Re: [lkp] [x86/copy_user] adb402cd14: will-it-scale.per_process_ops -12.7% regression

[Borislav Petkov]

On Mon, Nov 07, 2016 at 10:50:38AM +0800, kernel test robot wrote:
> 
> Greeting,
> 
> FYI, we noticed a -12.7% regression of will-it-scale.per_process_ops due to commit:
> 
> 
> commit adb402cd1461eef6e1a21db4532a3b9e6a6be853 ("x86/copy_user: Unify the code by removing the 64-bit asm _copy_*_user() variants")
> https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git x86/asm
> 
> in testcase: will-it-scale
> on test machine: 8 threads Intel(R) Core(TM) i7 CPU 870 @ 2.93GHz with 4G memory
> with following parameters:
> 
> 	test: poll1
> 	cpufreq_governor: performance

...

> # Lock Debugging (spinlocks, mutexes, etc...)
> #
> # CONFIG_DEBUG_RT_MUTEXES is not set
> # CONFIG_DEBUG_SPINLOCK is not set
> # CONFIG_DEBUG_MUTEXES is not set
> # CONFIG_DEBUG_WW_MUTEX_SLOWPATH is not set
> # CONFIG_DEBUG_LOCK_ALLOC is not set
> # CONFIG_PROVE_LOCKING is not set
> # CONFIG_LOCK_STAT is not set
> CONFIG_DEBUG_ATOMIC_SLEEP=y
  ^^^^^^^^^^^^^^^^^^^^^^^^^

So Mel says that this might be the culprit for the observed
change in perf. Can you please rerun your test without that
CONFIG_DEBUG_ATOMIC_SLEEP thing?

Thanks.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.