From 698b8223631472bf982ed570b0812faa61955683 Mon Sep 17 00:00:00 2001 From: Dave Chiluk Date: Tue, 28 May 2013 16:06:08 -0500 Subject: ncpfs: fix rmdir returns Device or resource busy 1d2ef5901483004d74947bbf78d5146c24038fe7 caused a regression in ncpfs such that directories could no longer be removed. This was because ncp_rmdir checked to see if a dentry could be unhashed before allowing it to be removed. Since 1d2ef5901483004d74947bbf78d5146c24038fe7 introduced a change that incremented dentry->d_count causing it to always be greater than 1 unhash would always fail. Thus causing the error path in ncp_rmdir to always be taken. Removing this error path is safe as unhashing is still accomplished by calls to dput from vfs_rmdir. Signed-off-by: Dave Chiluk Signed-off-by: Petr Vandrovec Signed-off-by: Al Viro --- fs/ncpfs/dir.c | 9 --------- 1 file changed, 9 deletions(-) (limited to 'fs') diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c index 8163260..6792ce1 100644 --- a/fs/ncpfs/dir.c +++ b/fs/ncpfs/dir.c @@ -1029,15 +1029,6 @@ static int ncp_rmdir(struct inode *dir, struct dentry *dentry) DPRINTK("ncp_rmdir: removing %s/%s\n", dentry->d_parent->d_name.name, dentry->d_name.name); - /* - * fail with EBUSY if there are still references to this - * directory. - */ - dentry_unhash(dentry); - error = -EBUSY; - if (!d_unhashed(dentry)) - goto out; - len = sizeof(__name); error = ncp_io2vol(server, __name, &len, dentry->d_name.name, dentry->d_name.len, !ncp_preserve_case(dir)); -- cgit v1.1 From 7b5ff90ed081787ec0765ceb4fe5ccf5677493a6 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 6 Jun 2013 10:29:40 -0400 Subject: Btrfs: don't delete fs_roots until after we cleanup the transaction We get a use after free if we had a transaction to cleanup since there could be delayed inodes which refer to their respective fs_root. Thanks Reported-by: David Sterba Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index e7b3cb5..bdaa092 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -2859,8 +2859,8 @@ fail_qgroup: btrfs_free_qgroup_config(fs_info); fail_trans_kthread: kthread_stop(fs_info->transaction_kthread); - del_fs_roots(fs_info); btrfs_cleanup_transaction(fs_info->tree_root); + del_fs_roots(fs_info); fail_cleaner: kthread_stop(fs_info->cleaner_kthread); -- cgit v1.1 From 6379ef9fb2482a92b5fe09f927d6ce1f989c0c6d Mon Sep 17 00:00:00 2001 From: Naohiro Aota Date: Thu, 6 Jun 2013 09:56:34 +0000 Subject: btrfs: Drop inode if inode root is NULL There is a path where btrfs_drop_inode() is called with its inode's root is NULL: In btrfs_new_inode(), when btrfs_set_inode_index() fails, iput() is called. We should handle this case before taking look at the root->root_item. Signed-off-by: Naohiro Aota Reviewed-by: Miao Xie Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'fs') diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 23c596c..c931a4d 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -8011,6 +8011,9 @@ int btrfs_drop_inode(struct inode *inode) { struct btrfs_root *root = BTRFS_I(inode)->root; + if (root == NULL) + return 1; + /* the snap/subvol tree is on deleting */ if (btrfs_root_refs(&root->root_item) == 0 && root != root->fs_info->tree_root) -- cgit v1.1 From a9995eece39a0630ebbfc1ab38570bce6c8a8f5b Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 31 May 2013 13:04:36 -0400 Subject: Btrfs: init relocate extent_io_tree with a mapping Dave reported a NULL pointer deref. This is caused because he thought he'd be smart and add sanity checks to the extent_io bit operations, but he didn't expect a tree to have a NULL mapping. To fix this we just need to init the relocation's processed_blocks with the btree_inode->i_mapping. Thanks, Reported-by: David Sterba Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/relocation.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 395b820..4febca4 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -4082,7 +4082,7 @@ out: return inode; } -static struct reloc_control *alloc_reloc_control(void) +static struct reloc_control *alloc_reloc_control(struct btrfs_fs_info *fs_info) { struct reloc_control *rc; @@ -4093,7 +4093,8 @@ static struct reloc_control *alloc_reloc_control(void) INIT_LIST_HEAD(&rc->reloc_roots); backref_cache_init(&rc->backref_cache); mapping_tree_init(&rc->reloc_root_tree); - extent_io_tree_init(&rc->processed_blocks, NULL); + extent_io_tree_init(&rc->processed_blocks, + fs_info->btree_inode->i_mapping); return rc; } @@ -4110,7 +4111,7 @@ int btrfs_relocate_block_group(struct btrfs_root *extent_root, u64 group_start) int rw = 0; int err = 0; - rc = alloc_reloc_control(); + rc = alloc_reloc_control(fs_info); if (!rc) return -ENOMEM; @@ -4311,7 +4312,7 @@ int btrfs_recover_relocation(struct btrfs_root *root) if (list_empty(&reloc_roots)) goto out; - rc = alloc_reloc_control(); + rc = alloc_reloc_control(root->fs_info); if (!rc) { err = -ENOMEM; goto out; -- cgit v1.1 From 2932505abe7c56477315a3d93ffb3c27c5182e9d Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Sun, 