto out; + } btrfs_release_path(path); ret = insert_one_name(trans, root, path, key->objectid, key->offset, name, name_len, log_type, &log_key); - if (ret && ret != -ENOENT) + if (ret && ret != -ENOENT && ret != -EEXIST) goto out; update_size = false; ret = 0; @@ -2164,7 +2208,7 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, parent = path->nodes[*level]; root_owner = btrfs_header_owner(parent); - next = btrfs_find_create_tree_block(root, bytenr, blocksize); + next = btrfs_find_create_tree_block(root, bytenr); if (!next) return -ENOMEM; @@ -2416,8 +2460,8 @@ static void wait_for_writer(struct btrfs_trans_handle *trans, mutex_unlock(&root->log_mutex); if (atomic_read(&root->log_writers)) schedule(); - mutex_lock(&root->log_mutex); finish_wait(&root->log_writer_wait, &wait); + mutex_lock(&root->log_mutex); } } @@ -3219,7 +3263,8 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans, static void fill_inode_item(struct btrfs_trans_handle *trans, struct extent_buffer *leaf, struct btrfs_inode_item *item, - struct inode *inode, int log_inode_only) + struct inode *inode, int log_inode_only, + u64 logged_isize) { struct btrfs_map_token token; @@ -3232,7 +3277,7 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, * to say 'update this inode with these values' */ btrfs_set_token_inode_generation(leaf, item, 0, &token); - btrfs_set_token_inode_size(leaf, item, 0, &token); + btrfs_set_token_inode_size(leaf, item, logged_isize, &token); } else { btrfs_set_token_inode_generation(leaf, item, BTRFS_I(inode)->generation, @@ -3245,19 +3290,19 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, btrfs_set_token_inode_mode(leaf, item, inode->i_mode, &token); btrfs_set_token_inode_nlink(leaf, item, inode->i_nlink, &token); - btrfs_set_token_timespec_sec(leaf, btrfs_inode_atime(item), + btrfs_set_token_timespec_sec(leaf, &item->atime, inode->i_atime.tv_sec, &token); - btrfs_set_token_timespec_nsec(leaf, btrfs_inode_atime(item), + btrfs_set_token_timespec_nsec(leaf, &item->atime, inode->i_atime.tv_nsec, &token); - btrfs_set_token_timespec_sec(leaf, btrfs_inode_mtime(item), + btrfs_set_token_timespec_sec(leaf, &item->mtime, inode->i_mtime.tv_sec, &token); - btrfs_set_token_timespec_nsec(leaf, btrfs_inode_mtime(item), + btrfs_set_token_timespec_nsec(leaf, &item->mtime, inode->i_mtime.tv_nsec, &token); - btrfs_set_token_timespec_sec(leaf, btrfs_inode_ctime(item), + btrfs_set_token_timespec_sec(leaf, &item->ctime, inode->i_ctime.tv_sec, &token); - btrfs_set_token_timespec_nsec(leaf, btrfs_inode_ctime(item), + btrfs_set_token_timespec_nsec(leaf, &item->ctime, inode->i_ctime.tv_nsec, &token); btrfs_set_token_inode_nbytes(leaf, item, inode_get_bytes(inode), @@ -3284,7 +3329,7 @@ static int log_inode_item(struct btrfs_trans_handle *trans, return ret; inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0], struct btrfs_inode_item); - fill_inode_item(trans, path->nodes[0], inode_item, inode, 0); + fill_inode_item(trans, path->nodes[0], inode_item, inode, 0, 0); btrfs_release_path(path); return 0; } @@ -3293,7 +3338,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, struct inode *inode, struct btrfs_path *dst_path, struct btrfs_path *src_path, u64 *last_extent, - int start_slot, int nr, int inode_only) + int start_slot, int nr, int inode_only, + u64 logged_isize) { unsigned long src_offset; unsigned long dst_offset; @@ -3350,7 +3396,8 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, dst_path->slots[0], struct btrfs_inode_item); fill_inode_item(trans, dst_path->nodes[0], inode_item, - inode, inode_only == LOG_INODE_EXISTS); + inode, inode_only == LOG_INODE_EXISTS, + logged_isize); } else { copy_extent_buffer(dst_path->nodes[0], src, dst_offset, src_offset, ins_sizes[i]); @@ -3902,6 +3949,33 @@ process: return ret; } +static int logged_inode_size(struct btrfs_root *log, struct inode *inode, + struct btrfs_path *path, u64 *size_ret) +{ + struct btrfs_key key; + int ret; + + key.objectid = btrfs_ino(inode); + key.type = BTRFS_INODE_ITEM_KEY; + key.offset = 0; + + ret = btrfs_search_slot(NULL, log, &key, path, 0, 0); + if (ret < 0) { + return ret; + } else if (ret > 0) { + *size_ret = i_size_read(inode); + } else { + struct btrfs_inode_item *item; + + item = btrfs_item_ptr(path->nodes[0], path->slots[0], + struct btrfs_inode_item); + *size_ret = btrfs_inode_size(path->nodes[0], item); + } + + btrfs_release_path(path); + return 0; +} + /* log a single inode in the tree log. * At least one parent directory for this inode must exist in the tree * or be logged already. @@ -3939,6 +4013,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, bool fast_search = false; u64 ino = btrfs_ino(inode); struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; + u64 logged_isize = 0; path = btrfs_alloc_path(); if (!path) @@ -3966,15 +4041,22 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, max_key.type = (u8)-1; max_key.offset = (u64)-1; - /* Only run delayed items if we are a dir or a new file */ + /* + * Only run delayed items if we are a dir or a new file. + * Otherwise commit the delayed inode only, which is needed in + * order for the log replay code to mark inodes for link count + * fixup (create temporary BTRFS_TREE_LOG_FIXUP_OBJECTID items). + */ if (S_ISDIR(inode->i_mode) || - BTRFS_I(inode)->generation > root->fs_info->last_trans_committed) { + BTRFS_I(inode)->generation > root->fs_info->last_trans_committed) ret = btrfs_commit_inode_delayed_items(trans, inode); - if (ret) { - btrfs_free_path(path); - btrfs_free_path(dst_path); - return ret; - } + else + ret = btrfs_commit_inode_delayed_inode(inode); + + if (ret) { + btrfs_free_path(path); + btrfs_free_path(dst_path); + return ret; } mutex_lock(&BTRFS_I(inode)->log_mutex); @@ -3988,22 +4070,56 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, if (S_ISDIR(inode->i_mode)) { int max_key_type = BTRFS_DIR_LOG_INDEX_KEY; - if (inode_only == LOG_INODE_EXISTS) - max_key_type = BTRFS_XATTR_ITEM_KEY; + if (inode_only == LOG_INODE_EXISTS) { + max_key_type = BTRFS_INODE_EXTREF_KEY; + max_key.type = max_key_type; + } ret = drop_objectid_items(trans, log, path, ino, max_key_type); } else { - if (test_and_clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC, - &BTRFS_I(inode)->runtime_flags)) { - clear_bit(BTRFS_INODE_COPY_EVERYTHING, - &BTRFS_I(inode)->runtime_flags); - ret = btrfs_truncate_inode_items(trans, log, - inode, 0, 0); - } else if (test_and_clear_bit(BTRFS_INODE_COPY_EVERYTHING, - &BTRFS_I(inode)->runtime_flags) || + if (inode_only == LOG_INODE_EXISTS) { + /* + * Make sure the new inode item we write to the log has + * the same isize as the current one (if it exists). + * This is necessary to prevent data loss after log + * replay, and also to prevent doing a wrong expanding + * truncate - for e.g. create file, write 4K into offset + * 0, fsync, write 4K into offset 4096, add hard link, + * fsync some other file (to sync log), power fail - if + * we use the inode's current i_size, after log replay + * we get a 8Kb file, with the last 4Kb extent as a hole + * (zeroes), as if an expanding truncate happened, + * instead of getting a file of 4Kb only. + */ + err = logged_inode_size(log, inode, path, + &logged_isize); + if (err) + goto out_unlock; + } + if (test_bit(BTRFS_INODE_NEEDS_FULL_SYNC, + &BTRFS_I(inode)->runtime_flags)) { + if (inode_only == LOG_INODE_EXISTS) { + max_key.type = BTRFS_INODE_EXTREF_KEY; + ret = drop_objectid_items(trans, log, path, ino, + max_key.type); + } else { + clear_bit(BTRFS_INODE_NEEDS_FULL_SYNC, + &BTRFS_I(inode)->runtime_flags); + clear_bit(BTRFS_INODE_COPY_EVERYTHING, + &BTRFS_I(inode)->runtime_flags); + ret = btrfs_truncate_inode_items(trans, log, + inode, 0, 0); + } + } else if (test_bit(BTRFS_INODE_COPY_EVERYTHING, + &BTRFS_I(inode)->runtime_flags) || inode_only == LOG_INODE_EXISTS) { - if (inode_only == LOG_INODE_ALL) + if (inode_only == LOG_INODE_ALL) { + clear_bit(BTRFS_INODE_COPY_EVERYTHING, + &BTRFS_I(inode)->runtime_flags); fast_search = true; - max_key.type = BTRFS_XATTR_ITEM_KEY; + max_key.type = BTRFS_XATTR_ITEM_KEY; + } else { + max_key.type = BTRFS_INODE_EXTREF_KEY; + } ret = drop_objectid_items(trans, log, path, ino, max_key.type); } else { @@ -4047,7 +4163,8 @@ again: } ret = copy_items(trans, inode, dst_path, path, &last_extent, - ins_start_slot, ins_nr, inode_only); + ins_start_slot, ins_nr, inode_only, + logged_isize); if (ret < 0) { err = ret; goto out_unlock; @@ -4071,7 +4188,7 @@ next_slot: if (ins_nr) { ret = copy_items(trans, inode, dst_path, path, &last_extent, ins_start_slot, - ins_nr, inode_only); + ins_nr, inode_only, logged_isize); if (ret < 0) { err = ret; goto out_unlock; @@ -4092,7 +4209,8 @@ next_slot: } if (ins_nr) { ret = copy_items(trans, inode, dst_path, path, &last_extent, - ins_start_slot, ins_nr, inode_only); + ins_start_slot, ins_nr, inode_only, + logged_isize); if (ret < 0) { err = ret; goto out_unlock; @@ -4273,6 +4391,9 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, struct dentry *old_parent = NULL; int ret = 0; u64 last_committed = root->fs_info->last_trans_committed; + const struct dentry * const first_parent = parent; + const bool did_unlink = (BTRFS_I(inode)->last_unlink_trans > + last_committed); sb = inode->i_sb; @@ -4328,7 +4449,6 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, goto end_trans; } - inode_only = LOG_INODE_EXISTS; while (1) { if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb) break; @@ -4337,8 +4457,22 @@ static int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, if (root != BTRFS_I(inode)->root) break; + /* + * On unlink we must make sure our immediate parent directory + * inode is fully logged. This is to prevent leaving dangling + * directory index entries and a wrong directory inode's i_size. + * Not doing so can result in a directory being impossible to + * delete after log replay (rmdir will always fail with error + * -ENOTEMPTY). + */ + if (did_unlink && parent == first_parent) + inode_only = LOG_INODE_ALL; + else + inode_only = LOG_INODE_EXISTS; + if (BTRFS_I(inode)->generation > - root->fs_info->last_trans_committed) { + root->fs_info->last_trans_committed || + inode_only == LOG_INODE_ALL) { ret = btrfs_log_inode(trans, root, inode, inode_only, 0, LLONG_MAX, ctx); if (ret) diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 50c5a87..8222f6f 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1310,6 +1310,8 @@ again: if (ret) { btrfs_error(root->fs_info, ret, "Failed to remove dev extent item"); + } else { + trans->transaction->have_free_bgs = 1; } out: btrfs_free_path(path); @@ -4196,7 +4198,7 @@ static u32 find_raid56_stripe_len(u32 data_devices, u32 dev_stripe_target) static void check_raid56_incompat_flag(struct btrfs_fs_info *info, u64 type) { - if (!(type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6))) + if (!(type & BTRFS_BLOCK_GROUP_RAID56_MASK)) return; btrfs_set_fs_incompat(info, RAID56); @@ -4803,10 +4805,8 @@ unsigned long btrfs_full_stripe_len(struct btrfs_root *root, BUG_ON(em->start > logical || em->start + em->len < logical); map = (struct map_lookup *)em->bdev; - if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) len = map->stripe_len * nr_data_stripes(map); - } free_extent_map(em); return len; } @@ -4826,8 +4826,7 @@ int btrfs_is_parity_mirror(struct btrfs_mapping_tree *map_tree, BUG_ON(em->start > logical || em->start + em->len < logical); map = (struct map_lookup *)em->bdev; - if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) ret = 1; free_extent_map(em); return ret; @@ -4876,32 +4875,24 @@ static inline int parity_smaller(u64 a, u64 b) } /* Bubble-sort the stripe set to put the parity/syndrome stripes last */ -static void sort_parity_stripes(struct btrfs_bio *bbio, u64 *raid_map) +static void sort_parity_stripes(struct btrfs_bio *bbio, int num_stripes) { struct btrfs_bio_stripe s; - int real_stripes = bbio->num_stripes - bbio->num_tgtdevs; int i; u64 l; int again = 1; - int m; while (again) { again = 0; - for (i = 0; i < real_stripes - 1; i++) { - if (parity_smaller(raid_map[i], raid_map[i+1])) { + for (i = 0; i < num_stripes - 1; i++) { + if (parity_smaller(bbio->raid_map[i], + bbio->raid_map[i+1])) { s = bbio->stripes[i]; - l = raid_map[i]; + l = bbio->raid_map[i]; bbio->stripes[i] = bbio->stripes[i+1]; - raid_map[i] = raid_map[i+1]; + bbio->raid_map[i] = bbio->raid_map[i+1]; bbio->stripes[i+1] = s; - raid_map[i+1] = l; - - if (bbio->tgtdev_map) { - m = bbio->tgtdev_map[i]; - bbio->tgtdev_map[i] = - bbio->tgtdev_map[i + 1]; - bbio->tgtdev_map[i + 1] = m; - } + bbio->raid_map[i+1] = l; again = 1; } @@ -4909,10 +4900,48 @@ static void sort_parity_stripes(struct btrfs_bio *bbio, u64 *raid_map) } } +static struct btrfs_bio *alloc_btrfs_bio(int total_stripes, int real_stripes) +{ + struct btrfs_bio *bbio = kzalloc( + /* the size of the btrfs_bio */ + sizeof(struct btrfs_bio) + + /* plus the variable array for the stripes */ + sizeof(struct btrfs_bio_stripe) * (total_stripes) + + /* plus the variable array for the tgt dev */ + sizeof(int) * (real_stripes) + + /* + * plus the raid_map, which includes both the tgt dev + * and the stripes + */ + sizeof(u64) * (total_stripes), + GFP_NOFS); + if (!bbio) + return NULL; + + atomic_set(&bbio->error, 0); + atomic_set(&bbio->refs, 1); + + return bbio; +} + +void btrfs_get_bbio(struct btrfs_bio *bbio) +{ + WARN_ON(!atomic_read(&bbio->refs)); + atomic_inc(&bbio->refs); +} + +void btrfs_put_bbio(struct btrfs_bio *bbio) +{ + if (!bbio) + return; + if (atomic_dec_and_test(&bbio->refs)) + kfree(bbio); +} + static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, u64 logical, u64 *length, struct btrfs_bio **bbio_ret, - int mirror_num, u64 **raid_map_ret) + int mirror_num, int need_raid_map) { struct extent_map *em; struct map_lookup *map; @@ -4925,7 +4954,6 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, u64 stripe_nr_orig; u64 stripe_nr_end; u64 stripe_len; - u64 *raid_map = NULL; int stripe_index; int i; int ret = 0; @@ -4976,7 +5004,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, stripe_offset = offset - stripe_offset; /* if we're here for raid56, we need to know the stripe aligned start */ - if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { unsigned long full_stripe_len = stripe_len * nr_data_stripes(map); raid56_full_stripe_start = offset; @@ -4989,8 +5017,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, if (rw & REQ_DISCARD) { /* we don't discard raid56 yet */ - if (map->type & - (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { ret = -EOPNOTSUPP; goto out; } @@ -5000,7 +5027,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, /* For writes to RAID[56], allow a full stripeset across all disks. For other RAID types and for RAID[56] reads, just allow a single stripe (on a single disk). */ - if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6) && + if ((map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) && (rw & REQ_WRITE)) { max_len = stripe_len * nr_data_stripes(map) - (offset - raid56_full_stripe_start); @@ -5047,7 +5074,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, u64 physical_of_found = 0; ret = __btrfs_map_block(fs_info, REQ_GET_READ_MIRRORS, - logical, &tmp_length, &tmp_bbio, 0, NULL); + logical, &tmp_length, &tmp_bbio, 0, 0); if (ret) { WARN_ON(tmp_bbio != NULL); goto out; @@ -5061,7 +5088,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, * is not left of the left cursor */ ret = -EIO; - kfree(tmp_bbio); + btrfs_put_bbio(tmp_bbio); goto out; } @@ -5096,11 +5123,11 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, } else { WARN_ON(1); ret = -EIO; - kfree(tmp_bbio); + btrfs_put_bbio(tmp_bbio); goto out; } - kfree(tmp_bbio); + btrfs_put_bbio(tmp_bbio); } else if (mirror_num > map->num_stripes) { mirror_num = 0; } @@ -5166,15 +5193,10 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, mirror_num = stripe_index - old_stripe_index + 1; } - } else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) { - u64 tmp; - - if (raid_map_ret && + } else if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { + if (need_raid_map && ((rw & (REQ_WRITE | REQ_GET_READ_MIRRORS)) || mirror_num > 1)) { - int i, rot; - /* push stripe_nr back to the start of the full stripe */ stripe_nr = raid56_full_stripe_start; do_div(stripe_nr, stripe_len * nr_data_stripes(map)); @@ -5183,32 +5205,12 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, num_stripes = map->num_stripes; max_errors = nr_parity_stripes(map); - raid_map = kmalloc_array(num_stripes, sizeof(u64), - GFP_NOFS); - if (!raid_map) { - ret = -ENOMEM; - goto out; - } - - /* Work out the disk rotation on this stripe-set */ - tmp = stripe_nr; - rot = do_div(tmp, num_stripes); - - /* Fill in the logical address of each stripe */ - tmp = stripe_nr * nr_data_stripes(map); - for (i = 0; i < nr_data_stripes(map); i++) - raid_map[(i+rot) % num_stripes] = - em->start + (tmp + i) * map->stripe_len; - - raid_map[(i+rot) % map->num_stripes] = RAID5_P_STRIPE; - if (map->type & BTRFS_BLOCK_GROUP_RAID6) - raid_map[(i+rot+1) % num_stripes] = - RAID6_Q_STRIPE; - *length = map->stripe_len; stripe_index = 0; stripe_offset = 0; } else { + u64 tmp; + /* * Mirror #0 or #1 means the original data block. * Mirror #2 is RAID5 parity block. @@ -5246,17 +5248,42 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, tgtdev_indexes = num_stripes; } - bbio = kzalloc(btrfs_bio_size(num_alloc_stripes, tgtdev_indexes), - GFP_NOFS); + bbio = alloc_btrfs_bio(num_alloc_stripes, tgtdev_indexes); if (!bbio) { - kfree(raid_map); ret = -ENOMEM; goto out; } - atomic_set(&bbio->error, 0); if (dev_replace_is_ongoing) bbio->tgtdev_map = (int *)(bbio->stripes + num_alloc_stripes); + /* build raid_map */ + if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK && + need_raid_map && ((rw & (REQ_WRITE | REQ_GET_READ_MIRRORS)) || + mirror_num > 1)) { + u64 tmp; + int i, rot; + + bbio->raid_map = (u64 *)((void *)bbio->stripes + + sizeof(struct btrfs_bio_stripe) * + num_alloc_stripes + + sizeof(int) * tgtdev_indexes); + + /* Work out the disk rotation on this stripe-set */ + tmp = stripe_nr; + rot = do_div(tmp, num_stripes); + + /* Fill in the logical address of each stripe */ + tmp = stripe_nr * nr_data_stripes(map); + for (i = 0; i < nr_data_stripes(map); i++) + bbio->raid_map[(i+rot) % num_stripes] = + em->start + (tmp + i) * map->stripe_len; + + bbio->raid_map[(i+rot) % map->num_stripes] = RAID5_P_STRIPE; + if (map->type & BTRFS_BLOCK_GROUP_RAID6) + bbio->raid_map[(i+rot+1) % num_stripes] = + RAID6_Q_STRIPE; + } + if (rw & REQ_DISCARD) { int factor = 0; int sub_stripes = 0; @@ -5340,6 +5367,9 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, if (rw & (REQ_WRITE | REQ_GET_READ_MIRRORS)) max_errors = btrfs_chunk_max_errors(map); + if (bbio->raid_map) + sort_parity_stripes(bbio, num_stripes); + tgtdev_indexes = 0; if (dev_replace_is_ongoing && (rw & (REQ_WRITE | REQ_DISCARD)) && dev_replace->tgtdev != NULL) { @@ -5427,6 +5457,7 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, } *bbio_ret = bbio; + bbio->map_type = map->type; bbio->num_stripes = num_stripes; bbio->max_errors = max_errors; bbio->mirror_num = mirror_num; @@ -5443,10 +5474,6 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, bbio->stripes[0].physical = physical_to_patch_in_first_stripe; bbio->mirror_num = map->num_stripes + 1; } - if (raid_map) { - sort_parity_stripes(bbio, raid_map); - *raid_map_ret = raid_map; - } out: if (dev_replace_is_ongoing) btrfs_dev_replace_unlock(dev_replace); @@ -5459,17 +5486,17 @@ int btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, struct btrfs_bio **bbio_ret, int mirror_num) { return __btrfs_map_block(fs_info, rw, logical, length, bbio_ret, - mirror_num, NULL); + mirror_num, 0); } /* For Scrub/replace */ int btrfs_map_sblock(struct btrfs_fs_info *fs_info, int rw, u64 logical, u64 *length, struct btrfs_bio **bbio_ret, int mirror_num, - u64 **raid_map_ret) + int need_raid_map) { return __btrfs_map_block(fs_info, rw, logical, length, bbio_ret, - mirror_num, raid_map_ret); + mirror_num, need_raid_map); } int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, @@ -5511,8 +5538,7 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, do_div(length, map->num_stripes / map->sub_stripes); else if (map->type & BTRFS_BLOCK_GROUP_RAID0) do_div(length, map->num_stripes); - else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | - BTRFS_BLOCK_GROUP_RAID6)) { + else if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) { do_div(length, nr_data_stripes(map)); rmap_len = map->stripe_len * nr_data_stripes(map); } @@ -5565,7 +5591,7 @@ static inline void btrfs_end_bbio(struct btrfs_bio *bbio, struct bio *bio, int e bio_endio_nodec(bio, err); else bio_endio(bio, err); - kfree(bbio); + btrfs_put_bbio(bbio); } static void btrfs_end_bio(struct bio *bio, int err) @@ -5808,7 +5834,6 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, u64 logical = (u64)bio->bi_iter.bi_sector << 9; u64 length = 0; u64 map_length; - u64 *raid_map = NULL; int ret; int dev_nr = 0; int total_devs = 1; @@ -5819,7 +5844,7 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, btrfs_bio_counter_inc_blocked(root->fs_info); ret = __btrfs_map_block(root->fs_info, rw, logical, &map_length, &bbio, - mirror_num, &raid_map); + mirror_num, 1); if (ret) { btrfs_bio_counter_dec(root->fs_info); return ret; @@ -5832,15 +5857,13 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, bbio->fs_info = root->fs_info; atomic_set(&bbio->stripes_pending, bbio->num_stripes); - if (raid_map) { + if (bbio->raid_map) { /* In this case, map_length has been set to the length of a single stripe; not the whole write */ if (rw & WRITE) { - ret = raid56_parity_write(root, bio, bbio, - raid_map, map_length); + ret = raid56_parity_write(root, bio, bbio, map_length); } else { - ret = raid56_parity_recover(root, bio, bbio, - raid_map, map_length, + ret = raid56_parity_recover(root, bio, bbio, map_length, mirror_num, 1); } @@ -6238,17 +6261,22 @@ int btrfs_read_sys_array(struct btrfs_root *root) struct extent_buffer *sb; struct btrfs_disk_key *disk_key; struct btrfs_chunk *chunk; - u8 *ptr; - unsigned long sb_ptr; + u8 *array_ptr; + unsigned long sb_array_offset; int ret = 0; u32 num_stripes; u32 array_size; u32 len = 0; - u32 cur; + u32 cur_offset; struct btrfs_key key; - sb = btrfs_find_create_tree_block(root, BTRFS_SUPER_INFO_OFFSET, - BTRFS_SUPER_INFO_SIZE); + ASSERT(BTRFS_SUPER_INFO_SIZE <= root->nodesize); + /* + * This will create extent buffer of nodesize, superblock size is + * fixed to BTRFS_SUPER_INFO_SIZE. If nodesize > sb size, this will + * overallocate but we can keep it as-is, only the first page is used. + */ + sb = btrfs_find_create_tree_block(root, BTRFS_SUPER_INFO_OFFSET); if (!sb) return -ENOMEM; btrfs_set_buffer_uptodate(sb); @@ -6271,35 +6299,56 @@ int btrfs_read_sys_array(struct btrfs_root *root) write_extent_buffer(sb, super_copy, 0, BTRFS_SUPER_INFO_SIZE); array_size = btrfs_super_sys_array_size(super_copy); - ptr = super_copy->sys_chunk_array; - sb_ptr = offsetof(struct btrfs_super_block, sys_chunk_array); - cur = 0; + array_ptr = super_copy->sys_chunk_array; + sb_array_offset = offsetof(struct btrfs_super_block, sys_chunk_array); + cur_offset = 0; + + while (cur_offset < array_size) { + disk_key = (struct btrfs_disk_key *)array_ptr; + len = sizeof(*disk_key); + if (cur_offset + len > array_size) + goto out_short_read; - while (cur < array_size) { - disk_key = (struct btrfs_disk_key *)ptr; btrfs_disk_key_to_cpu(&key, disk_key); - len = sizeof(*disk_key); ptr += len; - sb_ptr += len; - cur += len; + array_ptr += len; + sb_array_offset += len; + cur_offset += len; if (key.type == BTRFS_CHUNK_ITEM_KEY) { - chunk = (struct btrfs_chunk *)sb_ptr; + chunk = (struct btrfs_chunk *)sb_array_offset; + /* + * At least one btrfs_chunk with one stripe must be + * present, exact stripe count check comes afterwards + */ + len = btrfs_chunk_item_size(1); + if (cur_offset + len > array_size) + goto out_short_read; + + num_stripes = btrfs_chunk_num_stripes(sb, chunk); + len = btrfs_chunk_item_size(num_stripes); + if (cur_offset + len > array_size) + goto out_short_read; + ret = read_one_chunk(root, &key, sb, chunk); if (ret) break; - num_stripes = btrfs_chunk_num_stripes(sb, chunk); - len = btrfs_chunk_item_size(num_stripes); } else { ret = -EIO; break; } - ptr += len; - sb_ptr += len; - cur += len; + array_ptr += len; + sb_array_offset += len; + cur_offset += len; } free_extent_buffer(sb); return ret; + +out_short_read: + printk(KERN_ERR "BTRFS: sys_array too short to read %u bytes at offset %u\n", + len, cur_offset); + free_extent_buffer(sb); + return -EIO; } int btrfs_read_chunk_tree(struct btrfs_root *root) diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index d6fe73c..83069de 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -295,8 +295,10 @@ 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; struct btrfs_fs_info *fs_info; + u64 map_type; /* get from map_lookup->type */ bio_end_io_t *end_io; struct bio *orig_bio; unsigned long flags; @@ -307,6 +309,12 @@ struct btrfs_bio { int mirror_num; int num_tgtdevs; int *tgtdev_map; + /* + * logical block numbers for the start of each stripe + * The last one or two are p/q. These are sorted, + * so raid_map[0] is the start of our full stripe + */ + u64 *raid_map; struct btrfs_bio_stripe stripes[]; }; @@ -388,19 +396,15 @@ struct btrfs_balance_control { int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start, u64 end, u64 *length); - -#define btrfs_bio_size(total_stripes, real_stripes) \ - (sizeof(struct btrfs_bio) + \ - (sizeof(struct btrfs_bio_stripe) * (total_stripes)) + \ - (sizeof(int) * (real_stripes))) - +void btrfs_get_bbio(struct btrfs_bio *bbio); +void btrfs_put_bbio(struct btrfs_bio *bbio); int btrfs_map_block(struct btrfs_fs_info *fs_info, int rw, u64 logical, u64 *length, struct btrfs_bio **bbio_ret, int mirror_num); int btrfs_map_sblock(struct btrfs_fs_info *fs_info, int rw, u64 logical, u64 *length, struct btrfs_bio **bbio_ret, int mirror_num, - u64 **raid_map_ret); + int need_raid_map); int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree, u64 chunk_start, u64 physical, u64 devid, u64 **logical, int *naddrs, int *stripe_len); diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c index 47b1946..883b936 100644 --- a/fs/btrfs/xattr.c +++ b/fs/btrfs/xattr.c @@ -111,6 +111,8 @@ static int do_setxattr(struct btrfs_trans_handle *trans, name, name_len, -1); if (!di && (flags & XATTR_REPLACE)) ret = -ENODATA; + else if (IS_ERR(di)) + ret = PTR_ERR(di); else if (di) ret = btrfs_delete_one_dir_name(trans, root, path, di); goto out; @@ -127,10 +129,12 @@ static int do_setxattr(struct btrfs_trans_handle *trans, ASSERT(mutex_is_locked(&inode->i_mutex)); di = btrfs_lookup_xattr(NULL, root, path, btrfs_ino(inode), name, name_len, 0); - if (!di) { + if (!di) ret = -ENODATA; + else if (IS_ERR(di)) + ret = PTR_ERR(di); + if (ret) goto out; - } btrfs_release_path(path); di = NULL; } diff --git a/fs/cachefiles/daemon.c b/fs/cachefiles/daemon.c index ce1b115..f601def 100644 --- a/fs/cachefiles/daemon.c +++ b/fs/cachefiles/daemon.c @@ -574,7 +574,7 @@ static int cachefiles_daemon_cull(struct cachefiles_cache *cache, char *args) /* extract the directory dentry from the cwd */ get_fs_pwd(current->fs, &path); - if (!S_ISDIR(path.dentry->d_inode->i_mode)) + if (!d_can_lookup(path.dentry)) goto notdir; cachefiles_begin_secure(cache, &saved_cred); @@ -646,7 +646,7 @@ static int cachefiles_daemon_inuse(struct cachefiles_cache *cache, char *args) /* extract the directory dentry from the cwd */ get_fs_pwd(current->fs, &path); - if (!S_ISDIR(path.dentry->d_inode->i_mode)) + if (!d_can_lookup(path.dentry)) goto notdir; cachefiles_begin_secure(cache, &saved_cred); diff --git a/fs/cachefiles/interface.c b/fs/cachefiles/interface.c index 1c7293c..2324262 100644 --- a/fs/cachefiles/interface.c +++ b/fs/cachefiles/interface.c @@ -437,7 +437,7 @@ static int cachefiles_attr_changed(struct fscache_object *_object) if (!object->backer) return -ENOBUFS; - ASSERT(S_ISREG(object->backer->d_inode->i_mode)); + ASSERT(d_is_reg(object->backer)); fscache_set_store_limit(&object->fscache, ni_size); @@ -501,7 +501,7 @@ static void cachefiles_invalidate_object(struct fscache_operation *op) op->object->debug_id, (unsigned long long)ni_size); if (object->backer) { - ASSERT(S_ISREG(object->backer->d_inode->i_mode)); + ASSERT(d_is_reg(object->backer)); fscache_set_store_limit(&object->fscache, ni_size); diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c index 7f8e83f..1e51714e 100644 --- a/fs/cachefiles/namei.c +++ b/fs/cachefiles/namei.c @@ -277,7 +277,7 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache, _debug("remove %p from %p", rep, dir); /* non-directories can just be unlinked */ - if (!S_ISDIR(rep->d_inode->i_mode)) { + if (!d_is_dir(rep)) { _debug("unlink stale object"); path.mnt = cache->mnt; @@ -323,7 +323,7 @@ try_again: return 0; } - if (!S_ISDIR(cache->graveyard->d_inode->i_mode)) { + if (!d_can_lookup(cache->graveyard)) { unlock_rename(cache->graveyard, dir); cachefiles_io_error(cache, "Graveyard no longer a directory"); return -EIO; @@ -475,7 +475,7 @@ int cachefiles_walk_to_object(struct cachefiles_object *parent, ASSERT(parent->dentry); ASSERT(parent->dentry->d_inode); - if (!(S_ISDIR(parent->dentry->d_inode->i_mode))) { + if (!