From: Matthew Wilcox <matthew@wil.cx>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: Christoph Lameter <clameter@sgi.com>,
Eric Dumazet <dada1@cosmosbay.com>,
linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org,
David Miller <davem@davemloft.net>,
Peter Zijlstra <peterz@infradead.org>,
Rusty Russell <rusty@rustcorp.com.au>,
Mike Travis <travis@sgi.com>
Subject: Re: [patch 00/41] cpu alloc / cpu ops v3: Optimize per cpu access
Date: Fri, 30 May 2008 05:32:59 -0600 [thread overview]
Message-ID: <20080530113259.GE22636@parisc-linux.org> (raw)
In-Reply-To: <20080529233758.7bd2c977.akpm@linux-foundation.org>
On Thu, May 29, 2008 at 11:37:58PM -0700, Andrew Morton wrote:
> And struct nfs_iostats is 264 bytes and nfs does an alloc_percpu() of
> one of those per server and mounting thousands of servers per client
> is, I believe, a real-world operation.
Another example, not as extreme, there's an alloc_percpu(struct
disk_stats) [80 bytes on 64-bit machines] for every disk and every
partition in the machine. The TPC system has 3000 disks, each with 14
partitions on it. That's 15 * 80 * 3000 = 3,600,000 bytes.
Even if you're only putting a pointer to each allocation in the percpu
area, that's still 360,000 bytes, 12x as much as you think is sufficient
for the entire system.
--
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours. We can't possibly take such
a retrograde step."
next prev parent reply other threads:[~2008-05-30 11:33 UTC|newest]
Thread overview: 163+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-05-30 3:56 [patch 00/41] cpu alloc / cpu ops v3: Optimize per cpu access Christoph Lameter
2008-05-30 3:56 ` [patch 01/41] cpu_alloc: Increase percpu area size to 128k Christoph Lameter
2008-06-02 17:58 ` Luck, Tony
2008-06-02 23:48 ` Rusty Russell
2008-06-10 17:22 ` Christoph Lameter
2008-06-10 17:22 ` Christoph Lameter
2008-06-10 19:54 ` Luck, Tony
2008-05-30 3:56 ` [patch 02/41] cpu alloc: The allocator Christoph Lameter
2008-05-30 4:58 ` Andrew Morton
2008-05-30 5:10 ` Christoph Lameter
2008-05-30 5:31 ` Andrew Morton
2008-06-02 9:29 ` Paul Jackson
2008-05-30 5:56 ` KAMEZAWA Hiroyuki
2008-05-30 6:16 ` Christoph Lameter
2008-06-04 14:48 ` Mike Travis
2008-05-30 5:04 ` Eric Dumazet
2008-05-30 5:20 ` Christoph Lameter
2008-05-30 5:52 ` Rusty Russell
2008-06-04 15:30 ` Mike Travis
2008-06-05 23:48 ` Rusty Russell
2008-05-30 5:54 ` Eric Dumazet
2008-06-04 14:58 ` Mike Travis
2008-06-04 15:11 ` Eric Dumazet
2008-06-06 0:32 ` Rusty Russell
2008-06-06 0:32 ` Rusty Russell
2008-06-10 17:33 ` Christoph Lameter
2008-06-10 18:05 ` Eric Dumazet
2008-06-10 18:28 ` Christoph Lameter
2008-05-30 5:46 ` Rusty Russell
2008-06-04 15:04 ` Mike Travis
2008-06-10 17:34 ` Christoph Lameter
2008-05-31 20:58 ` Pavel Machek
2008-05-30 3:56 ` [patch 03/41] cpu alloc: Use cpu allocator instead of the builtin modules per cpu allocator Christoph Lameter
2008-05-30 4:58 ` Andrew Morton
2008-05-30 5:14 ` Christoph Lameter
2008-05-30 5:34 ` Andrew Morton
2008-05-30 6:08 ` Rusty Russell
2008-05-30 6:21 ` Christoph Lameter
2008-05-30 3:56 ` [patch 04/41] cpu ops: Core piece for generic atomic per cpu operations Christoph Lameter
2008-05-30 4:58 ` Andrew Morton
2008-05-30 5:17 ` Christoph Lameter
2008-05-30 5:38 ` Andrew Morton
