All of lore.kernel.org
 help / color / mirror / Atom feed
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Kevin Hao <haokexin@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>, Torsten Duwe <duwe@lst.de>,
	linuxppc-dev@lists.ozlabs.org
Subject: Re: [RFC] powerpc: use ticket spin lock for !CONFIG_PPC_SPLPAR
Date: Fri, 13 Mar 2015 18:07:18 +1100	[thread overview]
Message-ID: <1426230438.17565.162.camel@kernel.crashing.org> (raw)
In-Reply-To: <20150313060913.GC24485@pek-khao-d1.corp.ad.wrs.com>

On Fri, 2015-03-13 at 14:09 +0800, Kevin Hao wrote:
> On Thu, Mar 12, 2015 at 04:24:10PM +0100, Torsten Duwe wrote:
> > But generally, which platforms would benefit most from this change?
> 
> In theory, the more cpus the platform has, the more serious the thundering
> herd problem is. So the latest platforms which has more cpus would benefit
> most.

In fact, pseries would be one to benefit ... it would be nice to find a
way to sort out the lock owner problem.

One idea I had was to keep the owners elsewhere, some kind of hash chain
of "taken locks" hashed on the lock address. There shouldn't be *that*
many locks held at a given point in time.

Trying to keep the owner in a separate & non-atomic part of the lock
proved being too hard due to how the lockref and generic code are
interwined, though maybe somebody smarter can find a trick :-)

Cheers,
Ben.

  reply	other threads:[~2015-03-13  7:07 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-12 10:55 [RFC] powerpc: use ticket spin lock for !CONFIG_PPC_SPLPAR Kevin Hao
2015-03-12 11:13 ` Benjamin Herrenschmidt
2015-03-12 15:24   ` Torsten Duwe
2015-03-13  6:09     ` Kevin Hao
2015-03-13  7:07       ` Benjamin Herrenschmidt [this message]
2015-03-13  5:59   ` Kevin Hao
2015-03-13  7:09   ` Michael Ellerman
2015-03-13  7:14     ` Benjamin Herrenschmidt
2015-03-16  0:25       ` Sam Bobroff

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=1426230438.17565.162.camel@kernel.crashing.org \
    --to=benh@kernel.crashing.org \
    --cc=duwe@lst.de \
    --cc=haokexin@gmail.com \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=paulus@samba.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 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.