From: Jeremy Higdon <jeremy@sgi.com>
To: Keith Owens <kaos@sgi.com>
Cc: paulmck@us.ibm.com, Andi Kleen <ak@suse.de>,
ajwade@cpe001346162bf9-cm0011ae8cd564.cpe.net.cable.rogers.com,
vatsa@in.ibm.com, Oleg Nesterov <oleg@tv-sign.ru>,
linux-kernel@vger.kernel.org,
Dipankar Sarma <dipankar@in.ibm.com>,
Andrew Morton <akpm@osdl.org>, Ingo Molnar <mingo@elte.hu>
Subject: Re: Semantics of smp_mb() [was : Re: [PATCH] Fix RCU race in access of nohz_cpu_mask ]
Date: Thu, 15 Dec 2005 23:46:26 -0800 [thread overview]
Message-ID: <20051216074626.GB201289@sgi.com> (raw)
In-Reply-To: <17158.1134512861@ocs3.ocs.com.au>
Roland Dreier got this right. The purpose of the mmiowb is
to ensure that writes to I/O devices while holding a spinlock
are ordered with respect to writes issued after the original
processor releases and a second processor acquires said
spinlock.
A MMIO read would be sufficient, but is much heavier weight.
On the SGI MIPS-based systems, the "sync" instruction was used.
On the Altix systems, a register on the hub chip is read.
>From comments by jejb, we're looking at modifying the mmiowb
API by adding an argument which would be a register to read
from if the architecture in question needs ordering in this
way but does not have a lighter weight mechanism like the Altix
mmiowb. Since there will now need to be a width indication,
mmiowb will be replaced with mmiowb[bwlq].
jeremy
next prev parent reply other threads:[~2005-12-16 7:46 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-12-08 19:31 [PATCH] Fix RCU race in access of nohz_cpu_mask Oleg Nesterov
2005-12-09 2:46 ` Srivatsa Vaddagiri
2005-12-09 19:17 ` Oleg Nesterov
2005-12-10 15:19 ` Srivatsa Vaddagiri
2005-12-10 18:55 ` Oleg Nesterov
2005-12-11 17:41 ` Semantics of smp_mb() [was : Re: [PATCH] Fix RCU race in access of nohz_cpu_mask ] Srivatsa Vaddagiri
2005-12-11 21:21 ` Andrew James Wade
2005-12-11 23:45 ` Rusty Russell
2005-12-12 0:49 ` Keith Owens
2005-12-12 8:41 ` Srivatsa Vaddagiri
2005-12-12 19:33 ` Oleg Nesterov
2005-12-13 5:20 ` Paul E. McKenney
2005-12-13 5:07 ` Andrew James Wade
2005-12-13 5:43 ` Paul E. McKenney
2005-12-13 11:20 ` Andi Kleen
2005-12-13 16:20 ` Paul E. McKenney
2005-12-13 22:27 ` Keith Owens
2005-12-13 22:50 ` Paul E. McKenney
2005-12-14 1:12 ` Andi Kleen
2005-12-14 1:46 ` Paul E. McKenney
2005-12-15 21:15 ` Semantics of smp_mb() Roland Dreier
2005-12-16 7:46 ` Jeremy Higdon [this message]
2006-03-13 18:39 ` Semantics of smp_mb() [was : Re: [PATCH] Fix RCU race in access of nohz_cpu_mask ] Paul E. McKenney
2006-03-31 4:56 ` Jeremy Higdon
2006-03-31 6:18 ` Paul E. McKenney
2006-03-31 23:38 ` Jesse Barnes
2005-12-12 3:10 ` [PATCH] Fix RCU race in access of nohz_cpu_mask Paul E. McKenney
2005-12-12 4:32 ` Andrew Morton
2005-12-12 4:38 ` David S. Miller
2005-12-12 4:47 ` Nick Piggin
2005-12-12 4:49 ` Paul Mackerras
2005-12-12 6:27 ` Keith Owens
2005-12-09 2:56 ` Srivatsa Vaddagiri
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=20051216074626.GB201289@sgi.com \
--to=jeremy@sgi.com \
--cc=ajwade@cpe001346162bf9-cm0011ae8cd564.cpe.net.cable.rogers.com \
--cc=ak@suse.de \
--cc=akpm@osdl.org \
--cc=dipankar@in.ibm.com \
--cc=kaos@sgi.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=oleg@tv-sign.ru \
--cc=paulmck@us.ibm.com \
--cc=vatsa@in.ibm.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox