Hi Andros,
your commit "pnfs: lookup new lseg at lseg boundary: triggers a warning
with the block layout driver under xfstests. I have hard time
parsing he commit log, but it changes this WARN_ON from a > to a >=
without mentioning anything related to this one-off which make
me suspicious.
generic/013 files ...[ 170.557305] ------------[ cut here ]------------
[ 170.557889] WARNING: CPU: 0 PID: 9755 at ../fs/nfs/pnfs.c:1792 pnfs_generic_pg_test+0xba/0xd0()
[ 170.558728] Modules linked in:
[ 170.559104] CPU: 0 PID: 9755 Comm: fsstress Not tainted 4.0.0+ #356
[ 170.559770] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
[ 170.560368] ffffffff822a6079 ffff8800770af9c8 ffffffff81e03a7f 0000000000000000
[ 170.561282] 0000000000000000 ffff8800770afa08 ffffffff810c4dc2 0000000000000000
[ 170.562303] ffff88007a730140 ffff8800770afcb8 ffff88007a730140 ffff88007a730140
[ 170.563218] Call Trace:
[ 170.563636] [<ffffffff81e03a7f>] dump_stack+0x45/0x57
[ 170.564254] [<ffffffff810c4dc2>] warn_slowpath_common+0x92/0xd0
[ 170.565283] [<ffffffff810c4e15>] warn_slowpath_null+0x15/0x20
[ 170.566273] [<ffffffff813b608a>] pnfs_generic_pg_test+0xba/0xd0
[ 170.567325] [<ffffffff813be7c5>] bl_pg_test_write+0x45/0x50
[ 170.568533] [<ffffffff8137b7ae>] __nfs_pageio_add_request+0x10e/0x470
[ 170.569616] [<ffffffff8137c043>] nfs_pageio_add_request+0x93/0x1c0
[ 170.570679] [<ffffffff8137faef>] nfs_do_writepage+0x11f/0x1e0
[ 170.571649] [<ffffffff8137fc68>] nfs_writepages_callback+0x18/0x30
[ 170.572668] [<ffffffff8119d50c>] write_cache_pages+0x22c/0x4a0
[ 170.573704] [<ffffffff8111132d>] ? trace_hardirqs_on_caller+0x10d/0x1d0
[ 170.574861] [<ffffffff811113fd>] ? trace_hardirqs_on+0xd/0x10
[ 170.575833] [<ffffffff811da881>] ? kmem_cache_free+0xe1/0x210
[ 170.576867] [<ffffffff8137fc50>] ? nfs_writepage_locked+0xa0/0xa0
[ 170.577876] [<ffffffff8137fd44>] nfs_writepages+0x94/0x130
[ 170.578805] [<ffffffff8119f34c>] do_writepages+0x1c/0x40
[ 170.579712] [<ffffffff811948f1>] __filemap_fdatawrite_range+0x51/0x60
[ 170.580774] [<ffffffff811949d5>] filemap_write_and_wait_range+0x35/0x80
[ 170.581817] [<ffffffff813a73ef>] nfs4_file_fsync+0x5f/0xb0
[ 170.582710] [<ffffffff81213bd1>] vfs_fsync_range+0x41/0xc0
[ 170.583630] [<ffffffff813a75f0>] ? nfs4_file_open+0x1b0/0x1b0
[ 170.584584] [<ffffffff81213c67>] vfs_fsync+0x17/0x20
[ 170.585461] [<ffffffff813724ce>] nfs_file_flush+0x6e/0x90
[ 170.586428] [<ffffffff811dd185>] filp_close+0x35/0x80
[ 170.587075] [<ffffffff811ff7ce>] __close_fd+0x9e/0x110
[ 170.587694] [<ffffffff811dd11e>] SyS_close+0x1e/0x50
[ 170.588291] [<ffffffff81e0e2ee>] system_call_fastpath+0x12/0x76
[ 170.588971] ---[ end trace c6bc30a03cc4fc6e ]---
ping?
On Sat, Apr 25, 2015 at 09:56:40AM -0700, Christoph Hellwig wrote:
> Hi Andros,
>
> your commit "pnfs: lookup new lseg at lseg boundary: triggers a warning
> with the block layout driver under xfstests. I have hard time
> parsing he commit log, but it changes this WARN_ON from a > to a >=
> without mentioning anything related to this one-off which make
> me suspicious.