26 May 2013 13:50:27 +0000 Subject: Btrfs: fix use-after-free bug during umount Commit be283b2e674a09457d4563729015adb637ce7cc1 ( Btrfs: use helper to cleanup tree roots) introduced the following bug, BUG: unable to handle kernel NULL pointer dereference at 0000000000000034 IP: [] extent_buffer_get+0x4/0xa [btrfs] [...] Pid: 2463, comm: btrfs-cache-1 Tainted: G O 3.9.0+ #4 innotek GmbH VirtualBox/VirtualBox RIP: 0010:[] [] extent_buffer_get+0x4/0xa [btrfs] Process btrfs-cache-1 (pid: 2463, threadinfo ffff880112d60000, task ffff880117679730) [...] Call Trace: [] btrfs_search_slot+0x104/0x64d [btrfs] [] btrfs_next_old_leaf+0xa7/0x334 [btrfs] [] btrfs_next_leaf+0x10/0x12 [btrfs] [] caching_thread+0x1a3/0x2e0 [btrfs] [] worker_loop+0x14b/0x48e [btrfs] [] ? btrfs_queue_worker+0x25c/0x25c [btrfs] [] kthread+0x8d/0x95 [] ? kthread_freezable_should_stop+0x43/0x43 [] ret_from_fork+0x7c/0xb0 [] ? kthread_freezable_should_stop+0x43/0x43 RIP [] extent_buffer_get+0x4/0xa [btrfs] We've free'ed commit_root before actually getting to free block groups where caching thread needs valid extent_root->commit_root. Signed-off-by: Liu Bo Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index bdaa092..7c66c23 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3512,10 +3512,10 @@ int close_ctree(struct btrfs_root *root) percpu_counter_sum(&fs_info->delalloc_bytes)); } - free_root_pointers(fs_info, 1); - btrfs_free_block_groups(fs_info); + free_root_pointers(fs_info, 1); + del_fs_roots(fs_info); iput(fs_info->btree_inode); -- cgit v1.1 From 13e6c37b989859e70b0d73d3f2cb0aa022159b17 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Thu, 30 May 2013 16:55:44 -0400 Subject: Btrfs: stop all workers before cleaning up roots Dave reported a panic because the extent_root->commit_root was NULL in the caching kthread. That is because we just unset it in free_root_pointers, which is not the correct thing to do, we have to either wait for the caching kthread to complete or hold the extent_commit_sem lock so we know the thread has exited. This patch makes the kthreads all stop first and then we do our cleanup. This should fix the race. Thanks, Reported-by: David Sterba Signed-off-by: Josef Bacik --- fs/btrfs/disk-io.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 7c66c23..b8b60b6 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3514,13 +3514,13 @@ int close_ctree(struct btrfs_root *root) btrfs_free_block_groups(fs_info); - free_root_pointers(fs_info, 1); + btrfs_stop_all_workers(fs_info); del_fs_roots(fs_info); - iput(fs_info->btree_inode); + free_root_pointers(fs_info, 1); - btrfs_stop_all_workers(fs_info); + iput(fs_info->btree_inode); #ifdef CONFIG_BTRFS_FS_CHECK_INTEGRITY if (btrfs_test_opt(root, CHECK_INTEGRITY)) -- cgit v1.1 From 47ad2fcba9ddd0630acccb13c71f19a818947751 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Mon, 27 May 2013 16:38:19 +1000 Subject: xfs: don't emit v5 superblock warnings on write We write the superblock every 30s or so which results in the verifier being called. Right now that results in this output every 30s: XFS (vda): Version 5 superblock detected. This kernel has EXPERIMENTAL support enabled! Use of these features in this kernel is at your own risk! And spamming the logs. We don't need to check for whether we support v5 superblocks or whether there are feature bits we don't support set as these are only relevant when we first mount the filesytem. i.e. on superblock read. Hence for the write verification we can just skip all the checks (and hence verbose output) altogether. Signed-off-by: Dave Chinner Reviewed-by: Brian Foster Signed-off-by: Ben Myers (cherry picked from commit 34510185abeaa5be9b178a41c0a03d30aec3db7e) --- fs/xfs/xfs_mount.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index f6bfbd7..e8e310c 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -314,7 +314,8 @@ STATIC int xfs_mount_validate_sb( xfs_mount_t *mp, xfs_sb_t *sbp, - bool check_inprogress) + bool check_inprogress, + bool check_version) { /* @@ -337,9 +338,10 @@ xfs_mount_validate_sb( /* * Version 5 superblock feature mask validation. Reject combinations the - * kernel cannot support up front before checking anything else. + * kernel cannot support up front before checking anything else. For + * write validation, we don't need to check feature masks. */ - if (XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) { + if (check_version && XFS_SB_VERSION_NUM(sbp) == XFS_SB_VERSION_5) { xfs_alert(mp, "Version 5 superblock detected. This kernel has EXPERIMENTAL support enabled!