The Linux Kernel Mailing List
 help / color / mirror / Atom feed
From: Hillf Danton <hdanton@sina.com>
To: Vlastimil Babka <vbabka@suse.cz>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v5 01/14] slab: add opt-in caching layer of percpu sheaves
Date: Mon, 15 Sep 2025 08:11:37 +0800	[thread overview]
Message-ID: <20250915001139.7101-1-hdanton@sina.com> (raw)
In-Reply-To: <bbdb72e1-ada9-44e2-9ec1-53fd6ed92e84@suse.cz>

On Sun, 14 Sep 2025 22:24:19 +0200 Vlastimil Babka wrote:
>On 9/14/25 04:22, Hillf Danton wrote:
>> On Wed, 23 Jul 2025 15:34:34 +0200 Vlastimil Babka wrote:
>>> +
>>> +/*
>>> + * Caller needs to make sure migration is disabled in order to fully flush
>>> + * single cpu's sheaves
>>> + *
>> This misguides, see below for a workqueue case.
>> 
>>> + * must not be called from an irq
>>> + *
>>> + * flushing operations are rare so let's keep it simple and flush to slabs
>>> + * directly, skipping the barn
>>> + */
>>> +static void pcs_flush_all(struct kmem_cache *s)
>>> +{
>>> +	struct slub_percpu_sheaves *pcs;
>>> +	struct slab_sheaf *spare;
>>> +
>>> +	local_lock(&s->cpu_sheaves->lock);
>>> +	pcs = this_cpu_ptr(s->cpu_sheaves);
>>> +
>>> +	spare = pcs->spare;
>>> +	pcs->spare = NULL;
>>> +
>>> +	local_unlock(&s->cpu_sheaves->lock);
>>> +
>>> +	if (spare) {
>>> +		sheaf_flush_unused(s, spare);
>>> +		free_empty_sheaf(s, spare);
>>> +	}
>>> +
>>> +	sheaf_flush_main(s);
>>> +}
>>> +
>>> @@ -3326,30 +3755,18 @@ struct slub_flush_work {
>>>  static void flush_cpu_slab(struct work_struct *w)
>>>  {
>>>  	struct kmem_cache *s;
>>> -	struct kmem_cache_cpu *c;
>>>  	struct slub_flush_work *sfw;
>>>  
>>>  	sfw = container_of(w, struct slub_flush_work, work);
>>>  
>>>  	s = sfw->s;
>>> -	c = this_cpu_ptr(s->cpu_slab);
>>>  
>>> -	if (c->slab)
>>> -		flush_slab(s, c);
>>> +	if (s->cpu_sheaves)
>>> +		pcs_flush_all(s);
>>>  
>> Migration is not disabled.
>
> Can you elaborate how it's not? There's a comment above the function saying
> "Called from CPU work handler with migration disabled." and we have relied
> on this before sheaves. queue_work_on() says it will run on the specific
> cpu. AFAIK the workqueue workers are bound which is effectively disabled
> migration (we hold the cpu hotplug lock).
>
CPU affinity and migration_disable() are two different things regardless of
queue_work_on(), no?
I think you are right in accident rather than by design in this case.

