Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:7 as permitted sender) client-ip=2620:137:e000::3:7;
Message-ID: <b2e7daff1e314d927cc4b4713afa31ac25787825.camel@linux.intel.com>
Subject: Re: [RFC PATCH v2 3/5] padata: dispatch works on different nodes
From:   Tim Chen <tim.c.chen@linux.intel.com>
To:     Gang Li <gang.li@linux.dev>, David Hildenbrand <david@redhat.com>,
        David Rientjes <rientjes@google.com>,
        Mike Kravetz <mike.kravetz@oracle.com>,
        Muchun Song <muchun.song@linux.dev>,
        Andrew Morton <akpm@linux-foundation.org>
Cc:     linux-mm@kvack.org, linux-kernel@vger.kernel.org,
        ligang.bdlg@bytedance.com
Date:   Tue, 12 Dec 2023 15:40:18 -0800
In-Reply-To: <20231208025240.4744-4-gang.li@linux.dev>
References: <20231208025240.4744-1-gang.li@linux.dev>
         <20231208025240.4744-4-gang.li@linux.dev>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4 (3.44.4-2.fc36) 
MIME-Version: 1.0
Precedence: bulk


> =20
>  	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 no=
de for work
from all memory nodes without CPU. Is this what you want?  Or you would
like to spread them between CPU nodes?

Tim

> +					system_unbound_wq, &pw->pw_work);
> +		else
> +			queue_work(system_unbound_wq, &pw->pw_work);
> =20
>  	/* Use the current thread, which saves starting a workqueue worker. */
>  	padata_work_init(&my_work, padata_mt_helper, &ps, PADATA_WORK_ONSTACK);