\n" "Use of these features in this kernel is at your own risk!"); @@ -675,7 +677,8 @@ xfs_sb_to_disk( static int xfs_sb_verify( - struct xfs_buf *bp) + struct xfs_buf *bp, + bool check_version) { struct xfs_mount *mp = bp->b_target->bt_mount; struct xfs_sb sb; @@ -686,7 +689,8 @@ xfs_sb_verify( * Only check the in progress field for the primary superblock as * mkfs.xfs doesn't clear it from secondary superblocks. */ - return xfs_mount_validate_sb(mp, &sb, bp->b_bn == XFS_SB_DADDR); + return xfs_mount_validate_sb(mp, &sb, bp->b_bn == XFS_SB_DADDR, + check_version); } /* @@ -719,7 +723,7 @@ xfs_sb_read_verify( goto out_error; } } - error = xfs_sb_verify(bp); + error = xfs_sb_verify(bp, true); out_error: if (error) { @@ -758,7 +762,7 @@ xfs_sb_write_verify( struct xfs_buf_log_item *bip = bp->b_fspriv; int error; - error = xfs_sb_verify(bp); + error = xfs_sb_verify(bp, false); if (error) { XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, bp->b_addr); xfs_buf_ioerror(bp, error); -- cgit v1.1 From 5170711df79b284cf95b3924322e8ac4c0fd6c76 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Wed, 12 Jun 2013 12:19:07 +1000 Subject: xfs: fix implicit padding in directory and attr CRC formats Michael L. Semon has been testing CRC patches on a 32 bit system and been seeing assert failures in the directory code from xfs/080. Thanks to Michael's heroic efforts with printk debugging, we found that the problem was that the last free space being left in the directory structure was too small to fit a unused tag structure and it was being corrupted and attempting to log a region out of bounds. Hence the assert failure looked something like: ..... #5 calling xfs_dir2_data_log_unused() 36 32 #1 4092 4095 4096 #2 8182 8183 4096 XFS: Assertion failed: first <= last && last < BBTOB(bp->b_length), file: fs/xfs/xfs_trans_buf.c, line: 568 Where #1 showed the first region of the dup being logged (i.e. the last 4 bytes of a directory buffer) and #2 shows the corrupt values being calculated from the length of the dup entry which overflowed the size of the buffer. It turns out that the problem was not in the logging code, nor in the freespace handling code. It is an initial condition bug that only shows up on 32 bit systems. When a new buffer is initialised, where's the freespace that is set up: [ 172.316249] calling xfs_dir2_leaf_addname() from xfs_dir_createname() [ 172.316346] #9 calling xfs_dir2_data_log_unused() [ 172.316351] #1 calling xfs_trans_log_buf() 60 63 4096 [ 172.316353] #2 calling xfs_trans_log_buf() 4094 4095 4096 Note the offset of the first region being logged? It's 60 bytes into the buffer. Once I saw that, I pretty much knew that the bug was going to be caused by this. Essentially, all direct entries are rounded to 8 bytes in length, and all entries start with an 8 byte alignment. This means that we can decode inplace as variables are naturally aligned. With the directory data supposedly starting on a 8 byte boundary, and all entries padded to 8 bytes, the minimum freespace in a directory block is supposed to be 8 bytes, which is large enough to fit a unused data entry structure (6 bytes in size). The fact we only have 4 bytes of free space indicates a directory data block alignment problem. And what do you know - there's an implicit hole in the directory data block header for the CRC format, which means the header is 60 byte on 32 bit intel systems and 64 bytes on 64 bit systems. Needs padding. And while looking at the structures, I found the same problem in the attr leaf header. Fix them both. Note that this only affects 32 bit systems with CRCs enabled. Everything else is just fine. Note that CRC enabled filesystems created before this fix on such systems will not be readable with this fix applied. Reported-by: Michael L. Semon Debugged-by: Michael L. Semon Signed-off-by: Dave Chinner Reviewed-by: Ben Myers Signed-off-by: Ben Myers (cherry picked from commit 8a1fd2950e1fe267e11fc8c85dcaa6b023b51b60) --- fs/xfs/xfs_attr_leaf.h | 1 + fs/xfs/xfs_dir2_format.h | 5 +++-- 2 files changed, 4 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_attr_leaf.h b/fs/xfs/xfs_attr_leaf.h index f9d7846..444a770 100644 --- a/fs/xfs/xfs_attr_leaf.h +++ b/fs/xfs/xfs_attr_leaf.h @@ -128,6 +128,7 @@ struct xfs_attr3_leaf_hdr { __u8 holes; __u8 pad1; struct xfs_attr_leaf_map freemap[XFS_ATTR_LEAF_MAPSIZE]; + __be32 pad2; /* 64 bit alignment */ }; #define XFS_ATTR3_LEAF_CRC_OFF (offsetof(struct xfs_attr3_leaf_hdr, info.crc)) diff --git a/fs/xfs/xfs_dir2_format.h b/fs/xfs/xfs_dir2_format.h index 995f1f5..7826782b 100644 --- a/fs/xfs/xfs_dir2_format.h +++ b/fs/xfs/xfs_dir2_format.h @@ -266,6 +266,7 @@ struct xfs_dir3_blk_hdr { struct xfs_dir3_data_hdr { struct xfs_dir3_blk_hdr hdr; xfs_dir2_data_free_t best_free[XFS_DIR2_DATA_FD_COUNT]; + __be32 pad; /* 64 bit alignment */ }; #define XFS_DIR3_DATA_CRC_OFF offsetof(struct xfs_dir3_data_hdr, hdr.crc) @@ -477,7 +478,7 @@ struct xfs_dir3_leaf_hdr { struct xfs_da3_blkinfo info; /* header for da routines */ __be16 count; /* count of entries */ __be16 stale; /* count of stale entries */ - __be32 pad; + __be32 pad; /* 64 bit alignment */ }; struct xfs_dir3_icleaf_hdr { @@ -715,7 +716,7 @@ struct xfs_dir3_free_hdr { __be32 firstdb; /* db of first entry */ __be32 nvalid; /* count of valid entries */ __be32 nused; /* count of used entries */ - __be32 pad; /* 64 bit alignment. */ + __be32 pad; /* 64 bit alignment */ }; struct