On Tue, 2009-09-01 at 18:11 -0400, J. Bruce Fields wrote:
> I'm getting this, running connectathon tests on a kernel which includes your
> latest nfs-for-2.6.32. Is it new?
No, you are by far the first one to report this. It dates back a couple
of years at least. The last time it surfaced was here:
http://thread.gmane.org/gmane.linux.nfs/27406
Peter Zijlstra has a fix, but it hasn't yet been merged into mainline:
http://programming.kicks-ass.net/kernel-patches/mmap-vs-nfs/
Cheers
Trond
> --b.
>
> =======================================================
> [ INFO: possible circular locking dependency detected ]
> 2.6.31-rc8-00162-gf7fcab6 #321
> -------------------------------------------------------
> bigfile/3763 is trying to acquire lock:
> (&sb->s_type->i_mutex_key#11){+.+.+.}, at: [<c11cd5f2>] nfs_revalidate_mapping+0x92/0xd0
>
> but task is already holding lock:
> (&mm->mmap_sem){++++++}, at: [<c100640e>] sys_mmap2+0x8e/0xc0
>
> which lock already depends on the new lock.
>
>
> the existing dependency chain (in reverse order) is:
>
> -> #1 (&mm->mmap_sem){++++++}:
> [<c1051476>] __lock_acquire+0xef6/0x1710
> [<c1051d5f>] lock_acquire+0xcf/0xf0
> [<c10951cb>] might_fault+0x8b/0xb0
> [<c13cb5f7>] copy_to_user+0x37/0x70
> [<c10bb059>] filldir64+0xa9/0xf0
> [<c11c7f82>] nfs_do_filldir+0x1e2/0x5b0
> [<c11c871e>] nfs_readdir+0x3ce/0x910
> [<c10bb2f5>] vfs_readdir+0x95/0xb0
> [<c10bb37e>] sys_getdents64+0x6e/0xc0
> [<c1002f49>] syscall_call+0x7/0xb
> [<ffffffff>] 0xffffffff
>
> -> #0 (&sb->s_type->i_mutex_key#11){+.+.+.}:
> [<c10515ab>] __lock_acquire+0x102b/0x1710
> [<c1051d5f>] lock_acquire+0xcf/0xf0
> [<c17292ed>] mutex_lock_nested+0x5d/0x310
> [<c11cd5f2>] nfs_revalidate_mapping+0x92/0xd0
> [<c11ca5c8>] nfs_file_mmap+0x48/0x80
> [<c109c436>] mmap_region+0x2b6/0x430
> [<c109c7fd>] do_mmap_pgoff+0x24d/0x350
> [<c100642d>] sys_mmap2+0xad/0xc0
> [<c1002f49>] syscall_call+0x7/0xb
> [<ffffffff>] 0xffffffff
>
> other info that might help us debug this:
>
> 1 lock held by bigfile/3763:
> #0: (&mm->mmap_sem){++++++}, at: [<c100640e>] sys_mmap2+0x8e/0xc0
>
> stack backtrace:
> Pid: 3763, comm: bigfile Not tainted 2.6.31-rc8-00162-gf7fcab6 #321
> Call Trace:
> [<c1727d29>] ? printk+0x1d/0x1f
> [<c10500ac>] print_circular_bug_tail+0xbc/0xc0
> [<c10515ab>] __lock_acquire+0x102b/0x1710
> [<c11cd5f2>] ? nfs_revalidate_mapping+0x92/0xd0
> [<c1051d5f>] lock_acquire+0xcf/0xf0
> [<c11cd5f2>] ? nfs_revalidate_mapping+0x92/0xd0
> [<c11cd5f2>] ? nfs_revalidate_mapping+0x92/0xd0
> [<c17292ed>] mutex_lock_nested+0x5d/0x310
> [<c11cd5f2>] ? nfs_revalidate_mapping+0x92/0xd0
> [<c11f26f0>] ? nfs_have_delegation+0x0/0xc0
> [<c11cd5f2>] nfs_revalidate_mapping+0x92/0xd0
> [<c11ca5c8>] nfs_file_mmap+0x48/0x80
> [<c109c436>] mmap_region+0x2b6/0x430
> [<c109c7fd>] do_mmap_pgoff+0x24d/0x350
> [<c100642d>] sys_mmap2+0xad/0xc0
> [<c1002f49>] syscall_call+0x7/0xb
>
--
Trond Myklebust
Linux NFS client maintainer
NetApp
Trond.Myklebust@xxxxxxxxxx
www.netapp.com
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
