[lkml]   [2022]   [Mar]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [mm/page_alloc] 8212a964ee: vm-scalability.throughput 30.5% improvement
On 3/13/22 00:26, Eric Dumazet wrote:
> On Sat, Mar 12, 2022 at 10:59 AM Vlastimil Babka <> wrote:
>> On 3/12/22 16:43, kernel test robot wrote:
>>> Greeting,
>>> FYI, we noticed a 30.5% improvement of vm-scalability.throughput due to commit:
>>> commit: 8212a964ee020471104e34dce7029dec33c218a9 ("Re: [PATCH v2] mm/page_alloc: call check_new_pages() while zone spinlock is not held")
>>> url:
>>> patch link:
>> Heh, that's weird. I would expect some improvement from Eric's patch,
>> but this seems to be actually about Mel's "mm/page_alloc: check
>> high-order pages for corruption during PCP operations" applied directly
>> on 5.17-rc7 per the github url above. This was rather expected to make
>> performance worse if anything, so maybe the improvement is due to some
>> unexpected side-effect of different inlining decisions or cache alignment...
> I doubt this has anything to do with inlining or cache alignment.
> I am not familiar with the benchmark, but its name
> (anon-w-rand-hugetlb) hints at hugetlb ?
> After Mel fix, we go over 512 'struct page' to perform sanity checks,
> thus loading into cpu caches the 512 cache lines.

Ah, that's true.

> This caching is done while no lock is held.

But I don't think this is. The test was AFAICS done without your patch,
so the lock is still held in rmqueue(). And it's also held in
rmqueue_bulk() -> check_pcp_refill().

> If after this huge page allocation some mm operation needs to access
> these 512 struct pages,
> while holding a lock, then sure, there is a huge gain.

 \ /
  Last update: 2022-03-13 10:29    [W:0.116 / U:0.300 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site