(d_is_dir(parent->dentry))) { // TODO: convert file to dir _leave("looking up in none directory"); return -ENOBUFS; @@ -539,7 +539,7 @@ lookup_again: _debug("mkdir -> %p{%p{ino=%lu}}", next, next->d_inode, next->d_inode->i_ino); - } else if (!S_ISDIR(next->d_inode->i_mode)) { + } else if (!d_can_lookup(next)) { pr_err("inode %lu is not a directory\n", next->d_inode->i_ino); ret = -ENOBUFS; @@ -568,8 +568,8 @@ lookup_again: _debug("create -> %p{%p{ino=%lu}}", next, next->d_inode, next->d_inode->i_ino); - } else if (!S_ISDIR(next->d_inode->i_mode) && - !S_ISREG(next->d_inode->i_mode) + } else if (!d_can_lookup(next) && + !d_is_reg(next) ) { pr_err("inode %lu is not a file or directory\n", next->d_inode->i_ino); @@ -642,7 +642,7 @@ lookup_again: /* open a file interface onto a data file */ if (object->type != FSCACHE_COOKIE_TYPE_INDEX) { - if (S_ISREG(object->dentry->d_inode->i_mode)) { + if (d_is_reg(object->dentry)) { const struct address_space_operations *aops; ret = -EPERM; @@ -763,7 +763,7 @@ struct dentry *cachefiles_get_directory(struct cachefiles_cache *cache, /* we need to make sure the subdir is a directory */ ASSERT(subdir->d_inode); - if (!S_ISDIR(subdir->d_inode->i_mode)) { + if (!d_can_lookup(subdir)) { pr_err("%s is not a directory\n", dirname); ret = -EIO; goto check_error; diff --git a/fs/cachefiles/rdwr.c b/fs/cachefiles/rdwr.c index 616db0e7..c6cd8d7 100644 --- a/fs/cachefiles/rdwr.c +++ b/fs/cachefiles/rdwr.c @@ -900,7 +900,7 @@ int cachefiles_write_page(struct fscache_storage *op, struct page *page) return -ENOBUFS; } - ASSERT(S_ISREG(object->backer->d_inode->i_mode)); + ASSERT(d_is_reg(object->backer)); cache = container_of(object->fscache.cache, struct cachefiles_cache, cache); diff --git a/fs/ceph/acl.c b/fs/ceph/acl.c index 5bd853b..64fa248 100644 --- a/fs/ceph/acl.c +++ b/fs/ceph/acl.c @@ -40,20 +40,6 @@ static inline void ceph_set_cached_acl(struct inode *inode, spin_unlock(&ci->i_ceph_lock); } -static inline struct posix_acl *ceph_get_cached_acl(struct inode *inode, - int type) -{ - struct ceph_inode_info *ci = ceph_inode(inode); - struct posix_acl *acl = ACL_NOT_CACHED; - - spin_lock(&ci->i_ceph_lock); - if (__ceph_caps_issued_mask(ci, CEPH_CAP_XATTR_SHARED, 0)) - acl = get_cached_acl(inode, type); - spin_unlock(&ci->i_ceph_lock); - - return acl; -} - struct posix_acl *ceph_get_acl(struct inode *inode, int type) { int size; diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 24be059..fd5599d 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -196,17 +196,22 @@ static int readpage_nounlock(struct file *filp, struct page *page) u64 len = PAGE_CACHE_SIZE; if (off >= i_size_read(inode)) { - zero_user_segment(page, err, PAGE_CACHE_SIZE); + zero_user_segment(page, 0, PAGE_CACHE_SIZE); SetPageUptodate(page); return 0; } - /* - * Uptodate inline data should have been added into page cache - * while getting Fcr caps. - */ - if (ci->i_inline_version != CEPH_INLINE_NONE) - return -EINVAL; + if (ci->i_inline_version != CEPH_INLINE_NONE) { + /* + * Uptodate inline data should have been added + * into page cache while getting Fcr caps. + */ + if (off == 0) + return -EINVAL; + zero_user_segment(page, 0, PAGE_CACHE_SIZE); + SetPageUptodate(page); + return 0; + } err = ceph_readpage_from_fscache(inode, page); if (err == 0) diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index b93c631..8172775 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -577,7 +577,6 @@ void ceph_add_cap(struct inode *inode, struct ceph_snap_realm *realm = ceph_lookup_snap_realm(mdsc, realmino); if (realm) { - ceph_get_snap_realm(mdsc, realm); spin_lock(&realm->inodes_with_caps_lock); ci->i_snap_realm = realm; list_add(&ci->i_snap_realm_item, @@ -1451,8 +1450,8 @@ static int __mark_caps_flushing(struct inode *inode, spin_lock(&mdsc->cap_dirty_lock); list_del_init(&ci->i_dirty_item); - ci->i_cap_flush_seq = ++mdsc->cap_flush_seq; 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, @@ -2073,17 +2072,16 @@ 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, struct page **pinned_page, - int *check_max, int *err) + loff_t endoff, int *got, int *check_max, int *err) { struct inode *inode = &ci->vfs_inode; int ret = 0; - int have, implemented, _got = 0; + int have, implemented; int file_wanted; 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 */ @@ -2138,50 +2136,34 @@ again: inode, ceph_cap_string(have), ceph_cap_string(not), ceph_cap_string(revoking)); if ((revoking & not) == 0) { - _got = need | (have & want); - __take_cap_refs(ci, _got); + *got = need | (have & want); + __take_cap_refs(ci, *got); ret = 1; } } else { + int session_readonly = false; + if ((need & CEPH_CAP_FILE_WR) && ci->i_auth_cap) { + struct ceph_mds_session *s = ci->i_auth_cap->session; + spin_lock(&s->s_cap_lock); + session_readonly = s->s_readonly; + spin_unlock(&s->s_cap_lock); + } + if (session_readonly) { + dout("get_cap_refs %p needed %s but mds%d readonly\n", + inode, ceph_cap_string(need), ci->i_auth_cap->mds); + *err = -EROFS; + ret = 1; + goto out_unlock; + } + dout("get_cap_refs %p have %s needed %s\n", inode, ceph_cap_string(have), ceph_cap_string(need)); } out_unlock: spin_unlock(&ci->i_ceph_lock); - if (ci->i_inline_version != CEPH_INLINE_NONE && - (_got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) && - i_size_read(inode) > 0) { - int ret1; - struct page *page = find_get_page(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); - _got = 0; - - /* getattr request will bring inline data into page cache */ - ret1 = __ceph_do_getattr(inode, NULL, - CEPH_STAT_CAP_INLINE_DATA, true); - if (ret1 >= 0) { - ret = 0; - goto again; - } - *err = ret1; - ret = 1; - } -out: dout("get_cap_refs %p ret %d got %s\n", inode, - ret, ceph_cap_string(_got)); - *got = _got; + ret, ceph_cap_string(*got)); return ret; } @@ -2221,22 +2203,52 @@ 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 check_max, ret, err; + int _got, check_max, ret, err = 0; retry: if (endoff > 0) check_max_size(&ci->vfs_inode, endoff); + _got = 0; check_max = 0; - err = 0; ret = wait_event_interruptible(ci->i_cap_wq, - try_get_cap_refs(ci, need, want, endoff, - got, pinned_page, - &check_max, &err)); + try_get_cap_refs(ci, need, want, endoff, + &_got, &check_max, &err)); if (err) ret = err; + if (ret < 0) + return ret; + if (check_max) goto retry; - return ret; + + 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); + _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; + goto retry; + } +out: + *got = _got; + return 0; } /* @@ -2432,13 +2444,13 @@ static void invalidate_aliases(struct inode *inode) */ static void handle_cap_grant(struct ceph_mds_client *mdsc, struct inode *inode, struct ceph_mds_caps *grant, - void *snaptrace, int snaptrace_len, u64 inline_version, void *inline_data, int inline_len, struct ceph_buffer *xattr_buf, struct ceph_mds_session *session, struct ceph_cap *cap, int issued) __releases(ci->i_ceph_lock) + __releases(mdsc->snap_rwsem) { struct ceph_inode_info *ci = ceph_inode(inode); int mds = session->s_mds; @@ -2639,10 +2651,6 @@ static void handle_cap_grant(struct ceph_mds_client *mdsc, spin_unlock(&ci->i_ceph_lock); if (le32_to_cpu(grant->op) == CEPH_CAP_OP_IMPORT) { - down_write(&mdsc->snap_rwsem); - ceph_update_snap_trace(mdsc, snaptrace, - snaptrace + snaptrace_len, false); - downgrade_write(&mdsc->snap_rwsem); kick_flushing_inode_caps(mdsc, session, inode); up_read(&mdsc->snap_rwsem); if (newcaps & ~issued) @@ -3052,6 +3060,7 @@ void ceph_handle_caps(struct ceph_mds_session *session, struct ceph_cap *cap; struct ceph_mds_caps *h; struct ceph_mds_cap_peer *peer = NULL; + struct ceph_snap_realm *realm; int mds = session->s_mds; int op, issued; u32 seq, mseq; @@ -3153,11 +3162,23 @@ void ceph_handle_caps(struct ceph_mds_session *session, goto done_unlocked; case CEPH_CAP_OP_IMPORT: + realm = NULL; + if (snaptrace_len) { + down_write(&mdsc->snap_rwsem); + ceph_update_snap_trace(mdsc, snaptrace, + snaptrace + snaptrace_len, + false, &realm); + downgrade_write(&mdsc->snap_rwsem); + } else { + down_read(&mdsc->snap_rwsem); + } handle_cap_import(mdsc, inode, h, peer, session, &cap, &issued); - handle_cap_grant(mdsc, inode, h, snaptrace, snaptrace_len, + handle_cap_grant(mdsc, inode, h, inline_version, inline_data, inline_len, msg->middle, session, cap, issued); + if (realm) + ceph_put_snap_realm(mdsc, realm); goto done_unlocked; } @@ -3177,7 +3198,7 @@ void ceph_handle_caps(struct ceph_mds_session *session, case CEPH_CAP_OP_GRANT: __ceph_caps_issued(ci, &issued); issued |= __ceph_caps_dirty(ci); - handle_cap_grant(mdsc, inode, h, NULL, 0, + handle_cap_grant(mdsc, inode, h, inline_version, inline_data, inline_len, msg->middle, session, cap, issued); goto done_unlocked; diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index c241603..83e9976 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -26,8 +26,6 @@ * point by name. */ -const struct inode_operations ceph_dir_iops; -const struct file_operations ceph_dir_fops; const struct dentry_operations ceph_dentry_ops; /* @@ -672,13 +670,17 @@ int ceph_handle_notrace_create(struct inode *dir, struct dentry *dentry) /* * We created the item, then did a lookup, and found * it was already linked to another inode we already - * had in our cache (and thus got spliced). Link our - * dentry to that inode, but don't hash it, just in - * case the VFS wants to dereference it. + * had in our cache (and thus got spliced). To not + * confuse VFS (especially when inode is a directory), + * we don't link our dentry to that inode, return an + * error instead. + * + * This event should be rare and it happens only when + * we talk to old MDS. Recent MDS does not send traceless + * reply for request that creates new inode. */ - BUG_ON(!result->d_inode); - d_instantiate(dentry, result->d_inode); - return 0; + d_drop(result); + return -ESTALE; } return PTR_ERR(result); } @@ -902,7 +904,7 @@ static int ceph_unlink(struct inode *dir, struct dentry *dentry) } else if (ceph_snap(dir) == CEPH_NOSNAP) { dout("unlink/rmdir dir %p dn %p inode %p\n", dir, dentry, inode); - op = S_ISDIR(dentry->d_inode->i_mode) ? + op = d_is_dir(dentry) ? CEPH_MDS_OP_RMDIR : CEPH_MDS_OP_UNLINK; } else goto out; @@ -1335,6 +1337,13 @@ const struct file_operations ceph_dir_fops = { .fsync = ceph_dir_fsync, }; +const struct file_operations ceph_snapdir_fops = { + .iterate = ceph_readdir, + .llseek = ceph_dir_llseek, + .open = ceph_open, + .release = ceph_release, +}; + const struct inode_operations ceph_dir_iops = { .lookup = ceph_lookup, .permission = ceph_permission, @@ -1357,6 +1366,14 @@ const struct inode_operations ceph_dir_iops = { .atomic_open = ceph_atomic_open, }; +const struct inode_operations ceph_snapdir_iops = { + .lookup = ceph_lookup, + .permission = ceph_permission, + .getattr = ceph_getattr, + .mkdir = ceph_mkdir, + .rmdir = ceph_unlink, +}; + const struct dentry_operations ceph_dentry_ops = { .d_revalidate = ceph_d_revalidate, .d_release = ceph_d_release, diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 98e257c..139f2fe 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -274,10 +274,10 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry, err = ceph_mdsc_do_request(mdsc, (flags & (O_CREAT|O_TRUNC)) ? dir : NULL, req); + err = ceph_handle_snapdir(req, dentry, err); if (err) goto out_req; - err = ceph_handle_snapdir(req, dentry, err); if (err == 0 && (flags & O_CREAT) && !req->r_reply_info.head->is_dentry) err = ceph_handle_notrace_create(dir, dentry); @@ -291,7 +291,7 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry, } if (err) goto out_req; - if (dn || dentry->d_inode == NULL || S_ISLNK(dentry->d_inode->i_mode)) { + if (dn || dentry->d_inode == NULL || d_is_symlink(dentry)) { /* make vfs retry on splice, ENOENT, or symlink */ dout("atomic_open finish_no_open on dn %p\n", dn); err = finish_no_open(file, dn); @@ -391,13 +391,14 @@ more: if (ret >= 0) { int didpages; if (was_short && (pos + ret < inode->i_size)) { - u64 tmp = min(this_len - ret, - inode->i_size - pos - ret); + int zlen = min(this_len - ret, + inode->i_size - pos - ret); + int zoff = (o_direct ? buf_align : io_align) + + read + ret; dout(" zero gap %llu to %llu\n", - pos + ret, pos + ret + tmp); - ceph_zero_page_vector_range(page_align + read + ret, - tmp, pages); - ret += tmp; + pos + ret, pos + ret + zlen); + ceph_zero_page_vector_range(zoff, zlen, pages); + ret += zlen; } didpages = (page_align + ret) >> PAGE_CACHE_SHIFT; @@ -877,28 +878,34 @@ again: i_size = i_size_read(inode); if (retry_op == READ_INLINE) { - /* does not support inline data > PAGE_SIZE */ - if (i_size > PAGE_CACHE_SIZE) { - ret = -EIO; - } else if (iocb->ki_pos < i_size) { + BUG_ON(ret > 0 || read > 0); + if (iocb->ki_pos < i_size && + iocb->ki_pos < PAGE_CACHE_SIZE) { loff_t end = min_t(loff_t, i_size, iocb->ki_pos + len); + end = min_t(loff_t, end, PAGE_CACHE_SIZE); if (statret < end) zero_user_segment(page, statret, end); ret = copy_page_to_iter(page, iocb->ki_pos & ~PAGE_MASK, end - iocb->ki_pos, to); iocb->ki_pos += ret; - } else { - ret = 0; + read += ret; + } + if (iocb->ki_pos < i_size && read < len) { + size_t zlen = min_t(size_t, len - read, + i_size - iocb->ki_pos); + ret = iov_iter_zero(zlen, to); + iocb->ki_pos += ret; + read += ret; } __free_pages(page, 0); - return ret; + return read; } /* hit EOF or hole? */ if (retry_op == CHECK_EOF && iocb->ki_pos < i_size && - ret < len) { + ret < len) { dout("sync_read hit hole, ppos %lld < size %lld" ", reading more\n", iocb->ki_pos, inode->i_size); diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index 6b51736..119c43c 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -82,8 +82,8 @@ struct inode *ceph_get_snapdir(struct inode *parent) inode->i_mode = parent->i_mode; inode->i_uid = parent->i_uid; inode->i_gid = parent->i_gid; - inode->i_op = &ceph_dir_iops; - inode->i_fop = &ceph_dir_fops; + inode->i_op = &ceph_snapdir_iops; + inode->i_fop = &ceph_snapdir_fops; ci->i_snap_caps = CEPH_CAP_PIN; /* so we can open */ ci->i_rbytes = 0; return inode; @@ -838,30 +838,31 @@ static int fill_inode(struct inode *inode, struct page *locked_page, ceph_vinop(inode), inode->i_mode); } - /* set dir completion flag? */ - if (S_ISDIR(inode->i_mode) && - ci->i_files == 0 && ci->i_subdirs == 0 && - ceph_snap(inode) == CEPH_NOSNAP && - (le32_to_cpu(info->cap.caps) & CEPH_CAP_FILE_SHARED) && - (issued & CEPH_CAP_FILE_EXCL) == 0 && - !__ceph_dir_is_complete(ci)) { - dout(" marking %p complete (empty)\n", inode); - __ceph_dir_set_complete(ci, atomic_read(&ci->i_release_count), - ci->i_ordered_count); - } - /* were we issued a capability? */ if (info->cap.caps) { if (ceph_snap(inode) == CEPH_NOSNAP) { + unsigned caps = le32_to_cpu(info->cap.caps); ceph_add_cap(inode, session, le64_to_cpu(info->cap.cap_id), - cap_fmode, - le32_to_cpu(info->cap.caps), + cap_fmode, caps, le32_to_cpu(info->cap.wanted), le32_to_cpu(info->cap.seq), le32_to_cpu(info->cap.mseq), le64_to_cpu(info->cap.realm), info->cap.flags, &new_cap); + + /* set dir completion flag? */ + if (S_ISDIR(inode->i_mode) && + ci->i_files == 0 && ci->i_subdirs == 0 && + (caps & CEPH_CAP_FILE_SHARED) && + (issued & CEPH_CAP_FILE_EXCL) == 0 && + !__ceph_dir_is_complete(ci)) { + dout(" marking %p complete (empty)\n", inode); + __ceph_dir_set_complete(ci, + atomic_read(&ci->i_release_count), + ci->i_ordered_count); + } + wake = true; } else { dout(" %p got snap_caps %s\n", inode, @@ -1446,12 +1447,14 @@ retry_lookup: } if (!dn->d_inode) { - dn = splice_dentry(dn, in, NULL); - if (IS_ERR(dn)) { - err = PTR_ERR(dn); + struct dentry *realdn = splice_dentry(dn, in, NULL); + if (IS_ERR(realdn)) { + err = PTR_ERR(realdn); + d_drop(dn); dn = NULL; goto next_item; } + dn = realdn; } di = dn->d_fsdata; diff --git a/fs/ceph/locks.c b/fs/ceph/locks.c index 06ea5cd..4347039 100644 --- a/fs/ceph/locks.c +++ b/fs/ceph/locks.c @@ -245,6 +245,7 @@ int ceph_flock(struct file *file, int cmd, struct file_lock *fl) */ void ceph_count_locks(struct inode *inode, int *fcntl_count, int *flock_count) { + struct file_lock *lock; struct file_lock_context *ctx; *fcntl_count = 0; @@ -252,8 +253,12 @@ void ceph_count_locks(struct inode *inode, int *fcntl_count, int *flock_count) ctx = inode->i_flctx; if (ctx) { - *fcntl_count = ctx->flc_posix_cnt; - *flock_count = ctx->flc_flock_cnt; + spin_lock(&ctx->flc_lock); + list_for_each_entry(lock, &ctx->flc_posix, fl_list) + ++(*fcntl_count); + list_for_each_entry(lock, &ctx->flc_flock, fl_list) + ++(*flock_count); + spin_unlock(&ctx->flc_lock); } dout("counted %d flock locks and %d fcntl locks", *flock_count, *fcntl_count); diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 5f62fb7..71c073f 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -480,6 +480,7 @@ static struct ceph_mds_session *register_session(struct ceph_mds_client *mdsc, mdsc->max_sessions = newmax; } mdsc->sessions[mds] = s; + atomic_inc(&mdsc->num_sessions); atomic_inc(&s->s_ref); /* one ref to sessions[], one to caller */ ceph_con_open(&s->s_con, CEPH_ENTITY_TYPE_MDS, mds, @@ -503,6 +504,7 @@ static void __unregister_session(struct ceph_mds_client *mdsc, mdsc->sessions[s->s_mds] = NULL; ceph_con_close(&s->s_con); ceph_put_mds_session(s); + atomic_dec(&mdsc->num_sessions); } /* @@ -842,8 +844,9 @@ static struct ceph_msg *create_session_open_msg(struct ceph_mds_client *mdsc, u6 struct ceph_options *opt = mdsc->fsc->client->options; void *p; - const char* metadata[3][2] = { + const char* metadata[][2] = { {"hostname", utsname()->nodename}, + {"kernel_version", utsname()->release}, {"entity_id", opt->name ? opt->name : ""}, {NULL, NULL} }; @@ -1464,19 +1467,33 @@ out_unlocked: return err; } +static int check_cap_flush(struct inode *inode, u64 want_flush_seq) +{ + 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); + return ret; +} + /* * flush all dirty inode data to disk. * * returns true if we've flushed through want_flush_seq */ -static int check_cap_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq) +static void wait_caps_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq) { - int mds, ret = 1; + int mds; dout("check_cap_flush want %lld\n", want_flush_seq); mutex_lock(&mdsc->mutex); - for (mds = 0; ret && mds < mdsc->max_sessions; mds++) { + for (mds = 0; mds < mdsc->max_sessions; mds++) { struct ceph_mds_session *session = mdsc->sessions[mds]; + struct inode *inode = NULL; if (!session) continue; @@ -1489,29 +1506,29 @@ static int check_cap_flush(struct ceph_mds_client *mdsc, u64 want_flush_seq) list_entry(session->s_cap_flushing.next, struct ceph_inode_info, i_flushing_item); - struct inode *inode = &ci->vfs_inode; - spin_lock(&ci->i_ceph_lock); - if (ci->i_cap_flush_seq <= want_flush_seq) { + if (!check_cap_flush(&ci->vfs_inode, want_flush_seq)) { dout("check_cap_flush still flushing %p " - "seq %lld <= %lld to mds%d\n", inode, - ci->i_cap_flush_seq, want_flush_seq, - session->s_mds); - ret = 0; + "seq %lld <= %lld to mds%d\n", + &ci->vfs_inode, ci->i_cap_flush_seq, + want_flush_seq, session->s_mds); + inode = igrab(&ci->vfs_inode); } - spin_unlock(&ci->i_ceph_lock); } mutex_unlock(&session->s_mutex); ceph_put_mds_session(session); - if (!ret) - return ret; + if (inode) { + wait_event(mdsc->cap_flushing_wq, + check_cap_flush(inode, want_flush_seq)); + iput(inode); + } + mutex_lock(&mdsc->mutex); } mutex_unlock(&mdsc->mutex); dout("check_cap_flush ok, flushed thru %lld\n", want_flush_seq); - return ret; } /* @@ -1923,7 +1940,11 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc, head->num_releases = cpu_to_le16(releases); /* time stamp */ - ceph_encode_copy(&p, &req->r_stamp, sizeof(req->r_stamp)); + { + struct ceph_timespec ts; + ceph_encode_timespec(&ts, &req->r_stamp); + ceph_encode_copy(&p, &ts, sizeof(ts)); + } BUG_ON(p > end); msg->front.iov_len = p - msg->front.iov_base; @@ -2012,7 +2033,11 @@ static int __prepare_send_request(struct ceph_mds_client *mdsc, /* time stamp */ p = msg->front.iov_base + req->r_request_release_offset; - ceph_encode_copy(&p, &req->r_stamp, sizeof(req->r_stamp)); + { + struct ceph_timespec ts; + ceph_encode_timespec(&ts, &req->r_stamp); + ceph_encode_copy(&p, &ts, sizeof(ts)); + } msg->front.iov_len = p - msg->front.iov_base; msg->hdr.front_len = cpu_to_le32(msg->front.iov_len); @@ -2159,6 +2184,8 @@ static void kick_requests(struct ceph_mds_client *mdsc, int mds) p = rb_next(p); if (req->r_got_unsafe) continue; + if (req->r_attempts > 0) + continue; /* only new requests */ if (req->r_session && req->r_session->s_mds == mds) { dout(" kicking tid %llu\n", req->r_tid); @@ -2286,6 +2313,7 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg) struct ceph_mds_request *req; struct ceph_mds_reply_head *head = msg->front.iov_base; struct ceph_mds_reply_info_parsed *rinfo; /* parsed reply info */ + struct ceph_snap_realm *realm; u64 tid; int err, result; int mds = session->s_mds; @@ -2401,11 +2429,13 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg) } /* snap trace */ + realm = NULL; if (rinfo->snapblob_len) { down_write(&mdsc->snap_rwsem); ceph_update_snap_trace(mdsc, rinfo->snapblob, - rinfo->snapblob + rinfo->snapblob_len, - le32_to_cpu(head->op) == CEPH_MDS_OP_RMSNAP); + rinfo->snapblob + rinfo->snapblob_len, + le32_to_cpu(head->op) == CEPH_MDS_OP_RMSNAP, + &realm); downgrade_write(&mdsc->snap_rwsem); } else { down_read(&mdsc->snap_rwsem); @@ -2423,6 +2453,8 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg) mutex_unlock(&req->r_fill_mutex); up_read(&mdsc->snap_rwsem); + if (realm) + ceph_put_snap_realm(mdsc, realm); out_err: mutex_lock(&mdsc->mutex); if (!req->r_aborted) { @@ -2487,6 +2519,7 @@ static void handle_forward(struct ceph_mds_client *mdsc, dout("forward tid %llu to mds%d (we resend)\n", tid, next_mds); BUG_ON(req->r_err); BUG_ON(req->r_got_result); + req->r_attempts = 0; req->r_num_fwd = fwd_seq; req->r_resend_mds = next_mds; put_request_session(req); @@ -2580,6 +2613,14 @@ static void handle_session(struct ceph_mds_session *session, send_flushmsg_ack(mdsc, session, seq); break; + case CEPH_SESSION_FORCE_RO: + dout("force_session_readonly %p\n", session); + spin_lock(&session->s_cap_lock); + session->s_readonly = true; + spin_unlock(&session->s_cap_lock); + wake_up_session_caps(session, 0); + break; + default: pr_err("mdsc_handle_session bad op %d mds%d\n", op, mds); WARN_ON(1); @@ -2610,6 +2651,7 @@ static void replay_unsafe_requests(struct ceph_mds_client *mdsc, struct ceph_mds_session *session) { struct ceph_mds_request *req, *nreq; + struct rb_node *p; int err; dout("replay_unsafe_requests mds%d\n", session->s_mds); @@ -2622,6 +2664,28 @@ static void replay_unsafe_requests(struct ceph_mds_client *mdsc, ceph_con_send(&session->s_con, req->r_request); } } + + /* + * also re-send old requests when MDS enters reconnect stage. So that MDS + * can process completed request in clientreplay stage. + */ + p = rb_first(&mdsc->request_tree); + while (p) { + req = rb_entry(p, struct ceph_mds_request, r_node); + p = rb_next(p); + if (req->r_got_unsafe) + continue; + if (req->r_attempts == 0) + continue; /* only old requests */ + if (req->r_session && + req->r_session->s_mds == session->s_mds) { + err = __prepare_send_request(mdsc, req, session->s_mds); + if (!err) { + ceph_msg_get(req->r_request); + ceph_con_send(&session->s_con, req->r_request); + } + } + } mutex_unlock(&mdsc->mutex); } @@ -2787,6 +2851,8 @@ static void send_mds_reconnect(struct ceph_mds_client *mdsc, spin_unlock(&session->s_gen_ttl_lock); spin_lock(&session->s_cap_lock); + /* don't know if session is readonly */ + session->s_readonly = 0; /* * notify __ceph_remove_cap() that we are composing cap reconnect. * If a cap get released before being added to the cap reconnect, @@ -2933,9 +2999,6 @@ static void check_new_map(struct ceph_mds_client *mdsc, mutex_unlock(&s->s_mutex); s->s_state = CEPH_MDS_SESSION_RESTARTING; } - - /* kick any requests waiting on the recovering mds */ - kick_requests(mdsc, i); } else if (oldstate == newstate) { continue; /* nothing new with this mds */ } @@ -3295,6 +3358,7 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) init_waitqueue_head(&mdsc->session_close_wq); INIT_LIST_HEAD(&mdsc->waiting_for_map); mdsc->sessions = NULL; + atomic_set(&mdsc->num_sessions, 0); mdsc->max_sessions = 0; mdsc->stopping = 0; init_rwsem(&mdsc->snap_rwsem); @@ -3428,14 +3492,17 @@ void ceph_mdsc_sync(struct ceph_mds_client *mdsc) dout("sync\n"); mutex_lock(&mdsc->mutex); want_tid = mdsc->last_tid; - want_flush = mdsc->cap_flush_seq; mutex_unlock(&mdsc->mutex); - dout("sync want tid %lld flush_seq %lld\n", want_tid, want_flush); ceph_flush_dirty_caps(mdsc); + spin_lock(&mdsc->cap_dirty_lock); + want_flush = mdsc->cap_flush_seq; + spin_unlock(&mdsc->cap_dirty_lock); + + dout("sync want tid %lld flush_seq %lld\n", want_tid, want_flush); wait_unsafe_requests(mdsc, want_tid); - wait_event(mdsc->cap_flushing_wq, check_cap_flush(mdsc, want_flush)); + wait_caps_flush(mdsc, want_flush); } /* @@ -3443,17 +3510,9 @@ void ceph_mdsc_sync(struct ceph_mds_client *mdsc) */ static bool done_closing_sessions(struct ceph_mds_client *mdsc) { - int i, n = 0; - if (mdsc->fsc->mount_state == CEPH_MOUNT_SHUTDOWN) return true; - - mutex_lock(&mdsc->mutex); - for (i = 0; i < mdsc->max_sessions; i++) - if (mdsc->sessions[i]) - n++; - mutex_unlock(&mdsc->mutex); - return n == 0; + return atomic_read(&mdsc->num_sessions) == 0; } /* diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index e2817d0..1875b5d 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -137,6 +137,7 @@ struct ceph_mds_session { int s_nr_caps, s_trim_caps; int s_num_cap_releases; 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; @@ -272,6 +273,7 @@ struct ceph_mds_client { struct list_head waiting_for_map; struct ceph_mds_session **sessions; /* NULL for mds if no session */ + atomic_t num_sessions; int max_sessions; /* len of s_mds_sessions */ int stopping; /* true if shutting down */ diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index ce35fbd..a97e39f 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -70,13 +70,11 @@ void ceph_get_snap_realm(struct ceph_mds_client *mdsc, * safe. we do need to protect against concurrent empty list * additions, however. */ - if (atomic_read(&realm->nref) == 0) { + if (atomic_inc_return(&realm->nref) == 1) { spin_lock(&mdsc->snap_empty_lock); list_del_init(&realm->empty_item); spin_unlock(&mdsc->snap_empty_lock); } - - atomic_inc(&realm->nref); } static void __insert_snap_realm(struct rb_root *root, @@ -116,7 +114,7 @@ static struct ceph_snap_realm *ceph_create_snap_realm( if (!realm) return ERR_PTR(-ENOMEM); - atomic_set(&realm->nref, 0); /* tree does not take a ref */ + atomic_set(&realm->nref, 1); /* for caller */ realm->ino = ino; INIT_LIST_HEAD(&realm->children); INIT_LIST_HEAD(&realm->child_item); @@ -134,8 +132,8 @@ static struct ceph_snap_realm *ceph_create_snap_realm( * * caller must hold snap_rwsem for write. */ -struct ceph_snap_realm *ceph_lookup_snap_realm(struct ceph_mds_client *mdsc, - u64 ino) +static struct ceph_snap_realm *__lookup_snap_realm(struct ceph_mds_client *mdsc, + u64 ino) { struct rb_node *n = mdsc->snap_realms.rb_node; struct ceph_snap_realm *r; @@ -154,6 +152,16 @@ struct ceph_snap_realm *ceph_lookup_snap_realm(struct ceph_mds_client *mdsc, return NULL; } +struct ceph_snap_realm *ceph_lookup_snap_realm(struct ceph_mds_client *mdsc, + u64 ino) +{ + struct ceph_snap_realm *r; + r = __lookup_snap_realm(mdsc, ino); + if (r) + ceph_get_snap_realm(mdsc, r); + return r; +} + static void __put_snap_realm(struct ceph_mds_client *mdsc, struct ceph_snap_realm *realm); @@ -273,7 +281,6 @@ static int adjust_snap_realm_parent(struct ceph_mds_client *mdsc, } realm->parent_ino = parentino; realm->parent = parent; - ceph_get_snap_realm(mdsc, parent); list_add(&realm->child_item, &parent->children); return 1; } @@ -631,12 +638,14 @@ static void queue_realm_cap_snaps(struct ceph_snap_realm *realm) * Caller must hold snap_rwsem for write. */ int ceph_update_snap_trace(struct ceph_mds_client *mdsc, - void *p, void *e, bool deletion) + void *p, void *e, bool deletion, + struct ceph_snap_realm **realm_ret) { struct ceph_mds_snap_realm *ri; /* encoded */ __le64 *snaps; /* encoded */ __le64 *prior_parent_snaps; /* encoded */ - struct ceph_snap_realm *realm; + struct ceph_snap_realm *realm = NULL; + struct ceph_snap_realm *first_realm = NULL; int invalidate = 0; int err = -ENOMEM; LIST_HEAD(dirty_realms); @@ -704,13 +713,18 @@ more: dout("done with %llx %p, invalidated=%d, %p %p\n", realm->ino, realm, invalidate, p, e); - if (p < e) - goto more; - /* invalidate when we reach the _end_ (root) of the trace */ - if (invalidate) + if (invalidate && p >= e) rebuild_snap_realms(realm); + if (!first_realm) + first_realm = realm; + else + ceph_put_snap_realm(mdsc, realm); + + if (p < e) + goto more; + /* * queue cap snaps _after_ we've built the new snap contexts, * so that i_head_snapc can be set appropriately. @@ -721,12 +735,21 @@ more: queue_realm_cap_snaps(realm); } + if (realm_ret) + *realm_ret = first_realm; + else + ceph_put_snap_realm(mdsc, first_realm); + __cleanup_empty_realms(mdsc); return 0; bad: err = -EINVAL; fail: + if (realm && !IS_ERR(realm)) + ceph_put_snap_realm(mdsc, realm); + if (first_realm) + ceph_put_snap_realm(mdsc, first_realm); pr_err("update_snap_trace error %d\n", err); return err; } @@ -844,7 +867,6 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc, if (IS_ERR(realm)) goto out; } - ceph_get_snap_realm(mdsc, realm); dout("splitting snap_realm %llx %p\n", realm->ino, realm); for (i = 0; i < num_split_inos; i++) { @@ -905,7 +927,7 @@ skip_inode: /* we may have taken some of the old realm's children. */ for (i = 0; i < num_split_realms; i++) { struct ceph_snap_realm *child = - ceph_lookup_snap_realm(mdsc, + __lookup_snap_realm(mdsc, le64_to_cpu(split_realms[i])); if (!child) continue; @@ -918,7 +940,7 @@ skip_inode: * snap, we can avoid queueing cap_snaps. */ ceph_update_snap_trace(mdsc, p, e, - op == CEPH_SNAP_OP_DESTROY); + op == CEPH_SNAP_OP_DESTROY, NULL); if (op == CEPH_SNAP_OP_SPLIT) /* we took a reference when we created the realm, above */ diff --git a/fs/ceph/super.c b/fs/ceph/super.c index 5ae6258..a63997b 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -414,6 +414,10 @@ static int ceph_show_options(struct seq_file *m, struct dentry *root) seq_puts(m, ",noshare"); if (opt->flags & CEPH_OPT_NOCRC) seq_puts(m, ",nocrc"); + if (opt->flags & CEPH_OPT_NOMSGAUTH) + seq_puts(m, ",nocephx_require_signatures"); + if ((opt->flags & CEPH_OPT_TCP_NODELAY) == 0) + seq_puts(m, ",notcp_nodelay"); if (opt->name) seq_printf(m, ",name=%s", opt->name); diff --git a/fs/ceph/super.h b/fs/ceph/super.h index e1aa32d..04c8124 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -693,7 +693,8 @@ extern void ceph_get_snap_realm(struct ceph_mds_client *mdsc, extern void ceph_put_snap_realm(struct ceph_mds_client *mdsc, struct ceph_snap_realm *realm); extern int ceph_update_snap_trace(struct ceph_mds_client *m, - void *p, void *e, bool deletion); + void *p, void *e, bool deletion, + struct ceph_snap_realm **realm_ret); extern void ceph_handle_snap(struct ceph_mds_client *mdsc, struct ceph_mds_session *session, struct ceph_msg *msg); @@ -892,7 +893,9 @@ extern void ceph_fill_inline_data(struct inode *inode, struct page *locked_page, 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; extern const struct inode_operations ceph_dir_iops; +extern const struct inode_operations ceph_snapdir_iops; extern const struct dentry_operations ceph_dentry_ops, ceph_snap_dentry_ops, ceph_snapdir_dentry_ops; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 8fe1f7a..a94b3e6 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -1129,7 +1129,7 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile) struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); struct file_lock *flock; struct file_lock_context *flctx = inode->i_flctx; - unsigned int i; + unsigned int count = 0, i; int rc = 0, xid, type; struct list_head locks_to_send, *el; struct lock_to_push *lck, *tmp; @@ -1140,14 +1140,20 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile) if (!flctx) goto out; + spin_lock(&flctx->flc_lock); + list_for_each(el, &flctx->flc_posix) { + count++; + } + spin_unlock(&flctx->flc_lock); + INIT_LIST_HEAD(&locks_to_send); /* - * Allocating flc_posix_cnt locks is enough because no FL_POSIX locks - * can be added to the list while we are holding cinode->lock_sem that + * Allocating count locks is enough because no FL_POSIX locks can be + * added to the list while we are holding cinode->lock_sem that * protects locking operations of this inode. */ - for (i = 0; i < flctx->flc_posix_cnt; i++) { + for (i = 0; i < count; i++) { lck = kmalloc(sizeof(struct lock_to_push), GFP_KERNEL); if (!lck) { rc = -ENOMEM; diff --git a/fs/coda/dir.c b/fs/coda/dir.c index 281ee01..60cb88c 100644 --- a/fs/coda/dir.c +++ b/fs/coda/dir.c @@ -304,7 +304,7 @@ static int coda_rename(struct inode *old_dir, struct dentry *old_dentry, (const char *) old_name, (const char *)new_name); if (!error) { if (new_dentry->d_inode) { - if (S_ISDIR(new_dentry->d_inode->i_mode)) { + if (d_is_dir(new_dentry)) { coda_dir_drop_nlink(old_dir); coda_dir_inc_nlink(new_dir); } diff --git a/fs/configfs/configfs_internal.h b/fs/configfs/configfs_internal.h index a315677..b65d1ef 100644 --- a/fs/configfs/configfs_internal.h +++ b/fs/configfs/configfs_internal.h @@ -69,14 +69,13 @@ extern struct kmem_cache *configfs_dir_cachep; extern int configfs_is_root(struct config_item *item); extern struct inode * configfs_new_inode(umode_t mode, struct configfs_dirent *, struct super_block *); -extern int configfs_create(struct dentry *, umode_t mode, int (*init)(struct inode *)); +extern int configfs_create(struct dentry *, umode_t mode, void (*init)(struct inode *)); extern int configfs_create_file(struct config_item *, const struct configfs_attribute *); extern int configfs_make_dirent(struct configfs_dirent *, struct dentry *, void *, umode_t, int); extern int configfs_dirent_is_ready(struct configfs_dirent *); -extern int configfs_add_file(struct dentry *, const struct configfs_attribute *, int); extern void configfs_hash_and_remove(struct dentry * dir, const char * name); extern const unsigned char * configfs_get_name(struct configfs_dirent *sd); diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index c9c298b..cf0db00 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -240,60 +240,26 @@ int configfs_make_dirent(struct configfs_dirent * parent_sd, return 0; } -static int init_dir(struct inode * inode) +static void init_dir(struct inode * inode) { inode->i_op = &configfs_dir_inode_operations; inode->i_fop = &configfs_dir_operations; /* directory inodes start off with i_nlink == 2 (for "." entry) */ inc_nlink(inode); - return 0; } -static int configfs_init_file(struct inode * inode) +static void configfs_init_file(struct inode * inode) { inode->i_size = PAGE_SIZE; inode->i_fop = &configfs_file_operations; - return 0; } -static int init_symlink(struct inode * inode) +static void init_symlink(struct inode * inode) { inode->i_op = &configfs_symlink_inode_operations; - return 0; -} - -static int create_dir(struct config_item *k, struct dentry *d) -{ - int error; - umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO; - struct dentry *p = d->d_parent; - - BUG_ON(!k); - - error = configfs_dirent_exists(p->d_fsdata, d->d_name.name); - if (!error) - error = configfs_make_dirent(p->d_fsdata, d, k, mode, - CONFIGFS_DIR | CONFIGFS_USET_CREATING); - if (!error) { - configfs_set_dir_dirent_depth(p->d_fsdata, d->d_fsdata); - error = configfs_create(d, mode, init_dir); - if (!error) { - inc_nlink(p->d_inode); - } else { - struct configfs_dirent *sd = d->d_fsdata; - if (sd) { - spin_lock(&configfs_dirent_lock); - list_del_init(&sd->s_sibling); - spin_unlock(&configfs_dirent_lock); - configfs_put(sd); - } - } - } - return error; } - /** * configfs_create_dir - create a directory for an config_item. * @item: config_itemwe're creating directory for. @@ -303,11 +269,37 @@ static int create_dir(struct config_item *k, struct dentry *d) * until it is validated by configfs_dir_set_ready() */ -static int configfs_create_dir(struct config_item * item, struct dentry *dentry) +static int configfs_create_dir(struct config_item *item, struct dentry *dentry) { - int error = create_dir(item, dentry); - if (!error) + int error; + umode_t mode = S_IFDIR| S_IRWXU | S_IRUGO | S_IXUGO; + struct dentry *p = dentry->d_parent; + + BUG_ON(!item); + + error = configfs_dirent_exists(p->d_fsdata, dentry->d_name.name); + if (unlikely(error)) + return error; + + error = configfs_make_dirent(p->d_fsdata, dentry, item, mode, + CONFIGFS_DIR | CONFIGFS_USET_CREATING); + if (unlikely(error)) + return error; + + configfs_set_dir_dirent_depth(p->d_fsdata, dentry->d_fsdata); + error = configfs_create(dentry, mode, init_dir); + if (!error) { + inc_nlink(p->d_inode); item->ci_dentry = dentry; + } else { + struct configfs_dirent *sd = dentry->d_fsdata; + if (sd) { + spin_lock(&configfs_dirent_lock); + list_del_init(&sd->s_sibling); + spin_unlock(&configfs_dirent_lock); + configfs_put(sd); + } + } return error; } diff --git a/fs/configfs/file.c b/fs/configfs/file.c index 1d1c41f..56d2cdc 100644 --- a/fs/configfs/file.c +++ b/fs/configfs/file.c @@ -313,21 +313,6 @@ const struct file_operations configfs_file_operations = { .release = configfs_release, }; - -int configfs_add_file(struct dentry * dir, const struct configfs_attribute * attr, int type) -{ - struct configfs_dirent * parent_sd = dir->d_fsdata; - umode_t mode = (attr->ca_mode & S_IALLUGO) | S_IFREG; - int error = 0; - - mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_NORMAL); - error = configfs_make_dirent(parent_sd, NULL, (void *) attr, mode, type); - mutex_unlock(&dir->d_inode->i_mutex); - - return error; -} - - /** * configfs_create_file - create an attribute file for an item. * @item: item we're creating for. @@ -336,9 +321,16 @@ int configfs_add_file(struct dentry * dir, const struct configfs_attribute * att int configfs_create_file(struct config_item * item, const struct configfs_attribute * attr) { - BUG_ON(!item || !item->ci_dentry || !attr); + struct dentry *dir = item->ci_dentry; + struct configfs_dirent *parent_sd = dir->d_fsdata; + umode_t mode = (attr->ca_mode & S_IALLUGO) | S_IFREG; + int error = 0; - return configfs_add_file(item->ci_dentry, attr, - CONFIGFS_ITEM_ATTR); + mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_NORMAL); + error = configfs_make_dirent(parent_sd, NULL, (void *) attr, mode, + CONFIGFS_ITEM_ATTR); + mutex_unlock(&dir->d_inode->i_mutex); + + return error; } diff --git a/fs/configfs/inode.c b/fs/configfs/inode.c index 65af861..5423a6a 100644 --- a/fs/configfs/inode.c +++ b/fs/configfs/inode.c @@ -176,7 +176,7 @@ static void configfs_set_inode_lock_class(struct configfs_dirent *sd, #endif /* CONFIG_LOCKDEP */ -int configfs_create(struct dentry * dentry, umode_t mode, int (*init)(struct inode *)) +int configfs_create(struct dentry * dentry, umode_t mode, void (*init)(struct inode *)) { int error = 0; struct inode *inode = NULL; @@ -198,13 +198,7 @@ int configfs_create(struct dentry * dentry, umode_t mode, int (*init)(struct ino p_inode->i_mtime = p_inode->i_ctime = CURRENT_TIME; configfs_set_inode_lock_class(sd, inode); - if (init) { - error = init(inode); - if (error) { - iput(inode); - return error; - } - } + init(inode); d_instantiate(dentry, inode); if (S_ISDIR(mode) || S_ISLNK(mode)) dget(dentry); /* pin link and directory dentries in core */ @@ -242,7 +236,7 @@ void configfs_drop_dentry(struct configfs_dirent * sd, struct dentry * parent) if (dentry) { spin_lock(&dentry->d_lock); - if (!