xfs_dir3_free { -- cgit v1.1 From 088c9f67c3f53339d2bc20b42a9cb904901fdc5d Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Wed, 12 Jun 2013 12:19:08 +1000 Subject: xfs: ensure btree root split sets blkno correctly For CRC enabled filesystems, the BMBT is rooted in an inode, so it passes through a different code path on root splits than the freespace and inode btrees. This is much less traversed by xfstests than the other trees. When testing on a 1k block size filesystem, I've been seeing ASSERT failures in generic/234 like: XFS: Assertion failed: cur->bc_btnum != XFS_BTNUM_BMAP || cur->bc_private.b.allocated == 0, file: fs/xfs/xfs_btree.c, line: 317 which are generally preceded by a lblock check failure. I noticed this in the bmbt stats: $ pminfo -f xfs.btree.block_map xfs.btree.block_map.lookup value 39135 xfs.btree.block_map.compare value 268432 xfs.btree.block_map.insrec value 15786 xfs.btree.block_map.delrec value 13884 xfs.btree.block_map.newroot value 2 xfs.btree.block_map.killroot value 0 ..... Very little coverage of root splits and merges. Indeed, on a 4k filesystem, block_map.newroot and block_map.killroot are both zero. i.e. the code is not exercised at all, and it's the only generic btree infrastructure operation that is not exercised by a default run of xfstests. Turns out that on a 1k filesystem, generic/234 accounts for one of those two root splits, and that is somewhat of a smoking gun. In fact, it's the same problem we saw in the directory/attr code where headers are memcpy()d from one block to another without updating the self describing metadata. Simple fix - when copying the header out of the root block, make sure the block number is updated correctly. Signed-off-by: Dave Chinner Reviewed-by: Ben Myers Signed-off-by: Ben Myers (cherry picked from commit ade1335afef556df6538eb02e8c0dc91fbd9cc37) --- fs/xfs/xfs_btree.c | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c index 8804b8a..0903960 100644 --- a/fs/xfs/xfs_btree.c +++ b/fs/xfs/xfs_btree.c @@ -2544,7 +2544,17 @@ xfs_btree_new_iroot( if (error) goto error0; + /* + * we can't just memcpy() the root in for CRC enabled btree blocks. + * In that case have to also ensure the blkno remains correct + */ memcpy(cblock, block, xfs_btree_block_len(cur)); + if (cur->bc_flags & XFS_BTREE_CRC_BLOCKS) { + if (cur->bc_flags & XFS_BTREE_LONG_PTRS) + cblock->bb_u.l.bb_blkno = cpu_to_be64(cbp->b_bn); + else + cblock->bb_u.s.bb_blkno = cpu_to_be64(cbp->b_bn); + } be16_add_cpu(&block->bb_level, 1); xfs_btree_set_numrecs(block, 1); -- cgit v1.1 From d302cf1d316dca5f567e89872cf5d475c9a55f74 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Wed, 12 Jun 2013 12:19:06 +1000 Subject: xfs: don't shutdown log recovery on validation errors Unfortunately, we cannot guarantee that items logged multiple times and replayed by log recovery do not take objects back in time. When they are taken back in time, the go into an intermediate state which is corrupt, and hence verification that occurs on this intermediate state causes log recovery to abort with a corruption shutdown. Instead of causing a shutdown and unmountable filesystem, don't verify post-recovery items before they are written to disk. This is less than optimal, but there is no way to detect this issue for non-CRC filesystems If log recovery successfully completes, this will be undone and the object will be consistent by subsequent transactions that are replayed, so in most cases we don't need to take drastic action. For CRC enabled filesystems, leave the verifiers in place - we need to call them to recalculate the CRCs on the objects anyway. This recovery problem can be solved for such filesystems - we have a LSN stamped in all metadata at writeback time that we can to determine whether the item should be replayed or not. This is a separate piece of work, so is not addressed by this patch. Signed-off-by: Dave Chinner Reviewed-by: Ben Myers Signed-off-by: Ben Myers (cherry picked from commit 9222a9cf86c0d64ffbedf567412b55da18763aa3) --- fs/xfs/xfs_log_recover.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 45a85ff..7cf5e4e 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -1845,7 +1845,13 @@ xlog_recover_do_inode_buffer( xfs_agino_t *buffer_nextp; trace_xfs_log_recover_buf_inode_buf(mp->m_log, buf_f); - bp->b_ops = &xfs_inode_buf_ops; + + /* + * Post recovery validation only works properly on CRC enabled + * filesystems. + */ + if (xfs_sb_version_hascrc(&mp->m_sb)) + bp->b_ops = &xfs_inode_buf_ops; inodes_per_buf = BBTOB(bp->b_io_length) >> mp->m_sb.sb_inodelog; for (i = 0; i < inodes_per_buf; i++) { @@ -2205,7 +2211,16 @@ xlog_recover_do_reg_buffer( /* Shouldn't be any more regions */ ASSERT(i == item->ri_total); - xlog_recovery_validate_buf_type(mp, bp, buf_f); + /* + * We can only do post recovery validation on items on CRC enabled + * fielsystems as we need to know when the buffer was written to be able + * to determine if we should have replayed the item. If we replay old + * metadata over