Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.31 as permitted sender) client-ip=23.128.96.31;
Message-ID: <8c470d4d-b972-3f43-9b0a-712ee882a402@bytedance.com>
Date:   Mon, 18 Sep 2023 15:48:39 +0800
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:102.0)
 Gecko/20100101 Thunderbird/102.15.1
Subject: Re: [RFC PATCH net-next 3/3] sock: Throttle pressure-aware sockets
 under pressure
To:     "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Shakeel Butt <shakeelb@google.com>,
        Roman Gushchin <roman.gushchin@linux.dev>,
        Michal Hocko <mhocko@suse.com>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Yosry Ahmed <yosryahmed@google.com>,
        Yu Zhao <yuzhao@google.com>,
        "Matthew Wilcox (Oracle)" <willy@infradead.org>,
        Kefeng Wang <wangkefeng.wang@huawei.com>,
        Yafang Shao <laoar.shao@gmail.com>,
        Kuniyuki Iwashima <kuniyu@amazon.com>,
        Martin KaFai Lau <martin.lau@kernel.org>,
        Breno Leitao <leitao@debian.org>,
        Alexander Mikhalitsyn <alexander@mihalicyn.com>,
        David Howells <dhowells@redhat.com>,
        Jason Xing <kernelxing@tencent.com>
Cc:     open list <linux-kernel@vger.kernel.org>,
        "open list:NETWORKING [GENERAL]" <netdev@vger.kernel.org>,
        "open list:MEMORY MANAGEMENT" <linux-mm@kvack.org>
References: <20230901062141.51972-1-wuyun.abel@bytedance.com>
 <20230901062141.51972-4-wuyun.abel@bytedance.com>
Content-Language: en-US
From:   Abel Wu <wuyun.abel@bytedance.com>
In-Reply-To: <20230901062141.51972-4-wuyun.abel@bytedance.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 9/1/23 2:21 PM, Abel Wu wrote:
> @@ -3087,8 +3100,20 @@ int __sk_mem_raise_allocated(struct sock *sk, int size, int amt, int kind)
>   	if (sk_has_memory_pressure(sk)) {
>   		u64 alloc;
>   
> -		if (!sk_under_memory_pressure(sk))
> +		/* Be more conservative if the socket's memcg (or its
> +		 * parents) is under reclaim pressure, try to possibly
> +		 * avoid further memstall.
> +		 */
> +		if (under_memcg_pressure)
> +			goto suppress_allocation;
> +
> +		if (!sk_under_global_memory_pressure(sk))
>   			return 1;
> +
> +		/* Trying to be fair among all the sockets of same
> +		 * protocal under global memory pressure, by allowing
> +		 * the ones that under average usage to raise.
> +		 */
>   		alloc = sk_sockets_allocated_read_positive(sk);
>   		if (sk_prot_mem_limits(sk, 2) > alloc *
>   		    sk_mem_pages(sk->sk_wmem_queued +

I totally agree with what Shakeel said in last reply and will try ebpf-
based solution to let userspace inject proper strategies. But IMHO the
above hunk is irrelevant to the idea of this patchset, and is the right
thing to do, so maybe worth a separate patch?

This hunk originally passes the allocation when this socket is below
average usage even under global and/or memcg pressure. It makes sense
to do so under global pressure, as the 'average' is in the scope of
global, but it's really weird from a memcg's point of view. Actually
this pass condition was present before memcg pressure was introduced.

Please correct me if I missed something, thanks!

Best,
	Abel