(d_unhashed(dentry) && dentry->d_inode)) { + if (!d_unhashed(dentry) && dentry->d_inode) { dget_dlock(dentry); __d_drop(dentry); spin_unlock(&dentry->d_lock); diff --git a/fs/coredump.c b/fs/coredump.c index b5c86ff..f319926 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -572,7 +572,7 @@ void do_coredump(const siginfo_t *siginfo) * * Normally core limits are irrelevant to pipes, since * we're not writing to the file system, but we use - * cprm.limit of 1 here as a speacial value, this is a + * cprm.limit of 1 here as a special value, this is a * consistent way to catch recursive crashes. * We can still crash if the core_pattern binary sets * RLIM_CORE = !1, but it runs as root, and can do diff --git a/fs/dcache.c b/fs/dcache.c index dc400fd..d99736a 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1659,9 +1659,25 @@ void d_set_d_op(struct dentry *dentry, const struct dentry_operations *op) } EXPORT_SYMBOL(d_set_d_op); + +/* + * d_set_fallthru - Mark a dentry as falling through to a lower layer + * @dentry - The dentry to mark + * + * Mark a dentry as falling through to the lower layer (as set with + * d_pin_lower()). This flag may be recorded on the medium. + */ +void d_set_fallthru(struct dentry *dentry) +{ + spin_lock(&dentry->d_lock); + dentry->d_flags |= DCACHE_FALLTHRU; + spin_unlock(&dentry->d_lock); +} +EXPORT_SYMBOL(d_set_fallthru); + static unsigned d_flags_for_inode(struct inode *inode) { - unsigned add_flags = DCACHE_FILE_TYPE; + unsigned add_flags = DCACHE_REGULAR_TYPE; if (!inode) return DCACHE_MISS_TYPE; @@ -1674,13 +1690,21 @@ static unsigned d_flags_for_inode(struct inode *inode) else inode->i_opflags |= IOP_LOOKUP; } - } else if (unlikely(!(inode->i_opflags & IOP_NOFOLLOW))) { - if (unlikely(inode->i_op->follow_link)) + goto type_determined; + } + + if (unlikely(!(inode->i_opflags & IOP_NOFOLLOW))) { + if (unlikely(inode->i_op->follow_link)) { add_flags = DCACHE_SYMLINK_TYPE; - else - inode->i_opflags |= IOP_NOFOLLOW; + goto type_determined; + } + inode->i_opflags |= IOP_NOFOLLOW; } + if (unlikely(!S_ISREG(inode->i_mode))) + add_flags = DCACHE_SPECIAL_TYPE; + +type_determined: if (unlikely(IS_AUTOMOUNT(inode))) add_flags |= DCACHE_NEED_AUTOMOUNT; return add_flags; @@ -1691,7 +1715,8 @@ static void __d_instantiate(struct dentry *dentry, struct inode *inode) unsigned add_flags = d_flags_for_inode(inode); spin_lock(&dentry->d_lock); - __d_set_type(dentry, add_flags); + dentry->d_flags &= ~(DCACHE_ENTRY_TYPE | DCACHE_FALLTHRU); + dentry->d_flags |= add_flags; if (inode) hlist_add_head(&dentry->d_u.d_alias, &inode->i_dentry); dentry->d_inode = inode; @@ -2665,7 +2690,7 @@ static int __d_unalias(struct inode *inode, struct dentry *dentry, struct dentry *alias) { struct mutex *m1 = NULL, *m2 = NULL; - int ret = -EBUSY; + int ret = -ESTALE; /* If alias and dentry share a parent, then no extra locks required */ if (alias->d_parent == dentry->d_parent) diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 45b18a5..96400ab 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -169,10 +169,19 @@ static int debugfs_show_options(struct seq_file *m, struct dentry *root) return 0; } +static void debugfs_evict_inode(struct inode *inode) +{ + truncate_inode_pages_final(&inode->i_data); + clear_inode(inode); + if (S_ISLNK(inode->i_mode)) + kfree(inode->i_private); +} + static const struct super_operations debugfs_super_operations = { .statfs = simple_statfs, .remount_fs = debugfs_remount, .show_options = debugfs_show_options, + .evict_inode = debugfs_evict_inode, }; static struct vfsmount *debugfs_automount(struct path *path) @@ -511,23 +520,14 @@ static int __debugfs_remove(struct dentry *dentry, struct dentry *parent) int ret = 0; if (debugfs_positive(dentry)) { - if (dentry->d_inode) { - dget(dentry); - switch (dentry->d_inode->i_mode & S_IFMT) { - case S_IFDIR: - ret = simple_rmdir(parent->d_inode, dentry); - break; - case S_IFLNK: - kfree(dentry->d_inode->i_private); - /* fall through */ - default: - simple_unlink(parent->d_inode, dentry); - break; - } - if (!ret) - d_delete(dentry); - dput(dentry); - } + dget(dentry); + if (S_ISDIR(dentry->d_inode->i_mode)) + ret = simple_rmdir(parent->d_inode, dentry); + else + simple_unlink(parent->d_inode, dentry); + if (!ret) + d_delete(dentry); + dput(dentry); } return ret; } @@ -690,7 +690,7 @@ struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry, } d_move(old_dentry, dentry); fsnotify_move(old_dir->d_inode, new_dir->d_inode, old_name, - S_ISDIR(old_dentry->d_inode->i_mode), + d_is_dir(old_dentry), NULL, old_dentry); fsnotify_oldname_free(old_name); unlock_rename(new_dir, old_dir); diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index 90d1882..5ba029e 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -124,7 +124,7 @@ ecryptfs_get_key_payload_data(struct key *key) } #define ECRYPTFS_MAX_KEYSET_SIZE 1024 -#define ECRYPTFS_MAX_CIPHER_NAME_SIZE 32 +#define ECRYPTFS_MAX_CIPHER_NAME_SIZE 31 #define ECRYPTFS_MAX_NUM_ENC_KEYS 64 #define ECRYPTFS_MAX_IV_BYTES 16 /* 128 bits */ #define ECRYPTFS_SALT_BYTES 2 @@ -237,7 +237,7 @@ struct ecryptfs_crypt_stat { struct crypto_ablkcipher *tfm; struct crypto_hash *hash_tfm; /* Crypto context for generating * the initialization vectors */ - unsigned char cipher[ECRYPTFS_MAX_CIPHER_NAME_SIZE]; + unsigned char cipher[ECRYPTFS_MAX_CIPHER_NAME_SIZE + 1]; unsigned char key[ECRYPTFS_MAX_KEY_BYTES]; unsigned char root_iv[ECRYPTFS_MAX_IV_BYTES]; struct list_head keysig_list; diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index 273d36e..7967508 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -223,7 +223,7 @@ static int ecryptfs_open(struct inode *inode, struct file *file) } ecryptfs_set_file_lower( file, ecryptfs_inode_to_private(inode)->lower_file); - if (S_ISDIR(ecryptfs_dentry->d_inode->i_mode)) { + if (d_is_dir(ecryptfs_dentry)) { ecryptfs_printk(KERN_DEBUG, "This is a directory\n"); mutex_lock(&crypt_stat->cs_mutex); crypt_stat->flags &= ~(ECRYPTFS_ENCRYPTED); @@ -296,9 +296,22 @@ ecryptfs_unlocked_ioctl(struct file *file, unsigned int cmd, unsigned long arg) struct file *lower_file = ecryptfs_file_to_lower(file); long rc = -ENOTTY; - if (lower_file->f_op->unlocked_ioctl) + if (!lower_file->f_op->unlocked_ioctl) + return rc; + + switch (cmd) { + case FITRIM: + case FS_IOC_GETFLAGS: + case FS_IOC_SETFLAGS: + case FS_IOC_GETVERSION: + case FS_IOC_SETVERSION: rc = lower_file->f_op->unlocked_ioctl(lower_file, cmd, arg); - return rc; + fsstack_copy_attr_all(file_inode(file), file_inode(lower_file)); + + return rc; + default: + return rc; + } } #ifdef CONFIG_COMPAT @@ -308,9 +321,22 @@ ecryptfs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) struct file *lower_file = ecryptfs_file_to_lower(file); long rc = -ENOIOCTLCMD; - if (lower_file->f_op->compat_ioctl) + if (!lower_file->f_op->compat_ioctl) + return rc; + + switch (cmd) { + case FITRIM: + case FS_IOC32_GETFLAGS: + case FS_IOC32_SETFLAGS: + case FS_IOC32_GETVERSION: + case FS_IOC32_SETVERSION: rc = lower_file->f_op->compat_ioctl(lower_file, cmd, arg); - return rc; + fsstack_copy_attr_all(file_inode(file), file_inode(lower_file)); + + return rc; + default: + return rc; + } } #endif diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index 34b36a5..b08b518 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -907,9 +907,9 @@ static int ecryptfs_setattr(struct dentry *dentry, struct iattr *ia) lower_inode = ecryptfs_inode_to_lower(inode); lower_dentry = ecryptfs_dentry_to_lower(dentry); mutex_lock(&crypt_stat->cs_mutex); - if (S_ISDIR(dentry->d_inode->i_mode)) + if (d_is_dir(dentry)) crypt_stat->flags &= ~(ECRYPTFS_ENCRYPTED); - else if (S_ISREG(dentry->d_inode->i_mode) + else if (d_is_reg(dentry) && (!(crypt_stat->flags & ECRYPTFS_POLICY_APPLIED) || !(crypt_stat->flags & ECRYPTFS_KEY_VALID))) { struct ecryptfs_mount_crypt_stat *mount_crypt_stat; diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index 917bd5c..6bd67e2 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -891,7 +891,7 @@ struct ecryptfs_parse_tag_70_packet_silly_stack { struct blkcipher_desc desc; char fnek_sig_hex[ECRYPTFS_SIG_SIZE_HEX + 1]; char iv[ECRYPTFS_MAX_IV_BYTES]; - char cipher_string[ECRYPTFS_MAX_CIPHER_NAME_SIZE]; + char cipher_string[ECRYPTFS_MAX_CIPHER_NAME_SIZE + 1]; }; /** diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index 1895d60..c095d32 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -407,7 +407,7 @@ static int ecryptfs_parse_options(struct ecryptfs_sb_info *sbi, char *options, if (!cipher_name_set) { int cipher_name_len = strlen(ECRYPTFS_DEFAULT_CIPHER); - BUG_ON(cipher_name_len >= ECRYPTFS_MAX_CIPHER_NAME_SIZE); + BUG_ON(cipher_name_len > ECRYPTFS_MAX_CIPHER_NAME_SIZE); strcpy(mount_crypt_stat->global_default_cipher_name, ECRYPTFS_DEFAULT_CIPHER); } diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c index fdfd206..714cd37 100644 --- a/fs/exportfs/expfs.c +++ b/fs/exportfs/expfs.c @@ -429,7 +429,7 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid, if (IS_ERR(result)) return result; - if (S_ISDIR(result->d_inode->i_mode)) { + if (d_is_dir(result)) { /* * This request is for a directory. * diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 982d934..f63c3d5 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -364,7 +364,8 @@ struct flex_groups { #define EXT4_DIRTY_FL 0x00000100 #define EXT4_COMPRBLK_FL 0x00000200 /* One or more compressed clusters */ #define EXT4_NOCOMPR_FL 0x00000400 /* Don't compress */ -#define EXT4_ECOMPR_FL 0x00000800 /* Compression error */ + /* nb: was previously EXT2_ECOMPR_FL */ +#define EXT4_ENCRYPT_FL 0x00000800 /* encrypted file */ /* End compression flags --- maybe not all used */ #define EXT4_INDEX_FL 0x00001000 /* hash-indexed directory */ #define EXT4_IMAGIC_FL 0x00002000 /* AFS directory */ @@ -421,7 +422,7 @@ enum { EXT4_INODE_DIRTY = 8, EXT4_INODE_COMPRBLK = 9, /* One or more compressed clusters */ EXT4_INODE_NOCOMPR = 10, /* Don't compress */ - EXT4_INODE_ECOMPR = 11, /* Compression error */ + EXT4_INODE_ENCRYPT = 11, /* Compression error */ /* End compression flags --- maybe not all used */ EXT4_INODE_INDEX = 12, /* hash-indexed directory */ EXT4_INODE_IMAGIC = 13, /* AFS directory */ @@ -466,7 +467,7 @@ static inline void ext4_check_flag_values(void) CHECK_FLAG_VALUE(DIRTY); CHECK_FLAG_VALUE(COMPRBLK); CHECK_FLAG_VALUE(NOCOMPR); - CHECK_FLAG_VALUE(ECOMPR); + CHECK_FLAG_VALUE(ENCRYPT); CHECK_FLAG_VALUE(INDEX); CHECK_FLAG_VALUE(IMAGIC); CHECK_FLAG_VALUE(JOURNAL_DATA); @@ -1048,6 +1049,12 @@ extern void ext4_set_bits(void *bm, int cur, int len); /* Metadata checksum algorithm codes */ #define EXT4_CRC32C_CHKSUM 1 +/* Encryption algorithms */ +#define EXT4_ENCRYPTION_MODE_INVALID 0 +#define EXT4_ENCRYPTION_MODE_AES_256_XTS 1 +#define EXT4_ENCRYPTION_MODE_AES_256_GCM 2 +#define EXT4_ENCRYPTION_MODE_AES_256_CBC 3 + /* * Structure of the super block */ @@ -1161,7 +1168,8 @@ struct ext4_super_block { __le32 s_grp_quota_inum; /* inode for tracking group quota */ __le32 s_overhead_clusters; /* overhead blocks/clusters in fs */ __le32 s_backup_bgs[2]; /* groups with sparse_super2 SBs */ - __le32 s_reserved[106]; /* Padding to the end of the block */ + __u8 s_encrypt_algos[4]; /* Encryption algorithms in use */ + __le32 s_reserved[105]; /* Padding to the end of the block */ __le32 s_checksum; /* crc32c(superblock) */ }; @@ -1527,6 +1535,7 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei) * GDT_CSUM bits are mutually exclusive. */ #define EXT4_FEATURE_RO_COMPAT_METADATA_CSUM 0x0400 +#define EXT4_FEATURE_RO_COMPAT_READONLY 0x1000 #define EXT4_FEATURE_INCOMPAT_COMPRESSION 0x0001 #define EXT4_FEATURE_INCOMPAT_FILETYPE 0x0002 @@ -1542,6 +1551,7 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei) #define EXT4_FEATURE_INCOMPAT_BG_USE_META_CSUM 0x2000 /* use crc32c for bg */ #define EXT4_FEATURE_INCOMPAT_LARGEDIR 0x4000 /* >2GB or 3-lvl htree */ #define EXT4_FEATURE_INCOMPAT_INLINE_DATA 0x8000 /* data in inode */ +#define EXT4_FEATURE_INCOMPAT_ENCRYPT 0x10000 #define EXT2_FEATURE_COMPAT_SUPP EXT4_FEATURE_COMPAT_EXT_ATTR #define EXT2_FEATURE_INCOMPAT_SUPP (EXT4_FEATURE_INCOMPAT_FILETYPE| \ diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 8611640..740c787 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -1401,10 +1401,7 @@ end_range: * to free. Everything was covered by the start * of the range. */ - return 0; - } else { - /* Shared branch grows from an indirect block */ - partial2--; + goto do_indirects; } } else { /* @@ -1435,56 +1432,96 @@ end_range: /* Punch happened within the same level (n == n2) */ partial = ext4_find_shared(inode, n, offsets, chain, &nr); partial2 = ext4_find_shared(inode, n2, offsets2, chain2, &nr2); - /* - * ext4_find_shared returns Indirect structure which - * points to the last element which should not be - * removed by truncate. But this is end of the range - * in punch_hole so we need to point to the next element - */ - partial2->p++; - while ((partial > chain) || (partial2 > chain2)) { - /* We're at the same block, so we're almost finished */ - if ((partial->bh && partial2->bh) && - (partial->bh->b_blocknr == partial2->bh->b_blocknr)) { - if ((partial > chain) && (partial2 > chain2)) { + + /* Free top, but only if partial2 isn't its subtree. */ + if (nr) { + int level = min(partial - chain, partial2 - chain2); + int i; + int subtree = 1; + + for (i = 0; i <= level; i++) { + if (offsets[i] != offsets2[i]) { + subtree = 0; + break; + } + } + + if (!subtree) { + if (partial == chain) { + /* Shared branch grows from the inode */ + ext4_free_branches(handle, inode, NULL, + &nr, &nr+1, + (chain+n-1) - partial); + *partial->p = 0; + } else { + /* Shared branch grows from an indirect block */ + BUFFER_TRACE(partial->bh, "get_write_access"); ext4_free_branches(handle, inode, partial->bh, - partial->p + 1, - partial2->p, + partial->p, + partial->p+1, (chain+n-1) - partial); - BUFFER_TRACE(partial->bh, "call brelse"); - brelse(partial->bh); - BUFFER_TRACE(partial2->bh, "call brelse"); - brelse(partial2->bh); } - return 0; } + } + + if (!nr2) { /* - * Clear the ends of indirect blocks on the shared branch - * at the start of the range + * ext4_find_shared returns Indirect structure which + * points to the last element which should not be + * removed by truncate. But this is end of the range + * in punch_hole so we need to point to the next element */ - if (partial > chain) { + partial2->p++; + } + + while (partial > chain || partial2 > chain2) { + int depth = (chain+n-1) - partial; + int depth2 = (chain2+n2-1) - partial2; + + if (partial > chain && partial2 > chain2 && + partial->bh->b_blocknr == partial2->bh->b_blocknr) { + /* + * We've converged on the same block. Clear the range, + * then we're done. + */ ext4_free_branches(handle, inode, partial->bh, - partial->p + 1, - (__le32 *)partial->bh->b_data+addr_per_block, - (chain+n-1) - partial); + partial->p + 1, + partial2->p, + (chain+n-1) - partial); BUFFER_TRACE(partial->bh, "call brelse"); brelse(partial->bh); - partial--; + BUFFER_TRACE(partial2->bh, "call brelse"); + brelse(partial2->bh); + return 0; } + /* - * Clear the ends of indirect blocks on the shared branch - * at the end of the range + * The start and end partial branches may not be at the same + * level even though the punch happened within one level. So, we + * give them a chance to arrive at the same level, then walk + * them in step with each other until we converge on the same + * block. */ - if (partial2 > chain2) { + if (partial > chain && depth <= depth2) { + ext4_free_branches(handle, inode, partial->bh, + partial->p + 1, + (__le32 *)partial->bh->b_data+addr_per_block, + (chain+n-1) - partial); + BUFFER_TRACE(partial->bh, "call brelse"); + brelse(partial->bh); + partial--; + } + if (partial2 > chain2 && depth2 <= depth) { ext4_free_branches(handle, inode, partial2->bh, (__le32 *)partial2->bh->b_data, partial2->p, - (chain2+n-1) - partial2); + (chain2+n2-1) - partial2); BUFFER_TRACE(partial2->bh, "call brelse"); brelse(partial2->bh); partial2--; } } + return 0; do_indirects: /* Kill the remaining (whole) subtrees */ diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 6325d2c..a3f4513 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1023,6 +1023,7 @@ static int ext4_write_end(struct file *file, { handle_t *handle = ext4_journal_current_handle(); struct inode *inode = mapping->host; + loff_t old_size = inode->i_size; int ret = 0, ret2; int i_size_changed = 0; @@ -1053,6 +1054,8 @@ static int ext4_write_end(struct file *file, unlock_page(page); page_cache_release(page); + if (old_size < pos) + pagecache_isize_extended(inode, old_size, pos); /* * Don't mark the inode dirty under page lock. First, it unnecessarily * makes the holding time of page lock longer. Second, it forces lock @@ -1094,6 +1097,7 @@ static int ext4_journalled_write_end(struct file *file, { handle_t *handle = ext4_journal_current_handle(); struct inode *inode = mapping->host; + loff_t old_size = inode->i_size; int ret = 0, ret2; int partial = 0; unsigned from, to; @@ -1126,6 +1130,9 @@ static int ext4_journalled_write_end(struct file *file, unlock_page(page); page_cache_release(page); + if (old_size < pos) + pagecache_isize_extended(inode, old_size, pos); + if (size_changed) { ret2 = ext4_mark_inode_dirty(handle, inode); if (!ret) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 1adac68..e061e66 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -2779,6 +2779,12 @@ static int ext4_feature_set_ok(struct super_block *sb, int readonly) if (readonly) return 1; + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_READONLY)) { + ext4_msg(sb, KERN_INFO, "filesystem is read-only"); + sb->s_flags |= MS_RDONLY; + return 1; + } + /* Check that feature set is OK for a read-write mount */ if (EXT4_HAS_RO_COMPAT_FEATURE(sb, ~EXT4_FEATURE_RO_COMPAT_SUPP)) { ext4_msg(sb, KERN_ERR, "couldn't mount RDWR because of " @@ -3936,9 +3942,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) get_random_bytes(&sbi->s_next_generation, sizeof(u32)); spin_lock_init(&sbi->s_next_gen_lock); - init_timer(&sbi->s_err_report); - sbi->s_err_report.function = print_daily_error_info; - sbi->s_err_report.data = (unsigned long) sb; + setup_timer(&sbi->s_err_report, print_daily_error_info, + (unsigned long) sb); /* Register extent status tree shrinker */ if (ext4_es_register_shrinker(sbi)) @@ -4866,9 +4871,6 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) if (sbi->s_journal && sbi->s_journal->j_task->io_context) journal_ioprio = sbi->s_journal->j_task->io_context->ioprio; - /* - * Allow the "check" option to be passed as a remount option. - */ if (!parse_options(data, sb, NULL, &journal_ioprio, 1)) { err = -EINVAL; goto restore_opts; @@ -4877,17 +4879,8 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ test_opt(sb, JOURNAL_CHECKSUM)) { ext4_msg(sb, KERN_ERR, "changing journal_checksum " - "during remount not supported"); - err = -EINVAL; - goto restore_opts; - } - - if ((old_opts.s_mount_opt & EXT4_MOUNT_JOURNAL_CHECKSUM) ^ - test_opt(sb, JOURNAL_CHECKSUM)) { - ext4_msg(sb, KERN_ERR, "changing journal_checksum " - "during remount not supported"); - err = -EINVAL; - goto restore_opts; + "during remount not supported; ignoring"); + sbi->s_mount_opt ^= EXT4_MOUNT_JOURNAL_CHECKSUM; } if (test_opt(sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA) { @@ -4963,7 +4956,9 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) ext4_mark_recovery_complete(sb, es); } else { /* Make sure we can mount this feature set readwrite */ - if (!ext4_feature_set_ok(sb, 0)) { + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, + EXT4_FEATURE_RO_COMPAT_READONLY) || + !ext4_feature_set_ok(sb, 0)) { err = -EROFS; goto restore_opts; } diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 073657f..e907052 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -769,9 +769,9 @@ static long __writeback_inodes_wb(struct bdi_writeback *wb, struct inode *inode = wb_inode(wb->b_io.prev); struct super_block *sb = inode->i_sb; - if (!grab_super_passive(sb)) { + if (!trylock_super(sb)) { /* - * grab_super_passive() may fail consistently due to + * trylock_super() may fail consistently due to * s_umount being grabbed by someone else. Don't use * requeue_io() to avoid busy retrying the inode/sb. */ @@ -779,7 +779,7 @@ static long __writeback_inodes_wb(struct bdi_writeback *wb, continue; } wrote += writeback_sb_inodes(sb, wb, work); - drop_super(sb); + up_read(&sb->s_umount); /* refer to the same tests at the end of writeback_sb_inodes */ if (wrote) { diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 8c92c72..95a2797 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -889,8 +889,8 @@ static int fuse_try_move_page(struct fuse_copy_state *cs, struct page **pagep) newpage = buf->page; - if (WARN_ON(!PageUptodate(newpage))) - return -EIO; + if (!PageUptodate(newpage)) + SetPageUptodate(newpage); ClearPageMappedToDisk(newpage); @@ -1352,6 +1352,17 @@ static ssize_t fuse_dev_do_read(struct fuse_conn *fc, struct file *file, return err; } +static int fuse_dev_open(struct inode *inode, struct file *file) +{ + /* + * The fuse device's file's private_data is used to hold + * the fuse_conn(ection) when it is mounted, and is used to + * keep track of whether the file has been mounted already. + */ + file->private_data = NULL; + return 0; +} + static ssize_t fuse_dev_read(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { @@ -1796,6 +1807,9 @@ copy_finish: static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, unsigned int size, struct fuse_copy_state *cs) { + /* Don't try to move pages (yet) */ + cs->move_pages = 0; + switch (code) { case FUSE_NOTIFY_POLL: return fuse_notify_poll(fc, size, cs); @@ -2216,6 +2230,7 @@ static int fuse_dev_fasync(int fd, struct file *file, int on) const struct file_operations fuse_dev_operations = { .owner = THIS_MODULE, + .open = fuse_dev_open, .llseek = no_llseek, .read = do_sync_read, .aio_read = fuse_dev_read, diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 08e7b1a..1545b71 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -971,7 +971,7 @@ int fuse_reverse_inval_entry(struct super_block *sb, u64 parent_nodeid, err = -EBUSY; goto badentry; } - if (S_ISDIR(entry->d_inode->i_mode)) { + if (d_is_dir(entry)) { shrink_dcache_parent(entry); if (!simple_empty(entry)) { err = -ENOTEMPTY; diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 6371192..487527b 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -1809,7 +1809,7 @@ int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry) gfs2_consist_inode(dip); dip->i_entries--; dip->i_inode.i_mtime = dip->i_inode.i_ctime = tv; - if (S_ISDIR(dentry->d_inode->i_mode)) + if (d_is_dir(dentry)) drop_nlink(&dip->i_inode); mark_inode_dirty(&dip->i_inode); diff --git a/fs/hfsplus/dir.c b/fs/hfsplus/dir.c index 435bea2..f0235c1 100644 --- a/fs/hfsplus/dir.c +++ b/fs/hfsplus/dir.c @@ -530,7 +530,7 @@ static int hfsplus_rename(struct inode *old_dir, struct dentry *old_dentry, /* Unlink destination if it already exists */ if (new_dentry->d_inode) { - if (S_ISDIR(new_dentry->d_inode->i_mode)) + if (d_is_dir(new_dentry)) res = hfsplus_rmdir(new_dir, new_dentry); else res = hfsplus_unlink(new_dir, new_dentry); diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c index 5f27551..043ac9d 100644 --- a/fs/hppfs/hppfs.c +++ b/fs/hppfs/hppfs.c @@ -678,10 +678,10 @@ static struct inode *get_inode(struct super_block *sb, struct dentry *dentry) return NULL; } - if (S_ISDIR(dentry->d_inode->i_mode)) { + if (d_is_dir(dentry)) { inode->i_op = &hppfs_dir_iops; inode->i_fop = &hppfs_dir_fops; - } else if (S_ISLNK(dentry->d_inode->i_mode)) { + } else if (d_is_symlink(dentry)) { inode->i_op = &hppfs_link_iops; inode->i_fop = &hppfs_file_fops; } else { diff --git a/fs/internal.h b/fs/internal.h index 30459da..01dce1d 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -84,7 +84,7 @@ extern struct file *get_empty_filp(void); * super.c */ extern int do_remount_sb(struct super_block *, int, void *, int); -extern bool grab_super_passive(struct super_block *sb); +extern bool trylock_super(struct super_block *sb); extern struct dentry *mount_fs(struct file_system_type *, int, const char *, void *); extern struct super_block *user_get_super(dev_t); diff --git a/fs/jbd2/recovery.c b/fs/jbd2/recovery.c index bcbef08..b5128c6 100644 --- a/fs/jbd2/recovery.c +++ b/fs/jbd2/recovery.c @@ -524,6 +524,9 @@ static int do_one_pass(journal_t *journal, if (descr_csum_size > 0 && !jbd2_descr_block_csum_verify(journal, bh->b_data)) { + printk(KERN_ERR "JBD2: Invalid checksum " + "recovering block %lu in log\n", + next_log_block); err = -EIO; brelse(bh); goto failed; diff --git a/fs/jffs2/compr_rubin.c b/fs/jffs2/compr_rubin.c index 92e0644b..556de10 100644 --- a/fs/jffs2/compr_rubin.c +++ b/fs/jffs2/compr_rubin.c @@ -84,11 +84,6 @@ static inline int pullbit(struct pushpull *pp) return bit; } -static inline int pulledbits(struct pushpull *pp) -{ - return pp->ofs; -} - static void init_rubin(struct rubin_state *rs, int div, int *bits) { diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index 9385560..f21b6fb 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -252,7 +252,7 @@ static int jffs2_link (struct dentry *old_dentry, struct inode *dir_i, struct de if (!f->inocache) return -EIO; - if (S_ISDIR(old_dentry->d_inode->i_mode)) + if (d_is_dir(old_dentry)) return -EPERM; /* XXX: This is ugly */ @@ -772,7 +772,7 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry, */ if (new_dentry->d_inode) { victim_f = JFFS2_INODE_INFO(new_dentry->d_inode); - if (S_ISDIR(new_dentry->d_inode->i_mode)) { + if (d_is_dir(new_dentry)) { struct jffs2_full_dirent *fd; mutex_lock(&victim_f->sem); @@ -807,7 +807,7 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry, if (victim_f) { /* There was a victim. Kill it off nicely */ - if (S_ISDIR(new_dentry->d_inode->i_mode)) + if (d_is_dir(new_dentry)) clear_nlink(new_dentry->d_inode); else drop_nlink(new_dentry->d_inode); @@ -815,7 +815,7 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry, inode which didn't exist. */ if (victim_f->inocache) { mutex_lock(&victim_f->sem); - if (S_ISDIR(new_dentry->d_inode->i_mode)) + if (d_is_dir(new_dentry)) victim_f->inocache->pino_nlink = 0; else victim_f->inocache->pino_nlink--; @@ -825,7 +825,7 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry, /* If it was a directory we moved, and there was no victim, increase i_nlink on its new parent */ - if (S_ISDIR(old_dentry->d_inode->i_mode) && !victim_f) + if (d_is_dir(old_dentry) && !victim_f) inc_nlink(new_dir_i); /* Unlink the original */ @@ -839,7 +839,7 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry, struct jffs2_inode_info *f = JFFS2_INODE_INFO(old_dentry->d_inode); mutex_lock(&f->sem); inc_nlink(old_dentry->d_inode); - if (f->inocache && !S_ISDIR(old_dentry->d_inode->i_mode)) + if (f->inocache && !d_is_dir(old_dentry)) f->inocache->pino_nlink++; mutex_unlock(&f->sem); @@ -852,7 +852,7 @@ static int jffs2_rename (struct inode *old_dir_i, struct dentry *old_dentry, return ret; } - if (S_ISDIR(old_dentry->d_inode->i_mode)) + if (d_is_dir(old_dentry)) drop_nlink(old_dir_i); new_dir_i->i_mtime = new_dir_i->i_ctime = old_dir_i->i_mtime = old_dir_i->i_ctime = ITIME(now); diff --git a/fs/jffs2/scan.c b/fs/jffs2/scan.c index 7654e87..9ad5ba4 100644 --- a/fs/jffs2/scan.c +++ b/fs/jffs2/scan.c @@ -510,6 +510,10 @@ static int jffs2_scan_eraseblock (struct jffs2_sb_info *c, struct jffs2_eraseblo sumlen = c->sector_size - je32_to_cpu(sm->offset); sumptr = buf + buf_size - sumlen; + /* sm->offset maybe wrong but MAGIC maybe right */ + if (sumlen > c->sector_size) + goto full_scan; + /* Now, make sure the summary itself is available */ if (sumlen > buf_size) { /* Need to kmalloc for this. */ @@ -544,6 +548,7 @@ static int jffs2_scan_eraseblock (struct jffs2_sb_info *c, struct jffs2_eraseblo } } +full_scan: buf_ofs = jeb->offset; if (!buf_size) { diff --git a/fs/jffs2/super.c b/fs/jffs2/super.c index 0918f0e..3d76f28 100644 --- a/fs/jffs2/super.c +++ b/fs/jffs2/super.c @@ -138,7 +138,7 @@ static struct dentry *jffs2_get_parent(struct dentry *child) struct jffs2_inode_info *f; uint32_t pino; - BUG_ON(!S_ISDIR(child->d_inode->i_mode)); + BUG_ON(!d_is_dir(child)); f = JFFS2_INODE_INFO(child->d_inode); diff --git a/fs/kernfs/file.c b/fs/kernfs/file.c index b684e8a..2bacb99 100644 --- a/fs/kernfs/file.c +++ b/fs/kernfs/file.c @@ -207,6 +207,7 @@ static ssize_t kernfs_file_direct_read(struct kernfs_open_file *of, goto out_free; } + of->event = atomic_read(&of->kn->attr.open->event); ops = kernfs_ops(of->kn); if (ops->read) len = ops->read(of, buf, len, *ppos); @@ -329,7 +329,7 @@ int simple_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { struct inode *inode = old_dentry->d_inode; - int they_are_dirs = S_ISDIR(old_dentry->d_inode->i_mode); + int they_are_dirs = d_is_dir(old_dentry); if (!simple_empty(new_dentry)) return -ENOTEMPTY; @@ -681,21 +681,18 @@ static void locks_wake_up_blocks(struct file_lock *blocker) } static void -locks_insert_lock_ctx(struct file_lock *fl, int *counter, - struct list_head *before) +locks_insert_lock_ctx(struct file_lock *fl, struct list_head *before) { fl->fl_nspid = get_pid(task_tgid(current)); list_add_tail(&fl->fl_list, before); - ++*counter; locks_insert_global_locks(fl); } static void -locks_unlink_lock_ctx(struct file_lock *fl, int *counter) +locks_unlink_lock_ctx(struct file_lock *fl) { locks_delete_global_locks(fl); list_del_init(&fl->fl_list); - --*counter; if (fl->fl_nspid) { put_pid(fl->fl_nspid); fl->fl_nspid = NULL; @@ -704,10 +701,9 @@ locks_unlink_lock_ctx(struct file_lock *fl, int *counter) } static void -locks_delete_lock_ctx(struct file_lock *fl, int *counter, - struct list_head *dispose) +locks_delete_lock_ctx(struct file_lock *fl, struct list_head *dispose) { - locks_unlink_lock_ctx(fl, counter); + locks_unlink_lock_ctx(fl); if (dispose) list_add(&fl->fl_list, dispose); else @@ -895,7 +891,7 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) if (request->fl_type == fl->fl_type) goto out; found = true; - locks_delete_lock_ctx(fl, &ctx->flc_flock_cnt, &dispose); + locks_delete_lock_ctx(fl, &dispose); break; } @@ -905,16 +901,6 @@ static int flock_lock_file(struct file *filp, struct file_lock *request) goto out; } - /* - * If a higher-priority process was blocked on the old file lock, - * give it the opportunity to lock the file. - */ - if (found) { - spin_unlock(&ctx->flc_lock); - cond_resched(); - spin_lock(&ctx->flc_lock); - } - find_conflict: list_for_each_entry(fl, &ctx->flc_flock, fl_list) { if (!flock_locks_conflict(request, fl)) @@ -929,7 +915,7 @@ find_conflict: if (request->fl_flags & FL_ACCESS) goto out; locks_copy_lock(new_fl, request); - locks_insert_lock_ctx(new_fl, &ctx->flc_flock_cnt, &ctx->flc_flock); + locks_insert_lock_ctx(new_fl, &ctx->flc_flock); new_fl = NULL; error = 0; @@ -1046,8 +1032,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str else request->fl_end = fl->fl_end; if (added) { - locks_delete_lock_ctx(fl, &ctx->flc_posix_cnt, - &dispose); + locks_delete_lock_ctx(fl, &dispose); continue; } request = fl; @@ -1076,8 +1061,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str * one (This may happen several times). */ if (added) { - locks_delete_lock_ctx(fl, - &ctx->flc_posix_cnt, &dispose); + locks_delete_lock_ctx(fl, &dispose); continue; } /* @@ -1093,10 +1077,8 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str locks_copy_lock(new_fl, request); request = new_fl; new_fl = NULL; - locks_insert_lock_ctx(request, - &ctx->flc_posix_cnt, &fl->fl_list); - locks_delete_lock_ctx(fl, - &ctx->flc_posix_cnt, &dispose); + locks_insert_lock_ctx(request, &fl->fl_list); + locks_delete_lock_ctx(fl, &dispose); added = true; } } @@ -1124,8 +1106,8 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str goto out; } locks_copy_lock(new_fl, request); - locks_insert_lock_ctx(new_fl, &ctx->flc_posix_cnt, - &fl->fl_list); + locks_insert_lock_ctx(new_fl, &fl->fl_list); + fl = new_fl; new_fl = NULL; } if (right) { @@ -1136,8 +1118,7 @@ static int __posix_lock_file(struct inode *inode, struct file_lock *request, str left = new_fl2; new_fl2 = NULL; locks_copy_lock(left, right); - locks_insert_lock_ctx(left, &ctx->flc_posix_cnt, - &fl->fl_list); + locks_insert_lock_ctx(left, &fl->fl_list); } right->fl_start = request->fl_end + 1; locks_wake_up_blocks(right); @@ -1321,7 +1302,6 @@ static void lease_clear_pending(struct file_lock *fl, int arg) /* We already had a lease on this file; just change its type */ int lease_modify(struct file_lock *fl, int arg, struct list_head *dispose) { - struct file_lock_context *flctx; int error = assign_type(fl, arg); if (error) @@ -1331,7 +1311,6 @@ int lease_modify(struct file_lock *fl, int arg, struct list_head *dispose) if (arg == F_UNLCK) { struct file *filp = fl->fl_file; - flctx = file_inode(filp)->i_flctx; f_delown(filp); filp->f_owner.signum = 0; fasync_helper(0, fl->fl_file, 0, &fl->fl_fasync); @@ -1339,7 +1318,7 @@ int lease_modify(struct file_lock *fl, int arg, struct list_head *dispose) printk(KERN_ERR "locks_delete_lock: fasync == %p\n", fl->fl_fasync); fl->fl_fasync = NULL; } - locks_delete_lock_ctx(fl, &flctx->flc_lease_cnt, dispose); + locks_delete_lock_ctx(fl, dispose); } return 0; } @@ -1456,8 +1435,7 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type) fl->fl_downgrade_time = break_time; } if (fl->fl_lmops->lm_break(fl)) - locks_delete_lock_ctx(fl, &ctx->flc_lease_cnt, - &dispose); + locks_delete_lock_ctx(fl, &dispose); } if (list_empty(&ctx->flc_lease)) @@ -1687,7 +1665,8 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr } if (my_fl != NULL) { - error = lease->fl_lmops->lm_change(my_fl, arg, &dispose); + lease = my_fl; + error = lease->fl_lmops->lm_change(lease, arg, &dispose); if (error) goto out; goto out_setup; @@ -1697,7 +1676,7 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr if (!leases_enable) goto out; - locks_insert_lock_ctx(lease, &ctx->flc_lease_cnt, &ctx->flc_lease); + locks_insert_lock_ctx(lease, &ctx->flc_lease); /* * The check in break_lease() is lockless. It's possible for another * open to race in after we did the earlier check for a conflicting @@ -1710,7 +1689,7 @@ generic_add_lease(struct file *filp, long arg, struct file_lock **flp, void **pr smp_mb(); error = check_conflicting_open(dentry, arg, lease->fl_flags); if (error) { - locks_unlink_lock_ctx(lease, &ctx->flc_lease_cnt); + locks_unlink_lock_ctx(lease); goto out; } @@ -1749,7 +1728,7 @@ static int generic_delete_lease(struct file *filp, void *owner) break; } } - trace_generic_delete_lease(inode, fl); + trace_generic_delete_lease(inode, victim); if (victim) error = fl->fl_lmops->lm_change(victim, F_UNLCK, &dispose); spin_unlock(&ctx->flc_lock); @@ -2448,7 +2427,8 @@ locks_remove_lease(struct file *filp) spin_lock(&ctx->flc_lock); list_for_each_entry_safe(fl, tmp, &ctx->flc_lease, fl_list) - lease_modify(fl, F_UNLCK, &dispose); + if (filp == fl->fl_file) + lease_modify(fl, F_UNLCK, &dispose); spin_unlock(&ctx->flc_lock); locks_dispose_list(&dispose); } @@ -119,15 +119,14 @@ * PATH_MAX includes the nul terminator --RR. */ -#define EMBEDDED_NAME_MAX (PATH_MAX - sizeof(struct filename)) +#define EMBEDDED_NAME_MAX (PATH_MAX - offsetof(struct filename, iname)) struct filename * getname_flags(const char __user *filename, int flags, int *empty) { - struct filename *result, *err; - int len; - long max; + struct filename *result; char *kname; + int len; result = audit_reusename(filename); if (result) @@ -136,22 +135,18 @@ getname_flags(const char __user *filename, int flags, int *empty) result = __getname(); if (unlikely(!result)) return ERR_PTR(-ENOMEM); - result->refcnt = 1; /* * First, try to embed the struct filename inside the names_cache * allocation */ - kname = (char *)result + sizeof(*result); + kname = (char *)result->iname; result->name = kname; - result->separate = false; - max = EMBEDDED_NAME_MAX; -recopy: - len = strncpy_from_user(kname, filename, max); + len = strncpy_from_user(kname, filename, EMBEDDED_NAME_MAX); if (unlikely(len < 0)) { - err = ERR_PTR(len); - goto error; + __putname(result); + return ERR_PTR(len); } /* @@ -160,43 +155,49 @@ recopy: * names_cache allocation for the pathname, and re-do the copy from * userland. */ - if (len == EMBEDDED_NAME_MAX && max == EMBEDDED_NAME_MAX) { + if (unlikely(len == EMBEDDED_NAME_MAX)) { + const size_t size = offsetof(struct filename, iname[1]); kname = (char *)result; - result = kzalloc(sizeof(*result), GFP_KERNEL); - if (!result) { - err = ERR_PTR(-ENOMEM); - result = (struct filename *)kname; - goto error; + /* + * size is chosen that way we to guarantee that + * result->iname[0] is within the same object and that + * kname can't be equal to result->iname, no matter what. + */ + result = kzalloc(size, GFP_KERNEL); + if (unlikely(!result)) { + __putname(kname); + return ERR_PTR(-ENOMEM); } result->name = kname; - result->separate = true; - result->refcnt = 1; - max = PATH_MAX; - goto recopy; + len = strncpy_from_user(kname, filename, PATH_MAX); + if (unlikely(len < 0)) { + __putname(kname); + kfree(result); + return ERR_PTR(len); + } + if (unlikely(len == PATH_MAX)) { + __putname(kname); + kfree(result); + return ERR_PTR(-ENAMETOOLONG); + } } + result->refcnt = 1; /* The empty path is special. */ if (unlikely(!len)) { if (empty) *empty = 1; - err = ERR_PTR(-ENOENT); - if (!(flags & LOOKUP_EMPTY)) - goto error; + if (!(flags & LOOKUP_EMPTY)) { + putname(result); + return ERR_PTR(-ENOENT); + } } - err = ERR_PTR(-ENAMETOOLONG); - if (unlikely(len >= PATH_MAX)) - goto error; - result->uptr = filename; result->aname = NULL; audit_getname(result); return result; - -error: - putname(result); - return err; } struct filename * @@ -216,8 +217,7 @@ getname_kernel(const char * filename) return ERR_PTR(-ENOMEM); if (len <= EMBEDDED_NAME_MAX) { - result->name = (char *)(result) + sizeof(*result); - result->separate = false; + result->name = (char *)result->iname; } else if (len <= PATH_MAX) { struct filename *tmp; @@ -227,7 +227,6 @@ getname_kernel(const char * filename) return ERR_PTR(-ENOMEM); } tmp->name = (char *)result; - tmp->separate = true; result = tmp; } else { __putname(result); @@ -249,7 +248,7 @@ void putname(struct filename *name) if (--name->refcnt > 0) return; - if (name->separate) { + if (name->name != name->iname) { __putname(name->name); kfree(name); } else @@ -1851,10 +1850,11 @@ static int link_path_walk(const char *name, struct nameidata *nd) return err; } -static int path_init(int dfd, const char *name, unsigned int flags, +static int path_init(int dfd, const struct filename *name, unsigned int flags, struct nameidata *nd) { int retval = 0; + const char *s = name->name; nd->last_type = LAST_ROOT; /* if there are only slashes... */ nd->flags = flags | LOOKUP_JUMPED | LOOKUP_PARENT; @@ -1863,7 +1863,7 @@ static int path_init(int dfd, const char *name, unsigned int flags, if (flags & LOOKUP_ROOT) { struct dentry *root = nd->root.dentry; struct inode *inode = root->d_inode; - if (*name) { + if (*s) { if (!d_can_lookup(root)) return -ENOTDIR; retval = inode_permission(inode, MAY_EXEC); @@ -1885,7 +1885,7 @@ static int path_init(int dfd, const char *name, unsigned int flags, nd->root.mnt = NULL; nd->m_seq = read_seqbegin(&mount_lock); - if (*name=='/') { + if (*s == '/') { if (flags & LOOKUP_RCU) { rcu_read_lock(); nd->seq = set_root_rcu(nd); @@ -1919,7 +1919,7 @@ static int path_init(int dfd, const char *name, unsigned int flags, dentry = f.file->f_path.dentry; - if (*name) { + if (*s) { if (!d_can_lookup(dentry)) { fdput(f); return -ENOTDIR; @@ -1949,7 +1949,7 @@ static int path_init(int dfd, const char *name, unsigned int flags, return -ECHILD; done: current->total_link_count = 0; - return link_path_walk(name, nd); + return link_path_walk(s, nd); } static void path_cleanup(struct nameidata *nd) @@ -1972,7 +1972,7 @@ static inline int lookup_last(struct nameidata *nd, struct path *path) } /* Returns 0 and nd will be valid on success; Retuns error, otherwise. */ -static int path_lookupat(int dfd, const char *name, +static int path_lookupat(int dfd, const struct filename *name, unsigned int flags, struct nameidata *nd) { struct path path; @@ -2027,31 +2027,17 @@ static int path_lookupat(int dfd, const char *name, static int filename_lookup(int dfd, struct filename *name, unsigned int flags, struct nameidata *nd) { - int retval = path_lookupat(dfd, name->name, flags | LOOKUP_RCU, nd); + int retval = path_lookupat(dfd, name, flags | LOOKUP_RCU, nd); if (unlikely(retval == -ECHILD)) - retval = path_lookupat(dfd, name->name, flags, nd); + retval = path_lookupat(dfd, name, flags, nd); if (unlikely(retval == -ESTALE)) - retval = path_lookupat(dfd, name->name, - flags | LOOKUP_REVAL, nd); + retval = path_lookupat(dfd, name, flags | LOOKUP_REVAL, nd); if (likely(!retval)) audit_inode(name, nd->path.dentry, flags & LOOKUP_PARENT); return retval; } -static int do_path_lookup(int dfd, const char *name, - unsigned int flags, struct nameidata *nd) -{ - struct filename *filename = getname_kernel(name); - int retval = PTR_ERR(filename); - - if (!IS_ERR(filename)) { - retval = filename_lookup(dfd, filename, flags, nd); - putname(filename); - } - return retval; -} - /* does lookup, returns the object with parent locked */ struct dentry *kern_path_locked(const char *name, struct path *path) { @@ -2089,9 +2075,15 @@ out: int kern_path(const char *name, unsigned int flags, struct path *path) { struct nameidata nd; - int res = do_path_lookup(AT_FDCWD, name, flags, &nd); - if (!res) - *path = nd.path; + struct filename *filename = getname_kernel(name); + int res = PTR_ERR(filename); + + if (!IS_ERR(filename)) { + res = filename_lookup(AT_FDCWD, filename, flags, &nd); + putname(filename); + if (!res) + *path = nd.path; + } return res; } EXPORT_SYMBOL(kern_path); @@ -2108,15 +2100,22 @@ int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt, const char *name, unsigned int flags, struct path *path) { - struct nameidata nd; - int err; - nd.root.dentry = dentry; - nd.root.mnt = mnt; + struct filename *filename = getname_kernel(name); + int err = PTR_ERR(filename); + BUG_ON(flags & LOOKUP_PARENT); - /* the first argument of do_path_lookup() is ignored with LOOKUP_ROOT */ - err = do_path_lookup(AT_FDCWD, name, flags | LOOKUP_ROOT, &nd); - if (!err) - *path = nd.path; + + /* the first argument of filename_lookup() is ignored with LOOKUP_ROOT */ + if (!IS_ERR(filename)) { + struct nameidata nd; + nd.root.dentry = dentry; + nd.root.mnt = mnt; + err = filename_lookup(AT_FDCWD, filename, + flags | LOOKUP_ROOT, &nd); + if (!err) + *path = nd.path; + putname(filename); + } return err; } EXPORT_SYMBOL(vfs_path_lookup); @@ -2138,9 +2137,7 @@ static struct dentry *lookup_hash(struct nameidata *nd) * @len: maximum length @len should be interpreted to * * Note that this routine is purely a helper for filesystem usage and should - * not be called by generic code. Also note that by using this function the - * nameidata argument is passed to the filesystem methods and a filesystem - * using this helper needs to be prepared for that. + * not be called by generic code. */ struct dentry *lookup_one_len(const char *name, struct dentry *base, int len) { @@ -2341,7 +2338,8 @@ out: * Returns 0 and "path" will be valid on success; Returns error otherwise. */ static int -path_mountpoint(int dfd, const char *name, struct path *path, unsigned int flags) +path_mountpoint(int dfd, const struct filename *name, struct path *path, + unsigned int flags) { struct nameidata nd; int err; @@ -2370,20 +2368,20 @@ out: } static int -filename_mountpoint(int dfd, struct filename *s, struct path *path, +filename_mountpoint(int dfd, struct filename *name, struct path *path, unsigned int flags) { int error; - if (IS_ERR(s)) - return PTR_ERR(s); - error = path_mountpoint(dfd, s->name, path, flags | LOOKUP_RCU); + if (IS_ERR(name)) + return PTR_ERR(name); + error = path_mountpoint(dfd, name, path, flags | LOOKUP_RCU); if (unlikely(error == -ECHILD)) - error = path_mountpoint(dfd, s->name, path, flags); + error = path_mountpoint(dfd, name, path, flags); if (unlikely(error == -ESTALE)) - error = path_mountpoint(dfd, s->name, path, flags | LOOKUP_REVAL); + error = path_mountpoint(dfd, name, path, flags | LOOKUP_REVAL); if (likely(!error)) - audit_inode(s, path->dentry, 0); - putname(s); + audit_inode(name, path->dentry, 0); + putname(name); return error; } @@ -2814,7 +2812,7 @@ no_open: } else if (!dentry->d_inode) { goto out; } else if ((open_flag & O_TRUNC) && - S_ISREG(dentry->d_inode->i_mode)) { + d_is_reg(dentry)) { goto out; } /* will fail later, go on to get the right error */ @@ -3156,7 +3154,7 @@ static int do_tmpfile(int dfd, struct filename *pathname, static const struct qstr name = QSTR_INIT("/", 1); struct dentry *dentry, *child; struct inode *dir; - int error = path_lookupat(dfd, pathname->name, + int error = path_lookupat(dfd, pathname, flags | LOOKUP_DIRECTORY, nd); if (unlikely(error)) return error; @@ -3229,7 +3227,7 @@ static struct file *path_openat(int dfd, struct filename *pathname, goto out; } - error = path_init(dfd, pathname->name, flags, nd); + error = path_init(dfd, pathname, flags, nd); if (unlikely(error)) goto out; diff --git a/fs/namespace.c b/fs/namespace.c index 72a286e..82ef140 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1907,8 +1907,8 @@ static int graft_tree(struct mount *mnt, struct mount *p, struct mountpoint *mp) if (mnt->mnt.mnt_sb->s_flags & MS_NOUSER) return -EINVAL; - if (S_ISDIR(mp->m_dentry->d_inode->i_mode) != - S_ISDIR(mnt->mnt.mnt_root->d_inode->i_mode)) + if (d_is_dir(mp->m_dentry) != + d_is_dir(mnt->mnt.mnt_root)) return -ENOTDIR; return attach_recursive_mnt(mnt, p, mp, NULL); @@ -2180,8 +2180,8 @@ static int do_move_mount(struct path *path, const char *old_name) if (!mnt_has_parent(old)) goto out1; - if (S_ISDIR(path->dentry->d_inode->i_mode) != - S_ISDIR(old_path.dentry->d_inode->i_mode)) + if (d_is_dir(path->dentry) != + d_is_dir(old_path.dentry)) goto out1; /* * Don't move a mount residing in a shared parent. @@ -2271,7 +2271,7 @@ static int do_add_mount(struct mount *newmnt, struct path *path, int mnt_flags) goto unlock; err = -EINVAL; - if (S_ISLNK(newmnt->mnt.mnt_root->d_inode->i_mode)) + if (d_is_symlink(newmnt->mnt.mnt_root)) goto unlock; newmnt->mnt.mnt_flags = mnt_flags; diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index e36a9d7..197806f 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -427,6 +427,8 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args, if (clp == NULL) goto out; + if (!(clp->cl_session->flags & SESSION4_BACK_CHAN)) + goto out; tbl = &clp->cl_session->bc_slot_table; spin_lock(&tbl->slot_tbl_lock); diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index f4ccfe6..19ca95c 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -313,7 +313,7 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp, goto out; } - args->devs = kmalloc(n * sizeof(*args->devs), GFP_KERNEL); + args->devs = kmalloc_array(n, sizeof(*args->devs), GFP_KERNEL); if (!args->devs) { status = htonl(NFS4ERR_DELAY); goto out; @@ -415,7 +415,7 @@ static __be32 decode_rc_list(struct xdr_stream *xdr, rc_list->rcl_nrefcalls * 2 * sizeof(uint32_t)); if (unlikely(p == NULL)) goto out; - rc_list->rcl_refcalls = kmalloc(rc_list->rcl_nrefcalls * + rc_list->rcl_refcalls = kmalloc_array(rc_list->rcl_nrefcalls, sizeof(*rc_list->rcl_refcalls), GFP_KERNEL); if (unlikely(rc_list->rcl_refcalls == NULL)) @@ -464,8 +464,10 @@ static __be32 decode_cb_sequence_args(struct svc_rqst *rqstp, for (i = 0; i < args->csa_nrclists; i++) { status = decode_rc_list(xdr, &args->csa_rclists[i]); - if (status) + if (status) { + args->csa_nrclists = i; goto out_free; + } } } status = 0; diff --git a/fs/nfs/client.c b/fs/nfs/client.c index f9f4845..1987415 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -433,7 +433,7 @@ static struct nfs_client *nfs_match_client(const struct nfs_client_initdata *dat static bool nfs_client_init_is_complete(const struct nfs_client *clp) { - return clp->cl_cons_state != NFS_CS_INITING; + return clp->cl_cons_state <= NFS_CS_READY; } int nfs_wait_client_init_complete(const struct nfs_client *clp) diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index da54332..a6ad688 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -180,10 +180,9 @@ void nfs_inode_reclaim_delegation(struct inode *inode, struct rpc_cred *cred, delegation->cred = get_rpccred(cred); clear_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags); - NFS_I(inode)->delegation_state = delegation->type; spin_unlock(&delegation->lock); - put_rpccred(oldcred); rcu_read_unlock(); + put_rpccred(oldcred); trace_nfs4_reclaim_delegation(inode, res->delegation_type); } else { /* We appear to have raced with a delegation return. */ @@ -275,7 +274,6 @@ nfs_detach_delegation_locked(struct nfs_inode *nfsi, set_bit(NFS_DELEGATION_RETURNING, &delegation->flags); list_del_rcu(&delegation->super_list); delegation->inode = NULL; - nfsi->delegation_state = 0; rcu_assign_pointer(nfsi->delegation, NULL); spin_unlock(&delegation->lock); return delegation; @@ -355,7 +353,6 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct &delegation->stateid)) { nfs_update_inplace_delegation(old_delegation, delegation); - nfsi->delegation_state = old_delegation->type; goto out; } /* @@ -373,13 +370,15 @@ int nfs_inode_set_delegation(struct inode *inode, struct rpc_cred *cred, struct delegation = NULL; goto out; } - freeme = nfs_detach_delegation_locked(nfsi, + if (test_and_set_bit(NFS_DELEGATION_RETURNING, + &old_delegation->flags)) + goto out; + freeme = nfs_detach_delegation_locked(nfsi, old_delegation, clp); if (freeme == NULL) goto out; } list_add_rcu(&delegation->super_list, &server->delegations); - nfsi->delegation_state = delegation->type; rcu_assign_pointer(nfsi->delegation, delegation); delegation = NULL; @@ -437,6 +436,8 @@ static bool nfs_delegation_need_return(struct nfs_delegation *delegation) { bool ret = false; + if (test_bit(NFS_DELEGATION_RETURNING, &delegation->flags)) + goto out; if (test_and_clear_bit(NFS_DELEGATION_RETURN, &delegation->flags)) ret = true; if (test_and_clear_bit(NFS_DELEGATION_RETURN_IF_CLOSED, &delegation->flags) && !ret) { @@ -448,6 +449,7 @@ static bool nfs_delegation_need_return(struct nfs_delegation *delegation) ret = true; spin_unlock(&delegation->lock); } +out: return ret; } @@ -475,14 +477,20 @@ restart: super_list) { if (!nfs_delegation_need_return(delegation)) continue; - inode = nfs_delegation_grab_inode(delegation); - if (inode == NULL) + if (!nfs_sb_active(server->super)) continue; + inode = nfs_delegation_grab_inode(delegation); + if (inode == NULL) { + rcu_read_unlock(); + nfs_sb_deactive(server->super); + goto restart; + } delegation = nfs_start_delegation_return_locked(NFS_I(inode)); rcu_read_unlock(); err = nfs_end_delegation_return(inode, delegation, 0); iput(inode); + nfs_sb_deactive(server->super); if (!err) goto restart; set_bit(NFS4CLNT_DELEGRETURN, &clp->cl_state); @@ -813,19 +821,30 @@ restart: list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) { list_for_each_entry_rcu(delegation, &server->delegations, super_list) { + if (test_bit(NFS_DELEGATION_RETURNING, + &delegation->flags)) + continue; if (test_bit(NFS_DELEGATION_NEED_RECLAIM, &delegation->flags) == 0) continue; - inode = nfs_delegation_grab_inode(delegation); - if (inode == NULL) + if (!nfs_sb_active(server->super)) continue; - delegation = nfs_detach_delegation(NFS_I(inode), - delegation, server); + inode = nfs_delegation_grab_inode(delegation); + if (inode == NULL) { + rcu_read_unlock(); + nfs_sb_deactive(server->super); + goto restart; + } + delegation = nfs_start_delegation_return_locked(NFS_I(inode)); rcu_read_unlock(); - - if (delegation != NULL) - nfs_free_delegation(delegation); + if (delegation != NULL) { + delegation = nfs_detach_delegation(NFS_I(inode), + delegation, server); + if (delegation != NULL) + nfs_free_delegation(delegation); + } iput(inode); + nfs_sb_deactive(server->super); goto restart; } } diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 9b0c55c..c19e16f 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -408,14 +408,22 @@ static int xdr_decode(nfs_readdir_descriptor_t *desc, return 0; } +/* Match file and dirent using either filehandle or fileid + * Note: caller is responsible for checking the fsid + */ static int nfs_same_file(struct dentry *dentry, struct nfs_entry *entry) { + struct nfs_inode *nfsi; + if (dentry->d_inode == NULL) goto different; - if (nfs_compare_fh(entry->fh, NFS_FH(dentry->d_inode)) != 0) - goto different; - return 1; + + nfsi = NFS_I(dentry->d_inode); + if (entry->fattr->fileid == nfsi->fileid) + return 1; + if (nfs_compare_fh(entry->fh, &nfsi->fh) == 0) + return 1; different: return 0; } @@ -469,6 +477,10 @@ void nfs_prime_dcache(struct dentry *parent, struct nfs_entry *entry) struct inode *inode; int status; + if (!(entry->fattr->valid & NFS_ATTR_FATTR_FILEID)) + return; + if (!(entry->fattr->valid & NFS_ATTR_FATTR_FSID)) + return; if (filename.name[0] == '.') { if (filename.len == 1) return; @@ -479,6 +491,10 @@ void nfs_prime_dcache(struct dentry *parent, struct nfs_entry *entry) dentry = d_lookup(parent, &filename); if (dentry != NULL) { + /* Is there a mountpoint here? If so, just exit */ + if (!nfs_fsid_equal(&NFS_SB(dentry->d_sb)->fsid, + &entry->fattr->fsid)) + goto out; if (nfs_same_file(dentry, entry)) { nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); status = nfs_refresh_inode(dentry->d_inode, entry->fattr); diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index 5db3385..c3929fb 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -283,7 +283,7 @@ static void nfs_direct_release_pages(struct page **pages, unsigned int npages) void nfs_init_cinfo_from_dreq(struct nfs_commit_info *cinfo, struct nfs_direct_req *dreq) { - cinfo->lock = &dreq->lock; + cinfo->lock = &dreq->inode->i_lock; cinfo->mds = &dreq->mds_cinfo; cinfo->ds = &dreq->ds_cinfo; cinfo->dreq = dreq; diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 5d8b89c..37b1558 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -177,7 +177,7 @@ nfs_file_read(struct kiocb *iocb, struct iov_iter *to) iocb->ki_filp, iov_iter_count(to), (unsigned long) iocb->ki_pos); - result = nfs_revalidate_mapping(inode, iocb->ki_filp->f_mapping); + result = nfs_revalidate_mapping_protected(inode, iocb->ki_filp->f_mapping); if (!result) { result = generic_file_read_iter(iocb, to); if (result > 0) @@ -198,7 +198,7 @@ nfs_file_splice_read(struct file *filp, loff_t *ppos, dprintk("NFS: splice_read(%pD2, %lu@%Lu)\n", filp, (unsigned long) count, (unsigned long long) *ppos); - res = nfs_revalidate_mapping(inode, filp->f_mapping); + res = nfs_revalidate_mapping_protected(inode, filp->f_mapping); if (!res) { res = generic_file_splice_read(filp, ppos, pipe, count, flags); if (res > 0) @@ -371,6 +371,10 @@ start: nfs_wait_bit_killable, TASK_KILLABLE); if (ret) return ret; + /* + * Wait for O_DIRECT to complete + */ + nfs_inode_dio_wait(mapping->host); page = grab_cache_page_write_begin(mapping, index, flags); if (!page) @@ -618,6 +622,9 @@ static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) /* make sure the cache has finished storing the page */ nfs_fscache_wait_on_page_write(NFS_I(inode), page); + wait_on_bit_action(&NFS_I(inode)->flags, NFS_INO_INVALIDATING, + nfs_wait_bit_killable, TASK_KILLABLE); + lock_page(page); mapping = page_file_mapping(page); if (mapping != inode->i_mapping) diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c index 7ae1c26..91e88a7 100644 --- a/fs/nfs/filelayout/filelayout.c +++ b/fs/nfs/filelayout/filelayout.c @@ -960,52 +960,19 @@ filelayout_mark_request_commit(struct nfs_page *req, { struct nfs4_filelayout_segment *fl = FILELAYOUT_LSEG(lseg); u32 i, j; - struct list_head *list; - struct pnfs_commit_bucket *buckets; if (fl->commit_through_mds) { - list = &cinfo->mds->list; - spin_lock(cinfo->lock); - goto mds_commit; - } - - /* Note that we are calling nfs4_fl_calc_j_index on each page - * that ends up being committed to a data server. An attractive - * alternative is to add a field to nfs_write_data and nfs_page - * to store the value calculated in filelayout_write_pagelist - * and just use that here. - */ - j = nfs4_fl_calc_j_index(lseg, req_offset(req)); - i = select_bucket_index(fl, j); - spin_lock(cinfo->lock); - buckets = cinfo->ds->buckets; - list = &buckets[i].written; - if (list_empty(list)) { - /* Non-empty buckets hold a reference on the lseg. That ref - * is normally transferred to the COMMIT call and released - * there. It could also be released if the last req is pulled - * off due to a rewrite, in which case it will be done in - * pnfs_generic_clear_request_commit + nfs_request_add_commit_list(req, &cinfo->mds->list, cinfo); + } else { + /* Note that we are calling nfs4_fl_calc_j_index on each page + * that ends up being committed to a data server. An attractive + * alternative is to add a field to nfs_write_data and nfs_page + * to store the value calculated in filelayout_write_pagelist + * and just use that here. */ - buckets[i].wlseg = pnfs_get_lseg(lseg); - } - set_bit(PG_COMMIT_TO_DS, &req->wb_flags); - cinfo->ds->nwritten++; - -mds_commit: - /* nfs_request_add_commit_list(). We need to add req to list without - * dropping cinfo lock. - */ - set_bit(PG_CLEAN, &(req)->wb_flags); - nfs_list_add_request(req, list); - cinfo->mds->ncommit++; - spin_unlock(cinfo->lock); - if (!cinfo->dreq) { - inc_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); - inc_bdi_stat(inode_to_bdi(page_file_mapping(req->wb_page)->host), - BDI_RECLAIMABLE); - __mark_inode_dirty(req->wb_context->dentry->d_inode, - I_DIRTY_DATASYNC); + j = nfs4_fl_calc_j_index(lseg, req_offset(req)); + i = select_bucket_index(fl, j); + pnfs_layout_mark_request_commit(req, lseg, cinfo, i); } } diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c index c22ecaa..315cc68 100644 --- a/fs/nfs/flexfilelayout/flexfilelayout.c +++ b/fs/nfs/flexfilelayout/flexfilelayout.c @@ -1332,47 +1332,6 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync) return PNFS_ATTEMPTED; } -static void -ff_layout_mark_request_commit(struct nfs_page *req, - struct pnfs_layout_segment *lseg, - struct nfs_commit_info *cinfo, - u32 ds_commit_idx) -{ - struct list_head *list; - struct pnfs_commit_bucket *buckets; - - spin_lock(cinfo->lock); - buckets = cinfo->ds->buckets; - list = &buckets[ds_commit_idx].written; - if (list_empty(list)) { - /* Non-empty buckets hold a reference on the lseg. That ref - * is normally transferred to the COMMIT call and released - * there. It could also be released if the last req is pulled - * off due to a rewrite, in which case it will be done in - * pnfs_common_clear_request_commit - */ - WARN_ON_ONCE(buckets[ds_commit_idx].wlseg != NULL); - buckets[ds_commit_idx].wlseg = pnfs_get_lseg(lseg); - } - set_bit(PG_COMMIT_TO_DS, &req->wb_flags); - cinfo->ds->nwritten++; - - /* nfs_request_add_commit_list(). We need to add req to list without - * dropping cinfo lock. - */ - set_bit(PG_CLEAN, &(req)->wb_flags); - nfs_list_add_request(req, list); - cinfo->mds->ncommit++; - spin_unlock(cinfo->lock); - if (!cinfo->dreq) { - inc_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); - inc_bdi_stat(inode_to_bdi(page_file_mapping(req->wb_page)->host), - BDI_RECLAIMABLE); - __mark_inode_dirty(req->wb_context->dentry->d_inode, - I_DIRTY_DATASYNC); - } -} - static u32 calc_ds_index_from_commit(struct pnfs_layout_segment *lseg, u32 i) { return i; @@ -1540,7 +1499,7 @@ static struct pnfs_layoutdriver_type flexfilelayout_type = { .pg_write_ops = &ff_layout_pg_write_ops, .get_ds_info = ff_layout_get_ds_info, .free_deviceid_node = ff_layout_free_deveiceid_node, - .mark_request_commit = ff_layout_mark_request_commit, + .mark_request_commit = pnfs_layout_mark_request_commit, .clear_request_commit = pnfs_generic_clear_request_commit, .scan_commit_lists = pnfs_generic_scan_commit_lists, .recover_commit_reqs = pnfs_generic_recover_commit_reqs, diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index e4f0dce..d42dff6 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -556,6 +556,7 @@ EXPORT_SYMBOL_GPL(nfs_setattr); * This is a copy of the common vmtruncate, but with the locking * corrected to take into account the fact that NFS requires * inode->i_size to be updated under the inode->i_lock. + * Note: must be called with inode->i_lock held! */ static int nfs_vmtruncate(struct inode * inode, loff_t offset) { @@ -565,14 +566,14 @@ static int nfs_vmtruncate(struct inode * inode, loff_t offset) if (err) goto out; - spin_lock(&inode->i_lock); i_size_write(inode, offset); /* Optimisation */ if (offset == 0) NFS_I(inode)->cache_validity &= ~NFS_INO_INVALID_DATA; - spin_unlock(&inode->i_lock); + spin_unlock(&inode->i_lock); truncate_pagecache(inode, offset); + spin_lock(&inode->i_lock); out: return err; } @@ -585,10 +586,15 @@ out: * Note: we do this in the *proc.c in order to ensure that * it works for things like exclusive creates too. */ -void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr) +void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr, + struct nfs_fattr *fattr) { + /* Barrier: bump the attribute generation count. */ + nfs_fattr_set_barrier(fattr); + + spin_lock(&inode->i_lock); + NFS_I(inode)->attr_gencount = fattr->gencount; if ((attr->ia_valid & (ATTR_MODE|ATTR_UID|ATTR_GID)) != 0) { - spin_lock(&inode->i_lock); if ((attr->ia_valid & ATTR_MODE) != 0) { int mode = attr->ia_mode & S_IALLUGO; mode |= inode->i_mode & ~S_IALLUGO; @@ -600,12 +606,13 @@ void nfs_setattr_update_inode(struct inode *inode, struct iattr *attr) inode->i_gid = attr->ia_gid; nfs_set_cache_invalid(inode, NFS_INO_INVALID_ACCESS | NFS_INO_INVALID_ACL); - spin_unlock(&inode->i_lock); } if ((attr->ia_valid & ATTR_SIZE) != 0) { nfs_inc_stats(inode, NFSIOS_SETATTRTRUNC); nfs_vmtruncate(inode, attr->ia_size); } + nfs_update_inode(inode, fattr); + spin_unlock(&inode->i_lock); } EXPORT_SYMBOL_GPL(nfs_setattr_update_inode); @@ -1028,6 +1035,7 @@ static int nfs_invalidate_mapping(struct inode *inode, struct address_space *map if (mapping->nrpages != 0) { if (S_ISREG(inode->i_mode)) { + unmap_mapping_range(mapping, 0, 0, 0); ret = nfs_sync_mapping(mapping); if (ret < 0) return ret; @@ -1060,11 +1068,14 @@ static bool nfs_mapping_need_revalidate_inode(struct inode *inode) } /** - * nfs_revalidate_mapping - Revalidate the pagecache + * __nfs_revalidate_mapping - Revalidate the pagecache * @inode - pointer to host inode * @mapping - pointer to mapping + * @may_lock - take inode->i_mutex? */ -int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) +static int __nfs_revalidate_mapping(struct inode *inode, + struct address_space *mapping, + bool may_lock) { struct nfs_inode *nfsi = NFS_I(inode); unsigned long *bitlock = &nfsi->flags; @@ -1113,7 +1124,12 @@ int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) nfsi->cache_validity &= ~NFS_INO_INVALID_DATA; spin_unlock(&inode->i_lock); trace_nfs_invalidate_mapping_enter(inode); - ret = nfs_invalidate_mapping(inode, mapping); + if (may_lock) { + mutex_lock(&inode->i_mutex); + ret = nfs_invalidate_mapping(inode, mapping); + mutex_unlock(&inode->i_mutex); + } else + ret = nfs_invalidate_mapping(inode, mapping); trace_nfs_invalidate_mapping_exit(inode, ret); clear_bit_unlock(NFS_INO_INVALIDATING, bitlock); @@ -1123,6 +1139,29 @@ out: return ret; } +/** + * nfs_revalidate_mapping - Revalidate the pagecache + * @inode - pointer to host inode + * @mapping - pointer to mapping + */ +int nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) +{ + return __nfs_revalidate_mapping(inode, mapping, false); +} + +/** + * nfs_revalidate_mapping_protected - Revalidate the pagecache + * @inode - pointer to host inode + * @mapping - pointer to mapping + * + * Differs from nfs_revalidate_mapping() in that it grabs the inode->i_mutex + * while invalidating the mapping. + */ +int nfs_revalidate_mapping_protected(struct inode *inode, struct address_space *mapping) +{ + return __nfs_revalidate_mapping(inode, mapping, true); +} + static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr) { struct nfs_inode *nfsi = NFS_I(inode); @@ -1231,13 +1270,6 @@ static int nfs_ctime_need_update(const struct inode *inode, const struct nfs_fat return timespec_compare(&fattr->ctime, &inode->i_ctime) > 0; } -static int nfs_size_need_update(const struct inode *inode, const struct nfs_fattr *fattr) -{ - if (!