2008-05-30 6:12 ` Christoph Lameter
2008-05-30 7:08 ` Rusty Russell
2008-05-30 18:00 ` Christoph Lameter
2008-06-02 2:00 ` Rusty Russell
2008-06-04 18:18 ` Mike Travis
2008-06-05 23:59 ` Rusty Russell
2008-06-09 19:00 ` Christoph Lameter
2008-06-09 23:27 ` Rusty Russell
2008-06-09 23:54 ` Christoph Lameter
2008-06-10 2:56 ` Rusty Russell
2008-06-10 3:18 ` Christoph Lameter
2008-06-11 0:03 ` Rusty Russell
2008-06-11 0:15 ` Christoph Lameter
2008-06-09 23:09 ` Christoph Lameter
2008-06-10 17:42 ` Christoph Lameter
2008-06-11 11:10 ` Rusty Russell
2008-06-11 23:39 ` Christoph Lameter
2008-06-12 0:58 ` Nick Piggin
2008-06-12 2:44 ` Rusty Russell
2008-06-12 3:40 ` Nick Piggin
2008-06-12 9:37 ` Martin Peschke
2008-06-12 11:21 ` Nick Piggin
2008-06-12 17:19 ` Christoph Lameter
2008-06-13 0:38 ` Rusty Russell
2008-06-13 2:27 ` Christoph Lameter
2008-06-15 10:33 ` Rusty Russell
2008-06-15 10:33 ` Rusty Russell
2008-06-16 14:52 ` Christoph Lameter
2008-06-17 0:24 ` Rusty Russell
2008-06-17 2:29 ` Christoph Lameter
2008-06-17 14:21 ` Mike Travis
2008-05-30 7:05 ` Rusty Russell
2008-05-30 6:32 ` Rusty Russell
2008-05-30 3:56 ` [patch 05/41] cpu alloc: Percpu_counter conversion Christoph Lameter
2008-05-30 6:47 ` Rusty Russell
2008-05-30 17:54 ` Christoph Lameter
2008-05-30 3:56 ` [patch 06/41] cpu alloc: crash_notes conversion Christoph Lameter
2008-05-30 3:56 ` [patch 07/41] cpu alloc: Workqueue conversion Christoph Lameter
2008-05-30 3:56 ` [patch 08/41] cpu alloc: ACPI cstate handling conversion Christoph Lameter
2008-05-30 3:56 ` [patch 09/41] cpu alloc: Genhd statistics conversion Christoph Lameter
2008-05-30 3:56 ` [patch 10/41] cpu alloc: blktrace conversion Christoph Lameter
2008-05-30 3:56 ` [patch 11/41] cpu alloc: SRCU cpu alloc conversion Christoph Lameter
2008-05-30 3:56 ` [patch 12/41] cpu alloc: XFS counter conversion Christoph Lameter
2008-05-30 3:56 ` [patch 13/41] cpu alloc: NFS statistics Christoph Lameter
2008-05-30 3:56 ` [patch 14/41] cpu alloc: Neigbour statistics Christoph Lameter
2008-05-30 3:56 ` [patch 15/41] cpu_alloc: Convert ip route statistics Christoph Lameter
2008-05-30 3:56 ` [patch 16/41] cpu alloc: Tcp statistics conversion Christoph Lameter
2008-05-30 3:56 ` [patch 17/41] cpu alloc: Convert scratches to cpu alloc Christoph Lameter
2008-05-30 3:56 ` [patch 18/41] cpu alloc: Dmaengine conversion Christoph Lameter
2008-05-30 3:56 ` [patch 19/41] cpu alloc: Convert loopback statistics Christoph Lameter
2008-05-30 3:56 ` [patch 20/41] cpu alloc: Veth conversion Christoph Lameter
2008-05-30 3:56 ` [patch 21/41] cpu alloc: Chelsio statistics conversion Christoph Lameter
2008-05-30 3:56 ` [patch 22/41] cpu alloc: Convert network sockets inuse counter Christoph Lameter
2008-05-30 3:56 ` [patch 23/41] cpu alloc: Use it for infiniband Christoph Lameter
2008-05-30 3:56 ` [patch 24/41] cpu alloc: Use in the crypto subsystem Christoph Lameter
2008-05-30 3:56 ` [patch 25/41] cpu alloc: scheduler: Convert cpuusage to cpu_alloc Christoph Lameter
2008-05-30 3:56 ` [patch 26/41] cpu alloc: Convert mib handling to cpu alloc Christoph Lameter
2008-05-30 6:47 ` Eric Dumazet
2008-05-30 18:01 ` Christoph Lameter
2008-05-30 3:56 ` [patch 27/41] cpu alloc: Remove the allocpercpu functionality Christoph Lameter
2008-05-30 4:58 ` Andrew Morton
2008-05-30 3:56 ` [patch 28/41] Module handling: Use CPU_xx ops to dynamically allocate counters Christoph Lameter
2008-05-30 3:56 ` [patch 29/41] x86_64: Use CPU ops for nmi alert counter Christoph Lameter
2008-05-30 3:56 ` [patch 30/41] Remove local_t support Christoph Lameter
2008-05-30 3:56 ` [patch 31/41] VM statistics: Use CPU ops Christoph Lameter
2008-05-30 3:56 ` [patch 32/41] cpu alloc: Use in slub Christoph Lameter
2008-05-30 3:56 ` [patch 33/41] cpu alloc: Remove slub fields Christoph Lameter
2008-05-30 3:56 ` [patch 34/41] cpu alloc: Page allocator conversion Christoph Lameter
2008-05-30 3:56 ` [patch 35/41] Support for CPU ops Christoph Lameter
2008-05-30 4:58 ` Andrew Morton
2008-05-30 5:18 ` Christoph Lameter
2008-05-30 3:56 ` [patch 36/41] Zero based percpu: Infrastructure to rebase the per cpu area to zero Christoph Lameter
2008-05-30 3:56 ` [patch 37/41] x86_64: Fold pda into per cpu area Christoph Lameter
2008-05-30 3:56 ` [patch 38/41] x86: Extend percpu ops to 64 bit Christoph Lameter
2008-05-30 3:56 ` [patch 39/41] x86: Replace cpu_pda() using percpu logic and get rid of _cpu_pda() Christoph Lameter
2008-05-30 3:57 ` [patch 40/41] x86: Replace xxx_pda() operations with x86_xx_percpu() Christoph Lameter
2008-05-30 3:57 ` [patch 41/41] x86_64: Support for cpu ops Christoph Lameter
2008-05-30 4:58 ` [patch 00/41] cpu alloc / cpu ops v3: Optimize per cpu access Andrew Morton
2008-05-30 5:03 ` Christoph Lameter
2008-05-30 5:21 ` Andrew Morton
2008-05-30 5:27 ` Christoph Lameter
2008-05-30 5:49 ` Andrew Morton
2008-05-30 6:16 ` Christoph Lameter
2008-05-30 6:51 ` KAMEZAWA Hiroyuki
2008-05-30 14:38 ` Mike Travis
2008-05-30 17:50 ` Christoph Lameter
2008-05-30 18:00 ` Matthew Wilcox
2008-05-30 18:12 ` Christoph Lameter
2008-05-30 6:01 ` Eric Dumazet
2008-05-30 6:16 ` Andrew Morton
2008-05-30 6:22 ` Christoph Lameter
2008-05-30 6:37 ` Andrew Morton
2008-05-30 11:32 ` Matthew Wilcox [this message]
2008-06-04 15:07 ` Mike Travis
2008-06-06 5:33 ` Eric Dumazet
2008-06-06 13:08 ` Mike Travis
2008-06-08 6:00 ` Rusty Russell
2008-06-09 18:44 ` Christoph Lameter
2008-06-09 19:11 ` Andi Kleen
2008-06-09 20:15 ` Eric Dumazet
2008-05-30 9:12 ` Peter Zijlstra
2008-05-30 9:18 ` Ingo Molnar
2008-05-30 18:11 ` Christoph Lameter
2008-05-30 18:40 ` Peter Zijlstra
2008-05-30 18:56 ` Christoph Lameter
2008-05-30 19:13 ` Peter Zijlstra
2008-06-01 3:25 ` Christoph Lameter
2008-06-01 8:19 ` Peter Zijlstra
2008-05-30 18:06 ` Christoph Lameter
2008-05-30 18:19 ` Peter Zijlstra
2008-05-30 18:26 ` Christoph Lameter
2008-05-30 18:47 ` Peter Zijlstra
2008-05-30 19:10 ` Christoph Lameter
2008-05-30 19:21 ` Peter Zijlstra
2008-05-30 19:35 ` Peter Zijlstra
2008-06-01 3:27 ` Christoph Lameter
2008-05-30 18:08 ` Christoph Lameter
2008-05-30 18:39 ` Peter Zijlstra
2008-05-30 18:51 ` Christoph Lameter
2008-05-30 19:00 ` Peter Zijlstra
2008-05-30 19:11 ` 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=20080530113259.GE22636@parisc-linux.org \
--to=matthew@wil.cx \
--cc=akpm@linux-foundation.org \
--cc=clameter@sgi.com \
--cc=dada1@cosmosbay.com \
--cc=davem@davemloft.net \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=rusty@rustcorp.com.au \
--cc=travis@sgi.com \
/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