Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:2 as permitted sender) client-ip=2620:137:e000::3:2;
Message-ID: <675cb5c3-d228-3254-2f11-9d269b0ef6c6@suse.cz>
Date:   Wed, 22 Nov 2023 15:32:50 +0100
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.15.1
Subject: Re: [PATCH v5 6/9] slub: Delay freezing of partial slabs
Content-Language: en-US
To:     Chengming Zhou <chengming.zhou@linux.dev>,
        Mark Brown <broonie@kernel.org>
Cc:     cl@linux.com, penberg@kernel.org, rientjes@google.com,
        iamjoonsoo.kim@lge.com, akpm@linux-foundation.org,
        roman.gushchin@linux.dev, 42.hyeyoo@gmail.com, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org,
        Chengming Zhou <zhouchengming@bytedance.com>,
        Matthew Wilcox <willy@infradead.org>
References: <20231102032330.1036151-1-chengming.zhou@linux.dev>
 <20231102032330.1036151-7-chengming.zhou@linux.dev>
 <4f3bc1bd-ea87-465d-b58a-0ed57b15187b@sirena.org.uk>
 <83ff4b9e-94f1-8b35-1233-3dd414ea4dfe@suse.cz>
 <b62a6eee-6823-47e0-a18a-964b60d247cd@linux.dev>
 <4ebc67be-8286-17e9-da33-225ed75509a6@suse.cz>
 <2af8c92f-0de8-4528-af43-6c6e8c1ebdf3@linux.dev>
 <42867716-5d3d-0252-5fd2-0f8b62498523@suse.cz>
 <325f38f2-1c09-49a0-a882-d1818c2312ae@linux.dev>
From:   Vlastimil Babka <vbabka@suse.cz>
In-Reply-To: <325f38f2-1c09-49a0-a882-d1818c2312ae@linux.dev>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 11/22/23 15:28, Chengming Zhou wrote:
> 
> Yes, it looks so. There are some background services on the 128 CPUs machine.
> Although "stress-ng --rawpkt 128 --rawpkt-ops 100000000" has so much regression,
> I tried other less contented testcases:
> 
> 1. stress-ng --rawpkt 64 --rawpkt-ops 100000000
> 2. perf bench sched messaging -g 5 -t -l 100000
> 
> The performance numbers of this atomic version are pretty much the same.
> 
> So this atomic version should be good in most cases IMHO.

OK will fold the fix using full atomic version.

>>> And I also tested the atomic-optional version like below, found the
>>> performance numbers are much stable.
>> 
>> This gets rather ugly and fragile so I'd maybe rather go back to the
>> __unused field approach :/
>> 
> 
> Agree. If we don't want this atomic version, the __unused field approach
> seems better.
> 
> Thanks!
>