(fattr->valid & NFS_ATTR_FATTR_SIZE)) - return 0; - return nfs_size_to_loff_t(fattr->size) > i_size_read(inode); -} - static atomic_long_t nfs_attr_generation_counter; static unsigned long nfs_read_attr_generation_counter(void) @@ -1249,6 +1281,7 @@ unsigned long nfs_inc_attr_generation_counter(void) { return atomic_long_inc_return(&nfs_attr_generation_counter); } +EXPORT_SYMBOL_GPL(nfs_inc_attr_generation_counter); void nfs_fattr_init(struct nfs_fattr *fattr) { @@ -1260,6 +1293,22 @@ void nfs_fattr_init(struct nfs_fattr *fattr) } EXPORT_SYMBOL_GPL(nfs_fattr_init); +/** + * nfs_fattr_set_barrier + * @fattr: attributes + * + * Used to set a barrier after an attribute was updated. This + * barrier ensures that older attributes from RPC calls that may + * have raced with our update cannot clobber these new values. + * Note that you are still responsible for ensuring that other + * operations which change the attribute on the server do not + * collide. + */ +void nfs_fattr_set_barrier(struct nfs_fattr *fattr) +{ + fattr->gencount = nfs_inc_attr_generation_counter(); +} + struct nfs_fattr *nfs_alloc_fattr(void) { struct nfs_fattr *fattr; @@ -1370,7 +1419,6 @@ static int nfs_inode_attrs_need_update(const struct inode *inode, const struct n return ((long)fattr->gencount - (long)nfsi->attr_gencount) > 0 || nfs_ctime_need_update(inode, fattr) || - nfs_size_need_update(inode, fattr) || ((long)nfsi->attr_gencount - (long)nfs_read_attr_generation_counter() > 0); } @@ -1460,6 +1508,7 @@ int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) int status; spin_lock(&inode->i_lock); + nfs_fattr_set_barrier(fattr); status = nfs_post_op_update_inode_locked(inode, fattr); spin_unlock(&inode->i_lock); @@ -1468,7 +1517,7 @@ int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) EXPORT_SYMBOL_GPL(nfs_post_op_update_inode); /** - * nfs_post_op_update_inode_force_wcc - try to update the inode attribute cache + * nfs_post_op_update_inode_force_wcc_locked - update the inode attribute cache * @inode - pointer to inode * @fattr - updated attributes * @@ -1478,11 +1527,10 @@ EXPORT_SYMBOL_GPL(nfs_post_op_update_inode); * * This function is mainly designed to be used by the ->write_done() functions. */ -int nfs_post_op_update_inode_force_wcc(struct inode *inode, struct nfs_fattr *fattr) +int nfs_post_op_update_inode_force_wcc_locked(struct inode *inode, struct nfs_fattr *fattr) { int status; - spin_lock(&inode->i_lock); /* Don't do a WCC update if these attributes are already stale */ if ((fattr->valid & NFS_ATTR_FATTR) == 0 || !nfs_inode_attrs_need_update(inode, fattr)) { @@ -1514,6 +1562,27 @@ int nfs_post_op_update_inode_force_wcc(struct inode *inode, struct nfs_fattr *fa } out_noforce: status = nfs_post_op_update_inode_locked(inode, fattr); + return status; +} + +/** + * nfs_post_op_update_inode_force_wcc - try to update the inode attribute cache + * @inode - pointer to inode + * @fattr - updated attributes + * + * After an operation that has changed the inode metadata, mark the + * attribute cache as being invalid, then try to update it. Fake up + * weak cache consistency data, if none exist. + * + * This function is mainly designed to be used by the ->write_done() functions. + */ +int nfs_post_op_update_inode_force_wcc(struct inode *inode, struct nfs_fattr *fattr) +{ + int status; + + spin_lock(&inode->i_lock); + nfs_fattr_set_barrier(fattr); + status = nfs_post_op_update_inode_force_wcc_locked(inode, fattr); spin_unlock(&inode->i_lock); return status; } @@ -1715,6 +1784,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) nfs_inc_stats(inode, NFSIOS_ATTRINVALIDATE); nfsi->attrtimeo = NFS_MINATTRTIMEO(inode); nfsi->attrtimeo_timestamp = now; + /* Set barrier to be more recent than all outstanding updates */ nfsi->attr_gencount = nfs_inc_attr_generation_counter(); } else { if (!time_in_range_open(now, nfsi->attrtimeo_timestamp, nfsi->attrtimeo_timestamp + nfsi->attrtimeo)) { @@ -1722,6 +1792,9 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) nfsi->attrtimeo = NFS_MAXATTRTIMEO(inode); nfsi->attrtimeo_timestamp = now; } + /* Set the barrier to be more recent than this fattr */ + if ((long)fattr->gencount - (long)nfsi->attr_gencount > 0) + nfsi->attr_gencount = fattr->gencount; } invalid &= ~NFS_INO_INVALID_ATTR; /* Don't invalidate the data if we were to blame */ @@ -1775,7 +1848,6 @@ static inline void nfs4_init_once(struct nfs_inode *nfsi) #if IS_ENABLED(CONFIG_NFS_V4) INIT_LIST_HEAD(&nfsi->open_states); nfsi->delegation = NULL; - nfsi->delegation_state = 0; init_rwsem(&nfsi->rwsem); nfsi->layout = NULL; #endif diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 212b8c8..9e6475b 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -459,6 +459,7 @@ void nfs_mark_request_commit(struct nfs_page *req, struct nfs_commit_info *cinfo, u32 ds_commit_idx); int nfs_write_need_commit(struct nfs_pgio_header *); +void nfs_writeback_update_inode(struct nfs_pgio_header *hdr); int nfs_generic_commit_list(struct inode *inode, struct list_head *head, int how, struct nfs_commit_info *cinfo); void nfs_retry_commit(struct list_head *page_list, @@ -598,6 +599,19 @@ void nfs_super_set_maxbytes(struct super_block *sb, __u64 maxfilesize) } /* + * Record the page as unstable and mark its inode as dirty. + */ +static inline +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); + __mark_inode_dirty(inode, I_DIRTY_DATASYNC); +} + +/* * Determine the number of bytes of data the page contains */ static inline diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index 78e557c..1f11d25 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -138,7 +138,7 @@ nfs3_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (status == 0) - nfs_setattr_update_inode(inode, sattr); + nfs_setattr_update_inode(inode, sattr, fattr); dprintk("NFS reply setattr: %d\n", status); return status; } @@ -834,7 +834,7 @@ static int nfs3_write_done(struct rpc_task *task, struct nfs_pgio_header *hdr) if (nfs3_async_handle_jukebox(task, inode)) return -EAGAIN; if (task->tk_status >= 0) - nfs_post_op_update_inode_force_wcc(inode, hdr->res.fattr); + nfs_writeback_update_inode(hdr); return 0; } diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 2a932fd..53852a4 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -1987,6 +1987,11 @@ int nfs3_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, if (entry->fattr->valid & NFS_ATTR_FATTR_V3) entry->d_type = nfs_umode_to_dtype(entry->fattr->mode); + if (entry->fattr->fileid != entry->ino) { + entry->fattr->mounted_on_fileid = entry->ino; + entry->fattr->valid |= NFS_ATTR_FATTR_MOUNTED_ON_FILEID; + } + /* In fact, a post_op_fh3: */ p = xdr_inline_decode(xdr, 4); if (unlikely(p == NULL)) diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c index 8646af9..86d6214 100644 --- a/fs/nfs/nfs4client.c +++ b/fs/nfs/nfs4client.c @@ -621,6 +621,9 @@ int nfs41_walk_client_list(struct nfs_client *new, spin_lock(&nn->nfs_client_lock); list_for_each_entry(pos, &nn->nfs_client_list, cl_share_link) { + if (pos == new) + goto found; + if (pos->rpc_ops != new->rpc_ops) continue; @@ -639,10 +642,6 @@ int nfs41_walk_client_list(struct nfs_client *new, prev = pos; status = nfs_wait_client_init_complete(pos); - if (pos->cl_cons_state == NFS_CS_SESSION_INITING) { - nfs4_schedule_lease_recovery(pos); - status = nfs4_wait_clnt_recover(pos); - } spin_lock(&nn->nfs_client_lock); if (status < 0) break; @@ -668,7 +667,7 @@ int nfs41_walk_client_list(struct nfs_client *new, */ if (!nfs4_match_client_owner_id(pos, new)) continue; - +found: atomic_inc(&pos->cl_count); *result = pos; status = 0; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 2e7c9f7..627f37c 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -901,6 +901,7 @@ static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo) if (!cinfo->atomic || cinfo->before != dir->i_version) nfs_force_lookup_revalidate(dir); dir->i_version = cinfo->after; + nfsi->attr_gencount = nfs_inc_attr_generation_counter(); nfs_fscache_invalidate(dir); spin_unlock(&dir->i_lock); } @@ -1552,6 +1553,9 @@ static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, fmode_t fmod opendata->o_arg.open_flags = 0; opendata->o_arg.fmode = fmode; + opendata->o_arg.share_access = nfs4_map_atomic_open_share( + NFS_SB(opendata->dentry->d_sb), + fmode, 0); memset(&opendata->o_res, 0, sizeof(opendata->o_res)); memset(&opendata->c_res, 0, sizeof(opendata->c_res)); nfs4_init_opendata_res(opendata); @@ -2413,8 +2417,8 @@ static int _nfs4_do_open(struct inode *dir, opendata->o_res.f_attr, sattr, state, label, olabel); if (status == 0) { - nfs_setattr_update_inode(state->inode, sattr); - nfs_post_op_update_inode(state->inode, opendata->o_res.f_attr); + nfs_setattr_update_inode(state->inode, sattr, + opendata->o_res.f_attr); nfs_setsecurity(state->inode, opendata->o_res.f_attr, olabel); } } @@ -2651,7 +2655,7 @@ static void nfs4_close_done(struct rpc_task *task, void *data) case -NFS4ERR_BAD_STATEID: case -NFS4ERR_EXPIRED: if (!nfs4_stateid_match(&calldata->arg.stateid, - &state->stateid)) { + &state->open_stateid)) { rpc_restart_call_prepare(task); goto out_release; } @@ -2687,7 +2691,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data) is_rdwr = test_bit(NFS_O_RDWR_STATE, &state->flags); is_rdonly = test_bit(NFS_O_RDONLY_STATE, &state->flags); is_wronly = test_bit(NFS_O_WRONLY_STATE, &state->flags); - nfs4_stateid_copy(&calldata->arg.stateid, &state->stateid); + nfs4_stateid_copy(&calldata->arg.stateid, &state->open_stateid); /* Calculate the change in open mode */ calldata->arg.fmode = 0; if (state->n_rdwr == 0) { @@ -3288,7 +3292,7 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, status = nfs4_do_setattr(inode, cred, fattr, sattr, state, NULL, label); if (status == 0) { - nfs_setattr_update_inode(inode, sattr); + nfs_setattr_update_inode(inode, sattr, fattr); nfs_setsecurity(inode, fattr, label); } nfs4_label_free(label); @@ -4234,7 +4238,7 @@ static int nfs4_write_done_cb(struct rpc_task *task, } if (task->tk_status >= 0) { renew_lease(NFS_SERVER(inode), hdr->timestamp); - nfs_post_op_update_inode_force_wcc(inode, &hdr->fattr); + nfs_writeback_update_inode(hdr); } return 0; } @@ -6648,47 +6652,47 @@ nfs41_same_server_scope(struct nfs41_server_scope *a, int nfs4_proc_bind_conn_to_session(struct nfs_client *clp, struct rpc_cred *cred) { int status; + struct nfs41_bind_conn_to_session_args args = { + .client = clp, + .dir = NFS4_CDFC4_FORE_OR_BOTH, + }; struct nfs41_bind_conn_to_session_res res; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_BIND_CONN_TO_SESSION], - .rpc_argp = clp, + .rpc_argp = &args, .rpc_resp = &res, .rpc_cred = cred, }; dprintk("--> %s\n", __func__); - res.session = kzalloc(sizeof(struct nfs4_session), GFP_NOFS); - if (unlikely(res.session == NULL)) { - status = -ENOMEM; - goto out; - } + nfs4_copy_sessionid(&args.sessionid, &clp->cl_session->sess_id); + if (!(clp->cl_session->flags & SESSION4_BACK_CHAN)) + args.dir = NFS4_CDFC4_FORE; status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); trace_nfs4_bind_conn_to_session(clp, status); if (status == 0) { - if (memcmp(res.session->sess_id.data, + if (memcmp(res.sessionid.data, clp->cl_session->sess_id.data, NFS4_MAX_SESSIONID_LEN)) { dprintk("NFS: %s: Session ID mismatch\n", __func__); status = -EIO; - goto out_session; + goto out; } - if (res.dir != NFS4_CDFS4_BOTH) { + if ((res.dir & args.dir) != res.dir || res.dir == 0) { dprintk("NFS: %s: Unexpected direction from server\n", __func__); status = -EIO; - goto out_session; + goto out; } - if (res.use_conn_in_rdma_mode) { + if (res.use_conn_in_rdma_mode != args.use_conn_in_rdma_mode) { dprintk("NFS: %s: Server returned RDMA mode = true\n", __func__); status = -EIO; - goto out_session; + goto out; } } -out_session: - kfree(res.session); out: dprintk("<-- %s status= %d\n", __func__, status); return status; @@ -6893,9 +6897,13 @@ static int _nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred, if (status == 0) { clp->cl_clientid = res.clientid; - clp->cl_exchange_flags = (res.flags & ~EXCHGID4_FLAG_CONFIRMED_R); - if (!(res.flags & EXCHGID4_FLAG_CONFIRMED_R)) + clp->cl_exchange_flags = res.flags; + /* Client ID is not confirmed */ + if (!(res.flags & EXCHGID4_FLAG_CONFIRMED_R)) { + clear_bit(NFS4_SESSION_ESTABLISHED, + &clp->cl_session->session_state); clp->cl_seqid = res.seqid; + } kfree(clp->cl_serverowner); clp->cl_serverowner = res.server_owner; @@ -7166,10 +7174,11 @@ static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args) args->bc_attrs.max_reqs); } -static int nfs4_verify_fore_channel_attrs(struct nfs41_create_session_args *args, struct nfs4_session *session) +static int nfs4_verify_fore_channel_attrs(struct nfs41_create_session_args *args, + struct nfs41_create_session_res *res) { struct nfs4_channel_attrs *sent = &args->fc_attrs; - struct nfs4_channel_attrs *rcvd = &session->fc_attrs; + struct nfs4_channel_attrs *rcvd = &res->fc_attrs; if (rcvd->max_resp_sz > sent->max_resp_sz) return -EINVAL; @@ -7188,11 +7197,14 @@ static int nfs4_verify_fore_channel_attrs(struct nfs41_create_session_args *args return 0; } -static int nfs4_verify_back_channel_attrs(struct nfs41_create_session_args *args, struct nfs4_session *session) +static int nfs4_verify_back_channel_attrs(struct nfs41_create_session_args *args, + struct nfs41_create_session_res *res) { struct nfs4_channel_attrs *sent = &args->bc_attrs; - struct nfs4_channel_attrs *rcvd = &session->bc_attrs; + struct nfs4_channel_attrs *rcvd = &res->bc_attrs; + if (!(res->flags & SESSION4_BACK_CHAN)) + goto out; if (rcvd->max_rqst_sz > sent->max_rqst_sz) return -EINVAL; if (rcvd->max_resp_sz < sent->max_resp_sz) @@ -7204,18 +7216,33 @@ static int nfs4_verify_back_channel_attrs(struct nfs41_create_session_args *args return -EINVAL; if (rcvd->max_reqs != sent->max_reqs) return -EINVAL; +out: return 0; } static int nfs4_verify_channel_attrs(struct nfs41_create_session_args *args, - struct nfs4_session *session) + struct nfs41_create_session_res *res) { int ret; - ret = nfs4_verify_fore_channel_attrs(args, session); + ret = nfs4_verify_fore_channel_attrs(args, res); if (ret) return ret; - return nfs4_verify_back_channel_attrs(args, session); + return nfs4_verify_back_channel_attrs(args, res); +} + +static void nfs4_update_session(struct nfs4_session *session, + struct nfs41_create_session_res *res) +{ + nfs4_copy_sessionid(&session->sess_id, &res->sessionid); + /* Mark client id and session as being confirmed */ + session->clp->cl_exchange_flags |= EXCHGID4_FLAG_CONFIRMED_R; + set_bit(NFS4_SESSION_ESTABLISHED, &session->session_state); + session->flags = res->flags; + memcpy(&session->fc_attrs, &res->fc_attrs, sizeof(session->fc_attrs)); + if (res->flags & SESSION4_BACK_CHAN) + memcpy(&session->bc_attrs, &res->bc_attrs, + sizeof(session->bc_attrs)); } static int _nfs4_proc_create_session(struct nfs_client *clp, @@ -7224,11 +7251,12 @@ static int _nfs4_proc_create_session(struct nfs_client *clp, struct nfs4_session *session = clp->cl_session; struct nfs41_create_session_args args = { .client = clp, + .clientid = clp->cl_clientid, + .seqid = clp->cl_seqid, .cb_program = NFS4_CALLBACK, }; - struct nfs41_create_session_res res = { - .client = clp, - }; + struct nfs41_create_session_res res; + struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CREATE_SESSION], .rpc_argp = &args, @@ -7245,11 +7273,15 @@ static int _nfs4_proc_create_session(struct nfs_client *clp, if (!status) { /* Verify the session's negotiated channel_attrs values */ - status = nfs4_verify_channel_attrs(&args, session); + status = nfs4_verify_channel_attrs(&args, &res); /* Increment the clientid slot sequence id */ - clp->cl_seqid++; + if (clp->cl_seqid == res.seqid) + clp->cl_seqid++; + if (status) + goto out; + nfs4_update_session(session, &res); } - +out: return status; } @@ -7301,8 +7333,8 @@ int nfs4_proc_destroy_session(struct nfs4_session *session, dprintk("--> nfs4_proc_destroy_session\n"); /* session is still being setup */ - if (session->clp->cl_cons_state != NFS_CS_READY) - return status; + if (!test_and_clear_bit(NFS4_SESSION_ESTABLISHED, &session->session_state)) + return 0; status = rpc_call_sync(session->clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); trace_nfs4_destroy_session(session->clp, status); diff --git a/fs/nfs/nfs4session.c b/fs/nfs/nfs4session.c index e799dc3..e23366e 100644 --- a/fs/nfs/nfs4session.c +++ b/fs/nfs/nfs4session.c @@ -450,7 +450,7 @@ int nfs4_setup_session_slot_tables(struct nfs4_session *ses) tbl = &ses->fc_slot_table; tbl->session = ses; status = nfs4_realloc_slot_table(tbl, ses->fc_attrs.max_reqs, 1); - if (status) /* -ENOMEM */ + if (status || !(ses->flags & SESSION4_BACK_CHAN)) /* -ENOMEM */ return status; /* Back channel */ tbl = &ses->bc_slot_table; diff --git a/fs/nfs/nfs4session.h b/fs/nfs/nfs4session.h index b34ada9..e3ea2c5 100644 --- a/fs/nfs/nfs4session.h +++ b/fs/nfs/nfs4session.h @@ -70,6 +70,7 @@ struct nfs4_session { enum nfs4_session_state { NFS4_SESSION_INITING, + NFS4_SESSION_ESTABLISHED, }; extern int nfs4_setup_slot_table(struct nfs4_slot_table *tbl, @@ -118,6 +119,12 @@ static inline int nfs4_has_persistent_session(const struct nfs_client *clp) return 0; } +static inline void nfs4_copy_sessionid(struct nfs4_sessionid *dst, + const struct nfs4_sessionid *src) +{ + memcpy(dst->data, src->data, NFS4_MAX_SESSIONID_LEN); +} + #ifdef CONFIG_CRC32 /* * nfs_session_id_hash - calculate the crc32 hash for the session id diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 5ad908e..f95e3b5 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -346,9 +346,23 @@ int nfs41_discover_server_trunking(struct nfs_client *clp, status = nfs4_proc_exchange_id(clp, cred); if (status != NFS4_OK) return status; - set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); - return nfs41_walk_client_list(clp, result, cred); + status = nfs41_walk_client_list(clp, result, cred); + if (status < 0) + return status; + if (clp != *result) + return 0; + + /* Purge state if the client id was established in a prior instance */ + if (clp->cl_exchange_flags & EXCHGID4_FLAG_CONFIRMED_R) + set_bit(NFS4CLNT_PURGE_STATE, &clp->cl_state); + else + set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); + nfs4_schedule_state_manager(clp); + status = nfs_wait_client_init_complete(clp); + if (status < 0) + nfs_put_client(clp); + return status; } #endif /* CONFIG_NFS_V4_1 */ diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index e23a0a6..5c399ec 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -1715,17 +1715,17 @@ static void encode_secinfo(struct xdr_stream *xdr, const struct qstr *name, stru #if defined(CONFIG_NFS_V4_1) /* NFSv4.1 operations */ static void encode_bind_conn_to_session(struct xdr_stream *xdr, - struct nfs4_session *session, + struct nfs41_bind_conn_to_session_args *args, struct compound_hdr *hdr) { __be32 *p; encode_op_hdr(xdr, OP_BIND_CONN_TO_SESSION, decode_bind_conn_to_session_maxsz, hdr); - encode_opaque_fixed(xdr, session->sess_id.data, NFS4_MAX_SESSIONID_LEN); + encode_opaque_fixed(xdr, args->sessionid.data, NFS4_MAX_SESSIONID_LEN); p = xdr_reserve_space(xdr, 8); - *p++ = cpu_to_be32(NFS4_CDFC4_BACK_OR_BOTH); - *p = 0; /* use_conn_in_rdma_mode = False */ + *p++ = cpu_to_be32(args->dir); + *p = (args->use_conn_in_rdma_mode) ? cpu_to_be32(1) : cpu_to_be32(0); } static void encode_op_map(struct xdr_stream *xdr, struct nfs4_op_map *op_map) @@ -1806,8 +1806,8 @@ static void encode_create_session(struct xdr_stream *xdr, encode_op_hdr(xdr, OP_CREATE_SESSION, decode_create_session_maxsz, hdr); p = reserve_space(xdr, 16 + 2*28 + 20 + clnt->cl_nodelen + 12); - p = xdr_encode_hyper(p, clp->cl_clientid); - *p++ = cpu_to_be32(clp->cl_seqid); /*Sequence id */ + p = xdr_encode_hyper(p, args->clientid); + *p++ = cpu_to_be32(args->seqid); /*Sequence id */ *p++ = cpu_to_be32(args->flags); /*flags */ /* Fore Channel */ @@ -2734,14 +2734,14 @@ static void nfs4_xdr_enc_fsid_present(struct rpc_rqst *req, */ static void nfs4_xdr_enc_bind_conn_to_session(struct rpc_rqst *req, struct xdr_stream *xdr, - struct nfs_client *clp) + struct nfs41_bind_conn_to_session_args *args) { struct compound_hdr hdr = { - .minorversion = clp->cl_mvops->minor_version, + .minorversion = args->client->cl_mvops->minor_version, }; encode_compound_hdr(xdr, req, &hdr); - encode_bind_conn_to_session(xdr, clp->cl_session, &hdr); + encode_bind_conn_to_session(xdr, args, &hdr); encode_nops(&hdr); } @@ -5613,7 +5613,7 @@ static int decode_bind_conn_to_session(struct xdr_stream *xdr, status = decode_op_hdr(xdr, OP_BIND_CONN_TO_SESSION); if (!status) - status = decode_sessionid(xdr, &res->session->sess_id); + status = decode_sessionid(xdr, &res->sessionid); if (unlikely(status)) return status; @@ -5641,12 +5641,10 @@ static int decode_create_session(struct xdr_stream *xdr, { __be32 *p; int status; - struct nfs_client *clp = res->client; - struct nfs4_session *session = clp->cl_session; status = decode_op_hdr(xdr, OP_CREATE_SESSION); if (!status) - status = decode_sessionid(xdr, &session->sess_id); + status = decode_sessionid(xdr, &res->sessionid); if (unlikely(status)) return status; @@ -5654,13 +5652,13 @@ static int decode_create_session(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) goto out_overflow; - clp->cl_seqid = be32_to_cpup(p++); - session->flags = be32_to_cpup(p); + res->seqid = be32_to_cpup(p++); + res->flags = be32_to_cpup(p); /* Channel attributes */ - status = decode_chan_attrs(xdr, &session->fc_attrs); + status = decode_chan_attrs(xdr, &res->fc_attrs); if (!status) - status = decode_chan_attrs(xdr, &session->bc_attrs); + status = decode_chan_attrs(xdr, &res->bc_attrs); return status; out_overflow: print_overflow_msg(__func__, xdr); diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index 797cd62..635f086 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -344,6 +344,10 @@ void nfs4_pnfs_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds, struct nfs4_pnfs_ds_addr *nfs4_decode_mp_ds_addr(struct net *net, struct xdr_stream *xdr, gfp_t gfp_flags); +void pnfs_layout_mark_request_commit(struct nfs_page *req, + struct pnfs_layout_segment *lseg, + struct nfs_commit_info *cinfo, + u32 ds_commit_idx); static inline bool nfs_have_layout(struct inode *inode) { diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c index fdc4f65..54e36b3 100644 --- a/fs/nfs/pnfs_nfs.c +++ b/fs/nfs/pnfs_nfs.c @@ -838,3 +838,33 @@ out_err: return NULL; } EXPORT_SYMBOL_GPL(nfs4_decode_mp_ds_addr); + +void +pnfs_layout_mark_request_commit(struct nfs_page *req, + struct pnfs_layout_segment *lseg, + struct nfs_commit_info *cinfo, + u32 ds_commit_idx) +{ + struct list_head *list; + struct pnfs_commit_bucket *buckets; + + spin_lock(cinfo->lock); + buckets = cinfo->ds->buckets; + list = &buckets[ds_commit_idx].written; + if (list_empty(list)) { + /* Non-empty buckets hold a reference on the lseg. That ref + * is normally transferred to the COMMIT call and released + * there. It could also be released if the last req is pulled + * off due to a rewrite, in which case it will be done in + * pnfs_common_clear_request_commit + */ + WARN_ON_ONCE(buckets[ds_commit_idx].wlseg != NULL); + buckets[ds_commit_idx].wlseg = pnfs_get_lseg(lseg); + } + set_bit(PG_COMMIT_TO_DS, &req->wb_flags); + cinfo->ds->nwritten++; + spin_unlock(cinfo->lock); + + nfs_request_add_commit_list(req, list, cinfo); +} +EXPORT_SYMBOL_GPL(pnfs_layout_mark_request_commit); diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index b09cc23..c63189a 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -139,7 +139,7 @@ nfs_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (status == 0) - nfs_setattr_update_inode(inode, sattr); + nfs_setattr_update_inode(inode, sattr, fattr); dprintk("NFS reply setattr: %d\n", status); return status; } @@ -609,10 +609,8 @@ static int nfs_proc_pgio_rpc_prepare(struct rpc_task *task, static int nfs_write_done(struct rpc_task *task, struct nfs_pgio_header *hdr) { - struct inode *inode = hdr->inode; - if (task->tk_status >= 0) - nfs_post_op_update_inode_force_wcc(inode, hdr->res.fattr); + nfs_writeback_update_inode(hdr); return 0; } diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 88a6d21..849ed78 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -789,13 +789,8 @@ nfs_request_add_commit_list(struct nfs_page *req, struct list_head *dst, nfs_list_add_request(req, dst); cinfo->mds->ncommit++; spin_unlock(cinfo->lock); - if (!cinfo->dreq) { - inc_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); - inc_bdi_stat(inode_to_bdi(page_file_mapping(req->wb_page)->host), - BDI_RECLAIMABLE); - __mark_inode_dirty(req->wb_context->dentry->d_inode, - I_DIRTY_DATASYNC); - } + if (!cinfo->dreq) + nfs_mark_page_unstable(req->wb_page); } EXPORT_SYMBOL_GPL(nfs_request_add_commit_list); @@ -1382,6 +1377,36 @@ static int nfs_should_remove_suid(const struct inode *inode) return 0; } +static void nfs_writeback_check_extend(struct nfs_pgio_header *hdr, + struct nfs_fattr *fattr) +{ + struct nfs_pgio_args *argp = &hdr->args; + struct nfs_pgio_res *resp = &hdr->res; + + if (!(fattr->valid & NFS_ATTR_FATTR_SIZE)) + return; + if (argp->offset + resp->count != fattr->size) + return; + if (nfs_size_to_loff_t(fattr->size) < i_size_read(hdr->inode)) + return; + /* Set attribute barrier */ + nfs_fattr_set_barrier(fattr); +} + +void nfs_writeback_update_inode(struct nfs_pgio_header *hdr) +{ + struct nfs_fattr *fattr = hdr->res.fattr; + struct inode *inode = hdr->inode; + + if (fattr == NULL) + return; + spin_lock(&inode->i_lock); + nfs_writeback_check_extend(hdr, fattr); + nfs_post_op_update_inode_force_wcc_locked(inode, fattr); + spin_unlock(&inode->i_lock); +} +EXPORT_SYMBOL_GPL(nfs_writeback_update_inode); + /* * This function is called when the WRITE call is complete. */ @@ -1605,11 +1630,8 @@ void nfs_retry_commit(struct list_head *page_list, req = nfs_list_entry(page_list->next); nfs_list_remove_request(req); nfs_mark_request_commit(req, lseg, cinfo, ds_commit_idx); - if (!cinfo->dreq) { - dec_zone_page_state(req->wb_page, NR_UNSTABLE_NFS); - dec_bdi_stat(inode_to_bdi(page_file_mapping(req->wb_page)->host), - BDI_RECLAIMABLE); - } + if (!cinfo->dreq) + nfs_clear_page_commit(req->wb_page); nfs_unlock_and_release_request(req); } } diff --git a/fs/nfsd/nfs4layouts.c b/fs/nfsd/nfs4layouts.c index 3c1bfa1..1028a06 100644 --- a/fs/nfsd/nfs4layouts.c +++ b/fs/nfsd/nfs4layouts.c @@ -587,8 +587,6 @@ nfsd4_cb_layout_fail(struct nfs4_layout_stateid *ls) rpc_ntop((struct sockaddr *)&clp->cl_addr, addr_str, sizeof(addr_str)); - nfsd4_cb_layout_fail(ls); - printk(KERN_WARNING "nfsd: client %s failed to respond to layout recall. " " Fencing..\n", addr_str); diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index cc6a760..1c307f0 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -583,7 +583,7 @@ nfs4_reset_recoverydir(char *recdir) if (status) return status; status = -ENOTDIR; - if (S_ISDIR(path.dentry->d_inode->i_mode)) { + if (d_is_dir(path.dentry)) { strcpy(user_recovery_dirname, recdir); status = 0; } @@ -1426,7 +1426,7 @@ nfsd4_client_tracking_init(struct net *net) nn->client_tracking_ops = &nfsd4_legacy_tracking_ops; status = kern_path(nfs4_recoverydir(), LOOKUP_FOLLOW, &path); if (!status) { - status = S_ISDIR(path.dentry->d_inode->i_mode); + status = d_is_dir(path.dentry); path_put(&path); if (status) goto do_init; diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index f6b2a09..d2f2c37 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -1638,7 +1638,7 @@ __destroy_client(struct nfs4_client *clp) nfs4_put_stid(&dp->dl_stid); } while (!list_empty(&clp->cl_revoked)) { - dp = list_entry(reaplist.next, struct nfs4_delegation, dl_recall_lru); + dp = list_entry(clp->cl_revoked.next, struct nfs4_delegation, dl_recall_lru); list_del_init(&dp->dl_recall_lru); nfs4_put_stid(&dp->dl_stid); } diff --git a/fs/nfsd/nfsfh.c b/fs/nfsd/nfsfh.c index 965b478..e9fa966 100644 --- a/fs/nfsd/nfsfh.c +++ b/fs/nfsd/nfsfh.c @@ -114,8 +114,8 @@ static inline __be32 check_pseudo_root(struct svc_rqst *rqstp, * We're exposing only the directories and symlinks that have to be * traversed on the way to real exports: */ - if (unlikely(!S_ISDIR(dentry->d_inode->i_mode) && - !S_ISLNK(dentry->d_inode->i_mode))) + if (unlikely(!d_is_dir(dentry) && + !d_is_symlink(dentry))) return nfserr_stale; /* * A pseudoroot export gives permission to access only one @@ -259,7 +259,7 @@ static __be32 nfsd_set_fh_dentry(struct svc_rqst *rqstp, struct svc_fh *fhp) goto out; } - if (S_ISDIR(dentry->d_inode->i_mode) && + if (d_is_dir(dentry) && (dentry->d_flags & DCACHE_DISCONNECTED)) { printk("nfsd: find_fh_dentry returned a DISCONNECTED directory: %pd2\n", dentry); @@ -414,7 +414,7 @@ static inline void _fh_update_old(struct dentry *dentry, { fh->ofh_ino = ino_t_to_u32(dentry->d_inode->i_ino); fh->ofh_generation = dentry->d_inode->i_generation; - if (S_ISDIR(dentry->d_inode->i_mode) || + if (d_is_dir(dentry) || (exp->ex_flags & NFSEXP_NOSUBTREECHECK)) fh->ofh_dirino = 0; } diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 5685c67..3685265 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -615,9 +615,9 @@ nfsd_access(struct svc_rqst *rqstp, struct svc_fh *fhp, u32 *access, u32 *suppor export = fhp->fh_export; dentry = fhp->fh_dentry; - if (S_ISREG(dentry->d_inode->i_mode)) + if (d_is_reg(dentry)) map = nfs3_regaccess; - else if (S_ISDIR(dentry->d_inode->i_mode)) + else if (d_is_dir(dentry)) map = nfs3_diraccess; else map = nfs3_anyaccess; @@ -1402,7 +1402,7 @@ do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, switch (createmode) { case NFS3_CREATE_UNCHECKED: - if (! S_ISREG(dchild->d_inode->i_mode)) + if (! d_is_reg(dchild)) goto out; else if (truncp) { /* in nfsv4, we need to treat this case a little @@ -1615,7 +1615,7 @@ nfsd_link(struct svc_rqst *rqstp, struct svc_fh *ffhp, if (err) goto out; err = nfserr_isdir; - if (S_ISDIR(tfhp->fh_dentry->d_inode->i_mode)) + if (d_is_dir(tfhp->fh_dentry)) goto out; err = nfserr_perm; if (!len) diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index b2e3ff3..ecdbae1 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -31,6 +31,8 @@ #include "alloc.h" #include "dat.h" +static void __nilfs_btree_init(struct nilfs_bmap *bmap); + static struct nilfs_btree_path *nilfs_btree_alloc_path(void) { struct nilfs_btree_path *path; @@ -368,6 +370,34 @@ static int nilfs_btree_node_broken(const struct nilfs_btree_node *node, return ret; } +/** + * nilfs_btree_root_broken - verify consistency of btree root node + * @node: btree root node to be examined + * @ino: inode number + * + * Return Value: If node is broken, 1 is returned. Otherwise, 0 is returned. + */ +static int nilfs_btree_root_broken(const struct nilfs_btree_node *node, + unsigned long ino) +{ + int level, flags, nchildren; + int ret = 0; + + level = nilfs_btree_node_get_level(node); + flags = nilfs_btree_node_get_flags(node); + nchildren = nilfs_btree_node_get_nchildren(node); + + if (unlikely(level < NILFS_BTREE_LEVEL_NODE_MIN || + level > NILFS_BTREE_LEVEL_MAX || + nchildren < 0 || + nchildren > NILFS_BTREE_ROOT_NCHILDREN_MAX)) { + pr_crit("NILFS: bad btree root (inode number=%lu): level = %d, flags = 0x%x, nchildren = %d\n", + ino, level, flags, nchildren); + ret = 1; + } + return ret; +} + int nilfs_btree_broken_node_block(struct buffer_head *bh) { int ret; @@ -1713,7 +1743,7 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *btree, /* convert and insert */ dat = NILFS_BMAP_USE_VBN(btree) ? nilfs_bmap_get_dat(btree) : NULL; - nilfs_btree_init(btree); + __nilfs_btree_init(btree); if (nreq != NULL) { nilfs_bmap_commit_alloc_ptr(btree, dreq, dat); nilfs_bmap_commit_alloc_ptr(btree, nreq, dat); @@ -2294,12 +2324,23 @@ static const struct nilfs_bmap_operations nilfs_btree_ops_gc = { .bop_gather_data = NULL, }; -int nilfs_btree_init(struct nilfs_bmap *bmap) +static void __nilfs_btree_init(struct nilfs_bmap *bmap) { bmap->b_ops = &nilfs_btree_ops; bmap->b_nchildren_per_block = NILFS_BTREE_NODE_NCHILDREN_MAX(nilfs_btree_node_size(bmap)); - return 0; +} + +int nilfs_btree_init(struct nilfs_bmap *bmap) +{ + int ret = 0; + + __nilfs_btree_init(bmap); + + if (nilfs_btree_root_broken(nilfs_btree_get_root(bmap), + bmap->b_inode->i_ino)) + ret = -EIO; + return ret; } void nilfs_btree_init_gc(struct nilfs_bmap *bmap) diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 469086b..0c3f303 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -1907,6 +1907,7 @@ static void nilfs_segctor_drop_written_files(struct nilfs_sc_info *sci, struct the_nilfs *nilfs) { struct nilfs_inode_info *ii, *n; + int during_mount = !