>
> generic/013 files ...[ 170.557305] ------------[ cut here ]------------
> [ 170.557889] WARNING: CPU: 0 PID: 9755 at ../fs/nfs/pnfs.c:1792 pnfs_generic_pg_test+0xba/0xd0()
> [ 170.558728] Modules linked in:
> [ 170.559104] CPU: 0 PID: 9755 Comm: fsstress Not tainted 4.0.0+ #356
> [ 170.559770] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
> [ 170.560368] ffffffff822a6079 ffff8800770af9c8 ffffffff81e03a7f 0000000000000000
> [ 170.561282] 0000000000000000 ffff8800770afa08 ffffffff810c4dc2 0000000000000000
> [ 170.562303] ffff88007a730140 ffff8800770afcb8 ffff88007a730140 ffff88007a730140
> [ 170.563218] Call Trace:
> [ 170.563636] [<ffffffff81e03a7f>] dump_stack+0x45/0x57
> [ 170.564254] [<ffffffff810c4dc2>] warn_slowpath_common+0x92/0xd0
> [ 170.565283] [<ffffffff810c4e15>] warn_slowpath_null+0x15/0x20
> [ 170.566273] [<ffffffff813b608a>] pnfs_generic_pg_test+0xba/0xd0
> [ 170.567325] [<ffffffff813be7c5>] bl_pg_test_write+0x45/0x50
> [ 170.568533] [<ffffffff8137b7ae>] __nfs_pageio_add_request+0x10e/0x470
> [ 170.569616] [<ffffffff8137c043>] nfs_pageio_add_request+0x93/0x1c0
> [ 170.570679] [<ffffffff8137faef>] nfs_do_writepage+0x11f/0x1e0
> [ 170.571649] [<ffffffff8137fc68>] nfs_writepages_callback+0x18/0x30
> [ 170.572668] [<ffffffff8119d50c>] write_cache_pages+0x22c/0x4a0
> [ 170.573704] [<ffffffff8111132d>] ? trace_hardirqs_on_caller+0x10d/0x1d0
> [ 170.574861] [<ffffffff811113fd>] ? trace_hardirqs_on+0xd/0x10
> [ 170.575833] [<ffffffff811da881>] ? kmem_cache_free+0xe1/0x210
> [ 170.576867] [<ffffffff8137fc50>] ? nfs_writepage_locked+0xa0/0xa0
> [ 170.577876] [<ffffffff8137fd44>] nfs_writepages+0x94/0x130
> [ 170.578805] [<ffffffff8119f34c>] do_writepages+0x1c/0x40
> [ 170.579712] [<ffffffff811948f1>] __filemap_fdatawrite_range+0x51/0x60
> [ 170.580774] [<ffffffff811949d5>] filemap_write_and_wait_range+0x35/0x80
> [ 170.581817] [<ffffffff813a73ef>] nfs4_file_fsync+0x5f/0xb0
> [ 170.582710] [<ffffffff81213bd1>] vfs_fsync_range+0x41/0xc0
> [ 170.583630] [<ffffffff813a75f0>] ? nfs4_file_open+0x1b0/0x1b0
> [ 170.584584] [<ffffffff81213c67>] vfs_fsync+0x17/0x20
> [ 170.585461] [<ffffffff813724ce>] nfs_file_flush+0x6e/0x90
> [ 170.586428] [<ffffffff811dd185>] filp_close+0x35/0x80
> [ 170.587075] [<ffffffff811ff7ce>] __close_fd+0x9e/0x110
> [ 170.587694] [<ffffffff811dd11e>] SyS_close+0x1e/0x50
> [ 170.588291] [<ffffffff81e0e2ee>] system_call_fastpath+0x12/0x76
> [ 170.588971] ---[ end trace c6bc30a03cc4fc6e ]---
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
---end quoted text---
Hey Christoph,
Sorry this got buried in my inbox :-/
- WARN_ON_ONCE(req_start > seg_end);
+ WARN_ON_ONCE(req_start >= seg_end);
I think you’re right, I shouldn’t have changed that WARN_ON.
It’s been a while since I looked at this, but I’m wondering if that WARN_ON
even makes sense anymore now that we do handle multiple segments.
-dros
> On May 7, 2015, at 3:42 AM, Christoph Hellwig <[email protected]> wrote:
>
> ping?
>
> On Sat, Apr 25, 2015 at 09:56:40AM -0700, Christoph Hellwig wrote:
>> Hi Andros,
>>
>> your commit "pnfs: lookup new lseg at lseg boundary: triggers a warning
>> with the block layout driver under xfstests. I have hard time
>> parsing he commit log, but it changes this WARN_ON from a > to a >=
>> without mentioning anything related to this one-off which make
>> me suspicious.
