linux-xfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ross Zwisler <ross.zwisler@linux.intel.com>
To: Dave Chinner <david@fromorbit.com>
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>, linux-xfs@vger.kernel.org
Subject: Re: XFS: Assertion failed: !rwsem_is_locked(&inode->i_rwsem)
Date: Mon, 9 Jul 2018 16:04:02 -0600	[thread overview]
Message-ID: <20180709220402.GA26254@linux.intel.com> (raw)
In-Reply-To: <20180619232953.GK19934@dastard>

On Wed, Jun 20, 2018 at 09:29:54AM +1000, Dave Chinner wrote:
> On Tue, Jun 19, 2018 at 10:44:20AM -0600, Ross Zwisler wrote:
> > On Tue, Jun 19, 2018 at 12:32:46PM +1000, Dave Chinner wrote:
> > > On Mon, Jun 18, 2018 at 08:17:46PM -0600, Ross Zwisler wrote:
> > > > During some xfstest runs on next-20180615 I hit the following with DAX +
> > > > generic/388:
> > > > 
> > > > ================================================
> > > > WARNING: lock held when returning to user space!
> > > > 4.17.0-next-20180615-00001-gf09d99951966 #2 Not tainted
> > > > ------------------------------------------------
> > > > fsstress/6598 is leaving the kernel with locks still held!
> > > > 2 locks held by fsstress/6598:
> > > >  #0: 00000000d8f89e14 (&sb->s_type->i_mutex_key#13){++++}, at: xfs_ilock+0x211/0x310
> > > >  #1: 000000005cc93137 (&(&ip->i_mmaplock)->mr_lock){++++}, at: xfs_ilock+0x1eb/0x310
> > > 
> > > What errors occurred before this? generic/388 is testing all sorts
> > > of error paths by randomly shutting down the filesystem, so it'e
> > > entirely possible that we've leaked those locks (XFS_IOLOCK and
> > > XFS_MMAPLOCK) on some rarely travelled error path. The prior errors
> > > might help identify that path.
> > 
> > Here is the full output from another reproduction:
> ....
> >  XFS (pmem0p2): DAX enabled. Warning: EXPERIMENTAL, use at your own risk
> >  XFS (pmem0p2): Mounting V5 Filesystem
> >  XFS (pmem0p2): Starting recovery (logdev: internal)
> >  XFS (pmem0p2): Ending recovery (logdev: internal)
> >  XFS (pmem0p2): xfs_imap_lookup: xfs_ialloc_read_agi() returned error -5, agno 0
> >  
> >  ================================================
> >  WARNING: lock held when returning to user space!
> >  4.17.0-next-20180615 #1 Not tainted
> >  ------------------------------------------------
> 
> Ok, nothing extra to go on there. can you get lockdep to dump the
> stack or oops so we at least know what syscall was being run when
> this is detected?

Well, I can't seem to reproduce this reliably anymore.  :(

I did reproduce it once with the debug you requested, and the syscall that wsa
being run when we found the lock imbalance was __x64_sys_ioctl, for whatever
that's worth.

      reply	other threads:[~2018-07-09 22:04 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-06-19  2:17 XFS: Assertion failed: !rwsem_is_locked(&inode->i_rwsem) Ross Zwisler
2018-06-19  2:32 ` Dave Chinner
2018-06-19 16:44   ` Ross Zwisler
2018-06-19 23:29     ` Dave Chinner
2018-07-09 22:04       ` Ross Zwisler [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=20180709220402.GA26254@linux.intel.com \
    --to=ross.zwisler@linux.intel.com \
    --cc=david@fromorbit.com \
    --cc=linux-xfs@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).