(sci->sc_super->s_flags & MS_ACTIVE); int defer_iput = false; spin_lock(&nilfs->ns_inode_lock); @@ -1919,10 +1920,10 @@ static void nilfs_segctor_drop_written_files(struct nilfs_sc_info *sci, brelse(ii->i_bh); ii->i_bh = NULL; list_del_init(&ii->i_dirty); - if (!ii->vfs_inode.i_nlink) { + if (!ii->vfs_inode.i_nlink || during_mount) { /* - * Defer calling iput() to avoid a deadlock - * over I_SYNC flag for inodes with i_nlink == 0 + * Defer calling iput() to avoid deadlocks if + * i_nlink == 0 or mount is not yet finished. */ list_add_tail(&ii->i_dirty, &sci->sc_iput_queue); defer_iput = true; diff --git a/fs/notify/fanotify/fanotify.c b/fs/notify/fanotify/fanotify.c index 51ceb81..d2f97ec 100644 --- a/fs/notify/fanotify/fanotify.c +++ b/fs/notify/fanotify/fanotify.c @@ -115,8 +115,8 @@ static bool fanotify_should_send_event(struct fsnotify_mark *inode_mark, return false; /* sorry, fanotify only gives a damn about files and dirs */ - if (!S_ISREG(path->dentry->d_inode->i_mode) && - !S_ISDIR(path->dentry->d_inode->i_mode)) + if (!d_is_reg(path->dentry) && + !d_can_lookup(path->dentry)) return false; if (inode_mark && vfsmnt_mark) { @@ -139,11 +139,12 @@ static bool fanotify_should_send_event(struct fsnotify_mark *inode_mark, BUG(); } - if (S_ISDIR(path->dentry->d_inode->i_mode) && + if (d_is_dir(path->dentry) && !(marks_mask & FS_ISDIR & ~marks_ignored_mask)) return false; - if (event_mask & marks_mask & ~marks_ignored_mask) + if (event_mask & FAN_ALL_OUTGOING_EVENTS & marks_mask & + ~marks_ignored_mask) return true; return false; diff --git a/fs/ntfs/Makefile b/fs/ntfs/Makefile index 36ae529..2ff263e 100644 --- a/fs/ntfs/Makefile +++ b/fs/ntfs/Makefile @@ -8,7 +8,7 @@ ntfs-y := aops.o attrib.o collate.o compress.o debug.o dir.o file.o \ ntfs-$(CONFIG_NTFS_RW) += bitmap.o lcnalloc.o logfile.o quota.o usnjrnl.o -ccflags-y := -DNTFS_VERSION=\"2.1.31\" +ccflags-y := -DNTFS_VERSION=\"2.1.32\" ccflags-$(CONFIG_NTFS_DEBUG) += -DDEBUG ccflags-$(CONFIG_NTFS_RW) += -DNTFS_RW diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c index f16f2d8..c1da78d 100644 --- a/fs/ntfs/file.c +++ b/fs/ntfs/file.c @@ -1,7 +1,7 @@ /* * file.c - NTFS kernel file operations. Part of the Linux-NTFS project. * - * Copyright (c) 2001-2014 Anton Altaparmakov and Tuxera Inc. + * Copyright (c) 2001-2015 Anton Altaparmakov and Tuxera Inc. * * This program/include file is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License as published @@ -328,62 +328,168 @@ err_out: return err; } -/** - * ntfs_fault_in_pages_readable - - * - * Fault a number of userspace pages into pagetables. - * - * Unlike include/linux/pagemap.h::fault_in_pages_readable(), this one copes - * with more than two userspace pages as well as handling the single page case - * elegantly. - * - * If you find this difficult to understand, then think of the while loop being - * the following code, except that we do without the integer variable ret: - * - * do { - * ret = __get_user(c, uaddr); - * uaddr += PAGE_SIZE; - * } while (!ret && uaddr < end); - * - * Note, the final __get_user() may well run out-of-bounds of the user buffer, - * but _not_ out-of-bounds of the page the user buffer belongs to, and since - * this is only a read and not a write, and since it is still in the same page, - * it should not matter and this makes the code much simpler. - */ -static inline void ntfs_fault_in_pages_readable(const char __user *uaddr, - int bytes) +static ssize_t ntfs_prepare_file_for_write(struct file *file, loff_t *ppos, + size_t *count) { - const char __user *end; - volatile char c; - - /* Set @end to the first byte outside the last page we care about. */ - end = (const char __user*)PAGE_ALIGN((unsigned long)uaddr + bytes); - - while (!__get_user(c, uaddr) && (uaddr += PAGE_SIZE, uaddr < end)) - ; -} - -/** - * ntfs_fault_in_pages_readable_iovec - - * - * Same as ntfs_fault_in_pages_readable() but operates on an array of iovecs. - */ -static inline void ntfs_fault_in_pages_readable_iovec(const struct iovec *iov, - size_t iov_ofs, int bytes) -{ - do { - const char __user *buf; - unsigned len; + loff_t pos; + s64 end, ll; + ssize_t err; + unsigned long flags; + struct inode *vi = file_inode(file); + ntfs_inode *base_ni, *ni = NTFS_I(vi); + ntfs_volume *vol = ni->vol; - buf = iov->iov_base + iov_ofs; - len = iov->iov_len - iov_ofs; - if (len > bytes) - len = bytes; - ntfs_fault_in_pages_readable(buf, len); - bytes -= len; - iov++; - iov_ofs = 0; - } while (bytes); + ntfs_debug("Entering for i_ino 0x%lx, attribute type 0x%x, pos " + "0x%llx, count 0x%lx.", vi->i_ino, + (unsigned)le32_to_cpu(ni->type), + (unsigned long long)*ppos, (unsigned long)*count); + /* We can write back this queue in page reclaim. */ + current->backing_dev_info = inode_to_bdi(vi); + err = generic_write_checks(file, ppos, count, S_ISBLK(vi->i_mode)); + if (unlikely(err)) + goto out; + /* + * All checks have passed. Before we start doing any writing we want + * to abort any totally illegal writes. + */ + BUG_ON(NInoMstProtected(ni)); + BUG_ON(ni->type != AT_DATA); + /* If file is encrypted, deny access, just like NT4. */ + if (NInoEncrypted(ni)) { + /* Only $DATA attributes can be encrypted. */ + /* + * Reminder for later: Encrypted files are _always_ + * non-resident so that the content can always be encrypted. + */ + ntfs_debug("Denying write access to encrypted file."); + err = -EACCES; + goto out; + } + if (NInoCompressed(ni)) { + /* Only unnamed $DATA attribute can be compressed. */ + BUG_ON(ni->name_len); + /* + * Reminder for later: If resident, the data is not actually + * compressed. Only on the switch to non-resident does + * compression kick in. This is in contrast to encrypted files + * (see above). + */ + ntfs_error(vi->i_sb, "Writing to compressed files is not " + "implemented yet. Sorry."); + err = -EOPNOTSUPP; + goto out; + } + if (*count == 0) + goto out; + base_ni = ni; + if (NInoAttr(ni)) + base_ni = ni->ext.base_ntfs_ino; + err = file_remove_suid(file); + if (unlikely(err)) + goto out; + /* + * Our ->update_time method always succeeds thus file_update_time() + * cannot fail either so there is no need to check the return code. + */ + file_update_time(file); + pos = *ppos; + /* The first byte after the last cluster being written to. */ + end = (pos + *count + vol->cluster_size_mask) & + ~(u64)vol->cluster_size_mask; + /* + * If the write goes beyond the allocated size, extend the allocation + * to cover the whole of the write, rounded up to the nearest cluster. + */ + read_lock_irqsave(&ni->size_lock, flags); + ll = ni->allocated_size; + read_unlock_irqrestore(&ni->size_lock, flags); + if (end > ll) { + /* + * Extend the allocation without changing the data size. + * + * Note we ensure the allocation is big enough to at least + * write some data but we do not require the allocation to be + * complete, i.e. it may be partial. + */ + ll = ntfs_attr_extend_allocation(ni, end, -1, pos); + if (likely(ll >= 0)) { + BUG_ON(pos >= ll); + /* If the extension was partial truncate the write. */ + if (end > ll) { + ntfs_debug("Truncating write to inode 0x%lx, " + "attribute type 0x%x, because " + "the allocation was only " + "partially extended.", + vi->i_ino, (unsigned) + le32_to_cpu(ni->type)); + *count = ll - pos; + } + } else { + err = ll; + read_lock_irqsave(&ni->size_lock, flags); + ll = ni->allocated_size; + read_unlock_irqrestore(&ni->size_lock, flags); + /* Perform a partial write if possible or fail. */ + if (pos < ll) { + ntfs_debug("Truncating write to inode 0x%lx " + "attribute type 0x%x, because " + "extending the allocation " + "failed (error %d).", + vi->i_ino, (unsigned) + le32_to_cpu(ni->type), + (int)-err); + *count = ll - pos; + } else { + if (err != -ENOSPC) + ntfs_error(vi->i_sb, "Cannot perform " + "write to inode " + "0x%lx, attribute " + "type 0x%x, because " + "extending the " + "allocation failed " + "(error %ld).", + vi->i_ino, (unsigned) + le32_to_cpu(ni->type), + (long)-err); + else + ntfs_debug("Cannot perform write to " + "inode 0x%lx, " + "attribute type 0x%x, " + "because there is not " + "space left.", + vi->i_ino, (unsigned) + le32_to_cpu(ni->type)); + goto out; + } + } + } + /* + * If the write starts beyond the initialized size, extend it up to the + * beginning of the write and initialize all non-sparse space between + * the old initialized size and the new one. This automatically also + * increments the vfs inode->i_size to keep it above or equal to the + * initialized_size. + */ + read_lock_irqsave(&ni->size_lock, flags); + ll = ni->initialized_size; + read_unlock_irqrestore(&ni->size_lock, flags); + if (pos > ll) { + /* + * Wait for ongoing direct i/o to complete before proceeding. + * New direct i/o cannot start as we hold i_mutex. + */ + inode_dio_wait(vi); + err = ntfs_attr_extend_initialized(ni, pos); + if (unlikely(err < 0)) + ntfs_error(vi->i_sb, "Cannot perform write to inode " + "0x%lx, attribute type 0x%x, because " + "extending the initialized size " + "failed (error %d).", vi->i_ino, + (unsigned)le32_to_cpu(ni->type), + (int)-err); + } +out: + return err; } /** @@ -420,8 +526,8 @@ static inline int __ntfs_grab_cache_pages(struct address_space *mapping, goto err_out; } } - err = add_to_page_cache_lru(*cached_page, mapping, index, - GFP_KERNEL); + err = add_to_page_cache_lru(*cached_page, mapping, + index, GFP_KERNEL); if (unlikely(err)) { if (err == -EEXIST) continue; @@ -1267,180 +1373,6 @@ rl_not_mapped_enoent: return err; } -/* - * Copy as much as we can into the pages and return the number of bytes which - * were successfully copied. If a fault is encountered then clear the pages - * out to (ofs + bytes) and return the number of bytes which were copied. - */ -static inline size_t ntfs_copy_from_user(struct page **pages, - unsigned nr_pages, unsigned ofs, const char __user *buf, - size_t bytes) -{ - struct page **last_page = pages + nr_pages; - char *addr; - size_t total = 0; - unsigned len; - int left; - - do { - len = PAGE_CACHE_SIZE - ofs; - if (len > bytes) - len = bytes; - addr = kmap_atomic(*pages); - left = __copy_from_user_inatomic(addr + ofs, buf, len); - kunmap_atomic(addr); - if (unlikely(left)) { - /* Do it the slow way. */ - addr = kmap(*pages); - left = __copy_from_user(addr + ofs, buf, len); - kunmap(*pages); - if (unlikely(left)) - goto err_out; - } - total += len; - bytes -= len; - if (!bytes) - break; - buf += len; - ofs = 0; - } while (++pages < last_page); -out: - return total; -err_out: - total += len - left; - /* Zero the rest of the target like __copy_from_user(). */ - while (++pages < last_page) { - bytes -= len; - if (!bytes) - break; - len = PAGE_CACHE_SIZE; - if (len > bytes) - len = bytes; - zero_user(*pages, 0, len); - } - goto out; -} - -static size_t __ntfs_copy_from_user_iovec_inatomic(char *vaddr, - const struct iovec *iov, size_t iov_ofs, size_t bytes) -{ - size_t total = 0; - - while (1) { - const char __user *buf = iov->iov_base + iov_ofs; - unsigned len; - size_t left; - - len = iov->iov_len - iov_ofs; - if (len > bytes) - len = bytes; - left = __copy_from_user_inatomic(vaddr, buf, len); - total += len; - bytes -= len; - vaddr += len; - if (unlikely(left)) { - total -= left; - break; - } - if (!bytes) - break; - iov++; - iov_ofs = 0; - } - return total; -} - -static inline void ntfs_set_next_iovec(const struct iovec **iovp, - size_t *iov_ofsp, size_t bytes) -{ - const struct iovec *iov = *iovp; - size_t iov_ofs = *iov_ofsp; - - while (bytes) { - unsigned len; - - len = iov->iov_len - iov_ofs; - if (len > bytes) - len = bytes; - bytes -= len; - iov_ofs += len; - if (iov->iov_len == iov_ofs) { - iov++; - iov_ofs = 0; - } - } - *iovp = iov; - *iov_ofsp = iov_ofs; -} - -/* - * This has the same side-effects and return value as ntfs_copy_from_user(). - * The difference is that on a fault we need to memset the remainder of the - * pages (out to offset + bytes), to emulate ntfs_copy_from_user()'s - * single-segment behaviour. - * - * We call the same helper (__ntfs_copy_from_user_iovec_inatomic()) both when - * atomic and when not atomic. This is ok because it calls - * __copy_from_user_inatomic() and it is ok to call this when non-atomic. In - * fact, the only difference between __copy_from_user_inatomic() and - * __copy_from_user() is that the latter calls might_sleep() and the former - * should not zero the tail of the buffer on error. And on many architectures - * __copy_from_user_inatomic() is just defined to __copy_from_user() so it - * makes no difference at all on those architectures. - */ -static inline size_t ntfs_copy_from_user_iovec(struct page **pages, - unsigned nr_pages, unsigned ofs, const struct iovec **iov, - size_t *iov_ofs, size_t bytes) -{ - struct page **last_page = pages + nr_pages; - char *addr; - size_t copied, len, total = 0; - - do { - len = PAGE_CACHE_SIZE - ofs; - if (len > bytes) - len = bytes; - addr = kmap_atomic(*pages); - copied = __ntfs_copy_from_user_iovec_inatomic(addr + ofs, - *iov, *iov_ofs, len); - kunmap_atomic(addr); - if (unlikely(copied != len)) { - /* Do it the slow way. */ - addr = kmap(*pages); - copied = __ntfs_copy_from_user_iovec_inatomic(addr + - ofs, *iov, *iov_ofs, len); - if (unlikely(copied != len)) - goto err_out; - kunmap(*pages); - } - total += len; - ntfs_set_next_iovec(iov, iov_ofs, len); - bytes -= len; - if (!bytes) - break; - ofs = 0; - } while (++pages < last_page); -out: - return total; -err_out: - BUG_ON(copied > len); - /* Zero the rest of the target like __copy_from_user(). */ - memset(addr + ofs + copied, 0, len - copied); - kunmap(*pages); - total += copied; - ntfs_set_next_iovec(iov, iov_ofs, copied); - while (++pages < last_page) { - bytes -= len; - if (!bytes) - break; - len = PAGE_CACHE_SIZE; - if (len > bytes) - len = bytes; - zero_user(*pages, 0, len); - } - goto out; -} - static inline void ntfs_flush_dcache_pages(struct page **pages, unsigned nr_pages) { @@ -1761,86 +1693,83 @@ err_out: return err; } -static void ntfs_write_failed(struct address_space *mapping, loff_t to) +/* + * Copy as much as we can into the pages and return the number of bytes which + * were successfully copied. If a fault is encountered then clear the pages + * out to (ofs + bytes) and return the number of bytes which were copied. + */ +static size_t ntfs_copy_from_user_iter(struct page **pages, unsigned nr_pages, + unsigned ofs, struct iov_iter *i, size_t bytes) { - struct inode *inode = mapping->host; + struct page **last_page = pages + nr_pages; + size_t total = 0; + struct iov_iter data = *i; + unsigned len, copied; - if (to > inode->i_size) { - truncate_pagecache(inode, inode->i_size); - ntfs_truncate_vfs(inode); - } + do { + len = PAGE_CACHE_SIZE - ofs; + if (len > bytes) + len = bytes; + copied = iov_iter_copy_from_user_atomic(*pages, &data, ofs, + len); + total += copied; + bytes -= copied; + if (!bytes) + break; + iov_iter_advance(&data, copied); + if (copied < len) + goto err; + ofs = 0; + } while (++pages < last_page); +out: + return total; +err: + /* Zero the rest of the target like __copy_from_user(). */ + len = PAGE_CACHE_SIZE - copied; + do { + if (len > bytes) + len = bytes; + zero_user(*pages, copied, len); + bytes -= len; + copied = 0; + len = PAGE_CACHE_SIZE; + } while (++pages < last_page); + goto out; } /** - * ntfs_file_buffered_write - - * - * Locking: The vfs is holding ->i_mutex on the inode. + * ntfs_perform_write - perform buffered write to a file + * @file: file to write to + * @i: iov_iter with data to write + * @pos: byte offset in file at which to begin writing to */ -static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, - const struct iovec *iov, unsigned long nr_segs, - loff_t pos, loff_t *ppos, size_t count) +static ssize_t ntfs_perform_write(struct file *file, struct iov_iter *i, + loff_t pos) { - struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; struct inode *vi = mapping->host; ntfs_inode *ni = NTFS_I(vi); ntfs_volume *vol = ni->vol; struct page *pages[NTFS_MAX_PAGES_PER_CLUSTER]; struct page *cached_page = NULL; - char __user *buf = NULL; - s64 end, ll; VCN last_vcn; LCN lcn; - unsigned long flags; - size_t bytes, iov_ofs = 0; /* Offset in the current iovec. */ - ssize_t status, written; + size_t bytes; + ssize_t status, written = 0; unsigned nr_pages; - int err; - ntfs_debug("Entering for i_ino 0x%lx, attribute type 0x%x, " - "pos 0x%llx, count 0x%lx.", - vi->i_ino, (unsigned)le32_to_cpu(ni->type), - (unsigned long long)pos, (unsigned long)count); - if (unlikely(!count)) - return 0; - BUG_ON(NInoMstProtected(ni)); - /* - * If the attribute is not an index root and it is encrypted or - * compressed, we cannot write to it yet. Note we need to check for - * AT_INDEX_ALLOCATION since this is the type of both directory and - * index inodes. - */ - if (ni->type != AT_INDEX_ALLOCATION) { - /* If file is encrypted, deny access, just like NT4. */ - if (NInoEncrypted(ni)) { - /* - * Reminder for later: Encrypted files are _always_ - * non-resident so that the content can always be - * encrypted. - */ - ntfs_debug("Denying write access to encrypted file."); - return -EACCES; - } - if (NInoCompressed(ni)) { - /* Only unnamed $DATA attribute can be compressed. */ - BUG_ON(ni->type != AT_DATA); - BUG_ON(ni->name_len); - /* - * Reminder for later: If resident, the data is not - * actually compressed. Only on the switch to non- - * resident does compression kick in. This is in - * contrast to encrypted files (see above). - */ - ntfs_error(vi->i_sb, "Writing to compressed files is " - "not implemented yet. Sorry."); - return -EOPNOTSUPP; - } - } + ntfs_debug("Entering for i_ino 0x%lx, attribute type 0x%x, pos " + "0x%llx, count 0x%lx.", vi->i_ino, + (unsigned)le32_to_cpu(ni->type), + (unsigned long long)pos, + (unsigned long)iov_iter_count(i)); /* * If a previous ntfs_truncate() failed, repeat it and abort if it * fails again. */ if (unlikely(NInoTruncateFailed(ni))) { + int err; + inode_dio_wait(vi); err = ntfs_truncate(vi); if (err || NInoTruncateFailed(ni)) { @@ -1854,81 +1783,6 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, return err; } } - /* The first byte after the write. */ - end = pos + count; - /* - * If the write goes beyond the allocated size, extend the allocation - * to cover the whole of the write, rounded up to the nearest cluster. - */ - read_lock_irqsave(&ni->size_lock, flags); - ll = ni->allocated_size; - read_unlock_irqrestore(&ni->size_lock, flags); - if (end > ll) { - /* Extend the allocation without changing the data size. */ - ll = ntfs_attr_extend_allocation(ni, end, -1, pos); - if (likely(ll >= 0)) { - BUG_ON(pos >= ll); - /* If the extension was partial truncate the write. */ - if (end > ll) { - ntfs_debug("Truncating write to inode 0x%lx, " - "attribute type 0x%x, because " - "the allocation was only " - "partially extended.", - vi->i_ino, (unsigned) - le32_to_cpu(ni->type)); - end = ll; - count = ll - pos; - } - } else { - err = ll; - read_lock_irqsave(&ni->size_lock, flags); - ll = ni->allocated_size; - read_unlock_irqrestore(&ni->size_lock, flags); - /* Perform a partial write if possible or fail. */ - if (pos < ll) { - ntfs_debug("Truncating write to inode 0x%lx, " - "attribute type 0x%x, because " - "extending the allocation " - "failed (error code %i).", - vi->i_ino, (unsigned) - le32_to_cpu(ni->type), err); - end = ll; - count = ll - pos; - } else { - ntfs_error(vol->sb, "Cannot perform write to " - "inode 0x%lx, attribute type " - "0x%x, because extending the " - "allocation failed (error " - "code %i).", vi->i_ino, - (unsigned) - le32_to_cpu(ni->type), err); - return err; - } - } - } - written = 0; - /* - * If the write starts beyond the initialized size, extend it up to the - * beginning of the write and initialize all non-sparse space between - * the old initialized size and the new one. This automatically also - * increments the vfs inode->i_size to keep it above or equal to the - * initialized_size. - */ - read_lock_irqsave(&ni->size_lock, flags); - ll = ni->initialized_size; - read_unlock_irqrestore(&ni->size_lock, flags); - if (pos > ll) { - err = ntfs_attr_extend_initialized(ni, pos); - if (err < 0) { - ntfs_error(vol->sb, "Cannot perform write to inode " - "0x%lx, attribute type 0x%x, because " - "extending the initialized size " - "failed (error code %i).", vi->i_ino, - (unsigned)le32_to_cpu(ni->type), err); - status = err; - goto err_out; - } - } /* * Determine the number of pages per cluster for non-resident * attributes. @@ -1936,10 +1790,7 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, nr_pages = 1; if (vol->cluster_size > PAGE_CACHE_SIZE && NInoNonResident(ni)) nr_pages = vol->cluster_size >> PAGE_CACHE_SHIFT; - /* Finally, perform the actual write. */ last_vcn = -1; - if (likely(nr_segs == 1)) - buf = iov->iov_base; do { VCN vcn; pgoff_t idx, start_idx; @@ -1964,10 +1815,10 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, vol->cluster_size_bits, false); up_read(&ni->runlist.lock); if (unlikely(lcn < LCN_HOLE)) { - status = -EIO; if (lcn == LCN_ENOMEM) status = -ENOMEM; - else + else { + status = -EIO; ntfs_error(vol->sb, "Cannot " "perform write to " "inode 0x%lx, " @@ -1976,6 +1827,7 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, "is corrupt.", vi->i_ino, (unsigned) le32_to_cpu(ni->type)); + } break; } if (lcn == LCN_HOLE) { @@ -1988,8 +1840,9 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, } } } - if (bytes > count) - bytes = count; + if (bytes > iov_iter_count(i)) + bytes = iov_iter_count(i); +again: /* * Bring in the user page(s) that we will copy from _first_. * Otherwise there is a nasty deadlock on copying from the same @@ -1998,10 +1851,10 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, * pages being swapped out between us bringing them into memory * and doing the actual copying. */ - if (likely(nr_segs == 1)) - ntfs_fault_in_pages_readable(buf, bytes); - else - ntfs_fault_in_pages_readable_iovec(iov, iov_ofs, bytes); + if (unlikely(iov_iter_fault_in_multipages_readable(i, bytes))) { + status = -EFAULT; + break; + } /* Get and lock @do_pages starting at index @start_idx. */ status = __ntfs_grab_cache_pages(mapping, start_idx, do_pages, pages, &cached_page); @@ -2017,56 +1870,57 @@ static ssize_t ntfs_file_buffered_write(struct kiocb *iocb, status = ntfs_prepare_pages_for_non_resident_write( pages, do_pages, pos, bytes); if (unlikely(status)) { - loff_t i_size; - do { unlock_page(pages[--do_pages]); page_cache_release(pages[do_pages]); } while (do_pages); - /* - * The write preparation may have instantiated - * allocated space outside i_size. Trim this - * off again. We can ignore any errors in this - * case as we will just be waisting a bit of - * allocated space, which is not a disaster. - */ - i_size = i_size_read(vi); - if (pos + bytes > i_size) { - ntfs_write_failed(mapping, pos + bytes); - } break; } } u = (pos >> PAGE_CACHE_SHIFT) - pages[0]->index; - if (likely(nr_segs == 1)) { - copied = ntfs_copy_from_user(pages + u, do_pages - u, - ofs, buf, bytes); - buf += copied; - } else - copied = ntfs_copy_from_user_iovec(pages + u, - do_pages - u, ofs, &iov, &iov_ofs, - bytes); + copied = ntfs_copy_from_user_iter(pages + u, do_pages - u, ofs, + i, bytes); ntfs_flush_dcache_pages(pages + u, do_pages - u); - status = ntfs_commit_pages_after_write(pages, do_pages, pos, - bytes); - if (likely(!status)) { - written += copied; - count -= copied; - pos += copied; - if (unlikely(copied != bytes)) - status = -EFAULT; + status = 0; + if (likely(copied == bytes)) { + status = ntfs_commit_pages_after_write(pages, do_pages, + pos, bytes); + if (!status) + status = bytes; } do { unlock_page(pages[--do_pages]); page_cache_release(pages[do_pages]); } while (do_pages); - if (unlikely(status)) + if (unlikely(status < 0)) break; - balance_dirty_pages_ratelimited(mapping); + copied = status; cond_resched(); - } while (count); -err_out: - *ppos = pos; + if (unlikely(!copied)) { + size_t sc; + + /* + * We failed to copy anything. Fall back to single + * segment length write. + * + * This is needed to avoid possible livelock in the + * case that all segments in the iov cannot be copied + * at once without a pagefault. + */ + sc = iov_iter_single_seg_count(i); + if (bytes > sc) + bytes = sc; + goto again; + } + iov_iter_advance(i, copied); + pos += copied; + written += copied; + balance_dirty_pages_ratelimited(mapping); + if (fatal_signal_pending(current)) { + status = -EINTR; + break; + } + } while (iov_iter_count(i)); if (cached_page) page_cache_release(cached_page); ntfs_debug("Done. Returning %s (written 0x%lx, status %li).", @@ -2076,59 +1930,56 @@ err_out: } /** - * ntfs_file_aio_write_nolock - + * ntfs_file_write_iter_nolock - write data to a file + * @iocb: IO state structure (file, offset, etc.) + * @from: iov_iter with data to write + * + * Basically the same as __generic_file_write_iter() except that it ends + * up calling ntfs_perform_write() instead of generic_perform_write() and that + * O_DIRECT is not implemented. */ -static ssize_t ntfs_file_aio_write_nolock(struct kiocb *iocb, - const struct iovec *iov, unsigned long nr_segs, loff_t *ppos) +static ssize_t ntfs_file_write_iter_nolock(struct kiocb *iocb, + struct iov_iter *from) { struct file *file = iocb->ki_filp; - struct address_space *mapping = file->f_mapping; - struct inode *inode = mapping->host; - loff_t pos; - size_t count; /* after file limit checks */ - ssize_t written, err; + loff_t pos = iocb->ki_pos; + ssize_t written = 0; + ssize_t err; + size_t count = iov_iter_count(from); - count = iov_length(iov, nr_segs); - pos = *ppos; - /* We can write back this queue in page reclaim. */ - current->backing_dev_info = inode_to_bdi(inode); - written = 0; - err = generic_write_checks(file, &pos, &count, S_ISBLK(inode->i_mode)); - if (err) - goto out; - if (!count) - goto out; - err = file_remove_suid(file); - if (err) - goto out; - err = file_update_time(file); - if (err) - goto out; - written = ntfs_file_buffered_write(iocb, iov, nr_segs, pos, ppos, - count); -out: + err = ntfs_prepare_file_for_write(file, &pos, &count); + if (count && !err) { + iov_iter_truncate(from, count); + written = ntfs_perform_write(file, from, pos); + if (likely(written >= 0)) + iocb->ki_pos = pos + written; + } current->backing_dev_info = NULL; return written ? written : err; } /** - * ntfs_file_aio_write - + * ntfs_file_write_iter - simple wrapper for ntfs_file_write_iter_nolock() + * @iocb: IO state structure + * @from: iov_iter with data to write + * + * Basically the same as generic_file_write_iter() except that it ends up + * calling ntfs_file_write_iter_nolock() instead of + * __generic_file_write_iter(). */ -static ssize_t ntfs_file_aio_write(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) +static ssize_t ntfs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *file = iocb->ki_filp; - struct address_space *mapping = file->f_mapping; - struct inode *inode = mapping->host; + struct inode *vi = file_inode(file); ssize_t ret; - BUG_ON(iocb->ki_pos != pos); - - mutex_lock(&inode->i_mutex); - ret = ntfs_file_aio_write_nolock(iocb, iov, nr_segs, &iocb->ki_pos); - mutex_unlock(&inode->i_mutex); + mutex_lock(&vi->i_mutex); + ret = ntfs_file_write_iter_nolock(iocb, from); + mutex_unlock(&vi->i_mutex); if (ret > 0) { - int err = generic_write_sync(file, iocb->ki_pos - ret, ret); + ssize_t err; + + err = generic_write_sync(file, iocb->ki_pos - ret, ret); if (err < 0) ret = err; } @@ -2196,37 +2047,17 @@ static int ntfs_file_fsync(struct file *filp, loff_t start, loff_t end, #endif /* NTFS_RW */ const struct file_operations ntfs_file_ops = { - .llseek = generic_file_llseek, /* Seek inside file. */ - .read = new_sync_read, /* Read from file. */ - .read_iter = generic_file_read_iter, /* Async read from file. */ + .llseek = generic_file_llseek, + .read = new_sync_read, + .read_iter = generic_file_read_iter, #ifdef NTFS_RW - .write = do_sync_write, /* Write to file. */ - .aio_write = ntfs_file_aio_write, /* Async write to file. */ - /*.release = ,*/ /* Last file is closed. See - fs/ext2/file.c:: - ext2_release_file() for - how to use this to discard - preallocated space for - write opened files. */ - .fsync = ntfs_file_fsync, /* Sync a file to disk. */ - /*.aio_fsync = ,*/ /* Sync all outstanding async - i/o operations on a - kiocb. */ + .write = new_sync_write, + .write_iter = ntfs_file_write_iter, + .fsync = ntfs_file_fsync, #endif /* NTFS_RW */ - /*.ioctl = ,*/ /* Perform function on the - mounted filesystem. */ - .mmap = generic_file_mmap, /* Mmap file. */ - .open = ntfs_file_open, /* Open file. */ - .splice_read = generic_file_splice_read /* Zero-copy data send with - the data source being on - the ntfs partition. We do - not need to care about the - data destination. */ - /*.sendpage = ,*/ /* Zero-copy data send with - the data destination being - on the ntfs partition. We - do not need to care about - the data source. */ + .mmap = generic_file_mmap, + .open = ntfs_file_open, + .splice_read = generic_file_splice_read, }; const struct inode_operations ntfs_file_inode_ops = { diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 8490c64..460c6c3 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -502,7 +502,7 @@ static inline int ocfs2_writes_unwritten_extents(struct ocfs2_super *osb) static inline int ocfs2_supports_append_dio(struct ocfs2_super *osb) { - if (osb->s_feature_ro_compat & OCFS2_FEATURE_RO_COMPAT_APPEND_DIO) + if (osb->s_feature_incompat & OCFS2_FEATURE_INCOMPAT_APPEND_DIO) return 1; return 0; } diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h index 20e37a3..db64ce2 100644 --- a/fs/ocfs2/ocfs2_fs.h +++ b/fs/ocfs2/ocfs2_fs.h @@ -102,11 +102,11 @@ | OCFS2_FEATURE_INCOMPAT_INDEXED_DIRS \ | OCFS2_FEATURE_INCOMPAT_REFCOUNT_TREE \ | OCFS2_FEATURE_INCOMPAT_DISCONTIG_BG \ - | OCFS2_FEATURE_INCOMPAT_CLUSTERINFO) + | OCFS2_FEATURE_INCOMPAT_CLUSTERINFO \ + | OCFS2_FEATURE_INCOMPAT_APPEND_DIO) #define OCFS2_FEATURE_RO_COMPAT_SUPP (OCFS2_FEATURE_RO_COMPAT_UNWRITTEN \ | OCFS2_FEATURE_RO_COMPAT_USRQUOTA \ - | OCFS2_FEATURE_RO_COMPAT_GRPQUOTA \ - | OCFS2_FEATURE_RO_COMPAT_APPEND_DIO) + | OCFS2_FEATURE_RO_COMPAT_GRPQUOTA) /* * Heartbeat-only devices are missing journals and other files. The @@ -179,6 +179,11 @@ #define OCFS2_FEATURE_INCOMPAT_CLUSTERINFO 0x4000 /* + * Append Direct IO support + */ +#define OCFS2_FEATURE_INCOMPAT_APPEND_DIO 0x8000 + +/* * backup superblock flag is used to indicate that this volume * has backup superblocks. */ @@ -200,10 +205,6 @@ #define OCFS2_FEATURE_RO_COMPAT_USRQUOTA 0x0002 #define OCFS2_FEATURE_RO_COMPAT_GRPQUOTA 0x0004 -/* - * Append Direct IO support - */ -#define OCFS2_FEATURE_RO_COMPAT_APPEND_DIO 0x0008 /* The byte offset of the first backup block will be 1G. * The following will be 4G, 16G, 64G, 256G and 1T. @@ -570,6 +570,7 @@ static int chown_common(struct path *path, uid_t user, gid_t group) uid = make_kuid(current_user_ns(), user); gid = make_kgid(current_user_ns(), group); +retry_deleg: newattrs.ia_valid = ATTR_CTIME; if (user != (uid_t) -1) { if (!uid_valid(uid)) @@ -586,7 +587,6 @@ static int chown_common(struct path *path, uid_t user, gid_t group) if (!S_ISDIR(inode->i_mode)) newattrs.ia_valid |= ATTR_KILL_SUID | ATTR_KILL_SGID | ATTR_KILL_PRIV; -retry_deleg: mutex_lock(&inode->i_mutex); error = security_path_chown(path, uid, gid); if (!error) @@ -988,9 +988,6 @@ struct file *file_open_root(struct dentry *dentry, struct vfsmount *mnt, return ERR_PTR(err); if (flags & O_CREAT) return ERR_PTR(-EINVAL); - if (!filename && (flags & O_DIRECTORY)) - if (!dentry->d_inode->i_op->lookup) - return ERR_PTR(-ENOTDIR); return do_file_open_root(dentry, mnt, filename, &op); } EXPORT_SYMBOL(file_open_root); diff --git a/fs/overlayfs/copy_up.c b/fs/overlayfs/copy_up.c index ea10a87..24f6404 100644 --- a/fs/overlayfs/copy_up.c +++ b/fs/overlayfs/copy_up.c @@ -191,7 +191,6 @@ int ovl_set_attr(struct dentry *upperdentry, struct kstat *stat) ovl_set_timestamps(upperdentry, stat); return err; - } static int ovl_copy_up_locked(struct dentry *workdir, struct dentry *upperdir, @@ -385,7 +384,7 @@ int ovl_copy_up(struct dentry *dentry) struct kstat stat; enum ovl_path_type type = ovl_path_type(dentry); - if (type != OVL_PATH_LOWER) + if (OVL_TYPE_UPPER(type)) break; next = dget(dentry); @@ -394,7 +393,7 @@ int ovl_copy_up(struct dentry *dentry) parent = dget_parent(next); type = ovl_path_type(parent); - if (type != OVL_PATH_LOWER) + if (OVL_TYPE_UPPER(type)) break; dput(next); diff --git a/fs/overlayfs/dir.c b/fs/overlayfs/dir.c index 8ffc4b9..d139405 100644 --- a/fs/overlayfs/dir.c +++ b/fs/overlayfs/dir.c @@ -19,7 +19,7 @@ void ovl_cleanup(struct inode *wdir, struct dentry *wdentry) int err; dget(wdentry); - if (S_ISDIR(wdentry->d_inode->i_mode)) + if (d_is_dir(wdentry)) err = ovl_do_rmdir(wdir, wdentry); else err = ovl_do_unlink(wdir, wdentry); @@ -118,14 +118,14 @@ int ovl_create_real(struct inode *dir, struct dentry *newdentry, static int ovl_set_opaque(struct dentry *upperdentry) { - return ovl_do_setxattr(upperdentry, ovl_opaque_xattr, "y", 1, 0); + return ovl_do_setxattr(upperdentry, OVL_XATTR_OPAQUE, "y", 1, 0); } static void ovl_remove_opaque(struct dentry *upperdentry) { int err; - err = ovl_do_removexattr(upperdentry, ovl_opaque_xattr); + err = ovl_do_removexattr(upperdentry, OVL_XATTR_OPAQUE); if (err) { pr_warn("overlayfs: failed to remove opaque from '%s' (%i)\n", upperdentry->d_name.name, err); @@ -152,7 +152,7 @@ static int ovl_dir_getattr(struct vfsmount *mnt, struct dentry *dentry, * correct link count. nlink=1 seems to pacify 'find' and * other utilities. */ - if (type == OVL_PATH_MERGE) + if (OVL_TYPE_MERGE(type)) stat->nlink = 1; return 0; @@ -506,7 +506,7 @@ static int ovl_remove_and_whiteout(struct dentry *dentry, bool is_dir) struct dentry *opaquedir = NULL; int err; - if (is_dir) { + 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)) @@ -630,7 +630,7 @@ static int ovl_do_remove(struct dentry *dentry, bool is_dir) goto out_drop_write; type = ovl_path_type(dentry); - if (type == OVL_PATH_PURE_UPPER) { + if (OVL_TYPE_PURE_UPPER(type)) { err = ovl_remove_upper(dentry, is_dir); } else { const struct cred *old_cred; @@ -693,7 +693,7 @@ static int ovl_rename2(struct inode *olddir, struct dentry *old, bool new_create = false; bool cleanup_whiteout = false; bool overwrite = !