All of lore.kernel.org
 help / color / mirror / Atom feed
From: Eric Dumazet <dada1@cosmosbay.com>
To: "Siddha, Suresh B" <suresh.b.siddha@intel.com>
Cc: Christoph Lameter <christoph@lameter.com>,
	Andi Kleen <andi@firstfloor.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux kernel <linux-kernel@vger.kernel.org>
Subject: Re: non-NUMA cache_free_alien() (was Re: [RFC] SLAB : NUMA cache_free_alien() very expensive because of virt_to_slab(objp); nodeid = slabp->nodeid;)
Date: Thu, 22 Mar 2007 23:12:39 +0100	[thread overview]
Message-ID: <4602FF57.7030001@cosmosbay.com> (raw)
In-Reply-To: <20070322212831.GA25665@linux-os.sc.intel.com>

Siddha, Suresh B a écrit :
> Christoph,
> 
> While we are at this topic, recently I had reports that
> cache_free_alien() is costly on non NUMA platforms too (similar
> to the cache miss issues that Eric was referring to on NUMA)
> and the appended patch seems to fix it for non NUMA atleast.
> 
> Appended patch gives a nice 1% perf improvement on non-NUMA platform
> with database workload.
> 
> Please comment or Ack for mainline :)

I have one comment :)

> @@ -1394,6 +1394,9 @@ void __init kmem_cache_init(void)
>  	int order;
>  	int node;
>  
> +	if (num_online_nodes() == 1)
> +		use_alien_caches = 0;
> +

Unfortunatly this part is wrong.

You should check num_possible_nodes(), or nr_node_ids (this one is cheaper, 
its a variable instead of a function call)

I wonder if we could add a new SLAB_NUMA_BYPASS, so that we can declare some 
kmem_cache as non NUMA aware (for example, I feel network skb dont need the 
NUMA overhead)



  parent reply	other threads:[~2007-03-22 22:12 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-03-20 17:12 [RFC] SLAB : NUMA cache_free_alien() very expensive because of virt_to_slab(objp); nodeid = slabp->nodeid; Eric Dumazet
2007-03-20 19:54 ` Christoph Lameter
2007-03-20 21:32   ` Andi Kleen
2007-03-20 22:09     ` Eric Dumazet
2007-03-21  0:16       ` Christoph Lameter
2007-03-21  6:27         ` Eric Dumazet
2007-03-21  6:57           ` [PATCH] SLAB : Use num_possible_cpus() in enable_cpucache() Eric Dumazet
2007-03-21  7:21             ` [PATCH] SLAB : Dont allocate empty shared caches Eric Dumazet
2007-03-21 13:13               ` Pekka Enberg
2007-03-21 13:02             ` [PATCH] SLAB : Use num_possible_cpus() in enable_cpucache() Pekka Enberg
2007-03-21 18:45             ` Christoph Lameter
2007-03-21  7:03           ` [RFC] SLAB : NUMA cache_free_alien() very expensive because of virt_to_slab(objp); nodeid = slabp->nodeid; Christoph Lameter
2007-03-21  7:14             ` Eric Dumazet
2007-03-21 14:35               ` Christoph Lameter
2007-03-21  0:18     ` Christoph Lameter
2007-03-21  2:44       ` Andi Kleen
2007-03-21  3:10         ` Christoph Lameter
2007-03-22 21:28           ` non-NUMA cache_free_alien() (was Re: [RFC] SLAB : NUMA cache_free_alien() very expensive because of virt_to_slab(objp); nodeid = slabp->nodeid;) Siddha, Suresh B
2007-03-22 22:10             ` Christoph Lameter
2007-03-22 22:12             ` Eric Dumazet [this message]
2007-03-22 22:40               ` Siddha, Suresh B
2007-03-22 22:56                 ` Eric Dumazet
2007-03-23  1:25                 ` Christoph Lameter
2007-03-23 14:14                   ` Andi Kleen
2007-03-23 14:12                 ` Andi Kleen
2007-04-02 22:55                   ` Siddha, Suresh B
2007-04-03  0:23                     ` Christoph Lameter
2007-04-03  0:31                       ` Siddha, Suresh B
2007-04-09 18:01                       ` [patch 1/2] x86_64: set node_possible_map at runtime Siddha, Suresh B
2007-04-09 18:07                         ` [patch 2/2] slab, x86_64: skip cache_free_alien() on non NUMA Siddha, Suresh B
2007-04-09 20:23                         ` [patch 1/2] x86_64: set node_possible_map at runtime Andrew Morton

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=4602FF57.7030001@cosmosbay.com \
    --to=dada1@cosmosbay.com \
    --cc=akpm@linux-foundation.org \
    --cc=andi@firstfloor.org \
    --cc=christoph@lameter.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=suresh.b.siddha@intel.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 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.