>>
>> generic/013 files ...[ 170.557305] ------------[ cut here ]------------
>> [ 170.557889] WARNING: CPU: 0 PID: 9755 at ../fs/nfs/pnfs.c:1792 pnfs_generic_pg_test+0xba/0xd0()
>> [ 170.558728] Modules linked in:
>> [ 170.559104] CPU: 0 PID: 9755 Comm: fsstress Not tainted 4.0.0+ #356
>> [ 170.559770] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2007
>> [ 170.560368] ffffffff822a6079 ffff8800770af9c8 ffffffff81e03a7f 0000000000000000
>> [ 170.561282] 0000000000000000 ffff8800770afa08 ffffffff810c4dc2 0000000000000000
>> [ 170.562303] ffff88007a730140 ffff8800770afcb8 ffff88007a730140 ffff88007a730140
>> [ 170.563218] Call Trace:
>> [ 170.563636] [<ffffffff81e03a7f>] dump_stack+0x45/0x57
>> [ 170.564254] [<ffffffff810c4dc2>] warn_slowpath_common+0x92/0xd0
>> [ 170.565283] [<ffffffff810c4e15>] warn_slowpath_null+0x15/0x20
>> [ 170.566273] [<ffffffff813b608a>] pnfs_generic_pg_test+0xba/0xd0
>> [ 170.567325] [<ffffffff813be7c5>] bl_pg_test_write+0x45/0x50
>> [ 170.568533] [<ffffffff8137b7ae>] __nfs_pageio_add_request+0x10e/0x470
>> [ 170.569616] [<ffffffff8137c043>] nfs_pageio_add_request+0x93/0x1c0
>> [ 170.570679] [<ffffffff8137faef>] nfs_do_writepage+0x11f/0x1e0
>> [ 170.571649] [<ffffffff8137fc68>] nfs_writepages_callback+0x18/0x30
>> [ 170.572668] [<ffffffff8119d50c>] write_cache_pages+0x22c/0x4a0
>> [ 170.573704] [<ffffffff8111132d>] ? trace_hardirqs_on_caller+0x10d/0x1d0
>> [ 170.574861] [<ffffffff811113fd>] ? trace_hardirqs_on+0xd/0x10
>> [ 170.575833] [<ffffffff811da881>] ? kmem_cache_free+0xe1/0x210
>> [ 170.576867] [<ffffffff8137fc50>] ? nfs_writepage_locked+0xa0/0xa0
>> [ 170.577876] [<ffffffff8137fd44>] nfs_writepages+0x94/0x130
>> [ 170.578805] [<ffffffff8119f34c>] do_writepages+0x1c/0x40
>> [ 170.579712] [<ffffffff811948f1>] __filemap_fdatawrite_range+0x51/0x60
>> [ 170.580774] [<ffffffff811949d5>] filemap_write_and_wait_range+0x35/0x80
>> [ 170.581817] [<ffffffff813a73ef>] nfs4_file_fsync+0x5f/0xb0
>> [ 170.582710] [<ffffffff81213bd1>] vfs_fsync_range+0x41/0xc0
>> [ 170.583630] [<ffffffff813a75f0>] ? nfs4_file_open+0x1b0/0x1b0
>> [ 170.584584] [<ffffffff81213c67>] vfs_fsync+0x17/0x20
>> [ 170.585461] [<ffffffff813724ce>] nfs_file_flush+0x6e/0x90
>> [ 170.586428] [<ffffffff811dd185>] filp_close+0x35/0x80
>> [ 170.587075] [<ffffffff811ff7ce>] __close_fd+0x9e/0x110
>> [ 170.587694] [<ffffffff811dd11e>] SyS_close+0x1e/0x50
>> [ 170.588291] [<ffffffff81e0e2ee>] system_call_fastpath+0x12/0x76
>> [ 170.588971] ---[ end trace c6bc30a03cc4fc6e ]---
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
>> the body of a message to [email protected]
>> More majordomo info at http://vger.kernel.org/majordomo-info.html
> ---end quoted text---
On Wed, May 13, 2015 at 11:17:52AM -0400, Weston Andros Adamson wrote:
> Hey Christoph,
>
> Sorry this got buried in my inbox :-/
>
> - WARN_ON_ONCE(req_start > seg_end);
> + WARN_ON_ONCE(req_start >= seg_end);
>
> I think you???re right, I shouldn???t have changed that WARN_ON.
>
> It???s been a while since I looked at this, but I???m wondering if that WARN_ON
> even makes sense anymore now that we do handle multiple segments.
Are you going to send a patch for this?