(flags & RENAME_EXCHANGE); - bool is_dir = S_ISDIR(old->d_inode->i_mode); + bool is_dir = d_is_dir(old); bool new_is_dir = false; struct dentry *opaquedir = NULL; const struct cred *old_cred = NULL; @@ -712,7 +712,7 @@ static int ovl_rename2(struct inode *olddir, struct dentry *old, /* Don't copy up directory trees */ old_type = ovl_path_type(old); err = -EXDEV; - if ((old_type == OVL_PATH_LOWER || old_type == OVL_PATH_MERGE) && is_dir) + if (OVL_TYPE_MERGE_OR_LOWER(old_type) && is_dir) goto out; if (new->d_inode) { @@ -720,30 +720,30 @@ static int ovl_rename2(struct inode *olddir, struct dentry *old, if (err) goto out; - if (S_ISDIR(new->d_inode->i_mode)) + if (d_is_dir(new)) new_is_dir = true; new_type = ovl_path_type(new); err = -EXDEV; - if (!overwrite && (new_type == OVL_PATH_LOWER || new_type == OVL_PATH_MERGE) && new_is_dir) + if (!overwrite && OVL_TYPE_MERGE_OR_LOWER(new_type) && new_is_dir) goto out; err = 0; - if (new_type == OVL_PATH_LOWER && old_type == OVL_PATH_LOWER) { + if (!OVL_TYPE_UPPER(new_type) && !OVL_TYPE_UPPER(old_type)) { if (ovl_dentry_lower(old)->d_inode == ovl_dentry_lower(new)->d_inode) goto out; } - if (new_type != OVL_PATH_LOWER && old_type != OVL_PATH_LOWER) { + if (OVL_TYPE_UPPER(new_type) && OVL_TYPE_UPPER(old_type)) { if (ovl_dentry_upper(old)->d_inode == ovl_dentry_upper(new)->d_inode) goto out; } } else { if (ovl_dentry_is_opaque(new)) - new_type = OVL_PATH_UPPER; + new_type = __OVL_PATH_UPPER; else - new_type = OVL_PATH_PURE_UPPER; + new_type = __OVL_PATH_UPPER | __OVL_PATH_PURE; } err = ovl_want_write(old); @@ -763,8 +763,8 @@ static int ovl_rename2(struct inode *olddir, struct dentry *old, goto out_drop_write; } - old_opaque = old_type != OVL_PATH_PURE_UPPER; - new_opaque = new_type != OVL_PATH_PURE_UPPER; + old_opaque = !OVL_TYPE_PURE_UPPER(old_type); + new_opaque = !OVL_TYPE_PURE_UPPER(new_type); if (old_opaque || new_opaque) { err = -ENOMEM; @@ -787,7 +787,7 @@ static int ovl_rename2(struct inode *olddir, struct dentry *old, old_cred = override_creds(override_cred); } - if (overwrite && (new_type == OVL_PATH_LOWER || new_type == OVL_PATH_MERGE) && new_is_dir) { + if (overwrite && OVL_TYPE_MERGE_OR_LOWER(new_type) && new_is_dir) { opaquedir = ovl_check_empty_and_clear(new); err = PTR_ERR(opaquedir); if (IS_ERR(opaquedir)) { diff --git a/fs/overlayfs/inode.c b/fs/overlayfs/inode.c index 07d74b2..04f1248 100644 --- a/fs/overlayfs/inode.c +++ b/fs/overlayfs/inode.c @@ -205,7 +205,7 @@ static int ovl_readlink(struct dentry *dentry, char __user *buf, int bufsiz) static bool ovl_is_private_xattr(const char *name) { - return strncmp(name, "trusted.overlay.", 14) == 0; + return strncmp(name, OVL_XATTR_PRE_NAME, OVL_XATTR_PRE_LEN) == 0; } int ovl_setxattr(struct dentry *dentry, const char *name, @@ -238,7 +238,10 @@ out: static bool ovl_need_xattr_filter(struct dentry *dentry, enum ovl_path_type type) { - return type == OVL_PATH_UPPER && S_ISDIR(dentry->d_inode->i_mode); + if ((type & (__OVL_PATH_PURE | __OVL_PATH_UPPER)) == __OVL_PATH_UPPER) + return S_ISDIR(dentry->d_inode->i_mode); + else + return false; } ssize_t ovl_getxattr(struct dentry *dentry, const char *name, @@ -299,7 +302,7 @@ int ovl_removexattr(struct dentry *dentry, const char *name) if (ovl_need_xattr_filter(dentry, type) && ovl_is_private_xattr(name)) goto out_drop_write; - if (type == OVL_PATH_LOWER) { + if (!OVL_TYPE_UPPER(type)) { err = vfs_getxattr(realpath.dentry, name, NULL, 0); if (err < 0) goto out_drop_write; @@ -321,7 +324,7 @@ out: static bool ovl_open_need_copy_up(int flags, enum ovl_path_type type, struct dentry *realdentry) { - if (type != OVL_PATH_LOWER) + if (OVL_TYPE_UPPER(type)) return false; if (special_file(realdentry->d_inode->i_mode)) @@ -430,5 +433,4 @@ struct inode *ovl_new_inode(struct super_block *sb, umode_t mode, } return inode; - } diff --git a/fs/overlayfs/overlayfs.h b/fs/overlayfs/overlayfs.h index 814bed3..17ac5af 100644 --- a/fs/overlayfs/overlayfs.h +++ b/fs/overlayfs/overlayfs.h @@ -12,13 +12,20 @@ struct ovl_entry; enum ovl_path_type { - OVL_PATH_PURE_UPPER, - OVL_PATH_UPPER, - OVL_PATH_MERGE, - OVL_PATH_LOWER, + __OVL_PATH_PURE = (1 << 0), + __OVL_PATH_UPPER = (1 << 1), + __OVL_PATH_MERGE = (1 << 2), }; -extern const char *ovl_opaque_xattr; +#define OVL_TYPE_UPPER(type) ((type) & __OVL_PATH_UPPER) +#define OVL_TYPE_MERGE(type) ((type) & __OVL_PATH_MERGE) +#define OVL_TYPE_PURE_UPPER(type) ((type) & __OVL_PATH_PURE) +#define OVL_TYPE_MERGE_OR_LOWER(type) \ + (OVL_TYPE_MERGE(type) || !OVL_TYPE_UPPER(type)) + +#define OVL_XATTR_PRE_NAME "trusted.overlay." +#define OVL_XATTR_PRE_LEN 16 +#define OVL_XATTR_OPAQUE OVL_XATTR_PRE_NAME"opaque" static inline int ovl_do_rmdir(struct inode *dir, struct dentry *dentry) { @@ -130,6 +137,7 @@ void ovl_dentry_version_inc(struct dentry *dentry); void ovl_path_upper(struct dentry *dentry, struct path *path); void ovl_path_lower(struct dentry *dentry, struct path *path); enum ovl_path_type ovl_path_real(struct dentry *dentry, struct path *path); +int ovl_path_next(int idx, struct dentry *dentry, struct path *path); struct dentry *ovl_dentry_upper(struct dentry *dentry); struct dentry *ovl_dentry_lower(struct dentry *dentry); struct dentry *ovl_dentry_real(struct dentry *dentry); diff --git a/fs/overlayfs/readdir.c b/fs/overlayfs/readdir.c index c020599..907870e 100644 --- a/fs/overlayfs/readdir.c +++ b/fs/overlayfs/readdir.c @@ -24,7 +24,6 @@ struct ovl_cache_entry { struct list_head l_node; struct rb_node node; bool is_whiteout; - bool is_cursor; char name[]; }; @@ -40,6 +39,7 @@ struct ovl_readdir_data { struct rb_root root; struct list_head *list; struct list_head middle; + struct dentry *dir; int count; int err; }; @@ -48,7 +48,7 @@ struct ovl_dir_file { bool is_real; bool is_upper; struct ovl_dir_cache *cache; - struct ovl_cache_entry cursor; + struct list_head *cursor; struct file *realfile; struct file *upperfile; }; @@ -79,23 +79,49 @@ static struct ovl_cache_entry *ovl_cache_entry_find(struct rb_root *root, return NULL; } -static struct ovl_cache_entry *ovl_cache_entry_new(const char *name, int len, +static struct ovl_cache_entry *ovl_cache_entry_new(struct dentry *dir, + const char *name, int len, u64 ino, unsigned int d_type) { struct ovl_cache_entry *p; size_t size = offsetof(struct ovl_cache_entry, name[len + 1]); p = kmalloc(size, GFP_KERNEL); - if (p) { - memcpy(p->name, name, len); - p->name[len] = '\0'; - p->len = len; - p->type = d_type; - p->ino = ino; - p->is_whiteout = false; - p->is_cursor = false; - } + if (!p) + return NULL; + + memcpy(p->name, name, len); + p->name[len] = '\0'; + p->len = len; + p->type = d_type; + p->ino = ino; + 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); + } return p; } @@ -122,7 +148,7 @@ static int ovl_cache_entry_add_rb(struct ovl_readdir_data *rdd, return 0; } - p = ovl_cache_entry_new(name, len, ino, d_type); + p = ovl_cache_entry_new(rdd->dir, name, len, ino, d_type); if (p == NULL) return -ENOMEM; @@ -143,7 +169,7 @@ 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(name, namelen, ino, d_type); + p = ovl_cache_entry_new(rdd->dir, name, namelen, ino, d_type); if (p == NULL) rdd->err = -ENOMEM; else @@ -168,7 +194,6 @@ static void ovl_cache_put(struct ovl_dir_file *od, struct dentry *dentry) { struct ovl_dir_cache *cache = od->cache; - list_del_init(&od->cursor.l_node); WARN_ON(cache->refcount <= 0); cache->refcount--; if (!cache->refcount) { @@ -204,6 +229,7 @@ static inline int ovl_dir_read(struct path *realpath, if (IS_ERR(realfile)) return PTR_ERR(realfile); + rdd->dir = realpath->dentry; rdd->ctx.pos = 0; do { rdd->count = 0; @@ -227,108 +253,58 @@ static void ovl_dir_reset(struct file *file) if (cache && ovl_dentry_version_get(dentry) != cache->version) { ovl_cache_put(od, dentry); od->cache = NULL; + od->cursor = NULL; } - WARN_ON(!od->is_real && type != OVL_PATH_MERGE); - if (od->is_real && type == OVL_PATH_MERGE) + WARN_ON(!od->is_real && !OVL_TYPE_MERGE(type)); + if (od->is_real && OVL_TYPE_MERGE(type)) od->is_real = false; } -static int ovl_dir_mark_whiteouts(struct dentry *dir, - struct ovl_readdir_data *rdd) -{ - struct ovl_cache_entry *p; - struct dentry *dentry; - const struct cred *old_cred; - struct cred *override_cred; - - override_cred = prepare_creds(); - if (!override_cred) { - ovl_cache_free(rdd->list); - return -ENOMEM; - } - - /* - * CAP_DAC_OVERRIDE for lookup - */ - cap_raise(override_cred->cap_effective, CAP_DAC_OVERRIDE); - old_cred = override_creds(override_cred); - - mutex_lock(&dir->d_inode->i_mutex); - list_for_each_entry(p, rdd->list, l_node) { - if (p->is_cursor) - continue; - - if (p->type != DT_CHR) - continue; - - dentry = lookup_one_len(p->name, dir, p->len); - if (IS_ERR(dentry)) - continue; - - 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 0; -} - static int ovl_dir_read_merged(struct dentry *dentry, struct list_head *list) { int err; - struct path lowerpath; - struct path upperpath; + struct path realpath; struct ovl_readdir_data rdd = { .ctx.actor = ovl_fill_merge, .list = list, .root = RB_ROOT, .is_merge = false, }; + int idx, next; - ovl_path_lower(dentry, &lowerpath); - ovl_path_upper(dentry, &upperpath); + for (idx = 0; idx != -1; idx = next) { + next = ovl_path_next(idx, dentry, &realpath); - if (upperpath.dentry) { - err = ovl_dir_read(&upperpath, &rdd); - if (err) - goto out; - - if (lowerpath.dentry) { - err = ovl_dir_mark_whiteouts(upperpath.dentry, &rdd); + if (next != -1) { + err = ovl_dir_read(&realpath, &rdd); if (err) - goto out; + break; + } else { + /* + * Insert lowest layer entries before upper ones, this + * allows offsets to be reasonably constant + */ + list_add(&rdd.middle, rdd.list); + rdd.is_merge = true; + err = ovl_dir_read(&realpath, &rdd); + list_del(&rdd.middle); } } - if (lowerpath.dentry) { - /* - * Insert lowerpath entries before upperpath ones, this allows - * offsets to be reasonably constant - */ - list_add(&rdd.middle, rdd.list); - rdd.is_merge = true; - err = ovl_dir_read(&lowerpath, &rdd); - list_del(&rdd.middle); - } -out: return err; } static void ovl_seek_cursor(struct ovl_dir_file *od, loff_t pos) { - struct ovl_cache_entry *p; + struct list_head *p; loff_t off = 0; - list_for_each_entry(p, &od->cache->entries, l_node) { - if (p->is_cursor) - continue; + list_for_each(p, &od->cache->entries) { if (off >= pos) break; off++; } - list_move_tail(&od->cursor.l_node, &p->l_node); + /* Cursor is safe since the cache is stable */ + od->cursor = p; } static struct ovl_dir_cache *ovl_cache_get(struct dentry *dentry) @@ -367,6 +343,7 @@ static int ovl_iterate(struct file *file, struct dir_context *ctx) { struct ovl_dir_file *od = file->private_data; struct dentry *dentry = file->f_path.dentry; + struct ovl_cache_entry *p; if (!ctx->pos) ovl_dir_reset(file); @@ -385,19 +362,13 @@ static int ovl_iterate(struct file *file, struct dir_context *ctx) ovl_seek_cursor(od, ctx->pos); } - while (od->cursor.l_node.next != &od->cache->entries) { - struct ovl_cache_entry *p; - - p = list_entry(od->cursor.l_node.next, struct ovl_cache_entry, l_node); - /* Skip cursors */ - if (!p->is_cursor) { - if (!p->is_whiteout) { - if (!dir_emit(ctx, p->name, p->len, p->ino, p->type)) - break; - } - ctx->pos++; - } - list_move(&od->cursor.l_node, &p->l_node); + while (od->cursor != &od->cache->entries) { + p = list_entry(od->cursor, struct ovl_cache_entry, l_node); + if (!p->is_whiteout) + if (!dir_emit(ctx, p->name, p->len, p->ino, p->type)) + break; + od->cursor = p->l_node.next; + ctx->pos++; } return 0; } @@ -452,7 +423,7 @@ static int ovl_dir_fsync(struct file *file, loff_t start, loff_t end, /* * Need to check if we started out being a lower dir, but got copied up */ - if (!od->is_upper && ovl_path_type(dentry) != OVL_PATH_LOWER) { + if (!od->is_upper && OVL_TYPE_UPPER(ovl_path_type(dentry))) { struct inode *inode = file_inode(file); realfile = lockless_dereference(od->upperfile); @@ -516,11 +487,9 @@ static int ovl_dir_open(struct inode *inode, struct file *file) kfree(od); return PTR_ERR(realfile); } - INIT_LIST_HEAD(&od->cursor.l_node); od->realfile = realfile; - od->is_real = (type != OVL_PATH_MERGE); - od->is_upper = (type != OVL_PATH_LOWER); - od->cursor.is_cursor = true; + od->is_real = !OVL_TYPE_MERGE(type); + od->is_upper = OVL_TYPE_UPPER(type); file->private_data = od; return 0; diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index f16d318..5f0d199 100644 --- a/fs/overlayfs/super.c +++ b/fs/overlayfs/super.c @@ -35,7 +35,8 @@ struct ovl_config { /* private information held for overlayfs's superblock */ struct ovl_fs { struct vfsmount *upper_mnt; - struct vfsmount *lower_mnt; + unsigned numlower; + struct vfsmount **lower_mnt; struct dentry *workdir; long lower_namelen; /* pathnames of lower and upper dirs, for show_options */ @@ -47,7 +48,6 @@ struct ovl_dir_cache; /* private information held for every overlayfs dentry */ struct ovl_entry { struct dentry *__upperdentry; - struct dentry *lowerdentry; struct ovl_dir_cache *cache; union { struct { @@ -56,30 +56,36 @@ struct ovl_entry { }; struct rcu_head rcu; }; + unsigned numlower; + struct path lowerstack[]; }; -const char *ovl_opaque_xattr = "trusted.overlay.opaque"; +#define OVL_MAX_STACK 500 +static struct dentry *__ovl_dentry_lower(struct ovl_entry *oe) +{ + return oe->numlower ? oe->lowerstack[0].dentry : NULL; +} enum ovl_path_type ovl_path_type(struct dentry *dentry) { struct ovl_entry *oe = dentry->d_fsdata; + enum ovl_path_type type = 0; if (oe->__upperdentry) { - if (oe->lowerdentry) { + type = __OVL_PATH_UPPER; + + if (oe->numlower) { if (S_ISDIR(dentry->d_inode->i_mode)) - return OVL_PATH_MERGE; - else - return OVL_PATH_UPPER; - } else { - if (oe->opaque) - return OVL_PATH_UPPER; - else - return OVL_PATH_PURE_UPPER; + type |= __OVL_PATH_MERGE; + } else if (!oe->opaque) { + type |= __OVL_PATH_PURE; } } else { - return OVL_PATH_LOWER; + if (oe->numlower > 1) + type |= __OVL_PATH_MERGE; } + return type; } static struct dentry *ovl_upperdentry_dereference(struct ovl_entry *oe) @@ -98,10 +104,9 @@ void ovl_path_upper(struct dentry *dentry, struct path *path) enum ovl_path_type ovl_path_real(struct dentry *dentry, struct path *path) { - enum ovl_path_type type = ovl_path_type(dentry); - if (type == OVL_PATH_LOWER) + if (!OVL_TYPE_UPPER(type)) ovl_path_lower(dentry, path); else ovl_path_upper(dentry, path); @@ -120,7 +125,7 @@ struct dentry *ovl_dentry_lower(struct dentry *dentry) { struct ovl_entry *oe = dentry->d_fsdata; - return oe->lowerdentry; + return __ovl_dentry_lower(oe); } struct dentry *ovl_dentry_real(struct dentry *dentry) @@ -130,7 +135,7 @@ struct dentry *ovl_dentry_real(struct dentry *dentry) realdentry = ovl_upperdentry_dereference(oe); if (!realdentry) - realdentry = oe->lowerdentry; + realdentry = __ovl_dentry_lower(oe); return realdentry; } @@ -143,7 +148,7 @@ struct dentry *ovl_entry_real(struct ovl_entry *oe, bool *is_upper) if (realdentry) { *is_upper = true; } else { - realdentry = oe->lowerdentry; + realdentry = __ovl_dentry_lower(oe); *is_upper = false; } return realdentry; @@ -165,11 +170,9 @@ void ovl_set_dir_cache(struct dentry *dentry, struct ovl_dir_cache *cache) void ovl_path_lower(struct dentry *dentry, struct path *path) { - struct ovl_fs *ofs = dentry->d_sb->s_fs_info; struct ovl_entry *oe = dentry->d_fsdata; - path->mnt = ofs->lower_mnt; - path->dentry = oe->lowerdentry; + *path = oe->numlower ? oe->lowerstack[0] : (struct path) { NULL, NULL }; } int ovl_want_write(struct dentry *dentry) @@ -249,7 +252,7 @@ static bool ovl_is_opaquedir(struct dentry *dentry) if (!S_ISDIR(inode->i_mode) || !inode->i_op->getxattr) return false; - res = inode->i_op->getxattr(dentry, ovl_opaque_xattr, &val, 1); + res = inode->i_op->getxattr(dentry, OVL_XATTR_OPAQUE, &val, 1); if (res == 1 && val == 'y') return true; @@ -261,8 +264,11 @@ static void ovl_dentry_release(struct dentry *dentry) struct ovl_entry *oe = dentry->d_fsdata; if (oe) { + unsigned int i; + dput(oe->__upperdentry); - dput(oe->lowerdentry); + for (i = 0; i < oe->numlower; i++) + dput(oe->lowerstack[i].dentry); kfree_rcu(oe, rcu); } } @@ -271,9 +277,15 @@ static const struct dentry_operations ovl_dentry_operations = { .d_release = ovl_dentry_release, }; -static struct ovl_entry *ovl_alloc_entry(void) +static struct ovl_entry *ovl_alloc_entry(unsigned int numlower) { - return kzalloc(sizeof(struct ovl_entry), GFP_KERNEL); + size_t size = offsetof(struct ovl_entry, lowerstack[numlower]); + struct ovl_entry *oe = kzalloc(size, GFP_KERNEL); + + if (oe) + oe->numlower = numlower; + + return oe; } static inline struct dentry *ovl_lookup_real(struct dentry *dir, @@ -295,82 +307,154 @@ static inline struct dentry *ovl_lookup_real(struct dentry *dir, return dentry; } +/* + * Returns next layer in stack starting from top. + * Returns -1 if this is the last layer. + */ +int ovl_path_next(int idx, struct dentry *dentry, struct path *path) +{ + struct ovl_entry *oe = dentry->d_fsdata; + + BUG_ON(idx < 0); + if (idx == 0) { + ovl_path_upper(dentry, path); + if (path->dentry) + return oe->numlower ? 1 : -1; + idx++; + } + BUG_ON(idx > oe->numlower); + *path = oe->lowerstack[idx - 1]; + + return (idx < oe->numlower) ? idx + 1 : -1; +} + struct dentry *ovl_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) { struct ovl_entry *oe; - struct dentry *upperdir; - struct dentry *lowerdir; - struct dentry *upperdentry = NULL; - struct dentry *lowerdentry = NULL; + struct ovl_entry *poe = dentry->d_parent->d_fsdata; + struct path *stack = NULL; + struct dentry *upperdir, *upperdentry = NULL; + unsigned int ctr = 0; struct inode *inode = NULL; + bool upperopaque = false; + struct dentry *this, *prev = NULL; + unsigned int i; int err; - err = -ENOMEM; - oe = ovl_alloc_entry(); - if (!oe) - goto out; - - upperdir = ovl_dentry_upper(dentry->d_parent); - lowerdir = ovl_dentry_lower(dentry->d_parent); - + upperdir = ovl_upperdentry_dereference(poe); if (upperdir) { - upperdentry = ovl_lookup_real(upperdir, &dentry->d_name); - err = PTR_ERR(upperdentry); - if (IS_ERR(upperdentry)) - goto out_put_dir; - - if (lowerdir && upperdentry) { - if (ovl_is_whiteout(upperdentry)) { - dput(upperdentry); - upperdentry = NULL; - oe->opaque = true; - } else if (ovl_is_opaquedir(upperdentry)) { - oe->opaque = true; + this = ovl_lookup_real(upperdir, &dentry->d_name); + err = PTR_ERR(this); + if (IS_ERR(this)) + goto out; + + if (this) { + if (ovl_is_whiteout(this)) { + dput(this); + this = NULL; + upperopaque = true; + } else if (poe->numlower && ovl_is_opaquedir(this)) { + upperopaque = true; } } + upperdentry = prev = this; } - if (lowerdir && !oe->opaque) { - lowerdentry = ovl_lookup_real(lowerdir, &dentry->d_name); - err = PTR_ERR(lowerdentry); - if (IS_ERR(lowerdentry)) - goto out_dput_upper; + + if (!upperopaque && poe->numlower) { + err = -ENOMEM; + stack = kcalloc(poe->numlower, sizeof(struct path), GFP_KERNEL); + if (!stack) + goto out_put_upper; } - if (lowerdentry && upperdentry && - (!S_ISDIR(upperdentry->d_inode->i_mode) || - !S_ISDIR(lowerdentry->d_inode->i_mode))) { - dput(lowerdentry); - lowerdentry = NULL; - oe->opaque = true; + for (i = 0; !upperopaque && i < poe->numlower; i++) { + bool opaque = false; + struct path lowerpath = poe->lowerstack[i]; + + this = ovl_lookup_real(lowerpath.dentry, &dentry->d_name); + err = PTR_ERR(this); + if (IS_ERR(this)) { + /* + * If it's positive, then treat ENAMETOOLONG as ENOENT. + */ + if (err == -ENAMETOOLONG && (upperdentry || ctr)) + continue; + goto out_put; + } + if (!this) + continue; + if (ovl_is_whiteout(this)) { + dput(this); + break; + } + /* + * Only makes sense to check opaque dir if this is not the + * lowermost layer. + */ + if (i < poe->numlower - 1 && ovl_is_opaquedir(this)) + opaque = true; + + if (prev && (!S_ISDIR(prev->d_inode->i_mode) || + !S_ISDIR(this->d_inode->i_mode))) { + /* + * FIXME: check for upper-opaqueness maybe better done + * in remove code. + */ + if (prev == upperdentry) + upperopaque = true; + dput(this); + break; + } + /* + * If this is a non-directory then stop here. + */ + if (!S_ISDIR(this->d_inode->i_mode)) + opaque = true; + + stack[ctr].dentry = this; + stack[ctr].mnt = lowerpath.mnt; + ctr++; + prev = this; + if (opaque) + break; } - if (lowerdentry || upperdentry) { + oe = ovl_alloc_entry(ctr); + err = -ENOMEM; + if (!oe) + goto out_put; + + if (upperdentry || ctr) { struct dentry *realdentry; - realdentry = upperdentry ? upperdentry : lowerdentry; + realdentry = upperdentry ? upperdentry : stack[0].dentry; + err = -ENOMEM; inode = ovl_new_inode(dentry->d_sb, realdentry->d_inode->i_mode, oe); if (!inode) - goto out_dput; + goto out_free_oe; ovl_copyattr(realdentry->d_inode, inode); } + oe->opaque = upperopaque; oe->__upperdentry = upperdentry; - oe->lowerdentry = lowerdentry; - + memcpy(oe->lowerstack, stack, sizeof(struct path) * ctr); + kfree(stack); dentry->d_fsdata = oe; d_add(dentry, inode); return NULL; -out_dput: - dput(lowerdentry); -out_dput_upper: - dput(upperdentry); -out_put_dir: +out_free_oe: kfree(oe); +out_put: + for (i = 0; i < ctr; i++) + dput(stack[i].dentry); + kfree(stack); +out_put_upper: + dput(upperdentry); out: return ERR_PTR(err); } @@ -383,10 +467,12 @@ struct file *ovl_path_open(struct path *path, int flags) static void ovl_put_super(struct super_block *sb) { struct ovl_fs *ufs = sb->s_fs_info; + unsigned i; dput(ufs->workdir); mntput(ufs->upper_mnt); - mntput(ufs->lower_mnt); + for (i = 0; i < ufs->numlower; i++) + mntput(ufs->lower_mnt[i]); kfree(ufs->config.lowerdir); kfree(ufs->config.upperdir); @@ -400,7 +486,7 @@ static void ovl_put_super(struct super_block *sb) * @buf: The struct kstatfs to fill in with stats * * Get the filesystem statistics. As writes always target the upper layer - * filesystem pass the statfs to the same filesystem. + * filesystem pass the statfs to the upper filesystem (if it exists) */ static int ovl_statfs(struct dentry *dentry, struct kstatfs *buf) { @@ -409,7 +495,7 @@ static int ovl_statfs(struct dentry *dentry, struct kstatfs *buf) struct path path; int err; - ovl_path_upper(root_dentry, &path); + ovl_path_real(root_dentry, &path); err = vfs_statfs(&path, buf); if (!err) { @@ -432,8 +518,20 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry) struct ovl_fs *ufs = sb->s_fs_info; seq_printf(m, ",lowerdir=%s", ufs->config.lowerdir); - seq_printf(m, ",upperdir=%s", ufs->config.upperdir); - seq_printf(m, ",workdir=%s", ufs->config.workdir); + if (ufs->config.upperdir) { + seq_printf(m, ",upperdir=%s", ufs->config.upperdir); + seq_printf(m, ",workdir=%s", ufs->config.workdir); + } + return 0; +} + +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) + return -EROFS; + return 0; } @@ -441,6 +539,7 @@ static const struct super_operations ovl_super_operations = { .put_super = ovl_put_super, .statfs = ovl_statfs, .show_options = ovl_show_options, + .remount_fs = ovl_remount, }; enum { @@ -515,9 +614,19 @@ static int ovl_parse_opt(char *opt, struct ovl_config *config) break; default: + pr_err("overlayfs: unrecognized mount option \"%s\" or missing value\n", p); return -EINVAL; } } + + /* Workdir is useless in non-upper mount */ + if (!config->upperdir && config->workdir) { + pr_info("overlayfs: option \"workdir=%s\" is useless in a non-upper mount, ignore\n", + config->workdir); + kfree(config->workdir); + config->workdir = NULL; + } + return 0; } @@ -585,24 +694,6 @@ static void ovl_unescape(char *s) } } -static int ovl_mount_dir(const char *name, struct path *path) -{ - int err; - char *tmp = kstrdup(name, GFP_KERNEL); - - if (!tmp) - return -ENOMEM; - - ovl_unescape(tmp); - err = kern_path(tmp, LOOKUP_FOLLOW, path); - if (err) { - pr_err("overlayfs: failed to resolve '%s': %i\n", tmp, err); - err = -EINVAL; - } - kfree(tmp); - return err; -} - static bool ovl_is_allowed_fs_type(struct dentry *root) { const struct dentry_operations *dop = root->d_op; @@ -622,6 +713,75 @@ static bool ovl_is_allowed_fs_type(struct dentry *root) return true; } +static int ovl_mount_dir_noesc(const char *name, struct path *path) +{ + int err = -EINVAL; + + if (!*name) { + pr_err("overlayfs: empty lowerdir\n"); + goto out; + } + err = kern_path(name, LOOKUP_FOLLOW, path); + if (err) { + pr_err("overlayfs: failed to resolve '%s': %i\n", name, err); + goto out; + } + err = -EINVAL; + if (!ovl_is_allowed_fs_type(path->dentry)) { + pr_err("overlayfs: filesystem on '%s' not supported\n", name); + goto out_put; + } + if (!S_ISDIR(path->dentry->d_inode->i_mode)) { + pr_err("overlayfs: '%s' not a directory\n", name); + goto out_put; + } + return 0; + +out_put: + path_put(path); +out: + return err; +} + +static int ovl_mount_dir(const char *name, struct path *path) +{ + int err = -ENOMEM; + char *tmp = kstrdup(name, GFP_KERNEL); + + if (tmp) { + ovl_unescape(tmp); + err = ovl_mount_dir_noesc(tmp, path); + kfree(tmp); + } + return err; +} + +static int ovl_lower_dir(const char *name, struct path *path, long *namelen, + int *stack_depth) +{ + int err; + struct kstatfs statfs; + + err = ovl_mount_dir_noesc(name, path); + if (err) + goto out; + + err = vfs_statfs(path, &statfs); + if (err) { + pr_err("overlayfs: statfs failed on '%s'\n", name); + goto out_put; + } + *namelen = max(*namelen, statfs.f_namelen); + *stack_depth = max(*stack_depth, path->mnt->mnt_sb->s_stack_depth); + + return 0; + +out_put: + path_put(path); +out: + return err; +} + /* Workdir should not be subdir of upperdir and vice versa */ static bool ovl_workdir_ok(struct dentry *workdir, struct dentry *upperdir) { @@ -634,16 +794,39 @@ static bool ovl_workdir_ok(struct dentry *workdir, struct dentry *upperdir) return ok; } +static unsigned int ovl_split_lowerdirs(char *str) +{ + unsigned int ctr = 1; + char *s, *d; + + for (s = d = str;; s++, d++) { + if (*s == '\\') { + s++; + } else if (*s == ':') { + *d = '\0'; + ctr++; + continue; + } + *d = *s; + if (!