public inbox for netdev@vger.kernel.org 
 help / color / mirror / Atom feed
From: Greg Banks <gnb-97jfqw80gc6171pxa8y+qA@public•gmane.org>
To: Eric Dumazet <eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public•gmane.org>
Cc: NeilBrown <neilb-l3A5Bk7waGM@public•gmane.org>,
	linux-nfs-u79uwXL29TY76Z2rM5mHXA@public•gmane.org,
	David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public•gmane.org>,
	linux-kernel
	<linux-kernel-u79uwXL29TY76Z2rM5mHXA@public•gmane.org>,
	netdev <netdev-u79uwXL29TY76Z2rM5mHXA@public•gmane.org>
Subject: Re: Fw: [PATCH] sunrpc: use better NUMA affinities
Date: Fri, 29 Jul 2011 16:53:21 +1000	[thread overview]
Message-ID: <4E3258E1.6020000@fastmail.fm> (raw)
In-Reply-To: <1311921035.7845.10.camel@edumazet-laptop>

On 29/07/11 16:30, Eric Dumazet wrote:
> Le vendredi 29 juillet 2011 à 16:05 +1000, Greg Banks a écrit :
>> On 29/07/11 15:32, NeilBrown wrote:
>>
>> I seem to remember coming to the conclusion that Jeff eventually
>> addressed this problem...am I misremembering or did something regress?
>>
> Currently, all nfsd kthreads use memory for their kernel stack and
> various initial data from a _single_ node, even if you use
> sunrpc.pool_mode=pernode  (or percpu)

That's just plain broken and I'm very pleased to see you fix it.

I was just surprised that it was still broken and wondering how that 
happened.  Looking at ToT I see that because I dropped the ball in 2008, 
Jeff's patches didn't address the problem.  In ToT 
svc_pool_map_set_cpumask() is called *after* kthread_create() and 
applies to the child thread, *after* it's stack has been allocated on 
the wrong node.  In the working SGI code, svc_pool_map_set_cpumask() is 
called by the parent node on itself *before* calling kernel_thread() or 
doing any of the data structure allocations, thus ensuring that 
everything gets allocated using the default memory allocation policy, 
which on SGI NFS servers was globally tuned to be "node-local".

> With my patch, we make sure each thread gets its stack from its local
> node.
>
> Check commit 94dcf29a11b3d20a (kthread: use kthread_create_on_node()) to
> see how this strategy already was adopted for ksoftirqd, kworker,
> migration, and pktgend kthreads.

Ah, I see.  It's unfortunate that the kthread_create() API ends up being 
passed a CPU number but that's only used to format the name and not for 
sensible things :(

-- 
Greg.

--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public•gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

  reply	other threads:[~2011-07-29  6:53 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20110729153207.17af3085@notabene.brown>
2011-07-29  6:05 ` Fw: [PATCH] sunrpc: use better NUMA affinities Greg Banks
     [not found]   ` <4E324DB4.7060600-97jfqw80gc6171pxa8y+qA@public.gmane.org>
2011-07-29  6:30     ` Eric Dumazet
2011-07-29  6:53       ` Greg Banks [this message]
     [not found]         ` <4E3258E1.6020000-97jfqw80gc6171pxa8y+qA@public.gmane.org>
2011-07-29 10:36           ` Christoph Hellwig
2011-07-29 11:58             ` Greg Banks
2011-07-29 12:11               ` Eric Dumazet
2011-07-29 13:30                 ` Greg Banks
2011-07-29 16:48                   ` J. Bruce Fields
2011-07-29 16:53                     ` J. Bruce Fields
     [not found]                       ` <20110729165345.GM23194-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2011-07-29 18:15                         ` Eric Dumazet
2011-07-29 20:34                         ` Greg Banks
     [not found]                           ` <F562C6F4-3FC2-4952-A903-8E87E9EF2D9D-97jfqw80gc6171pxa8y+qA@public.gmane.org>
2011-07-29 23:30                             ` NeilBrown
     [not found]                               ` <20110730093025.716f3f50-wvvUuzkyo1EYVZTmpyfIwg@public.gmane.org>
2011-07-29 23:48                                 ` J. Bruce Fields
     [not found]                                   ` <20110729234857.GA30941-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2011-07-30  4:08                                     ` Eric Dumazet
2011-07-30  6:06                                       ` NeilBrown
2011-07-30  6:23                                         ` Eric Dumazet
2011-07-31  6:58                                           ` Eric Dumazet
2011-08-27  0:02                                             ` J. Bruce Fields
     [not found]                                               ` <20110827000219.GF18699-uC3wQj2KruNg9hUCZPvPmw@public.gmane.org>
2011-08-28 10:02                                                 ` Eric Dumazet
2011-08-02  1:06                                 ` J. Bruce Fields
2011-07-29 16:45         ` Fw: " J. Bruce Fields
2011-07-29 20:24           ` Greg Banks
2011-07-29 16:48     ` Fw: " J. Bruce Fields

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=4E3258E1.6020000@fastmail.fm \
    --to=gnb-97jfqw80gc6171pxa8y+qa@public$(echo .)gmane.org \
    --cc=davem-fT/PcQaiUtIeIZ0/mPfg9Q@public$(echo .)gmane.org \
    --cc=eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public$(echo .)gmane.org \
    --cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public$(echo .)gmane.org \
    --cc=linux-nfs-u79uwXL29TY76Z2rM5mHXA@public$(echo .)gmane.org \
    --cc=neilb-l3A5Bk7waGM@public$(echo .)gmane.org \
    --cc=netdev-u79uwXL29TY76Z2rM5mHXA@public$(echo .)gmane.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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