From: Andi Kleen <ak@suse.de>
To: davidm@hpl.hp.com
Cc: Andi Kleen <ak@suse.de>, Rusty Russell <rusty@rustcorp.com.au>,
linux-kernel@vger.kernel.org, torvalds@transmeta.com,
rth@twiddle.net
Subject: Re: [PATCH] 2.5.1-pre5: per-cpu areas
Date: Thu, 14 Mar 2002 19:51:22 +0100 [thread overview]
Message-ID: <20020314195122.A30566@wotan.suse.de> (raw)
In-Reply-To: <15504.7958.677592.908691@napali.hpl.hp.com.suse.lists.linux.kernel> <E16lMzi-0002bb-00@wagner.rustcorp.com.au.suse.lists.linux.kernel> <p73bsdrsftu.fsf@oldwotan.suse.de> <15504.58901.82038.420706@napali.hpl.hp.com>
In-Reply-To: <15504.58901.82038.420706@napali.hpl.hp.com>
On Thu, Mar 14, 2002 at 10:04:05AM -0800, David Mosberger wrote:
> How about the following proposal:
>
> - taking the address of this_cpu(var) is never allowed (can be
> enforced with RELOC_HIDE())
>
> - taking the address of per_cpu(var, n) is always legal and
> will return a pointer which will access CPU n's version of
> the variable, no matter what CPU dereferences the pointer
>
> Andi, I think this would take care of the x86-64 problem as well, right?
Yes, it would.
It would be a bit more overhead for taking the address than a
this_cpu_address(), because one would need to fetch the CPU number first
and do the arithmetic and the array reference instead of fetching the
address directly. But I agree that per_cpu() has much cleaner semantics than
this_cpu_address() for addresses, so it is worth it.
When one considers preemptive kernels where you can lose your CPU anytime
then it makes even more sense. But then this_cpu is only safe there when
you turn off preemption or hold some lock or run in interrupt context.
Outside such regions per_cpu() seems to be safer.
Thanks,
-Andi
next prev parent reply other threads:[~2002-03-14 18:51 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <15504.7958.677592.908691@napali.hpl.hp.com.suse.lists.linux.kernel>
[not found] ` <E16lMzi-0002bb-00@wagner.rustcorp.com.au.suse.lists.linux.kernel>
2002-03-14 8:39 ` [PATCH] 2.5.1-pre5: per-cpu areas Andi Kleen
2002-03-14 11:09 ` Rusty Russell
2002-03-14 11:14 ` Andi Kleen
2002-03-14 19:48 ` H. Peter Anvin
2002-03-14 18:04 ` David Mosberger
2002-03-14 18:51 ` Andi Kleen [this message]
2002-03-15 4:07 ` Rusty Russell
2002-03-15 9:13 ` Andi Kleen
2002-03-17 7:17 ` Rusty Russell
2002-03-18 7:35 ` Andi Kleen
2002-03-19 0:02 ` Rusty Russell
2002-03-19 0:08 ` J.A. Magallon
2002-03-19 0:15 ` Andi Kleen
2002-03-19 17:05 ` Richard Henderson
2002-03-15 4:19 ` David Mosberger
2002-03-15 5:52 ` Rusty Russell
[not found] <15504.7958.677592.908691@napali.hpl.hp.com>
2002-03-14 4:37 ` Rusty Russell
2002-03-14 5:05 ` Jeff Garzik
2002-03-14 11:14 ` Rusty Russell
2002-03-14 11:26 ` Jeff Garzik
2002-03-14 12:16 ` Rusty Russell
2002-03-14 12:25 ` Jeff Garzik
2002-03-15 1:00 ` Richard Henderson
2002-03-14 9:37 ` Richard Henderson
2002-03-14 18:06 ` David Mosberger
2001-12-05 22:09 Rusty Russell
2001-12-06 7:21 ` Keith Owens
2001-12-06 8:07 ` David S. Miller
2001-12-06 9:18 ` Chris Wedgwood
2001-12-07 15:03 ` Pavel Machek
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=20020314195122.A30566@wotan.suse.de \
--to=ak@suse.de \
--cc=davidm@hpl.hp.com \
--cc=linux-kernel@vger.kernel.org \
--cc=rth@twiddle.net \
--cc=rusty@rustcorp.com.au \
--cc=torvalds@transmeta.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;
as well as URLs for NNTP newsgroup(s).