Re: [slub] WARNING: CPU: 1 PID: 1 at mm/slub.c:992deactivate_slab()

From: David Rientjes
Date: Wed Feb 05 2014 - 03:10:58 EST


On Wed, 5 Feb 2014, Fengguang Wu wrote:

> Greetings,
>
> I got the below dmesg and the first bad commit is in upstream
>
> commit c65c1877bd6826ce0d9713d76e30a7bed8e49f38
> Author: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> AuthorDate: Fri Jan 10 13:23:49 2014 +0100
> Commit: Pekka Enberg <penberg@xxxxxxxxxx>
> CommitDate: Mon Jan 13 21:34:39 2014 +0200
>
> slub: use lockdep_assert_held
>
> Instead of using comments in an attempt at getting the locking right,
> use proper assertions that actively warn you if you got it wrong.
>
> Also add extra braces in a few sites to comply with coding-style.
>
> Signed-off-by: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
> Signed-off-by: Pekka Enberg <penberg@xxxxxxxxxx>
>
> ===================================================
> PARENT COMMIT NOT CLEAN. LOOK OUT FOR WRONG BISECT!
> ===================================================
>
> +---------------------------------------------------------+--------------+--------------+
> | | 8afb1474db47 | 1738cc0ecc54 |
> +---------------------------------------------------------+--------------+--------------+
> | boot_successes | 166 | 6 |
> | boot_failures | 10 | 13 |
> | BUG:kernel_test_crashed | 9 | 1 |
> | WARNING:CPU:PID:at_arch/x86/kernel/cpu/amd.c:init_amd() | 1 | |
> | WARNING:CPU:PID:at_mm/slub.c:deactivate_slab() | 0 | 12 |
> +---------------------------------------------------------+--------------+--------------+
>
> [1868680.126265] netconsole: network logging started
> [1868680.135018] Unregister pv shared memory for cpu 0
> [1868680.523086] ------------[ cut here ]------------
> [1868680.526909] WARNING: CPU: 1 PID: 1 at mm/slub.c:992 deactivate_slab+0x4ce/0xa70()
> [1868680.537875] Modules linked in:
> [1868680.541340] CPU: 1 PID: 1 Comm: swapper/0 Not tainted 3.13.0-02621-g1738cc0 #8
> [1868680.555880] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
> [1868680.565937] ffffffff ce04dd64 c1a6323f 00000000 00000000 000003e0 ce04dd94 c106fbe1
> [1868680.572881] c1efb154 00000001 00000001 c1f09c28 000003e0 c11c2d0e c11c2d0e 00000001
> [1868680.582142] ce5db280 ce000640 ce04dda4 c106fc7d 00000009 00000000 ce04de0c c11c2d0e
> [1868680.589099] Call Trace:
> [1868680.591109] [<c1a6323f>] dump_stack+0x7a/0xdb
> [1868680.593887] [<c106fbe1>] warn_slowpath_common+0x91/0xb0
> [1868680.597430] [<c11c2d0e>] ? deactivate_slab+0x4ce/0xa70
> [1868680.600510] [<c11c2d0e>] ? deactivate_slab+0x4ce/0xa70
> [1868680.603588] [<c106fc7d>] warn_slowpath_null+0x1d/0x20
> [1868680.606728] [<c11c2d0e>] deactivate_slab+0x4ce/0xa70

Hi Fengguang,

I think this is the inlined add_full() and should be fixed with
http://marc.info/?l=linux-kernel&m=139147105027693 that has been added to
the -mm tree and should now be in next. Is this patch included for this
kernel?
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/