Re: 2.6.19-rc2-mm1 warning in invalidate_inode_pages2_range()

[Andrew Morton <akpm@osdl.org>]

On Thu, 19 Oct 2006 15:39:06 -0700
Badari Pulavarty <pbadari@us.ibm.com> wrote:

> Hi Zach,
> 
> While running IO tests I get following messages on 2.6.19-rc2-mm1
> 
> BUG: warning at mm/truncate.c:400/invalidate_inode_pages2_range()
> 
> Call Trace:
>  [<ffffffff8020b481>] show_trace+0x41/0x70
>  [<ffffffff8020b4c2>] dump_stack+0x12/0x20
>  [<ffffffff80257f17>] invalidate_inode_pages2_range+0x297/0x2e0
>  [<ffffffff8024fcb5>] generic_file_direct_IO+0xf5/0x130
>  [<ffffffff8024fd64>] generic_file_direct_write+0x74/0x140
>  [<ffffffff802507cc>] __generic_file_aio_write_nolock+0x36c/0x4b0
>  [<ffffffff80250977>] generic_file_aio_write+0x67/0xd0
>  [<ffffffff8808f6d3>] :ext4dev:ext4_file_write+0x23/0xc0
> DWARF2 unwinder stuck at ext4_file_write+0x23/0xc0 [ext4dev]
> Leftover inexact backtrace:
>  [<ffffffff8027379f>] do_sync_write+0xcf/0x120
>  [<ffffffff802772b7>] cp_new_stat+0xe7/0x100
>  [<ffffffff8023db90>] autoremove_wake_function+0x0/0x30
>  [<ffffffff804805bf>] __mutex_lock_slowpath+0x1df/0x1f0
>  [<ffffffff8027408d>] vfs_write+0xbd/0x180
>  [<ffffffff802747b3>] sys_write+0x53/0x90
>  [<ffffffff80209c1e>] system_call+0x7e/0x83
> 

that's warning that we weren't able to invalidate some pagecache.  That's
not really suprising.  Perhaps we should be more careful in deciding when
to fail the call (ie: leaving behind a clean page is ok, leaving behind a
dirty page is bad).

Probably it's leaving behind dirty pagecache and you're about to lose your
data.  Which I bet is your own darn fault for doing silly things.

What workload was in use?
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/