a newer buffer, then it will enter a temporarily + * inconsistent state resulting in verification failures. Hence for now + * just avoid the verification stage for non-crc filesystems + */ + if (xfs_sb_version_hascrc(&mp->m_sb)) + xlog_recovery_validate_buf_type(mp, bp, buf_f); } /* -- cgit v1.1 From e7b2c4069252732d52f1de6d1f7c82d99a156659 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Fri, 14 Jun 2013 21:09:47 +0200 Subject: fput: task_work_add() can fail if the caller has passed exit_task_work() fput() assumes that it can't be called after exit_task_work() but this is not true, for example free_ipc_ns()->shm_destroy() can do this. In this case fput() silently leaks the file. Change it to fallback to delayed_fput_work if task_work_add() fails. The patch looks complicated but it is not, it changes the code from if (PF_KTHREAD) { schedule_work(...); return; } task_work_add(...) to if (!PF_KTHREAD) { if (!task_work_add(...)) return; /* fallback */ } schedule_work(...); As for shm_destroy() in particular, we could make another fix but I think this change makes sense anyway. There could be another similar user, it is not safe to assume that task_work_add() can't fail. Reported-by: Andrey Vagin Signed-off-by: Oleg Nesterov Signed-off-by: Al Viro --- fs/file_table.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) (limited to 'fs') diff --git a/fs/file_table.c b/fs/file_table.c index cd4d87a..485dc0e 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -306,17 +306,18 @@ void fput(struct file *file) { if (atomic_long_dec_and_test(&file->f_count)) { struct task_struct *task = current; + unsigned long flags; + file_sb_list_del(file); - if (unlikely(in_interrupt() || task->flags & PF_KTHREAD)) { - unsigned long flags; - spin_lock_irqsave(&delayed_fput_lock, flags); - list_add(&file->f_u.fu_list, &delayed_fput_list); - schedule_work(&delayed_fput_work); - spin_unlock_irqrestore(&delayed_fput_lock, flags); - return; + if (likely(!in_interrupt() && !(task->flags & PF_KTHREAD))) { + init_task_work(&file->f_u.fu_rcuhead, ____fput); + if (!task_work_add(task, &file->f_u.fu_rcuhead, true)) + return; } - init_task_work(&file->f_u.fu_rcuhead, ____fput); - task_work_add(task, &file->f_u.fu_rcuhead, true); + spin_lock_irqsave(&delayed_fput_lock, flags); + list_add(&file->f_u.fu_list, &delayed_fput_list); + schedule_work(&delayed_fput_work); + spin_unlock_irqrestore(&delayed_fput_lock, flags); } } -- cgit v1.1 From 05252901199d886a68830befb135d1723730ca86 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 6 Jun 2013 19:33:47 -0400 Subject: use can_lookup() instead of direct checks of ->i_op->lookup a couple of places got missed back when Linus has introduced that one... Signed-off-by: Al Viro --- fs/namei.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/fs/namei.c b/fs/namei.c index 85e40d1..9ed9361 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1976,7 +1976,7 @@ static int path_lookupat(int dfd, const char *name, err = complete_walk(nd); if (!err && nd->flags & LOOKUP_DIRECTORY) { - if (!nd->inode->i_op->lookup) { + if (!can_lookup(nd->inode)) { path_put(&nd->path); err = -ENOTDIR; } @@ -2850,7 +2850,7 @@ finish_lookup: if ((open_flag & O_CREAT) && S_ISDIR(nd->inode->i_mode)) goto out; error = -ENOTDIR; - if ((nd->flags & LOOKUP_DIRECTORY) && !nd->inode->i_op->lookup) + if ((nd->flags & LOOKUP_DIRECTORY) && !can_lookup(nd->inode)) goto out; audit_inode(name, nd->path.dentry, 0); finish_open: -- cgit v1.1 From 14c14414d157ea851119c96c61a17306a2b4a035 Mon Sep 17 00:00:00 2001 From: Maxim Patlasov Date: Thu, 13 Jun 2013 12:16:39 +0400 Subject: fuse: hold i_mutex in fuse_file_fallocate() Changing size of a file on server and local update (fuse_write_update_size) should be always protected by inode->i_mutex. Otherwise a race like this is possible: 1. Process 'A' calls fallocate(2) to extend file (~FALLOC_FL_KEEP_SIZE). fuse_file_fallocate() sends FUSE_FALLOCATE request to the server. 2. Process 'B' calls ftruncate(2) shrinking the file. fuse_do_setattr() sends shrinking FUSE_SETATTR request to the server and updates local i_size by i_size_write(inode, outarg.attr.size). 3. Process 'A' resumes execution of fuse_file_fallocate() and calls fuse_write_update_size(inode, offset + length). But 'offset + length' was obsoleted by ftruncate from previous step. Changed in v2 (thanks Brian and Anand for suggestions): - made relation between mutex_lock() and fuse_set_nowrite(inode) more explicit and clear. - updated patch description to use ftruncate(2) in example Signed-off-by: Maxim V. Patlasov Reviewed-by: Brian Foster Signed-off-by: Miklos Szeredi --- fs/fuse/file.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'fs') diff --git a/fs/fuse/file.c b/fs/fuse/file.c index e570081..35f2810 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -2470,13 +2470,16 @@ static long fuse_file_fallocate(struct file *file, int mode, loff_t offset, .mode = mode }; int err; + bool lock_inode = !