| Message ID | 1360272524-19548-1-git-send-email-jbacik@fusionio.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, 7 Feb 2013 16:28:44 -0500, Josef Bacik wrote: > I noticed we were getting lots of warnings with xfstest 83 because we have > reservations outstanding. This is because we moved the orphan add outside > of the truncate, but we don't actually cleanup our reservation if something > fails. This fixes the problem and I no longer see warnings. Thanks, > > Signed-off-by: Josef Bacik <jbacik@fusionio.com> > --- > fs/btrfs/inode.c | 2 ++ > 1 files changed, 2 insertions(+), 0 deletions(-) > > diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c > index 73e9409..905297f 100644 > --- a/fs/btrfs/inode.c > +++ b/fs/btrfs/inode.c > @@ -2538,6 +2538,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) > goto out; > > ret = btrfs_truncate(inode); > + if (ret) > + btrfs_orphan_del(NULL, inode); > } else { > nr_unlink++; > } There are more warnings when running xfstest 083. With btrfs-next plus your two new patches ("[PATCH] Btrfs: steal from global reserve if we are cleaning up orphans V2" and "[PATCH] Btrfs: cleanup orphan reservation if truncate fails") I get a warning (see below) almost every time when I run xfstest 083. If I apply only the 1st patch, it happens in 50% of the cases. And if I apply only the 2nd patch, the warning and the "Busy inodes" message happens in 50% of the cases as well. With btrfs-next _without_ your new patches, the same thing happens in 3% of the cases. You made it happen more often with your new patches. run xfstest 083 btrfs: disk space caching is enabled Btrfs detected SSD devices, enabling SSD mode ------------[ cut here ]------------ WARNING: at fs/btrfs/disk-io.c:3227 free_fs_root+0x86/0x90 [btrfs]() Hardware name: X8SIL Modules linked in: btrfs raid1 mpt2sas scsi_transport_sas raid_class [last unloaded: btrfs] Pid: 24225, comm: umount Tainted: G B W 3.7.0+ #27 Call Trace: [<ffffffff8108642a>] warn_slowpath_common+0x7a/0xb0 [<ffffffff81086475>] warn_slowpath_null+0x15/0x20 [<ffffffffa03403a6>] free_fs_root+0x86/0x90 [btrfs] [<ffffffffa03438eb>] btrfs_free_fs_root+0x7b/0x90 [btrfs] [<ffffffffa03439cf>] del_fs_roots+0xcf/0x110 [btrfs] [<ffffffffa033809f>] ? btrfs_free_block_groups+0x28f/0x360 [btrfs] [<ffffffffa0345240>] close_ctree+0x1d0/0x350 [btrfs] [<ffffffff811b1dc4>] ? evict_inodes+0x104/0x120 [<ffffffffa031a284>] btrfs_put_super+0x14/0x20 [btrfs] [<ffffffff8119949c>] generic_shutdown_super+0x5c/0xe0 [<ffffffff811995b1>] kill_anon_super+0x11/0x20 [<ffffffffa031e3a5>] btrfs_kill_super+0x15/0x90 [btrfs] [<ffffffff8119a601>] ? deactivate_super+0x41/0x70 [<ffffffff811999bd>] deactivate_locked_super+0x3d/0x70 [<ffffffff8119a609>] deactivate_super+0x49/0x70 [<ffffffff811b5b3c>] mntput_no_expire+0xfc/0x160 [<ffffffff811b6af1>] sys_umount+0x71/0x3a0 [<ffffffff81963fd2>] system_call_fastpath+0x16/0x1b ---[ end trace 0742f5dfe09cd270 ]--- VFS: Busy inodes after unmount of sds. Self-destruct in 5 seconds. Have a nice day... The VFS style "Have a nice day" is much more friendly than the Btrfs style "cleanup orphan" or "truncate orphan" BTW. -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 73e9409..905297f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2538,6 +2538,8 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) goto out; ret = btrfs_truncate(inode); + if (ret) + btrfs_orphan_del(NULL, inode); } else { nr_unlink++; }
I noticed we were getting lots of warnings with xfstest 83 because we have reservations outstanding. This is because we moved the orphan add outside of the truncate, but we don't actually cleanup our reservation if something fails. This fixes the problem and I no longer see warnings. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com> --- fs/btrfs/inode.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)