From: Eric Dumazet <dada1@cosmosbay.com>
To: Andi Kleen <ak@suse.de>
Cc: Eric dumazet <abuse@cosmosbay.com>,
netdev@vger.kernel.org, netfilter-devel@lists.netfilter.org
Subject: Re: [PATCH, netfilter] NUMA aware ipv4/netfilter/ip_tables.c
Date: Mon, 19 Sep 2005 21:09:14 +0200 [thread overview]
Message-ID: <432F0CDA.40207@cosmosbay.com> (raw)
In-Reply-To: <200509191948.55333.ak@suse.de>
Andi Kleen a écrit :
> On Monday 19 September 2005 19:09, Eric dumazet wrote:
>
>>Hi
>>
>>Part of the performance problem we have with netfilter is memory
>>allocation is not NUMA aware, but 'only' SMP aware.
>
>
> How do you know? Did you measure it somehow?
Well, it seems that even a 1MB cache is not enough for this workload, on a
dual Opterons 275 machine.
CPU: AMD64 processors, speed 2205.54 MHz (estimated)
Counted DATA_CACHE_MISSES events (Data cache misses) with a unit mask of 0x00
(No unit mask) count 5000
samples % symbol name
86104 13.3027 ipt_do_table
21508 3.3229 copy_user_generic_c
18618 2.8764 try_to_wake_up
18180 2.8087 memset
18085 2.7941 tcp_v4_rcv
16949 2.6185 schedule
# iptables -nvL | wc -l
190
>
> Normally I would expect fire wall state to mostly fit
> into caches (1MB+) of modern Opterons and other NUMA systems.
>
>
>>What do you think of this patch ?
>>
>>Note : The allocation function is quite complex (copied from
>>drivers/pci/pci-driver.c pci_call_probe())
>>because current kernel doesnt have a NUMA aware vmalloc() wrapper
>
>
> Normal vmalloc is NUMA aware and allocates on the local node.
> Like all memory allocations
Yes, but I want to allocate percpu (and per node) data, using vmalloc(), not
memory on the local node.
>
>
>>, maybe
>>a future kernel will export
>>such a common function ?
>
>
> BTW if you want to put data onto a specific node you don't need
> to schedule there - it is enough to set the process mempolicy
> with sys_set_mempolicy. This is fine from kernel too, as long
> as you save/restore the old one (current->mempolicy)
OK thanks, I will try this.
Eric
next prev parent reply other threads:[~2005-09-19 19:09 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-09-19 17:09 [PATCH, netfilter] NUMA aware ipv4/netfilter/ip_tables.c Eric dumazet
2005-09-19 17:20 ` Eric Dumazet
2005-09-19 17:48 ` Andi Kleen
2005-09-19 19:09 ` Eric Dumazet [this message]
2005-09-20 9:47 ` Eric Dumazet
2005-09-20 16:30 ` Andi Kleen
2005-09-20 17:02 ` Eric Dumazet
2005-09-20 21:45 ` [PATCH] Adds sys_set_mempolicy() in include/linux/syscalls.h , " Eric Dumazet
2005-09-20 21:46 ` [PATCH] Adds sys_set_mempolicy() in include/linux/syscalls.h Eric Dumazet
2005-09-21 21:24 ` [PATCH 0/3] netfilter : 3 patches to boost ip_tables performance Eric Dumazet
2005-09-21 22:43 ` Christoph Lameter
2005-09-22 0:34 ` David S. Miller
2005-09-22 1:44 ` Christoph Lameter
2005-09-22 12:11 ` Eric Dumazet
2005-09-22 12:49 ` Christoph Hellwig
2005-09-22 12:54 ` Andi Kleen
2005-09-22 12:58 ` Christoph Hellwig
2005-09-22 13:05 ` Andi Kleen
2005-09-22 15:37 ` Christoph Lameter
2005-09-22 15:50 ` Eric Dumazet
2005-09-22 15:55 ` Christoph Lameter
2005-09-23 17:11 ` Harald Welte
2005-09-23 17:44 ` Christoph Lameter
2005-09-23 18:04 ` Dave Hansen
2005-09-23 17:47 ` Eric Dumazet
2005-09-23 18:00 ` Kyle Moffett
2005-09-22 4:18 ` James Morris
2005-09-22 5:07 ` Eric Dumazet
2005-09-22 13:03 ` Andi Kleen
2005-09-22 13:30 ` Eric Dumazet
2005-09-23 17:09 ` Harald Welte
2005-09-27 16:23 ` Andi Kleen
2005-09-28 0:25 ` Henrik Nordstrom
2005-09-28 8:32 ` Harald Welte
2005-09-28 8:37 ` Andi Kleen
2005-10-04 17:01 ` Patrick McHardy
2005-10-05 16:53 ` Andi Kleen
2005-10-07 2:38 ` Harald Welte
2005-10-06 17:59 ` Andi Kleen
2005-10-07 17:08 ` Patrick McHardy
2005-10-07 17:21 ` Andi Kleen
2005-10-07 17:50 ` Patrick McHardy
2005-09-28 10:34 ` Henrik Nordstrom
2005-11-25 11:23 ` [PATCH] netfilter : zap get_cpu()/put_cpu() calls from ip_tables Eric Dumazet
2005-11-25 11:28 ` [PATCH (resent with the attachment !)] " Eric Dumazet
2005-11-25 18:20 ` Patrick McHardy
2005-09-21 21:29 ` [PATCH 1/3] netfilter : 3 patches to boost ip_tables performance Eric Dumazet
2005-09-22 12:57 ` Harald Welte
2005-09-22 13:17 ` Eric Dumazet
2005-09-21 21:32 ` [PATCH 2/3] " Eric Dumazet
2005-09-22 12:48 ` Harald Welte
2005-09-22 13:05 ` Eric Dumazet
2005-09-23 4:02 ` Willy Tarreau
2005-09-23 5:14 ` Eric Dumazet
2005-09-23 11:33 ` Willy Tarreau
2005-09-23 14:00 ` Tim Mattox
2005-09-21 21:37 ` [PATCH 3/3] " Eric Dumazet
2005-09-22 12:50 ` Harald Welte
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=432F0CDA.40207@cosmosbay.com \
--to=dada1@cosmosbay.com \
--cc=abuse@cosmosbay.com \
--cc=ak@suse.de \
--cc=netdev@vger.kernel.org \
--cc=netfilter-devel@lists.netfilter.org \
/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;
as well as URLs for NNTP newsgroup(s).