(mode & FALLOC_FL_KEEP_SIZE) || + (mode & FALLOC_FL_PUNCH_HOLE); if (fc->no_fallocate) return -EOPNOTSUPP; - if (mode & FALLOC_FL_PUNCH_HOLE) { + if (lock_inode) { mutex_lock(&inode->i_mutex); - fuse_set_nowrite(inode); + if (mode & FALLOC_FL_PUNCH_HOLE) + fuse_set_nowrite(inode); } req = fuse_get_req_nopages(fc); @@ -2511,8 +2514,9 @@ static long fuse_file_fallocate(struct file *file, int mode, loff_t offset, fuse_invalidate_attr(inode); out: - if (mode & FALLOC_FL_PUNCH_HOLE) { - fuse_release_nowrite(inode); + if (lock_inode) { + if (mode & FALLOC_FL_PUNCH_HOLE) + fuse_release_nowrite(inode); mutex_unlock(&inode->i_mutex); } -- cgit v1.1 From 7995bd287134f6c8f80d94bebe7396f05a9bc42b Mon Sep 17 00:00:00 2001 From: Al Viro Date: Thu, 20 Jun 2013 18:58:36 +0400 Subject: splice: don't pass the address of ->f_pos to methods Signed-off-by: Al Viro --- fs/internal.h | 6 ++++++ fs/read_write.c | 24 ++++++++++++++++-------- fs/splice.c | 31 ++++++++++++++++++------------- 3 files changed, 40 insertions(+), 21 deletions(-) (limited to 'fs') diff --git a/fs/internal.h b/fs/internal.h index eaa75f7..6812158 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -132,6 +132,12 @@ extern struct dentry *__d_alloc(struct super_block *, const struct qstr *); extern ssize_t __kernel_write(struct file *, const char *, size_t, loff_t *); /* + * splice.c + */ +extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, + loff_t *opos, size_t len, unsigned int flags); + +/* * pipe.c */ extern const struct file_operations pipefifo_fops; diff --git a/fs/read_write.c b/fs/read_write.c index 0343000..2cefa41 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -1064,6 +1064,7 @@ static ssize_t do_sendfile(int out_fd, int in_fd, loff_t *ppos, struct fd in, out; struct inode *in_inode, *out_inode; loff_t pos; + loff_t out_pos; ssize_t retval; int fl; @@ -1077,12 +1078,14 @@ static ssize_t do_sendfile(int out_fd, int in_fd, loff_t *ppos, if (!(in.file->f_mode & FMODE_READ)) goto fput_in; retval = -ESPIPE; - if (!ppos) - ppos = &in.file->f_pos; - else + if (!ppos) { + pos = in.file->f_pos; + } else { + pos = *ppos; if (!(in.file->f_mode & FMODE_PREAD)) goto fput_in; - retval = rw_verify_area(READ, in.file, ppos, count); + } + retval = rw_verify_area(READ, in.file, &pos, count); if (retval < 0) goto fput_in; count = retval; @@ -1099,7 +1102,8 @@ static ssize_t do_sendfile(int out_fd, int in_fd, loff_t *ppos, retval = -EINVAL; in_inode = file_inode(in.file); out_inode = file_inode(out.file); - retval = rw_verify_area(WRITE, out.file, &out.file->f_pos, count); + out_pos = out.file->f_pos; + retval = rw_verify_area(WRITE, out.file, &out_pos, count); if (retval < 0) goto fput_out; count = retval; @@ -1107,7 +1111,6 @@ static ssize_t do_sendfile(int out_fd, int in_fd, loff_t *ppos, if (!max) max = min(in_inode->i_sb->s_maxbytes, out_inode->i_sb->s_maxbytes); - pos = *ppos; if (unlikely(pos + count > max)) { retval = -EOVERFLOW; if (pos >= max) @@ -1126,18 +1129,23 @@ static ssize_t do_sendfile(int out_fd, int in_fd, loff_t *ppos, if (in.file->f_flags & O_NONBLOCK) fl = SPLICE_F_NONBLOCK; #endif - retval = do_splice_direct(in.file, ppos, out.file, count, fl); + retval = do_splice_direct(in.file, &pos, out.file, &out_pos, count, fl); if (retval > 0) { add_rchar(current, retval); add_wchar(current, retval); fsnotify_access(in.file); fsnotify_modify(out.file); + out.file->f_pos = out_pos; + if (ppos) + *ppos = pos; + else + in.file->f_pos = pos; } inc_syscr(current); inc_syscw(current); - if (*ppos > max) + if (pos > max) retval = -EOVERFLOW; fput_out: diff --git a/fs/splice.c b/fs/splice.c index e6b2559..9eca476 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -1274,7 +1274,7 @@ static int direct_splice_actor(struct pipe_inode_info *pipe, { struct file *file = sd->u.file; - return do_splice_from(pipe, file, &file->f_pos, sd->total_len, + return do_splice_from(pipe, file, sd->opos, sd->total_len, sd->flags); } @@ -1294,7 +1294,7 @@ static int direct_splice_actor(struct pipe_inode_info *pipe, * */ long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, - size_t len, unsigned int flags) + loff_t *opos, size_t len, unsigned int flags) { struct splice_desc sd = { .len = len, @@ -1302,6 +1302,7 @@ long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, .flags = flags, .pos = *ppos, .u.file = out, + .opos = opos, }; long ret; @@ -1325,7 +1326,7 @@ static long do_splice(struct file *in, loff_t __user *off_in, { struct pipe_inode_info *ipipe; struct pipe_inode_info *opipe; - loff_t offset, *off; + loff_t offset; long ret; ipipe = get_pipe_info(in); @@ -1356,13 +1357,15 @@ static long do_splice(struct file *in, loff_t __user *off_in, return -EINVAL; if (copy_from_user(&offset, off_out, sizeof(loff_t))) return -EFAULT; - off = &offset; - } else - off = &out->f_pos; + } else { + offset = out->f_pos; + } - ret = do_splice_from(ipipe, out, off, len, flags); + ret = do_splice_from(ipipe, out, &offset, len, flags); - if (off_out && copy_to_user(off_out, off, sizeof(loff_t))) + if (!off_out) + out->f_pos = offset; + else if (copy_to_user(off_out, &offset, sizeof(loff_t))) ret = -EFAULT; return ret; @@ -1376,13 +1379,15 @@ static long do_splice(struct file *in, loff_t __user *off_in, return -EINVAL; if (copy_from_user(&offset, off_in, sizeof(loff_t))) return -EFAULT; - off = &offset; - } else - off = &in->f_pos; + } else { + offset = in->f_pos; + } - ret = do_splice_to(in, off, opipe, len, flags); + ret = do_splice_to(in, &offset, opipe, len, flags); - if (off_in && copy_to_user(off_in, off, sizeof(loff_t))) + if (!off_in) + in->f_pos = offset; + else if (copy_to_user(off_in, &offset, sizeof(loff_t))) ret = -EFAULT; return ret; -- cgit v1.1 From acdb37c361dc87e165889a504e291c1e82ae133c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 22 Jun 2013 19:44:08 -0700 Subject: fs: fix new splice.c kernel-doc warning Fix new kernel-doc warning in fs/splice.c: Warning(fs/splice.c:1298): No description found for parameter 'opos' Signed-off-by: Randy Dunlap Cc: Al Viro Signed-off-by: Linus Torvalds --- fs/splice.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/splice.c b/fs/splice.c index 9eca476..d37431d 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -1283,6 +1283,7 @@ static int direct_splice_actor(struct pipe_inode_info *pipe, * @in: file to splice from * @ppos: input file offset * @out: file to splice to + * @opos: output file offset * @len: number of bytes to splice * @flags: splice modifier flags * -- cgit v1.1 From 2976b10f05bd7f6dab9f9e7524451ddfed656a89 Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Thu, 20 Jun 2013 11:36:28 +0200 Subject: perf: Disable monitoring on setuid processes for regular users There was a a bug in setup_new_exec(), whereby the test to disabled perf monitoring was not correct because the new credentials for the process were not yet committed and therefore the get_dumpable() test was never firing. The patch fixes the problem by moving the perf_event test until after the credentials are committed. Signed-off-by: Stephane Eranian Tested-by: Jiri Olsa Acked-by: Peter Zijlstra Cc: Signed-off-by: Ingo Molnar --- fs/exec.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) (limited to 'fs') diff --git a/fs/exec.c b/fs/exec.c index 6430195..ffd7a81 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1135,13 +1135,6 @@ void setup_new_exec(struct linux_binprm * bprm) set_dumpable(current->mm, suid_dumpable); } - /* - * Flush performance counters when crossing a - * security domain: - */ - if (!get_dumpable(current->mm)) - perf_event_exit_task(current); - /* An exec changes our domain. We are no longer part of the thread group */ @@ -1205,6 +1198,15 @@ void install_exec_creds(struct linux_binprm *bprm) commit_creds(bprm->cred); bprm->cred = NULL; + + /* + * Disable monitoring for regular users + * when executing setuid binaries. Must + * wait until new credentials are committed + * by commit_creds() above + */ + if (get_dumpable(current->mm) != SUID_DUMP_USER) + perf_event_exit_task(current); /* * cred_guard_mutex must be held at least to this point to prevent * ptrace_attach() from altering our determination of the task's -- cgit v1.1 From 33f1a63ae84dfd9ad298cf275b8f1887043ced36 Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Fri, 28 Jun 2013 14:15:14 +0300 Subject: UBIFS: prepare to fix a horrid bug Al Viro pointed me to the fact that '->readdir()' and '->llseek()' have no mutual exclusion, which means the 'ubifs_dir_llseek()' can be run while we are in the middle of 'ubifs_readdir()'. First of all, this means that 'file->private_data' can be freed while 'ubifs_readdir()' uses it. But this particular patch does not fix the problem. This patch is only a preparation, and the fix will follow next. In this patch we make 'ubifs_readdir()' stop using 'file->f_pos' directly, because 'file->f_pos' can be changed by '->llseek()' at any point. This may lead 'ubifs_readdir()' to returning inconsistent data: directory entry names may correspond to incorrect file positions. So here we introduce a local variable 'pos', read 'file->f_pose' once at very the beginning, and then stick to 'pos'. The result of this is that when 'ubifs_dir_llseek()' changes 'file->f_pos' while we are in the middle of 'ubifs_readdir()', the latter "wins". Cc: stable@vger.kernel.org Reported-by: Al Viro Tested-by: Artem Bityutskiy Signed-off-by: Artem Bityutskiy Signed-off-by: Al Viro --- fs/ubifs/dir.