*s) + break; + } + return ctr; +} + static int ovl_fill_super(struct super_block *sb, void *data, int silent) { - struct path lowerpath; - struct path upperpath; - struct path workpath; - struct inode *root_inode; + struct path upperpath = { NULL, NULL }; + struct path workpath = { NULL, NULL }; struct dentry *root_dentry; struct ovl_entry *oe; struct ovl_fs *ufs; - struct kstatfs statfs; + struct path *stack = NULL; + char *lowertmp; + char *lower; + unsigned int numlower; + unsigned int stacklen = 0; + unsigned int i; int err; err = -ENOMEM; @@ -655,123 +838,147 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) if (err) goto out_free_config; - /* FIXME: workdir is not needed for a R/O mount */ err = -EINVAL; - if (!ufs->config.upperdir || !ufs->config.lowerdir || - !ufs->config.workdir) { - pr_err("overlayfs: missing upperdir or lowerdir or workdir\n"); + if (!ufs->config.lowerdir) { + pr_err("overlayfs: missing 'lowerdir'\n"); goto out_free_config; } - err = -ENOMEM; - oe = ovl_alloc_entry(); - if (oe == NULL) - goto out_free_config; - - err = ovl_mount_dir(ufs->config.upperdir, &upperpath); - if (err) - goto out_free_oe; + sb->s_stack_depth = 0; + if (ufs->config.upperdir) { + if (!ufs->config.workdir) { + pr_err("overlayfs: missing 'workdir'\n"); + goto out_free_config; + } - err = ovl_mount_dir(ufs->config.lowerdir, &lowerpath); - if (err) - goto out_put_upperpath; + err = ovl_mount_dir(ufs->config.upperdir, &upperpath); + if (err) + goto out_free_config; - err = ovl_mount_dir(ufs->config.workdir, &workpath); - if (err) - goto out_put_lowerpath; + /* Upper fs should not be r/o */ + if (upperpath.mnt->mnt_sb->s_flags & MS_RDONLY) { + pr_err("overlayfs: upper fs is r/o, try multi-lower layers mount\n"); + err = -EINVAL; + goto out_put_upperpath; + } - err = -EINVAL; - if (!S_ISDIR(upperpath.dentry->d_inode->i_mode) || - !S_ISDIR(lowerpath.dentry->d_inode->i_mode) || - !S_ISDIR(workpath.dentry->d_inode->i_mode)) { - pr_err("overlayfs: upperdir or lowerdir or workdir not a directory\n"); - goto out_put_workpath; - } + err = ovl_mount_dir(ufs->config.workdir, &workpath); + if (err) + goto out_put_upperpath; - if (upperpath.mnt != workpath.mnt) { - pr_err("overlayfs: workdir and upperdir must reside under the same mount\n"); - goto out_put_workpath; + err = -EINVAL; + if (upperpath.mnt != workpath.mnt) { + pr_err("overlayfs: workdir and upperdir must reside under the same mount\n"); + goto out_put_workpath; + } + if (!ovl_workdir_ok(workpath.dentry, upperpath.dentry)) { + pr_err("overlayfs: workdir and upperdir must be separate subtrees\n"); + goto out_put_workpath; + } + sb->s_stack_depth = upperpath.mnt->mnt_sb->s_stack_depth; } - if (!ovl_workdir_ok(workpath.dentry, upperpath.dentry)) { - pr_err("overlayfs: workdir and upperdir must be separate subtrees\n"); + err = -ENOMEM; + lowertmp = kstrdup(ufs->config.lowerdir, GFP_KERNEL); + if (!lowertmp) goto out_put_workpath; - } - if (!ovl_is_allowed_fs_type(upperpath.dentry)) { - pr_err("overlayfs: filesystem of upperdir is not supported\n"); - goto out_put_workpath; + err = -EINVAL; + stacklen = ovl_split_lowerdirs(lowertmp); + if (stacklen > OVL_MAX_STACK) { + pr_err("overlayfs: too many lower directries, limit is %d\n", + OVL_MAX_STACK); + goto out_free_lowertmp; + } else if (!ufs->config.upperdir && stacklen == 1) { + pr_err("overlayfs: at least 2 lowerdir are needed while upperdir nonexistent\n"); + goto out_free_lowertmp; } - if (!ovl_is_allowed_fs_type(lowerpath.dentry)) { - pr_err("overlayfs: filesystem of lowerdir is not supported\n"); - goto out_put_workpath; - } + stack = kcalloc(stacklen, sizeof(struct path), GFP_KERNEL); + if (!stack) + goto out_free_lowertmp; - err = vfs_statfs(&lowerpath, &statfs); - if (err) { - pr_err("overlayfs: statfs failed on lowerpath\n"); - goto out_put_workpath; - } - ufs->lower_namelen = statfs.f_namelen; + lower = lowertmp; + for (numlower = 0; numlower < stacklen; numlower++) { + err = ovl_lower_dir(lower, &stack[numlower], + &ufs->lower_namelen, &sb->s_stack_depth); + if (err) + goto out_put_lowerpath; - sb->s_stack_depth = max(upperpath.mnt->mnt_sb->s_stack_depth, - lowerpath.mnt->mnt_sb->s_stack_depth) + 1; + lower = strchr(lower, '\0') + 1; + } err = -EINVAL; + sb->s_stack_depth++; if (sb->s_stack_depth > FILESYSTEM_MAX_STACK_DEPTH) { pr_err("overlayfs: maximum fs stacking depth exceeded\n"); - goto out_put_workpath; + goto out_put_lowerpath; } - ufs->upper_mnt = clone_private_mount(&upperpath); - err = PTR_ERR(ufs->upper_mnt); - if (IS_ERR(ufs->upper_mnt)) { - pr_err("overlayfs: failed to clone upperpath\n"); - goto out_put_workpath; - } + if (ufs->config.upperdir) { + ufs->upper_mnt = clone_private_mount(&upperpath); + err = PTR_ERR(ufs->upper_mnt); + if (IS_ERR(ufs->upper_mnt)) { + pr_err("overlayfs: failed to clone upperpath\n"); + goto out_put_lowerpath; + } - ufs->lower_mnt = clone_private_mount(&lowerpath); - err = PTR_ERR(ufs->lower_mnt); - if (IS_ERR(ufs->lower_mnt)) { - pr_err("overlayfs: failed to clone lowerpath\n"); - goto out_put_upper_mnt; + 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; + } } - 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_lower_mnt; - } + err = -ENOMEM; + ufs->lower_mnt = kcalloc(numlower, sizeof(struct vfsmount *), GFP_KERNEL); + if (ufs->lower_mnt == NULL) + goto out_put_workdir; + for (i = 0; i < numlower; i++) { + struct vfsmount *mnt = clone_private_mount(&stack[i]); - /* - * Make lower_mnt R/O. That way fchmod/fchown on lower file - * will fail instead of modifying lower fs. - */ - ufs->lower_mnt->mnt_flags |= MNT_READONLY; + err = PTR_ERR(mnt); + if (IS_ERR(mnt)) { + pr_err("overlayfs: failed to clone lowerpath\n"); + goto out_put_lower_mnt; + } + /* + * Make lower_mnt R/O. That way fchmod/fchown on lower file + * will fail instead of modifying lower fs. + */ + mnt->mnt_flags |= MNT_READONLY; + + ufs->lower_mnt[ufs->numlower] = mnt; + ufs->numlower++; + } - /* If the upper fs is r/o, we mark overlayfs r/o too */ - if (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY) + /* If the upper fs is nonexistent, we mark overlayfs r/o too */ + if (!ufs->upper_mnt) sb->s_flags |= MS_RDONLY; sb->s_d_op = &ovl_dentry_operations; err = -ENOMEM; - root_inode = ovl_new_inode(sb, S_IFDIR, oe); - if (!root_inode) - goto out_put_workdir; + oe = ovl_alloc_entry(numlower); + if (!oe) + goto out_put_lower_mnt; - root_dentry = d_make_root(root_inode); + root_dentry = d_make_root(ovl_new_inode(sb, S_IFDIR, oe)); if (!root_dentry) - goto out_put_workdir; + goto out_free_oe; mntput(upperpath.mnt); - mntput(lowerpath.mnt); + for (i = 0; i < numlower; i++) + mntput(stack[i].mnt); path_put(&workpath); + kfree(lowertmp); oe->__upperdentry = upperpath.dentry; - oe->lowerdentry = lowerpath.dentry; + for (i = 0; i < numlower; i++) { + oe->lowerstack[i].dentry = stack[i].dentry; + oe->lowerstack[i].mnt = ufs->lower_mnt[i]; + } root_dentry->d_fsdata = oe; @@ -782,20 +989,26 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) return 0; +out_free_oe: + kfree(oe); +out_put_lower_mnt: + for (i = 0; i < ufs->numlower; i++) + mntput(ufs->lower_mnt[i]); + kfree(ufs->lower_mnt); out_put_workdir: dput(ufs->workdir); -out_put_lower_mnt: - mntput(ufs->lower_mnt); out_put_upper_mnt: mntput(ufs->upper_mnt); +out_put_lowerpath: + for (i = 0; i < numlower; i++) + path_put(&stack[i]); + kfree(stack); +out_free_lowertmp: + kfree(lowertmp); out_put_workpath: path_put(&workpath); -out_put_lowerpath: - path_put(&lowerpath); out_put_upperpath: path_put(&upperpath); -out_free_oe: - kfree(oe); out_free_config: kfree(ufs->config.lowerdir); kfree(ufs->config.upperdir); diff --git a/fs/posix_acl.c b/fs/posix_acl.c index 0855f77..3a48bb7 100644 --- a/fs/posix_acl.c +++ b/fs/posix_acl.c @@ -564,13 +564,11 @@ posix_acl_create(struct inode *dir, umode_t *mode, *acl = posix_acl_clone(p, GFP_NOFS); if (!*acl) - return -ENOMEM; + goto no_mem; ret = posix_acl_create_masq(*acl, mode); - if (ret < 0) { - posix_acl_release(*acl); - return -ENOMEM; - } + if (ret < 0) + goto no_mem_clone; if (ret == 0) { posix_acl_release(*acl); @@ -591,6 +589,12 @@ no_acl: *default_acl = NULL; *acl = NULL; return 0; + +no_mem_clone: + posix_acl_release(*acl); +no_mem: + posix_acl_release(p); + return -ENOMEM; } EXPORT_SYMBOL_GPL(posix_acl_create); @@ -772,7 +776,7 @@ posix_acl_xattr_get(struct dentry *dentry, const char *name, if (!IS_POSIXACL(dentry->d_inode)) return -EOPNOTSUPP; - if (S_ISLNK(dentry->d_inode->i_mode)) + if (d_is_symlink(dentry)) return -EOPNOTSUPP; acl = get_acl(dentry->d_inode, type); @@ -832,7 +836,7 @@ posix_acl_xattr_list(struct dentry *dentry, char *list, size_t list_size, if (!IS_POSIXACL(dentry->d_inode)) return -EOPNOTSUPP; - if (S_ISLNK(dentry->d_inode->i_mode)) + if (d_is_symlink(dentry)) return -EOPNOTSUPP; if (type == ACL_TYPE_ACCESS) diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 3309f59..be65b20 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -19,7 +19,6 @@ #include <linux/mount.h> #include <linux/init.h> #include <linux/idr.h> -#include <linux/namei.h> #include <linux/bitops.h> #include <linux/spinlock.h> #include <linux/completion.h> @@ -223,17 +222,6 @@ void proc_free_inum(unsigned int inum) spin_unlock_irqrestore(&proc_inum_lock, flags); } -static void *proc_follow_link(struct dentry *dentry, struct nameidata *nd) -{ - nd_set_link(nd, __PDE_DATA(dentry->d_inode)); - return NULL; -} - -static const struct inode_operations proc_link_inode_operations = { - .readlink = generic_readlink, - .follow_link = proc_follow_link, -}; - /* * Don't create negative dentries here, return -ENOENT by hand * instead. diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 13a50a3..7697b66 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -23,6 +23,7 @@ #include <linux/slab.h> #include <linux/mount.h> #include <linux/magic.h> +#include <linux/namei.h> #include <asm/uaccess.h> @@ -393,6 +394,26 @@ static const struct file_operations proc_reg_file_ops_no_compat = { }; #endif +static void *proc_follow_link(struct dentry *dentry, struct nameidata *nd) +{ + struct proc_dir_entry *pde = PDE(dentry->d_inode); + if (unlikely(!use_pde(pde))) + return ERR_PTR(-EINVAL); + nd_set_link(nd, pde->data); + return pde; +} + +static void proc_put_link(struct dentry *dentry, struct nameidata *nd, void *p) +{ + unuse_pde(p); +} + +const struct inode_operations proc_link_inode_operations = { + .readlink = generic_readlink, + .follow_link = proc_follow_link, + .put_link = proc_put_link, +}; + struct inode *proc_get_inode(struct super_block *sb, struct proc_dir_entry *de) { struct inode *inode = new_inode_pseudo(sb); diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 6fcdba5..c835b94 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -200,6 +200,7 @@ struct pde_opener { int closing; struct completion *c; }; +extern const struct inode_operations proc_link_inode_operations; extern const struct inode_operations proc_pid_link_inode_operations; diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 956b75d..6dee68d 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -1325,6 +1325,9 @@ out: static int pagemap_open(struct inode *inode, struct file *file) { + /* do not disclose physical addresses: attack vector */ + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; pr_warn_once("Bits 55-60 of /proc/PID/pagemap entries are about " "to stop being page-shift some time soon. See the " "linux/Documentation/vm/pagemap.txt for details.\n"); diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index 04b0614..4e781e6 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -266,7 +266,7 @@ static int reiserfs_for_each_xattr(struct inode *inode, for (i = 0; !err && i < buf.count && buf.dentries[i]; i++) { struct dentry *dentry = buf.dentries[i]; - if (!S_ISDIR(dentry->d_inode->i_mode)) + if (!d_is_dir(dentry)) err = action(dentry, data); dput(dentry); @@ -322,7 +322,7 @@ static int delete_one_xattr(struct dentry *dentry, void *data) struct inode *dir = dentry->d_parent->d_inode; /* This is the xattr dir, handle specially. */ - if (S_ISDIR(dentry->d_inode->i_mode)) + if (d_is_dir(dentry)) return xattr_rmdir(dir, dentry); return xattr_unlink(dir, dentry); @@ -66,7 +66,7 @@ int vfs_getattr(struct path *path, struct kstat *stat) { int retval; - retval = security_inode_getattr(path->mnt, path->dentry); + retval = security_inode_getattr(path); if (retval) return retval; return vfs_getattr_nosec(path, stat); @@ -71,7 +71,7 @@ static unsigned long super_cache_scan(struct shrinker *shrink, if (!(sc->gfp_mask & __GFP_FS)) return SHRINK_STOP; - if (!grab_super_passive(sb)) + if (!trylock_super(sb)) return SHRINK_STOP; if (sb->s_op->nr_cached_objects) @@ -105,7 +105,7 @@ static unsigned long super_cache_scan(struct shrinker *shrink, freed += sb->s_op->free_cached_objects(sb, sc); } - drop_super(sb); + up_read(&sb->s_umount); return freed; } @@ -118,7 +118,7 @@ static unsigned long super_cache_count(struct shrinker *shrink, sb = container_of(shrink, struct super_block, s_shrink); /* - * Don't call grab_super_passive as it is a potential + * Don't call trylock_super as it is a potential * scalability bottleneck. The counts could get updated * between super_cache_count and super_cache_scan anyway. * Call to super_cache_count with shrinker_rwsem held @@ -348,35 +348,31 @@ static int grab_super(struct super_block *s) __releases(sb_lock) } /* - * grab_super_passive - acquire a passive reference + * trylock_super - try to grab ->s_umount shared * @sb: reference we are trying to grab * - * Tries to acquire a passive reference. This is used in places where we + * Try to prevent fs shutdown. This is used in places where we * cannot take an active reference but we need to ensure that the - * superblock does not go away while we are working on it. It returns - * false if a reference was not gained, and returns true with the s_umount - * lock held in read mode if a reference is gained. On successful return, - * the caller must drop the s_umount lock and the passive reference when - * done. + * filesystem is not shut down while we are working on it. It returns + * false if we cannot acquire s_umount or if we lose the race and + * filesystem already got into shutdown, and returns true with the s_umount + * lock held in read mode in case of success. On successful return, + * the caller must drop the s_umount lock when done. + * + * Note that unlike get_super() et.al. this one does *not* bump ->s_count. + * The reason why it's safe is that we are OK with doing trylock instead + * of down_read(). There's a couple of places that are OK with that, but + * it's very much not a general-purpose interface. */ -bool grab_super_passive(struct super_block *sb) +bool trylock_super(struct super_block *sb) { - spin_lock(&sb_lock); - if (hlist_unhashed(&sb->s_instances)) { - spin_unlock(&sb_lock); - return false; - } - - sb->s_count++; - spin_unlock(&sb_lock); - if (down_read_trylock(&sb->s_umount)) { - if (sb->s_root && (sb->s_flags & MS_BORN)) + if (!hlist_unhashed(&sb->s_instances) && + sb->s_root && (sb->s_flags & MS_BORN)) return true; up_read(&sb->s_umount); } - put_super(sb); return false; } diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile index d617999..df68285 100644 --- a/fs/xfs/Makefile +++ b/fs/xfs/Makefile @@ -121,3 +121,4 @@ xfs-$(CONFIG_XFS_POSIX_ACL) += xfs_acl.o xfs-$(CONFIG_PROC_FS) += xfs_stats.o xfs-$(CONFIG_SYSCTL) += xfs_sysctl.o xfs-$(CONFIG_COMPAT) += xfs_ioctl32.o +xfs-$(CONFIG_NFSD_PNFS) += xfs_pnfs.o diff --git a/fs/xfs/xfs_export.c b/fs/xfs/xfs_export.c index 5eb4a14..b97359b 100644 --- a/fs/xfs/xfs_export.c +++ b/fs/xfs/xfs_export.c @@ -30,6 +30,7 @@ #include "xfs_trace.h" #include "xfs_icache.h" #include "xfs_log.h" +#include "xfs_pnfs.h" /* * Note that we only accept fileids which are long enough rather than allow @@ -245,4 +246,9 @@ const struct export_operations xfs_export_operations = { .fh_to_parent = xfs_fs_fh_to_parent, .get_parent = xfs_fs_get_parent, .commit_metadata = xfs_fs_nfs_commit_metadata, +#ifdef CONFIG_NFSD_PNFS + .get_uuid = xfs_fs_get_uuid, + .map_blocks = xfs_fs_map_blocks, + .commit_blocks = xfs_fs_commit_blocks, +#endif }; diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index f527618..f44212f 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -36,6 +36,7 @@ #include "xfs_trace.h" #include "xfs_log.h" #include "xfs_icache.h" +#include "xfs_pnfs.h" #include <linux/dcache.h> #include <linux/falloc.h> @@ -395,7 +396,8 @@ STATIC int /* error (positive) */ xfs_zero_last_block( struct xfs_inode *ip, xfs_fsize_t offset, - xfs_fsize_t isize) + xfs_fsize_t isize, + bool *did_zeroing) { struct xfs_mount *mp = ip->i_mount; xfs_fileoff_t last_fsb = XFS_B_TO_FSBT(mp, isize); @@ -423,6 +425,7 @@ xfs_zero_last_block( zero_len = mp->m_sb.sb_blocksize - zero_offset; if (isize + zero_len > offset) zero_len = offset - isize; + *did_zeroing = true; return xfs_iozero(ip, isize, zero_len); } @@ -441,7 +444,8 @@ int /* error (positive) */ xfs_zero_eof( struct xfs_inode *ip, xfs_off_t offset, /* starting I/O offset */ - xfs_fsize_t isize) /* current inode size */ + xfs_fsize_t isize, /* current inode size */ + bool *did_zeroing) { struct xfs_mount *mp = ip->i_mount; xfs_fileoff_t start_zero_fsb; @@ -463,7 +467,7 @@ xfs_zero_eof( * We only zero a part of that block so it is handled specially. */ if (XFS_B_FSB_OFFSET(mp, isize) != 0) { - error = xfs_zero_last_block(ip, offset, isize); + error = xfs_zero_last_block(ip, offset, isize, did_zeroing); if (error) return error; } @@ -523,6 +527,7 @@ xfs_zero_eof( if (error) return error; + *did_zeroing = true; start_zero_fsb = imap.br_startoff + imap.br_blockcount; ASSERT(start_zero_fsb <= (end_zero_fsb + 1)); } @@ -553,6 +558,10 @@ restart: if (error) return error; + error = xfs_break_layouts(inode, iolock); + if (error) + return error; + /* * If the offset is beyond the size of the file, we need to zero any * blocks that fall between the existing EOF and the start of this @@ -561,13 +570,15 @@ restart: * having to redo all checks before. */ if (*pos > i_size_read(inode)) { + bool zero = false; + if (*iolock == XFS_IOLOCK_SHARED) { xfs_rw_iunlock(ip, *iolock); *iolock = XFS_IOLOCK_EXCL; xfs_rw_ilock(ip, *iolock); goto restart; } - error = xfs_zero_eof(ip, *pos, i_size_read(inode)); + error = xfs_zero_eof(ip, *pos, i_size_read(inode), &zero); if (error) return error; } @@ -821,6 +832,7 @@ xfs_file_fallocate( struct xfs_inode *ip = XFS_I(inode); long error; enum xfs_prealloc_flags flags = 0; + uint iolock = XFS_IOLOCK_EXCL; loff_t new_size = 0; if (!S_ISREG(inode->i_mode)) @@ -829,7 +841,11 @@ xfs_file_fallocate( FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_ZERO_RANGE)) return -EOPNOTSUPP; - xfs_ilock(ip, XFS_IOLOCK_EXCL); + xfs_ilock(ip, iolock); + error = xfs_break_layouts(inode, &iolock); + if (error) + goto out_unlock; + if (mode & FALLOC_FL_PUNCH_HOLE) { error = xfs_free_file_space(ip, offset, len); if (error) @@ -893,7 +909,7 @@ xfs_file_fallocate( } out_unlock: - xfs_iunlock(ip, XFS_IOLOCK_EXCL); + xfs_iunlock(ip, iolock); return error; } diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index fba6532..74efe5b 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -602,6 +602,12 @@ xfs_growfs_data( if (!mutex_trylock(&mp->m_growlock)) return -EWOULDBLOCK; error = xfs_growfs_data_private(mp, in); + /* + * Increment the generation unconditionally, the error could be from + * updating the secondary superblocks, in which case the new size + * is live already. + */ + mp->m_generation++; mutex_unlock(&mp->m_growlock); return error; } diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index daafa1f..6163767 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -2867,6 +2867,10 @@ xfs_rename( * Handle RENAME_EXCHANGE flags */ if (flags & RENAME_EXCHANGE) { + if (target_ip == NULL) { + error = -EINVAL; + goto error_return; + } error = xfs_cross_rename(tp, src_dp, src_name, src_ip, target_dp, target_name, target_ip, &free_list, &first_block, spaceres); diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 86cd6b3..a1cd55f 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -384,10 +384,11 @@ enum xfs_prealloc_flags { XFS_PREALLOC_INVISIBLE = (1 << 4), }; -int xfs_update_prealloc_flags(struct xfs_inode *, - enum xfs_prealloc_flags); -int xfs_zero_eof(struct xfs_inode *, xfs_off_t, xfs_fsize_t); -int xfs_iozero(struct xfs_inode *, loff_t, size_t); +int xfs_update_prealloc_flags(struct xfs_inode *ip, + enum xfs_prealloc_flags flags); +int xfs_zero_eof(struct xfs_inode *ip, xfs_off_t offset, + xfs_fsize_t isize, bool *did_zeroing); +int xfs_iozero(struct xfs_inode *ip, loff_t pos, size_t count); #define IHOLD(ip) \ diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c index f7afb86..ac4feae 100644 --- a/fs/xfs/xfs_ioctl.c +++ b/fs/xfs/xfs_ioctl.c @@ -39,6 +39,7 @@ #include "xfs_icache.h" #include "xfs_symlink.h" #include "xfs_trans.h" +#include "xfs_pnfs.h" #include <linux/capability.h> #include <linux/dcache.h> @@ -286,7 +287,7 @@ xfs_readlink_by_handle( return PTR_ERR(dentry); /* Restrict this handle operation to symlinks only. */ - if (!S_ISLNK(dentry->d_inode->i_mode)) { + if (!d_is_symlink(dentry)) { error = -EINVAL; goto out_dput; } @@ -608,6 +609,7 @@ xfs_ioc_space( { struct iattr iattr; enum xfs_prealloc_flags flags = 0; + uint iolock = XFS_IOLOCK_EXCL; int error; /* @@ -636,7 +638,10 @@ xfs_ioc_space( if (error) return error; - xfs_ilock(ip, XFS_IOLOCK_EXCL); + xfs_ilock(ip, iolock); + error = xfs_break_layouts(inode, &iolock); + if (error) + goto out_unlock; switch (bf->l_whence) { case 0: /*SEEK_SET*/ @@ -725,7 +730,7 @@ xfs_ioc_space( error = xfs_update_prealloc_flags(ip, flags); out_unlock: - xfs_iunlock(ip, XFS_IOLOCK_EXCL); + xfs_iunlock(ip, iolock); mnt_drop_write_file(filp); return error; } diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index ce80eeb..e53a903 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -37,6 +37,7 @@ #include "xfs_da_btree.h" #include "xfs_dir2.h" #include "xfs_trans_space.h" +#include "xfs_pnfs.h" #include <linux/capability.h> #include <linux/xattr.h> @@ -505,7 +506,7 @@ xfs_setattr_mode( inode->i_mode |= mode & ~S_IFMT; } -static void +void xfs_setattr_time( struct xfs_inode *ip, struct iattr *iattr) @@ -750,6 +751,7 @@ xfs_setattr_size( int error; uint lock_flags = 0; uint commit_flags = 0; + bool did_zeroing = false; trace_xfs_setattr(ip); @@ -793,20 +795,16 @@ xfs_setattr_size( return error; /* - * Now we can make the changes. Before we join the inode to the - * transaction, take care of the part of the truncation that must be - * done without the inode lock. This needs to be done before joining - * the inode to the transaction, because the inode cannot be unlocked - * once it is a part of the transaction. + * File data changes must be complete before we start the transaction to + * modify the inode. This needs to be done before joining the inode to + * the transaction because the inode cannot be unlocked once it is a + * part of the transaction. + * + * Start with zeroing any data block beyond EOF that we may expose on + * file extension. */ if (newsize > oldsize) { - /* - * Do the first part of growing a file: zero any data in the - * last block that is beyond the old EOF. We need to do this - * before the inode is joined to the transaction to modify - * i_size. - */ - error = xfs_zero_eof(ip, newsize, oldsize); + error = xfs_zero_eof(ip, newsize, oldsize, &did_zeroing); if (error) return error; } @@ -816,23 +814,18 @@ xfs_setattr_size( * any previous writes that are beyond the on disk EOF and the new * EOF that have not been written out need to be written here. If we * do not write the data out, we expose ourselves to the null files - * problem. - * - * Only flush from the on disk size to the smaller of the in memory - * file size or the new size as that's the range we really care about - * here and prevents waiting for other data not within the range we - * care about here. + * problem. Note that this includes any block zeroing we did above; + * otherwise those blocks may not be zeroed after a crash. */ - if (oldsize != ip->i_d.di_size && newsize > ip->i_d.di_size) { + if (newsize > ip->i_d.di_size && + (oldsize != ip->i_d.di_size || did_zeroing)) { error = filemap_write_and_wait_range(VFS_I(ip)->i_mapping, ip->i_d.di_size, newsize); if (error) return error; } - /* - * Wait for all direct I/O to complete. - */ + /* Now wait for all direct I/O to complete. */ inode_dio_wait(inode); /* @@ -979,9 +972,13 @@ xfs_vn_setattr( int error; if (iattr->ia_valid & ATTR_SIZE) { - xfs_ilock(ip, XFS_IOLOCK_EXCL); - error = xfs_setattr_size(ip, iattr); - xfs_iunlock(ip, XFS_IOLOCK_EXCL); + uint iolock = XFS_IOLOCK_EXCL; + + xfs_ilock(ip, iolock); + error = xfs_break_layouts(dentry->d_inode, &iolock); + if (!error) + error = xfs_setattr_size(ip, iattr); + xfs_iunlock(ip, iolock); } else { error = xfs_setattr_nonsize(ip, iattr, 0); } diff --git a/fs/xfs/xfs_iops.h b/fs/xfs/xfs_iops.h index 1c34e43..ea7a98e 100644 --- a/fs/xfs/xfs_iops.h +++ b/fs/xfs/xfs_iops.h @@ -32,6 +32,7 @@ extern void xfs_setup_inode(struct xfs_inode *); */ #define XFS_ATTR_NOACL 0x01 /* Don't call posix_acl_chmod */ +extern void xfs_setattr_time(struct xfs_inode *ip, struct iattr *iattr); extern int xfs_setattr_nonsize(struct xfs_inode *ip, struct iattr *vap, int flags); extern int xfs_setattr_size(struct xfs_inode *ip, struct iattr *vap); diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index a5b2ff8..0d8abd6 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -174,6 +174,17 @@ typedef struct xfs_mount { struct workqueue_struct *m_reclaim_workqueue; struct workqueue_struct *m_log_workqueue; struct workqueue_struct *m_eofblocks_workqueue; + + /* + * Generation of the filesysyem layout. This is incremented by each + * growfs, and used by the pNFS server to ensure the client updates + * its view of the block device once it gets a layout that might + * reference the newly added blocks. Does not need to be persistent + * as long as we only allow file system size increments, but if we + * ever support shrinks it would have to be persisted in addition + * to various other kinds of pain inflicted on the pNFS server. + */ + __uint32_t m_generation; } xfs_mount_t; /* diff --git a/fs/xfs/xfs_pnfs.c b/fs/xfs/xfs_pnfs.c new file mode 100644 index 0000000..365dd57 --- /dev/null +++ b/fs/xfs/xfs_pnfs.c @@ -0,0 +1,324 @@ +/* + * Copyright (c) 2014 Christoph Hellwig. + */ +#include "xfs.h" +#include "xfs_format.h" +#include "xfs_log_format.h" +#include "xfs_trans_resv.h" +#include "xfs_sb.h" +#include "xfs_mount.h" +#include "xfs_inode.h" +#include "xfs_trans.h" +#include "xfs_log.h" +#include "xfs_bmap.h" +#include "xfs_bmap_util.h" +#include "xfs_error.h" +#include "xfs_iomap.h" +#include "xfs_shared.h" +#include "xfs_bit.h" +#include "xfs_pnfs.h" + +/* + * Ensure that we do not have any outstanding pNFS layouts that can be used by + * clients to directly read from or write to this inode. This must be called + * before every operation that can remove blocks from the extent map. + * Additionally we call it during the write operation, where aren't concerned + * about exposing unallocated blocks but just want to provide basic + * synchronization between a local writer and pNFS clients. mmap writes would + * also benefit from this sort of synchronization, but due to the tricky locking + * rules in the page fault path we don't bother. + */ +int +xfs_break_layouts( + struct inode *inode, + uint *iolock) +{ + struct xfs_inode *ip = XFS_I(inode); + int error; + + ASSERT(xfs_isilocked(ip, XFS_IOLOCK_SHARED|XFS_IOLOCK_EXCL)); + + while ((error = break_layout(inode, false) == -EWOULDBLOCK)) { + xfs_iunlock(ip, *iolock); + error = break_layout(inode, true); + *iolock = XFS_IOLOCK_EXCL; + xfs_ilock(ip, *iolock); + } + + return error; +} + +/* + * Get a unique ID including its location so that the client can identify + * the exported device. + */ +int +xfs_fs_get_uuid( + struct super_block *sb, + u8 *buf, + u32 *len, + u64 *offset) +{ + struct xfs_mount *mp = XFS_M(sb); + + printk_once(KERN_NOTICE +"XFS (%s): using experimental pNFS feature, use at your own risk!\n", + mp->m_fsname); + + if (*len < sizeof(uuid_t)) + return -EINVAL; + + memcpy(buf, &mp->m_sb.sb_uuid, sizeof(uuid_t)); + *len = sizeof(uuid_t); + *offset = offsetof(struct xfs_dsb, sb_uuid); + return 0; +} + +static void +xfs_bmbt_to_iomap( + struct xfs_inode *ip, + struct iomap *iomap, + struct xfs_bmbt_irec *imap) +{ + struct xfs_mount *mp = ip->i_mount; + + if (imap->br_startblock == HOLESTARTBLOCK) { + iomap->blkno = IOMAP_NULL_BLOCK; + iomap->type = IOMAP_HOLE; + } else if (imap->br_startblock == DELAYSTARTBLOCK) { + iomap->blkno = IOMAP_NULL_BLOCK; + iomap->type = IOMAP_DELALLOC; + } else { + iomap->blkno = + XFS_FSB_TO_DADDR(ip->i_mount, imap->br_startblock); + if (imap->br_state == XFS_EXT_UNWRITTEN) + iomap->type = IOMAP_UNWRITTEN; + else + iomap->type = IOMAP_MAPPED; + } + iomap->offset = XFS_FSB_TO_B(mp, imap->br_startoff); + iomap->length = XFS_FSB_TO_B(mp, imap->br_blockcount); +} + +/* + * Get a layout for the pNFS client. + */ +int +xfs_fs_map_blocks( + struct inode *inode, + loff_t offset, + u64 length, + struct iomap *iomap, + bool write, + u32 *device_generation) +{ + struct xfs_inode *ip = XFS_I(inode); + struct xfs_mount *mp = ip->i_mount; + struct xfs_bmbt_irec imap; + xfs_fileoff_t offset_fsb, end_fsb; + loff_t limit; + int bmapi_flags = XFS_BMAPI_ENTIRE; + int nimaps = 1; + uint lock_flags; + int error = 0; + + if (XFS_FORCED_SHUTDOWN(mp)) + return -EIO; + + /* + * We can't export inodes residing on the realtime device. The realtime + * device doesn't have a UUID to identify it, so the client has no way + * to find it. + */ + if (XFS_IS_REALTIME_INODE(ip)) + return -ENXIO; + + /* + * Lock out any other I/O before we flush and invalidate the pagecache, + * and then hand out a layout to the remote system. This is very + * similar to direct I/O, except that the synchronization is much more + * complicated. See the comment near xfs_break_layouts for a detailed + * explanation. + */ + xfs_ilock(ip, XFS_IOLOCK_EXCL); + + error = -EINVAL; + limit = mp->m_super->s_maxbytes; + if (!write) + limit = max(limit, round_up(i_size_read(inode), + inode->i_sb->s_blocksize)); + if (offset > limit) + goto out_unlock; + if (offset > limit - length) + length = limit - offset; + + error = filemap_write_and_wait(inode->i_mapping); + if (error) + goto out_unlock; + error = invalidate_inode_pages2(inode->i_mapping); + if (WARN_ON_ONCE(error)) + return error; + + end_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)offset + length); + offset_fsb = XFS_B_TO_FSBT(mp, offset); + + lock_flags = xfs_ilock_data_map_shared(ip); + error = xfs_bmapi_read(ip, offset_fsb, end_fsb - offset_fsb, + &imap, &nimaps, bmapi_flags); + xfs_iunlock(ip, lock_flags); + + if (error) + goto out_unlock; + + if (write) { + enum xfs_prealloc_flags flags = 0; + + ASSERT(imap.br_startblock != DELAYSTARTBLOCK); + + if (!nimaps || imap.br_startblock == HOLESTARTBLOCK) { + error = xfs_iomap_write_direct(ip, offset, length, + &imap, nimaps); + if (error) + goto out_unlock; + + /* + * Ensure the next transaction is committed + * synchronously so that the blocks allocated and + * handed out to the client are guaranteed to be + * present even after a server crash. + */ + flags |= XFS_PREALLOC_SET | XFS_PREALLOC_SYNC; + } + + error = xfs_update_prealloc_flags(ip, flags); + if (error) + goto out_unlock; + } + xfs_iunlock(ip, XFS_IOLOCK_EXCL); + + xfs_bmbt_to_iomap(ip, iomap, &imap); + *device_generation = mp->m_generation; + return error; +out_unlock: + xfs_iunlock(ip, XFS_IOLOCK_EXCL); + return error; +} + +/* + * Ensure the size update falls into a valid allocated block. + */ +static int +xfs_pnfs_validate_isize( + struct xfs_inode *ip, + xfs_off_t isize) +{ + struct xfs_bmbt_irec imap; + int nimaps = 1; + int error = 0; + + xfs_ilock(ip, XFS_ILOCK_SHARED); + error = xfs_bmapi_read(ip, XFS_B_TO_FSBT(ip->i_mount, isize - 1), 1, + &imap, &nimaps, 0); + xfs_iunlock(ip, XFS_ILOCK_SHARED); + if (error) + return error; + + if (imap.br_startblock == HOLESTARTBLOCK || + imap.br_startblock == DELAYSTARTBLOCK || + imap.br_state == XFS_EXT_UNWRITTEN) + return -EIO; + return 0; +} + +/* + * Make sure the blocks described by maps are stable on disk. This includes + * converting any unwritten extents, flushing the disk cache and updating the + * time stamps. + * + * Note that we rely on the caller to always send us a timestamp update so that + * we always commit a transaction here. If that stops being true we will have + * to manually flush the cache here similar to what the fsync code path does + * for datasyncs on files that have no dirty metadata. + */ +int +xfs_fs_commit_blocks( + struct inode *inode, + struct iomap *maps, + int nr_maps, + struct iattr *iattr) +{ + struct xfs_inode *ip = XFS_I(inode); + struct xfs_mount *mp = ip->i_mount; + struct xfs_trans *tp; + bool update_isize = false; + int error, i; + loff_t size; + + ASSERT(iattr->ia_valid & (ATTR_ATIME|ATTR_CTIME|ATTR_MTIME)); + + xfs_ilock(ip, XFS_IOLOCK_EXCL); + + size = i_size_read(inode); + if ((iattr->ia_valid & ATTR_SIZE) && iattr->ia_size > size) { + update_isize = true; + size = iattr->ia_size; + } + + for (i = 0; i < nr_maps; i++) { + u64 start, length, end; + + start = maps[i].offset; + if (start > size) + continue; + + end = start + maps[i].length; + if (end > size) + end = size; + + length = end - start; + if (!length) + continue; + + /* + * Make sure reads through the pagecache see the new data. + */ + error = invalidate_inode_pages2_range(inode->i_mapping, + start >> PAGE_CACHE_SHIFT, + (end - 1) >> PAGE_CACHE_SHIFT); + WARN_ON_ONCE(error); + + error = xfs_iomap_write_unwritten(ip, start, length); + if (error) + goto out_drop_iolock; + } + + if (update_isize) { + error = xfs_pnfs_validate_isize(ip, size); + if (error) + goto out_drop_iolock; + } + + 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); + goto out_drop_iolock; + } + + xfs_ilock(ip, XFS_ILOCK_EXCL); + xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL); + xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); + + xfs_setattr_time(ip, iattr); + if (update_isize) { + i_size_write(inode, iattr->ia_size); + ip->i_d.di_size = iattr->ia_size; + } + + xfs_trans_set_sync(tp); + error = xfs_trans_commit(tp, 0); + +out_drop_iolock: + xfs_iunlock(ip, XFS_IOLOCK_EXCL); + return error; +} diff --git a/fs/xfs/xfs_pnfs.h b/fs/xfs/xfs_pnfs.h new file mode 100644 index 0000000..b7fbfce --- /dev/null +++ b/fs/xfs/xfs_pnfs.h @@ -0,0 +1,18 @@ +#ifndef _XFS_PNFS_H +#define _XFS_PNFS_H 1 + +#ifdef CONFIG_NFSD_PNFS +int xfs_fs_get_uuid(struct super_block *sb, u8 *buf, u32 *len, u64 *offset); +int xfs_fs_map_blocks(struct inode *inode, loff_t offset, u64 length, + struct iomap *iomap, bool write, u32 *device_generation); +int xfs_fs_commit_blocks(struct inode *inode, struct iomap *maps, int nr_maps, + struct iattr *iattr); + +int xfs_break_layouts(struct inode *inode, uint *iolock); +#else +static inline int xfs_break_layouts(struct inode *inode, uint *iolock) +{ + return 0; +} +#endif /* CONFIG_NFSD_PNFS */ +#endif /* _XFS_PNFS_H */ diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 53cc2aa..fbbb9e6 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -836,6 +836,11 @@ xfs_qm_reset_dqcounts( */ xfs_dqcheck(mp, ddq, id+j, type, XFS_QMOPT_DQREPAIR, "xfs_quotacheck"); + /* + * Reset type in case we are reusing group quota file for + * project quotas or vice versa + */ + ddq->d_flags = type; ddq->d_bcount = 0; ddq->d_icount = 0; ddq->d_rtbcount = 0; |
