On Sat, May 13 2023 at 18:33, Tetsuo Handa wrote:
> On 2023/05/13 17:33, Thomas Gleixner wrote:
>> Now I just stared at __alloc_pages_slowpath() and looked at the
>> condition for wakeup_all_kswapds(). ALLOC_KSWAPD is set because
>> debugobject uses GFP_ATOMIC which contains __GFP_KSWAPD_RECLAIM.
>>
>> So debug objects needs to have s/GFP_ATOMIC/__GFP_HIGH/ to prevent that
>> wakeup path.
>
> Yes. That is exactly what my patch does.
Indeed. For some reason your patch (though you cc'ed me) did not show up
in my inbox. I've grabbed it from lore so no need to resend.
Actually we want both changes.
- Your's to fix the underlying ancient problem.
- The one I did which restores the performance behaviour
Thanks,
tglx
