Re: linux-next: Tree for Oct 25 (block ?)

[Jens Axboe]

On 2011-10-25 15:43, Sedat Dilek wrote:
> On Tue, Oct 25, 2011 at 3:38 PM, Jens Axboe <axboe@xxxxxxxxx> wrote:
>> On 2011-10-25 15:10, Sedat Dilek wrote:
>>> On Tue, Oct 25, 2011 at 3:04 PM, Sedat Dilek <sedat.dilek@xxxxxxxxxxxxxx> wrote:
>>>> On Tue, Oct 25, 2011 at 11:36 AM, Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx> wrote:
>>>>> Hi all,
>>>>>
>>>>> *Note well*
>>>>>
>>>>> This tree has nate had any build testing at all.  As such, it probably
>>>>> doesn't build :-) This tree is really just a roll up of the current state
>>>>> of the trees when the v3.2 merge window opened.  It will not be put into
>>>>> the build system referred to below.
>>>>>
>>>>
>>>> I have generated a single linux-next (next-20111025) patch on top of
>>>> v3.1, it's approx. 100M!
>>>>
>>>> $ du -h patch-v3.1-next-20111025.patch
>>>> 96M     patch-v3.1-next-20111025.patch
>>>>
>>>> I noticed this build-failure:
>>>>
>>>>  CC      block/blk-throttle.o
>>>>  CC [M]  fs/fuse/dir.o
>>>> /mnt/sdb3/linux-kernel/linux-3.1/debian/build/source_i386_none/block/blk-throttle.c:
>>>> In function 'blk_throtl_drain':
>>>> /mnt/sdb3/linux-kernel/linux-3.1/debian/build/source_i386_none/block/blk-throttle.c:1221:2:
>>>> error: implicit declaration of function 'lockdep_is_held'
>>>> [-Werror=implicit-function-declaration]
>>>> cc1: some warnings being treated as errors
>>>>
>>>> make[5]: *** [block/blk-throttle.o] Error 1
>>>> make[4]: *** [block] Error 2
>>>>
>>>> This happens with Debian's gcc-4.6 (4.6.1-16) and default
>>>> KBUILD_ENABLE_EXTRA_GCC_CHECKS value.
>>>>
>>>> - Sedat -
>>>>
>>>
>>> Looks like "#include <linux/lockdep.h>" is missing in block/blk-throttle.c?
>>
>> Hmm, I wonder why it isn't triggering for cfq-iosched.o or elevator.o as
>> well. Is blk-throttle modular? What is your .config?
>>
>> --
>> Jens Axboe
>>
>>
> 
> CONFIG_BLK_DEV_THROTTLING=y
> 
> My kernel-config is attached.

It's just a bug in blk-throttle, you can't use lockdep_is_help if
!CONFIG_LOCKDEP. I don't think you can reliably do this without either
wrapping a spinlock test inside CONFIG_SMP or not, or hide this in
CONFIG_LOCKDEP.

-- 
Jens Axboe

--
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/