| Message ID | 1391220136-22034-1-git-send-email-fdmanana@gmail.com (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
On Sat, Feb 1, 2014 at 2:02 AM, Filipe David Borba Manana <fdmanana@gmail.com> wrote: > The commit titled "Btrfs: fix infinite path build loops in incremental send" > didn't cover a particular case where the parent-child relationship inversion > of directories doesn't imply a rename of the new parent directory. This was > due to a simple logic mistake, a logical and instead of a logical or. > > Steps to reproduce: > > $ mkfs.btrfs -f /dev/sdb3 > $ mount /dev/sdb3 /mnt/btrfs > $ mkdir -p /mnt/btrfs/a/b/bar1/bar2/bar3/bar4 > $ btrfs subvol snapshot -r /mnt/btrfs /mnt/btrfs/snap1 > $ mv /mnt/btrfs/a/b/bar1/bar2/bar3/bar4 /mnt/btrfs/a/b/k44 > $ mv /mnt/btrfs/a/b/bar1/bar2/bar3 /mnt/btrfs/a/b/k44 > $ mv /mnt/btrfs/a/b/bar1/bar2 /mnt/btrfs/a/b/k44/bar3 > $ mv /mnt/btrfs/a/b/bar1 /mnt/btrfs/a/b/k44/bar3/bar2/k11 > $ btrfs subvol snapshot -r /mnt/btrfs /mnt/btrfs/snap2 > $ btrfs send -p /mnt/btrfs/snap1 /mnt/btrfs/snap2 > /tmp/incremental.send > > A patch to update the test btrfs/030 from xfstests, so that it covers > this case, will be submitted soon. > > Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com> > --- Hi, Should this go to 3.14 too? It is covered by the test case btrfs/030 for xfstests. thanks > fs/btrfs/send.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c > index e0b49f6..c47fcef 100644 > --- a/fs/btrfs/send.c > +++ b/fs/btrfs/send.c > @@ -3041,8 +3041,8 @@ static int wait_for_parent_move(struct send_ctx *sctx, > > len1 = fs_path_len(path_before); > len2 = fs_path_len(path_after); > - if ((parent_ino_before != parent_ino_after) && (len1 != len2 || > - memcmp(path_before->start, path_after->start, len1))) { > + if (parent_ino_before != parent_ino_after || len1 != len2 || > + memcmp(path_before->start, path_after->start, len1)) { > ret = 1; > goto out; > } > -- > 1.7.9.5 >
diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c index e0b49f6..c47fcef 100644 --- a/fs/btrfs/send.c +++ b/fs/btrfs/send.c @@ -3041,8 +3041,8 @@ static int wait_for_parent_move(struct send_ctx *sctx, len1 = fs_path_len(path_before); len2 = fs_path_len(path_after); - if ((parent_ino_before != parent_ino_after) && (len1 != len2 || - memcmp(path_before->start, path_after->start, len1))) { + if (parent_ino_before != parent_ino_after || len1 != len2 || + memcmp(path_before->start, path_after->start, len1)) { ret = 1; goto out; }
The commit titled "Btrfs: fix infinite path build loops in incremental send" didn't cover a particular case where the parent-child relationship inversion of directories doesn't imply a rename of the new parent directory. This was due to a simple logic mistake, a logical and instead of a logical or. Steps to reproduce: $ mkfs.btrfs -f /dev/sdb3 $ mount /dev/sdb3 /mnt/btrfs $ mkdir -p /mnt/btrfs/a/b/bar1/bar2/bar3/bar4 $ btrfs subvol snapshot -r /mnt/btrfs /mnt/btrfs/snap1 $ mv /mnt/btrfs/a/b/bar1/bar2/bar3/bar4 /mnt/btrfs/a/b/k44 $ mv /mnt/btrfs/a/b/bar1/bar2/bar3 /mnt/btrfs/a/b/k44 $ mv /mnt/btrfs/a/b/bar1/bar2 /mnt/btrfs/a/b/k44/bar3 $ mv /mnt/btrfs/a/b/bar1 /mnt/btrfs/a/b/k44/bar3/bar2/k11 $ btrfs subvol snapshot -r /mnt/btrfs /mnt/btrfs/snap2 $ btrfs send -p /mnt/btrfs/snap1 /mnt/btrfs/snap2 > /tmp/incremental.send A patch to update the test btrfs/030 from xfstests, so that it covers this case, will be submitted soon. Signed-off-by: Filipe David Borba Manana <fdmanana@gmail.com> --- fs/btrfs/send.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)