From: Pedro Fonseca <pfonseca@mpi-sws.org>
To: Chris Mason <clm@fb.com>, linux-btrfs@vger.kernel.org
Subject: Re: Warning in inode_tree_add() -- !(entry->vfs_inode.i_state & (I_WILL_FREE | I_FREEING)
Date: Tue, 10 Dec 2013 16:53:08 +0100 [thread overview]
Message-ID: <52A738E4.5070706@mpi-sws.org> (raw)
In-Reply-To: <20131210132529.16869.59608@ret>
On 12/10/13 14:25 , Chris Mason wrote:
> Quoting Pedro Fonseca (2013-12-09 16:15:46)
>> Hi,
>>
>> I've got multiple instances of warnings caused by the statement
>> "WARN_ON(!(entry->vfs_inode.i_state & (I_WILL_FREE | I_FREEING)));" in
>> inode_tree_add(). The tests were executed on the official 3.12.2 kernel.
>> Do you think it's something serious?
>
> Definitely. You're in the autodefrag code here and an iget is replacing
> an existing inode that wasn't ready to be replaced. What test are you
> running?
>
> -chris
>
I'm running a custom test based on fsstress and on a QEMU based VMM. Btrfs was statically compiled and the btrfs filesystem was mounted on one or two loop devices.
Bellow I've tried to annotate the last file system operations with their parameters (e.g., "filename", "[start, len]",...). Note that the log only specifies the starting
order of the operations and note that operations on different CPUs may overlap (i.e. execute concurrently).
Let me know if you need additional information.
Pedro
* Test 1:
CPU: 1 Op: dread (d16/f22 [233472,90112])
CPU: 1 Op: chown (d16/da6/c111)
CPU: 1 Op: dwrite (d16/d21/d38/fad [299008,65536])
CPU: 0 Op: chown (d16/da6/fe2)
CPU: 0 Op: rmdir (d16/d21/d6d/d74 39)
CPU: 0 Op: chown (fb)
CPU: 0 Op: chown (d16/d21/f96)
CPU: 0 Op: chown (d16/d21/d38/f66)
CPU: 0 Op: btrfs_subvol_list (d16)
CPU: 0 Op: btrfs_ioctl_filesystem_resize (+8m /mnt/dir)
CPU: 1 Op: write (d16/d21/d6d/feb [1581147,53815])
CPU: 1 Op: read (d16/d21/fc2 [159384,10768])
CPU: 1 Op: dwrite (f12 [1990656,90112])
CPU: 1 Op: dread (d16/d21/d6d/fa4 zero size)
CPU: 1 Op: btrfs_ioctl_filesystem_defragment (d16/f22)
* Test 2:
CPU: 1 Op: dread (d16/f22 [233472,90112])
CPU: 1 Op: chown (d16/da6/c111)
CPU: 1 Op: dwrite (d16/d21/d38/fad [299008,65536])
CPU: 1 Op: write (d16/d21/d6d/feb [1581147,53815])
CPU: 1 Op: read (d16/d21/fc2 [159384,10768])
CPU: 1 Op: dwrite (f12 [1990656,90112])
CPU: 0 Op: dread (d16/f22 [233472,90112])
CPU: 0 Op: chown (d16/da6/c111)
CPU: 0 Op: dwrite (d16/d21/d38/fad [299008,65536])
CPU: 1 Op: dread (d16/d21/d6d/fa4 zero size)
CPU: 1 Op: btrfs_ioctl_filesystem_defragment (d16/f22)
CPU: 0 Op: write (d16/d21/d6d/feb [2193555,53815])
CPU: 0 Op: read (d16/d21/fc2 [159384,10768])
CPU: 0 Op: dwrite (f12 [688128,90112])
CPU: 0 Op: dread (d16/d21/d6d/fa4 zero size)
CPU: 0 Op: btrfs_ioctl_filesystem_defragment (d16/f22)
* Test 3:
CPU: 1 Op: dread (d16/dce/fe1 zero size)
CPU: 1 Op: stat (stat d16/da6/f117)
CPU: 1 Op: btrfs_ioctl_filesystem_resize (-5m /mnt/dir)
CPU: 1 Op: write (ff [956340,50300])
CPU: 1 Op: mkdir (d16/d13b)
CPU: 1 Op: creat (d16/d21/d6d/d74/f13c x:0 0 0)
CPU: 1 Op: chown (d16/f116)
CPU: 1 Op: rename ( d16/d21/d38/l7c to d16/d21/d38/l13d)
CPU: 0 Op: dread (d16/f22 [233472,90112])
CPU: 0 Op: chown (d16/da6/c111)
CPU: 0 Op: dwrite (d16/d21/d38/fad [299008,65536])
CPU: 0 Op: write (d16/d21/d6d/feb [1581147,53815])
CPU: 0 Op: read (d16/d21/fc2 [159384,10768])
CPU: 0 Op: dwrite (f12 [1990656,90112])
CPU: 0 Op: dread (d16/d21/d6d/fa4 zero size)
CPU: 0 Op: btrfs_ioctl_filesystem_defragment (d16/f22)
prev parent reply other threads:[~2013-12-10 15:46 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-12-09 21:15 Warning in inode_tree_add() -- !(entry->vfs_inode.i_state & (I_WILL_FREE | I_FREEING) Pedro Fonseca
2013-12-10 13:25 ` Chris Mason
2013-12-10 15:53 ` Pedro Fonseca [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=52A738E4.5070706@mpi-sws.org \
--to=pfonseca@mpi-sws.org \
--cc=clm@fb.com \
--cc=linux-btrfs@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).