From: Vladimir Davydov <vdavydov@parallels.com>
To: Christoph Lameter <cl@gentwo.org>
Cc: hannes@cmpxchg.org, mhocko@suse.cz, akpm@linux-foundation.org,
linux-kernel@vger.kernel.org, linux-mm@kvack.org
Subject: Re: [PATCH RFC 3/3] slub: reparent memcg caches' slabs on memcg offline
Date: Fri, 23 May 2014 19:26:43 +0400 [thread overview]
Message-ID: <20140523152642.GD3147@esperanza> (raw)
In-Reply-To: <alpine.DEB.2.10.1405221422390.15766@gentwo.org>
On Thu, May 22, 2014 at 02:25:30PM -0500, Christoph Lameter wrote:
> slab_free calls __slab_free which can release slabs via
> put_cpu_partial()/unfreeze_partials()/discard_slab() to the page
> allocator. I'd rather have preemption enabled there.
Hmm, why? IMO, calling __free_pages with preempt disabled won't hurt
latency, because it proceeds really fast. BTW, we already call it for a
bunch of pages from __slab_free() -> put_cpu_partial() ->
unfreeze_partials() with irqs disabled, which is harder. FWIW, SLAB has
the whole obj free path executed under local_irq_save/restore, and it
doesn't bother enabling irqs for freeing pages.
IMO, the latency improvement we can achieve by enabling preemption while
calling __free_pages is rather minor, and it isn't worth complicating
the code.
Thanks.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
WARNING: multiple messages have this Message-ID (diff)
From: Vladimir Davydov <vdavydov@parallels.com>
To: Christoph Lameter <cl@gentwo.org>
Cc: <hannes@cmpxchg.org>, <mhocko@suse.cz>,
<akpm@linux-foundation.org>, <linux-kernel@vger.kernel.org>,
<linux-mm@kvack.org>
Subject: Re: [PATCH RFC 3/3] slub: reparent memcg caches' slabs on memcg offline
Date: Fri, 23 May 2014 19:26:43 +0400 [thread overview]
Message-ID: <20140523152642.GD3147@esperanza> (raw)
In-Reply-To: <alpine.DEB.2.10.1405221422390.15766@gentwo.org>
On Thu, May 22, 2014 at 02:25:30PM -0500, Christoph Lameter wrote:
> slab_free calls __slab_free which can release slabs via
> put_cpu_partial()/unfreeze_partials()/discard_slab() to the page
> allocator. I'd rather have preemption enabled there.
Hmm, why? IMO, calling __free_pages with preempt disabled won't hurt
latency, because it proceeds really fast. BTW, we already call it for a
bunch of pages from __slab_free() -> put_cpu_partial() ->
unfreeze_partials() with irqs disabled, which is harder. FWIW, SLAB has
the whole obj free path executed under local_irq_save/restore, and it
doesn't bother enabling irqs for freeing pages.
IMO, the latency improvement we can achieve by enabling preemption while
calling __free_pages is rather minor, and it isn't worth complicating
the code.
Thanks.
next prev parent reply other threads:[~2014-05-23 15:26 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-13 13:48 [PATCH RFC 0/3] kmemcg slab reparenting Vladimir Davydov
2014-05-13 13:48 ` Vladimir Davydov
2014-05-13 13:48 ` [PATCH RFC 1/3] slub: keep full slabs on list for per memcg caches Vladimir Davydov
2014-05-13 13:48 ` Vladimir Davydov
2014-05-14 16:16 ` Christoph Lameter
2014-05-14 16:16 ` Christoph Lameter
2014-05-15 6:34 ` Vladimir Davydov
2014-05-15 6:34 ` Vladimir Davydov
2014-05-15 15:15 ` Christoph Lameter
2014-05-15 15:15 ` Christoph Lameter
2014-05-16 13:06 ` Vladimir Davydov
2014-05-16 13:06 ` Vladimir Davydov
2014-05-16 15:05 ` Christoph Lameter
2014-05-16 15:05 ` Christoph Lameter
2014-05-13 13:48 ` [PATCH RFC 2/3] percpu-refcount: allow to get dead reference Vladimir Davydov
2014-05-13 13:48 ` Vladimir Davydov
2014-05-13 13:48 ` [PATCH RFC 3/3] slub: reparent memcg caches' slabs on memcg offline Vladimir Davydov
2014-05-13 13:48 ` Vladimir Davydov
2014-05-14 16:20 ` Christoph Lameter
2014-05-14 16:20 ` Christoph Lameter
2014-05-15 7:16 ` Vladimir Davydov
2014-05-15 7:16 ` Vladimir Davydov
2014-05-15 15:16 ` Christoph Lameter
2014-05-15 15:16 ` Christoph Lameter
2014-05-16 13:22 ` Vladimir Davydov
2014-05-16 13:22 ` Vladimir Davydov
2014-05-16 15:03 ` Christoph Lameter
2014-05-16 15:03 ` Christoph Lameter
2014-05-19 15:24 ` Vladimir Davydov
2014-05-19 15:24 ` Vladimir Davydov
2014-05-19 16:03 ` Christoph Lameter
2014-05-19 16:03 ` Christoph Lameter
2014-05-19 18:27 ` Vladimir Davydov
2014-05-19 18:27 ` Vladimir Davydov
2014-05-21 13:58 ` Vladimir Davydov
2014-05-21 13:58 ` Vladimir Davydov
2014-05-21 14:45 ` Christoph Lameter
2014-05-21 14:45 ` Christoph Lameter
2014-05-21 15:14 ` Vladimir Davydov
2014-05-21 15:14 ` Vladimir Davydov
2014-05-22 0:15 ` Christoph Lameter
2014-05-22 0:15 ` Christoph Lameter
2014-05-22 14:07 ` Vladimir Davydov
2014-05-22 14:07 ` Vladimir Davydov
2014-05-21 14:41 ` Christoph Lameter
2014-05-21 14:41 ` Christoph Lameter
2014-05-21 15:04 ` Vladimir Davydov
2014-05-21 15:04 ` Vladimir Davydov
2014-05-22 0:13 ` Christoph Lameter
2014-05-22 0:13 ` Christoph Lameter
2014-05-22 13:47 ` Vladimir Davydov
2014-05-22 13:47 ` Vladimir Davydov
2014-05-22 19:25 ` Christoph Lameter
2014-05-22 19:25 ` Christoph Lameter
2014-05-23 15:26 ` Vladimir Davydov [this message]
2014-05-23 15:26 ` Vladimir Davydov
2014-05-23 17:45 ` Christoph Lameter
2014-05-23 17:45 ` Christoph Lameter
2014-05-23 19:57 ` Vladimir Davydov
2014-05-23 19:57 ` Vladimir Davydov
2014-05-27 14:38 ` Christoph Lameter
2014-05-27 14:38 ` Christoph Lameter
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=20140523152642.GD3147@esperanza \
--to=vdavydov@parallels.com \
--cc=akpm@linux-foundation.org \
--cc=cl@gentwo.org \
--cc=hannes@cmpxchg.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.