From: "Paul E. McKenney" <paulmck@us.ibm.com>
To: Oliver Neukum <oliver@neukum.org>
Cc: Alan Stern <stern@rowland.harvard.edu>,
David Howells <dhowells@redhat.com>,
Kernel development list <linux-kernel@vger.kernel.org>
Subject: Re: Uses for memory barriers
Date: Mon, 11 Sep 2006 13:29:08 -0700 [thread overview]
Message-ID: <20060911202908.GF1295@us.ibm.com> (raw)
In-Reply-To: <200609112148.42302.oliver@neukum.org>
On Mon, Sep 11, 2006 at 09:48:42PM +0200, Oliver Neukum wrote:
> Am Montag, 11. September 2006 18:21 schrieb Paul E. McKenney:
> > 1. A given CPU will always perceive its own memory operations
> > as occuring in program order.
>
> Is this true for physical memory if virtually indexed caches are
> involved?
As I understand it, in systems with virtually indexed caches, the OS must
take care to ensure that a given cacheline appears only once in the cache,
even if it is mapped to multiple virtual addresses. If an OS failed to
do this, then, as far as I can see, all bets are off. Curt Schimmel's
book "UNIX(R) Systems for Modern Architectures: Symmetric Multiprocessing
and Caching for Kernel Programmers" is an excellent guide to the issues
posed by virtually indexed and virtually tagged caches.
In principle, one could construct a virtually indexed/tagged CPU cache
that automatically ejected any line with a conflicting physical address
(given that lookups are presumably much more frequent than loading new
cache lines), but I have no idea if any real hardware takes this approach.
I have had the good fortune to always work with physically tagged/indexed
caches. ;-)
Thanx, Paul
next prev parent reply other threads:[~2006-09-11 20:28 UTC|newest]
Thread overview: 92+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <200609082230.22225.oliver@neukum.org>
2006-09-08 21:26 ` Alan Stern
2006-09-08 21:46 ` Oliver Neukum
2006-09-08 22:25 ` Alan Stern
2006-09-08 22:49 ` Oliver Neukum
2006-09-09 2:25 ` Alan Stern
2006-09-11 16:21 ` Paul E. McKenney
2006-09-11 16:50 ` Alan Stern
2006-09-11 17:23 ` Segher Boessenkool
2006-09-11 19:04 ` Paul E. McKenney
2006-09-11 19:03 ` Paul E. McKenney
2006-09-11 17:21 ` Segher Boessenkool
2006-09-11 19:48 ` Oliver Neukum
2006-09-11 20:29 ` Paul E. McKenney [this message]
2006-09-12 8:57 ` David Howells
2006-09-12 9:01 ` David Howells
2006-09-12 10:22 ` Oliver Neukum
2006-09-12 14:55 ` Paul E. McKenney
2006-09-12 15:07 ` Oliver Neukum
2006-09-12 16:12 ` Paul E. McKenney
2006-09-12 17:50 ` Segher Boessenkool
2006-09-12 14:42 ` Paul E. McKenney
[not found] <20060911190005.GA1295@us.ibm.com>
2006-09-12 18:08 ` Alan Stern
2006-09-12 20:23 ` Paul E. McKenney
2006-09-14 14:58 ` Alan Stern
2006-09-15 5:16 ` Paul E. McKenney
2006-09-15 19:48 ` Alan Stern
2006-09-16 4:19 ` Paul E. McKenney
2006-09-16 15:28 ` Alan Stern
2006-09-18 19:13 ` Paul E. McKenney
2006-09-18 20:13 ` Alan Stern
2006-09-19 0:47 ` Paul E. McKenney
2006-09-19 16:04 ` Alan Stern
2006-09-19 16:38 ` Nick Piggin
2006-09-19 17:40 ` Alan Stern
2006-09-19 17:51 ` Nick Piggin
2006-09-19 18:19 ` Paul E. McKenney
2006-09-19 18:48 ` Nick Piggin
2006-09-19 19:36 ` Paul E. McKenney
2006-09-19 19:48 ` Nick Piggin
2006-09-19 20:01 ` Paul E. McKenney
2006-09-19 20:38 ` Alan Stern
2006-09-21 1:43 ` Paul E. McKenney
2006-09-19 18:16 ` Paul E. McKenney
2006-09-20 19:39 ` Alan Stern
2006-09-21 1:34 ` Paul E. McKenney
2006-09-21 20:59 ` Alan Stern
2006-09-22 5:02 ` Paul E. McKenney
2006-09-22 20:38 ` Alan Stern
2006-09-27 21:06 ` Alan Stern
2006-09-30 1:11 ` Paul E. McKenney
2006-09-30 21:01 ` Alan Stern
2006-10-02 0:06 ` Paul E. McKenney
2006-10-02 15:44 ` Alan Stern
2006-10-04 15:35 ` Paul E. McKenney
2006-10-04 18:04 ` Alan Stern
2006-10-13 16:51 ` Paul E. McKenney
2006-10-13 18:30 ` Alan Stern
2006-10-13 22:39 ` Paul E. McKenney
2006-10-14 2:27 ` Alan Stern
2006-10-17 1:24 ` Paul E. McKenney
2006-10-17 15:29 ` Alan Stern
2006-10-17 17:27 ` Paul E. McKenney
2006-10-17 19:42 ` Alan Stern
2006-10-17 20:15 ` Paul E. McKenney
2006-10-17 21:21 ` Alan Stern
2006-10-17 22:58 ` Paul E. McKenney
2006-10-18 19:05 ` Alan Stern
2006-10-18 23:01 ` Paul E. McKenney
2006-10-19 16:44 ` Alan Stern
2006-10-19 19:21 ` Paul E. McKenney
2006-10-19 20:55 ` Alan Stern
2006-10-19 22:46 ` Paul E. McKenney
2006-10-20 16:54 ` Alan Stern
2006-10-21 0:59 ` Paul E. McKenney
2006-10-21 19:47 ` Alan Stern
2006-10-21 22:52 ` Paul E. McKenney
2006-10-22 2:18 ` Alan Stern
2006-10-23 5:32 ` Paul E. McKenney
2006-10-23 14:07 ` Alan Stern
2006-10-24 17:52 ` Paul E. McKenney
[not found] <200609081929.33027.oliver@neukum.org>
2006-09-08 18:06 ` Alan Stern
2006-09-08 18:22 ` Oliver Neukum
2006-09-07 21:25 Alan Stern
2006-09-07 22:10 ` linux-os (Dick Johnson)
2006-09-08 18:39 ` Alan Stern
2006-09-08 0:14 ` Paul E. McKenney
2006-09-08 15:55 ` Alan Stern
2006-09-08 18:57 ` Paul E. McKenney
2006-09-08 21:23 ` Alan Stern
2006-09-09 0:44 ` Paul E. McKenney
2006-09-11 16:05 ` Alan Stern
2006-09-08 5:52 ` David Schwartz
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=20060911202908.GF1295@us.ibm.com \
--to=paulmck@us.ibm.com \
--cc=dhowells@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=oliver@neukum.org \
--cc=stern@rowland.harvard.edu \
/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
Powered by JetHome