From: David Miller <davem@davemloft.net>
To: clameter@sgi.com
Cc: akpm@linux-foundation.org, linux-arch@vger.kernel.org,
linux-kernel@vger.kernel.org, mathieu.desnoyers@polymtl.ca,
penberg@cs.helsinki.fi
Subject: Re: [patch 0/7] [RFC] SLUB: Improve allocpercpu to reduce per cpu access overhead
Date: Wed, 31 Oct 2007 22:38:36 -0700 (PDT) [thread overview]
Message-ID: <20071031.223836.162573733.davem@davemloft.net> (raw)
In-Reply-To: <Pine.LNX.4.64.0710312115520.26072@schroedinger.engr.sgi.com>
From: Christoph Lameter <clameter@sgi.com>
Date: Wed, 31 Oct 2007 21:16:59 -0700 (PDT)
> /*
> * Maximum allowed per cpu data per cpu
> */
> +#ifdef CONFIG_NUMA
> +#define PER_CPU_ALLOC_SIZE (32768 + MAX_NUMNODES * 512)
> +#else
> #define PER_CPU_ALLOC_SIZE 32768
> +#endif
> +
Christoph, as Rusty found out years ago when he first wrote this code,
you cannot put hard limits on the alloc_percpu() allocations.
They can be done by anyone, any module, and since there was no limit
before you cannot reasonably add one now.
As just one of many examples, several networking devices use
alloc_percpu() for each instance they bring up. This alone can
request arbitrary amounts of per-cpu data.
Therefore, you'll need to do your optimization without imposing any
size limits.
next prev parent reply other threads:[~2007-11-01 5:38 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-11-01 0:02 Christoph Lameter
2007-11-01 0:02 ` [patch 1/7] allocpercpu: Make it a true per cpu allocator by allocating from a per cpu array Christoph Lameter
2007-11-01 7:24 ` Eric Dumazet
2007-11-01 12:59 ` Christoph Lameter
2007-11-01 0:02 ` [patch 2/7] allocpercpu: Remove functions that are rarely used Christoph Lameter
2007-11-01 0:02 ` [patch 3/7] Allocpercpu: Do __percpu_disguise() only if CONFIG_DEBUG_VM is set Christoph Lameter
2007-11-01 7:25 ` Eric Dumazet
2007-11-01 0:02 ` [patch 4/7] Percpu: Add support for this_cpu_offset() to be able to create this_cpu_ptr() Christoph Lameter
2007-11-01 0:02 ` [patch 5/7] SLUB: Use allocpercpu to allocate per cpu data instead of running our own per cpu allocator Christoph Lameter
2007-11-01 0:02 ` [patch 6/7] SLUB: No need to cache kmem_cache data in kmem_cache_cpu anymore Christoph Lameter
2007-11-01 0:02 ` [patch 7/7] SLUB: Optimize per cpu access on the local cpu using this_cpu_ptr() Christoph Lameter
2007-11-01 0:24 ` [patch 0/7] [RFC] SLUB: Improve allocpercpu to reduce per cpu access overhead David Miller
2007-11-01 0:26 ` Christoph Lameter
2007-11-01 0:27 ` David Miller
2007-11-01 0:31 ` Christoph Lameter
2007-11-01 0:51 ` David Miller
2007-11-01 0:53 ` Christoph Lameter
2007-11-01 1:00 ` David Miller
2007-11-01 1:01 ` Christoph Lameter
2007-11-01 1:09 ` David Miller
2007-11-01 1:12 ` Christoph Lameter
2007-11-01 1:13 ` David Miller
2007-11-01 1:21 ` Christoph Lameter
2007-11-01 5:27 ` David Miller
2007-11-01 4:16 ` Christoph Lameter
2007-11-01 5:38 ` David Miller [this message]
2007-11-01 7:01 ` David Miller
2007-11-01 9:14 ` David Miller
2007-11-01 13:03 ` Christoph Lameter
2007-11-01 21:29 ` David Miller
2007-11-01 22:15 ` Christoph Lameter
2007-11-01 22:38 ` David Miller
2007-11-01 22:48 ` Christoph Lameter
2007-11-01 22:58 ` David Miller
2007-11-02 1:06 ` Christoph Lameter
2007-11-02 2:51 ` David Miller
2007-11-02 10:28 ` Peter Zijlstra
2007-11-02 14:35 ` Christoph Lameter
2007-11-02 15:20 ` Peter Zijlstra
2007-11-02 15:29 ` Christoph Lameter
2007-11-12 10:52 ` Herbert Xu
2007-11-12 19:14 ` Christoph Lameter
2007-11-12 19:48 ` Eric Dumazet
2007-11-12 19:56 ` Christoph Lameter
2007-11-12 20:18 ` Eric Dumazet
2007-11-12 22:46 ` David Miller
2007-11-12 19:57 ` Luck, Tony
2007-11-12 20:14 ` Eric Dumazet
2007-11-12 22:46 ` David Miller
2007-11-12 21:28 ` David Miller
2007-11-01 23:00 ` Eric Dumazet
2007-11-02 0:58 ` Christoph Lameter
2007-11-02 1:40 ` Christoph Lameter
2007-11-01 7:17 ` Eric Dumazet
2007-11-01 7:57 ` David Miller
2007-11-01 13:01 ` Christoph Lameter
2007-11-01 21:25 ` David Miller
2007-11-01 12:57 ` Christoph Lameter
2007-11-01 21:28 ` David Miller
2007-11-01 22:11 ` Christoph Lameter
2007-11-01 22:14 ` David Miller
2007-11-01 22:16 ` 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=20071031.223836.162573733.davem@davemloft.net \
--to=davem@davemloft.net \
--cc=akpm@linux-foundation.org \
--cc=clameter@sgi.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mathieu.desnoyers@polymtl.ca \
--cc=penberg@cs.helsinki.fi \
/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
Powered by JetHome