All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Andi Kleen <andi@firstfloor.org>,
	Huang Ying <ying.huang@intel.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 8/5] llist: Remove cpu_relax() usage in cmpxchg loops
Date: Mon, 12 Sep 2011 17:09:04 +0200	[thread overview]
Message-ID: <1315840144.26517.66.camel@twins> (raw)
In-Reply-To: <20110912144706.GA21716@Krystal>

On Mon, 2011-09-12 at 10:47 -0400, Mathieu Desnoyers wrote:
> * Andi Kleen (andi@firstfloor.org) wrote:
> > On Mon, Sep 12, 2011 at 04:05:58PM +0200, Peter Zijlstra wrote:
> > > Subject: llist: Remove cpu_relax() usage in cmpxchg loops
> > > From: Peter Zijlstra <a.p.zijlstra@chello.nl>
> > > Date: Mon Sep 12 15:50:49 CEST 2011
> > > 
> > > Initial benchmarks show they're a net loss (2 socket wsm):
> > > 
> > 
> > May still save power.
> 
> Looking at kernel/spinlock.c:
> 
> void __lockfunc __raw_##op##_lock(locktype##_t *lock) \
> {
> [...]
>                 while (!raw_##op##_can_lock(lock) && (lock)->break_lock)\
>                         arch_##op##_relax(&lock->raw_lock); \
> 
> so basically, in typical locking primitives (spinlock), it looks like
> lower power consumption is preferred over getting the raw maximal
> performance in fully contented scenarios.

Who says its about power consumption? That was a baseless claim made by
Andi which you propagate as a truth. Typically PAUSE is too short to
really save any power and the only gain of having it in loops is to
provide a window where another core on the cache domain can have a go.

If power consumption would be the prime concern Intel should fix their
damn mwait implementation so we can use that for locks. Local spinners +
mwait would give a very power efficient 'spin'-lock.

> So what is the rationale for making those lock-less lists retry scheme
> different from spinlocks here ?

PAUSE should help on the medium contended case (it would be pointless on
heavy contention), but hurt the lightly contended case. We have all
kinds of contention spinlocks in the kernel, but we don't as of yet have
very contended llist users in the kernel.

Furthermore, I would argue we should avoid growing them, significantly
contended atomic ops are bad, use a different scheme.


  reply	other threads:[~2011-09-12 15:09 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-09-08  6:00 [PATCH -mm -v2 0/5] irq_work, Use llist in irq_work Huang Ying
2011-09-08  6:00 ` [PATCH -mm -v2 1/5] llist, Make all llist functions inline Huang Ying
2011-09-08  6:00 ` [PATCH -mm -v2 2/5] llist, Define macro to check NMI safe cmpxchg Huang Ying
2011-09-08  6:00 ` [PATCH -mm -v2 3/5] llist, Move cpu_relax after cmpxchg Huang Ying
2011-09-08  6:00 ` [PATCH -mm -v2 4/5] llist, Return whether list is empty before adding in llist_add Huang Ying
2011-09-08  6:00 ` [PATCH -mm -v2 5/5] irq_work, Use llist in irq_work Huang Ying
2011-09-12 14:05 ` [PATCH 6/5] llist: Add llist_next() Peter Zijlstra
2011-09-12 14:05 ` [PATCH 7/5] sched: Convert to use llist Peter Zijlstra
2011-09-12 14:05 ` [PATCH 8/5] llist: Remove cpu_relax() usage in cmpxchg loops Peter Zijlstra
2011-09-12 14:23   ` Andi Kleen
2011-09-12 14:23     ` Peter Zijlstra
2011-09-12 14:47     ` Mathieu Desnoyers
2011-09-12 15:09       ` Peter Zijlstra [this message]
2011-09-12 15:24       ` Peter Zijlstra
2011-09-12 16:38       ` Andi Kleen
2011-09-12 18:53         ` Peter Zijlstra
2011-09-12 14:26   ` Avi Kivity
2011-09-12 14:32     ` Peter Zijlstra
2011-09-13 11:43       ` Avi Kivity
2011-09-13 14:22         ` Peter Zijlstra
2011-09-13 14:51           ` Avi Kivity
2011-09-13 14:53             ` Peter Zijlstra
2011-09-12 14:06 ` [PATCH -mm -v2 0/5] irq_work, Use llist in irq_work Peter Zijlstra

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=1315840144.26517.66.camel@twins \
    --to=peterz@infradead.org \
    --cc=akpm@linux-foundation.org \
    --cc=andi@firstfloor.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mathieu.desnoyers@efficios.com \
    --cc=ying.huang@intel.com \
    --subject='Re: [PATCH 8/5] llist: Remove cpu_relax() usage in cmpxchg loops' \
    /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

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.