From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Chris Metcalf <cmetcalf@tilera.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
Peter Zijlstra <peterz@infradead.org>,
"H. Peter Anvin" <hpa@zytor.com>,
Frederic Weisbecker <fweisbec@gmail.com>,
Thomas Gleixner <tglx@linutronix.de>,
LKML <linux-kernel@vger.kernel.org>,
Paul Mackerras <paulus@au1.ibm.com>,
Ingo Molnar <mingo@kernel.org>,
James Hogan <james.hogan@imgtec.com>,
"James E.J. Bottomley" <jejb@parisc-linux.org>,
Helge Deller <deller@gmx.de>,
Martin Schwidefsky <schwidefsky@de.ibm.com>,
Heiko Carstens <heiko.carstens@de.ibm.com>,
"David S. Miller" <davem@davemloft.net>,
Andrew Morton <akpm@linux-foundation.org>
Subject: Re: [RFC GIT PULL] softirq: Consolidation and stack overrun fix
Date: Tue, 24 Sep 2013 06:57:17 +1000 [thread overview]
Message-ID: <1379969837.11249.24.camel@pasglop> (raw)
In-Reply-To: <5240819F.4050903@tilera.com>
On Mon, 2013-09-23 at 13:59 -0400, Chris Metcalf wrote:
> We just came up against this on tilegx with a customer bug report in a
> PREEMPT environment. On tile, %tp is a GPR that points to the percpu area.
> The following seems to be the right abstraction -- though I'd also argue
> that letting barrier() clobber not just memory, but %tp, might be a better
> solution, but it's not clear what the best way is to do per-architecture
> overrides of per-compiler definitions like barrier(). See also the ARM v7
> code, which has to do something similar, though their percpu pointer is
> not a GPR, which changes the tradeoffs somewhat.
Hrm, if I read correctly what you did is that you read "tp" into another
register *and* also mark that action as clobbering the top int on the stack ?
I don't quite get what the stack clobber brings you here and how it works
around the fact that gcc might still cache that copy of tp into another
register accross preempt_enable/disable...
It's hard to tell with gcc ... the best I've had so far as an option was
something that would mark my per-cpu register (r13) *itself* as clobbered...
Cheers,
Ben.
next prev parent reply other threads:[~2013-09-23 20:58 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-09-19 19:51 [RFC GIT PULL] softirq: Consolidation and stack overrun fix Frederic Weisbecker
2013-09-19 19:51 ` [PATCH 1/3] irq: Consolidate do_softirq() arch overriden implementations Frederic Weisbecker
2013-09-19 19:51 ` [PATCH 2/3] irq: Execute softirq on its own stack on irq exit Frederic Weisbecker
2013-09-19 19:51 ` [PATCH 3/3] irq: Comment on the use of inline stack for ksoftirqd Frederic Weisbecker
2013-09-20 0:02 ` [RFC GIT PULL] softirq: Consolidation and stack overrun fix Linus Torvalds
2013-09-20 1:53 ` Benjamin Herrenschmidt
2013-09-20 11:03 ` Thomas Gleixner
2013-09-20 11:11 ` Peter Zijlstra
2013-09-21 0:55 ` Benjamin Herrenschmidt
2013-09-20 16:26 ` Frederic Weisbecker
2013-09-20 17:30 ` Thomas Gleixner
2013-09-20 18:37 ` Frederic Weisbecker
2013-09-20 22:14 ` Linus Torvalds
2013-09-21 7:47 ` Ingo Molnar
2013-09-21 18:58 ` Frederic Weisbecker
2013-09-21 21:45 ` Benjamin Herrenschmidt
2013-09-21 23:27 ` Frederic Weisbecker
2013-09-22 2:01 ` H. Peter Anvin
2013-09-22 4:39 ` Benjamin Herrenschmidt
2013-09-22 4:41 ` Benjamin Herrenschmidt
2013-09-22 16:24 ` Peter Zijlstra
2013-09-22 17:47 ` H. Peter Anvin
2013-09-22 22:00 ` Benjamin Herrenschmidt
2013-09-22 21:56 ` Benjamin Herrenschmidt
2013-09-22 22:22 ` Linus Torvalds
2013-09-22 22:38 ` Benjamin Herrenschmidt
2013-09-23 4:35 ` [PATCH] powerpc/irq: Run softirqs off the top of the irq stack Benjamin Herrenschmidt
2013-09-23 4:35 ` Benjamin Herrenschmidt
2013-09-23 7:56 ` Stephen Rothwell
2013-09-23 7:56 ` Stephen Rothwell
2013-09-23 10:13 ` Benjamin Herrenschmidt
2013-09-23 10:13 ` Benjamin Herrenschmidt
2013-09-23 16:47 ` Linus Torvalds
2013-09-23 16:47 ` Linus Torvalds
2013-09-23 20:51 ` Benjamin Herrenschmidt
2013-09-23 20:51 ` Benjamin Herrenschmidt
2013-09-24 5:42 ` [PATCH v2] " Benjamin Herrenschmidt
2013-09-24 5:42 ` Benjamin Herrenschmidt
2013-09-23 17:59 ` [RFC GIT PULL] softirq: Consolidation and stack overrun fix Chris Metcalf
2013-09-23 20:57 ` Benjamin Herrenschmidt [this message]
2013-09-24 19:27 ` Chris Metcalf
2013-09-24 20:58 ` Benjamin Herrenschmidt
2013-09-24 0:10 ` Benjamin Herrenschmidt
2013-09-24 1:19 ` Linus Torvalds
2013-09-24 1:52 ` Benjamin Herrenschmidt
2013-09-24 8:04 ` Peter Zijlstra
2013-09-24 8:16 ` Benjamin Herrenschmidt
2013-09-24 8:21 ` Peter Zijlstra
2013-09-24 9:31 ` Benjamin Herrenschmidt
2013-09-23 4:40 ` Benjamin Herrenschmidt
2013-09-23 5:01 ` David Miller
2013-09-24 2:44 ` Frederic Weisbecker
2013-09-24 4:42 ` Benjamin Herrenschmidt
2013-09-24 13:56 ` Frederic Weisbecker
2013-09-24 20:55 ` Benjamin Herrenschmidt
2013-09-25 8:46 ` Frederic Weisbecker
2013-09-21 0:52 ` Benjamin Herrenschmidt
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=1379969837.11249.24.camel@pasglop \
--to=benh@kernel.crashing.org \
--cc=akpm@linux-foundation.org \
--cc=cmetcalf@tilera.com \
--cc=davem@davemloft.net \
--cc=deller@gmx.de \
--cc=fweisbec@gmail.com \
--cc=heiko.carstens@de.ibm.com \
--cc=hpa@zytor.com \
--cc=james.hogan@imgtec.com \
--cc=jejb@parisc-linux.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=paulus@au1.ibm.com \
--cc=peterz@infradead.org \
--cc=schwidefsky@de.ibm.com \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.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.