Received-SPF: pass (google.com: domain of linux-kernel+bounces-11962-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Message-ID: <1a236102-0607-4952-8bc3-0cb45030bc6a@linux.dev>
Date: Wed, 27 Dec 2023 18:33:50 +0800
Precedence: bulk
MIME-Version: 1.0
Subject: Re: [RFC PATCH v2 3/5] padata: dispatch works on different nodes
Content-Language: en-US
To: Tim Chen <tim.c.chen@linux.intel.com>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
 Muchun Song <muchun.song@linux.dev>, Mike Kravetz <mike.kravetz@oracle.com>,
 David Rientjes <rientjes@google.com>, David Hildenbrand <david@redhat.com>,
 Andrew Morton <akpm@linux-foundation.org>
References: <20231208025240.4744-1-gang.li@linux.dev>
 <20231208025240.4744-4-gang.li@linux.dev>
 <b2e7daff1e314d927cc4b4713afa31ac25787825.camel@linux.intel.com>
 <642a3472-9570-409a-94e0-64f30966b86f@linux.dev>
From: Gang Li <gang.li@linux.dev>
In-Reply-To: <642a3472-9570-409a-94e0-64f30966b86f@linux.dev>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

Hi Tim,

According to queue_work_node, if there are no CPUs available on the
given node, it will schedule to any available CPU.

On 2023/12/18 14:46, Gang Li wrote:
> On 2023/12/13 07:40, Tim Chen wrote:
>>
>>>       list_for_each_entry(pw, &works, pw_list)
>>> -        queue_work(system_unbound_wq, &pw->pw_work);
>>> +        if (job->numa_aware)
>>> +            queue_work_node((++nid % num_node_state(N_MEMORY)),
>>
>> The nid may fall on a NUMA node with only memory but no CPU.  In that 
>> case you
>> may still put the work on the unbound queue. You could end up on one 
>> CPU node for work
>> from all memory nodes without CPU. Is this what you want?  Or you would
>> like to spread them between CPU nodes?
>>
>> Tim
> 
> Hi, thank you for your reminder. My intention was to fully utilize all
> memory bandwidth.
> 
> For memory nodes without CPUs, I also hope to be able to spread them on
> different CPUs.