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index de08c92f..8e587af 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -349,15 +349,16 @@ static unsigned int vfs_dent_type(uint8_t type) static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) { int err, over = 0; + loff_t pos = file->f_pos; struct qstr nm; union ubifs_key key; struct ubifs_dent_node *dent; struct inode *dir = file_inode(file); struct ubifs_info *c = dir->i_sb->s_fs_info; - dbg_gen("dir ino %lu, f_pos %#llx", dir->i_ino, file->f_pos); + dbg_gen("dir ino %lu, f_pos %#llx", dir->i_ino, pos); - if (file->f_pos > UBIFS_S_KEY_HASH_MASK || file->f_pos == 2) + if (pos > UBIFS_S_KEY_HASH_MASK || pos == 2) /* * The directory was seek'ed to a senseless position or there * are no more entries. @@ -365,15 +366,15 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) return 0; /* File positions 0 and 1 correspond to "." and ".." */ - if (file->f_pos == 0) { + if (pos == 0) { ubifs_assert(!file->private_data); over = filldir(dirent, ".", 1, 0, dir->i_ino, DT_DIR); if (over) return 0; - file->f_pos = 1; + file->f_pos = pos = 1; } - if (file->f_pos == 1) { + if (pos == 1) { ubifs_assert(!file->private_data); over = filldir(dirent, "..", 2, 1, parent_ino(file->f_path.dentry), DT_DIR); @@ -389,7 +390,7 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) goto out; } - file->f_pos = key_hash_flash(c, &dent->key); + file->f_pos = pos = key_hash_flash(c, &dent->key); file->private_data = dent; } @@ -397,17 +398,16 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) if (!dent) { /* * The directory was seek'ed to and is now readdir'ed. - * Find the entry corresponding to @file->f_pos or the - * closest one. + * Find the entry corresponding to @pos or the closest one. */ - dent_key_init_hash(c, &key, dir->i_ino, file->f_pos); + dent_key_init_hash(c, &key, dir->i_ino, pos); nm.name = NULL; dent = ubifs_tnc_next_ent(c, &key, &nm); if (IS_ERR(dent)) { err = PTR_ERR(dent); goto out; } - file->f_pos = key_hash_flash(c, &dent->key); + file->f_pos = pos = key_hash_flash(c, &dent->key); file->private_data = dent; } @@ -419,7 +419,7 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) ubifs_inode(dir)->creat_sqnum); nm.len = le16_to_cpu(dent->nlen); - over = filldir(dirent, dent->name, nm.len, file->f_pos, + over = filldir(dirent, dent->name, nm.len, pos, le64_to_cpu(dent->inum), vfs_dent_type(dent->type)); if (over) @@ -435,7 +435,7 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) } kfree(file->private_data); - file->f_pos = key_hash_flash(c, &dent->key); + file->f_pos = pos = key_hash_flash(c, &dent->key); file->private_data = dent; cond_resched(); } -- cgit v1.1 From 605c912bb843c024b1ed173dc427cd5c08e5d54d Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Fri, 28 Jun 2013 14:15:15 +0300 Subject: UBIFS: fix a horrid bug Al Viro pointed me to the fact that '->readdir()' and '->llseek()' have no mutual exclusion, which means the 'ubifs_dir_llseek()' can be run while we are in the middle of 'ubifs_readdir()'. This means that 'file->private_data' can be freed while 'ubifs_readdir()' uses it, and this is a very bad bug: not only 'ubifs_readdir()' can return garbage, but this may corrupt memory and lead to all kinds of problems like crashes an security holes. This patch fixes the problem by using the 'file->f_version' field, which '->llseek()' always unconditionally sets to zero. We set it to 1 in 'ubifs_readdir()' and whenever we detect that it became 0, we know there was a seek and it is time to clear the state saved in 'file->private_data'. I tested this patch by writing a user-space program which runds readdir and seek in parallell. I could easily crash the kernel without these patches, but could not crash it with these patches. Cc: stable@vger.kernel.org Reported-by: Al Viro Tested-by: Artem Bityutskiy Signed-off-by: Artem Bityutskiy Signed-off-by: Al Viro --- fs/ubifs/dir.c | 30 +++++++++++++++++++++++++++--- 1 file changed, 27 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 8e587af..605af51 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -365,6 +365,24 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) */ return 0; + if (file->f_version == 0) { + /* + * The file was seek'ed, which means that @file->private_data + * is now invalid. This may also be just the first + * 'ubifs_readdir()' invocation, in which case + * @file->private_data is NULL, and the below code is + * basically a no-op. + */ + kfree(file->private_data); + file->private_data = NULL; + } + + /* + * 'generic_file_llseek()' unconditionally sets @file->f_version to + * zero, and we use this for detecting whether the file was seek'ed. + */ + file->f_version = 1; + /* File positions 0 and 1 correspond to "." and ".." */ if (pos == 0) { ubifs_assert(!file->private_data); @@ -438,6 +456,14 @@ static int ubifs_readdir(struct file *file, void *dirent, filldir_t filldir) file->f_pos = pos = key_hash_flash(c, &dent->key); file->private_data = dent; cond_resched(); + + if (file->f_version == 0) + /* + * The file was seek'ed meanwhile, lets return and start + * reading direntries from the new position on the next + * invocation. + */ + return 0; } out: @@ -448,15 +474,13 @@ out: kfree(file->private_data); file->private_data = NULL; + /* 2 is a special value indicating that there are no more direntries */ file->f_pos = 2; return 0; } -/* If a directory is seeked, we have to free saved readdir() state */ static loff_t ubifs_dir_llseek(struct file *file, loff_t offset, int whence) { - kfree(file->private_data); - file->private_data = NULL; return generic_file_llseek(file, offset, whence); } -- cgit v1.1