  reply	other threads:[~2025-09-15  0:11 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-23 13:34 [PATCH v5 00/14] SLUB percpu sheaves Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 01/14] slab: add opt-in caching layer of " Vlastimil Babka
2025-08-18 10:09   ` Harry Yoo
2025-08-26  8:03     ` Vlastimil Babka
2025-08-19  4:19   ` Suren Baghdasaryan
2025-08-26  8:51     ` Vlastimil Babka
2025-09-13 14:35   ` Mateusz Guzik
2025-09-13 20:32     ` Vlastimil Babka
2025-09-14  2:22   ` Hillf Danton
2025-09-14 20:24     ` Vlastimil Babka
2025-09-15  0:11       ` Hillf Danton [this message]
2025-09-15  7:21         ` Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 02/14] slab: add sheaf support for batching kfree_rcu() operations Vlastimil Babka
2025-07-23 16:39   ` Uladzislau Rezki
2025-07-24 14:30     ` Vlastimil Babka
2025-07-24 17:36       ` Uladzislau Rezki
2025-07-23 13:34 ` [PATCH v5 03/14] slab: sheaf prefilling for guaranteed allocations Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 04/14] slab: determine barn status racily outside of lock Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 05/14] tools: Add testing support for changes to rcu and slab for sheaves Vlastimil Babka
2025-08-22 16:28   ` Suren Baghdasaryan
2025-08-26  9:32     ` Vlastimil Babka
2025-08-27  0:19       ` Suren Baghdasaryan
2025-07-23 13:34 ` [PATCH v5 06/14] tools: Add sheaves support to testing infrastructure Vlastimil Babka
2025-08-22 16:56   ` Suren Baghdasaryan
2025-08-26  9:59     ` Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 07/14] maple_tree: use percpu sheaves for maple_node_cache Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 08/14] mm, vma: use percpu sheaves for vm_area_struct cache Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 09/14] mm, slub: skip percpu sheaves for remote object freeing Vlastimil Babka
2025-08-25  5:22   ` Harry Yoo
2025-08-26 10:11     ` Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 10/14] mm, slab: allow NUMA restricted allocations to use percpu sheaves Vlastimil Babka
2025-08-22 19:58   ` Suren Baghdasaryan
2025-08-25  6:52   ` Harry Yoo
2025-08-26 10:49     ` Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 11/14] testing/radix-tree/maple: Increase readers and reduce delay for faster machines Vlastimil Babka
2025-07-23 13:34 ` [PATCH v5 12/14] maple_tree: Sheaf conversion Vlastimil Babka
2025-08-22 20:18   ` Suren Baghdasaryan
2025-08-26 14:22     ` Liam R. Howlett
2025-08-27  2:07       ` Suren Baghdasaryan
2025-08-28 14:27         ` Liam R. Howlett
2025-07-23 13:34 ` [PATCH v5 13/14] maple_tree: Add single node allocation support to maple state Vlastimil Babka
2025-08-22 20:25   ` Suren Baghdasaryan
2025-08-26 15:10     ` Liam R. Howlett
2025-08-27  2:03       ` Suren Baghdasaryan
2025-07-23 13:34 ` [PATCH v5 14/14] maple_tree: Convert forking to use the sheaf interface Vlastimil Babka
2025-08-22 20:29   ` Suren Baghdasaryan
2025-08-15 22:53 ` [PATCH v5 00/14] SLUB percpu sheaves Sudarsan Mahendran
2025-08-16  8:05   ` Harry Yoo
     [not found]     ` <CAA9mObAiQbAYvzhW---VoqDA6Zsb152p5ePMvbco0xgwyvaB2Q@mail.gmail.com>
2025-08-16 18:31       ` Vlastimil Babka
2025-08-16 18:33         ` Vlastimil Babka
2025-08-17  4:28           ` Sudarsan Mahendran
2025-09-13  0:09 ` Benchmarking " Sudarsan Mahendran
2025-09-15  7:51   ` Jan Engelhardt
2025-09-15 12:13     ` Paul E. McKenney
2025-09-15 15:22       ` Vlastimil Babka
2025-09-16 17:09         ` Suren Baghdasaryan
2025-09-17  5:19           ` Uladzislau Rezki
2025-09-17 16:14             ` Suren Baghdasaryan
2025-09-17 23:59               ` Suren Baghdasaryan
2025-09-18 11:50                 ` Uladzislau Rezki
2025-09-18 15:29                   ` Liam R. Howlett
2025-09-19 15:07                     ` Uladzislau Rezki

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250915001139.7101-1-hdanton@sina.com \
    --to=hdanton@sina.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=vbabka@suse.cz \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox