From 9c3b306e1c9e6be4be09e99a8fe2227d1005effc Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Thu, 31 Jul 2014 14:41:07 +0100 Subject: Btrfs: race free update of commit root for ro snapshots This is a better solution for the problem addressed in the following commit: Btrfs: update commit root on snapshot creation after orphan cleanup (3821f348889e506efbd268cc8149e0ebfa47c4e5) The previous solution wasn't the best because of 2 reasons: 1) It added another full transaction commit, which is more expensive than just swapping the commit root with the root; 2) If a reboot happened after the first transaction commit (the one that creates the snapshot) and before the second transaction commit, then we would end up with the same problem if a send using that snapshot was requested before the first transaction commit after the reboot. This change addresses those 2 issues. The second issue is addressed by switching the commit root in the dentry lookup VFS callback, which is also called by the snapshot/subvol creation ioctl and performs orphan cleanup if needed. Like the vfs, the ioctl locks the parent inode too, preventing race issues between a dentry lookup and snapshot creation. Cc: Alex Lyakas Signed-off-by: Filipe Manana Signed-off-by: Chris Mason --- fs/btrfs/ioctl.c | 33 --------------------------------- 1 file changed, 33 deletions(-) (limited to 'fs/btrfs/ioctl.c') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 47aceb494d1d..845287ca59c3 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -711,39 +711,6 @@ static int create_snapshot(struct btrfs_root *root, struct inode *dir, if (ret) goto fail; - ret = btrfs_orphan_cleanup(pending_snapshot->snap); - if (ret) - goto fail; - - /* - * If orphan cleanup did remove any orphans, it means the tree was - * modified and therefore the commit root is not the same as the - * current root anymore. This is a problem, because send uses the - * commit root and therefore can see inode items that don't exist - * in the current root anymore, and for example make calls to - * btrfs_iget, which will do tree lookups based on the current root - * and not on the commit root. Those lookups will fail, returning a - * -ESTALE error, and making send fail with that error. So make sure - * a send does not see any orphans we have just removed, and that it - * will see the same inodes regardless of whether a transaction - * commit happened before it started (meaning that the commit root - * will be the same as the current root) or not. - */ - if (readonly && pending_snapshot->snap->node != - pending_snapshot->snap->commit_root) { - trans = btrfs_join_transaction(pending_snapshot->snap); - if (IS_ERR(trans) && PTR_ERR(trans) != -ENOENT) { - ret = PTR_ERR(trans); - goto fail; - } - if (!IS_ERR(trans)) { - ret = btrfs_commit_transaction(trans, - pending_snapshot->snap); - if (ret) - goto fail; - } - } - inode = btrfs_lookup_dentry(dentry->d_parent->d_inode, dentry); if (IS_ERR(inode)) { ret = PTR_ERR(inode); -- cgit v1.2.3 From 62e2390e1ad78f956e96a6a831761adc6f2bf58a Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Fri, 8 Aug 2014 02:47:06 +0100 Subject: Btrfs: clone, don't create invalid hole extent map When cloning a file that consists of an inline extent, we were creating an extent map that represents a non-existing trailing hole starting at a file offset that isn't a multiple of the sector size. This happened because when processing an inline extent we weren't aligning the extent's length to the sector size, and therefore incorrectly treating the range [inline_extent_length; sector_size[ as a hole. Signed-off-by: Filipe Manana Reviewed-by: Satoru Takeuchi Signed-off-by: Chris Mason --- fs/btrfs/ioctl.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'fs/btrfs/ioctl.c') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 845287ca59c3..fce6fd0e3f50 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -3494,7 +3494,8 @@ process_slot: btrfs_mark_buffer_dirty(leaf); btrfs_release_path(path); - last_dest_end = new_key.offset + datal; + last_dest_end = ALIGN(new_key.offset + datal, + root->sectorsize); ret = clone_finish_inode_update(trans, inode, last_dest_end, destoff, olen); -- cgit v1.2.3 From e9512d72e8e61c750c90efacd720abe3c4569822 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 26 Aug 2014 13:55:54 -0700 Subject: Btrfs: fix autodefrag with compression The autodefrag code skips defrag when two extents are adjacent. But one big advantage for autodefrag is cutting down on the number of small extents, even when they are adjacent. This commit changes it to defrag all small extents. Signed-off-by: Chris Mason --- fs/btrfs/ioctl.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) (limited to 'fs/btrfs/ioctl.c') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index fce6fd0e3f50..a018ea484d39 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1019,8 +1019,10 @@ static bool defrag_check_next_extent(struct inode *inode, struct extent_map *em) return false; next = defrag_lookup_extent(inode, em->start + em->len); - if (!next || next->block_start >= EXTENT_MAP_LAST_BYTE || - (em->block_start + em->block_len == next->block_start)) + if (!next || next->block_start >= EXTENT_MAP_LAST_BYTE) + ret = false; + else if ((em->block_start + em->block_len == next->block_start) && + (em->block_len > 128 * 1024 && next->block_len > 128 * 1024)) ret = false; free_extent_map(next); @@ -1055,7 +1057,6 @@ static int should_defrag_range(struct inode *inode, u64 start, int thresh, } next_mergeable = defrag_check_next_extent(inode, em); - /* * we hit a real extent, if it is big or the next extent is not a * real extent, don't bother defragging it -- cgit v1.2.3 From c47ca32d3aadb234f73389a34c97574085bc9eda Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Thu, 4 Sep 2014 14:09:15 +0300 Subject: Btrfs: kfree()ing ERR_PTRs The "inherit" in btrfs_ioctl_snap_create_v2() and "vol_args" in btrfs_ioctl_rm_dev() are ERR_PTRs so we can't call kfree() on them. These kind of bugs are "One Err Bugs" where there is just one error label that does everything. I could set the "inherit = NULL" and keep the single out label but it ends up being more complicated that way. It makes the code simpler to re-order the unwind so it's in the mirror order of the allocation and introduce some new error labels. Signed-off-by: Dan Carpenter Signed-off-by: Chris Mason --- fs/btrfs/ioctl.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) (limited to 'fs/btrfs/ioctl.c') diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index a018ea484d39..8a8e29878c34 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1703,7 +1703,7 @@ static noinline int btrfs_ioctl_snap_create_v2(struct file *file, ~(BTRFS_SUBVOL_CREATE_ASYNC | BTRFS_SUBVOL_RDONLY | BTRFS_SUBVOL_QGROUP_INHERIT)) { ret = -EOPNOTSUPP; - goto out; + goto free_args; } if (vol_args->flags & BTRFS_SUBVOL_CREATE_ASYNC) @@ -1713,27 +1713,31 @@ static noinline int btrfs_ioctl_snap_create_v2(struct file *file, if (vol_args->flags & BTRFS_SUBVOL_QGROUP_INHERIT) { if (vol_args->size > PAGE_CACHE_SIZE) { ret = -EINVAL; - goto out; + goto free_args; } inherit = memdup_user(vol_args->qgroup_inherit, vol_args->size); if (IS_ERR(inherit)) { ret = PTR_ERR(inherit); - goto out; + goto free_args; } } ret = btrfs_ioctl_snap_create_transid(file, vol_args->name, vol_args->fd, subvol, ptr, readonly, inherit); + if (ret) + goto free_inherit; - if (ret == 0 && ptr && - copy_to_user(arg + - offsetof(struct btrfs_ioctl_vol_args_v2, - transid), ptr, sizeof(*ptr))) + if (ptr && copy_to_user(arg + + offsetof(struct btrfs_ioctl_vol_args_v2, + transid), + ptr, sizeof(*ptr))) ret = -EFAULT; -out: - kfree(vol_args); + +free_inherit: kfree(inherit); +free_args: + kfree(vol_args); return ret; } @@ -2653,7 +2657,7 @@ static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg) vol_args = memdup_user(arg, sizeof(*vol_args)); if (IS_ERR(vol_args)) { ret = PTR_ERR(vol_args); - goto out; + goto err_drop; } vol_args->name[BTRFS_PATH_NAME_MAX] = '\0'; @@ -2671,6 +2675,7 @@ static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg) out: kfree(vol_args); +err_drop: mnt_drop_write_file(file); return ret; } -- cgit v1.2.3