From: Rusty Russell <rusty@rustcorp.com.au>
To: torvalds@transmeta.com
Cc: akpm@zip.com.au, linux-kernel@vger.kernel.org
Subject: [PATCH] More barriers in module code.
Date: Thu, 09 Oct 2003 18:50:28 +1000 [thread overview]
Message-ID: <20031009085114.231B52C23B@lists.samba.org> (raw)
Paul McKenney convinced me that there's no *guarantee* that all archs
will refuse the speculate the atomic ops above the state test, eg:
CPU0 (stopref_set_state) CPU1 (stopref)
atomic_set(&ack, 0); if (state == XXX)
wmb(); atomic_inc(&ack);
state = XXX;
Certainly Alpha needs a rmb() inside stopref to guarantee it sees
the same ordering.
---
Linus, please apply.
Name: More Barriers for Module Stopref Threads
Author: Rusty Russell
Status: Trivial
D: Paul McKenney convinced me that there's no *guarantee* that all archs
D: will refuse the speculate the atomic ops above the state test, eg:
D:
D: CPU0 (stopref_set_state) CPU1 (stopref)
D:
D: atomic_set(&ack, 0); if (state == XXX)
D: wmb(); atomic_inc(&ack);
D: state = XXX;
D:
D: Certainly Alpha needs a rmb() inside stopref to guarantee it sees
D: the same ordering.
diff -urpN --exclude TAGS -X /home/rusty/devel/kernel/kernel-patches/current-dontdiff --minimal linux-2.5.69-bk6/kernel/module.c working-2.5.69-bk6-stopref/kernel/module.c
--- linux-2.5.69-bk6/kernel/module.c 2003-05-05 12:37:13.000000000 +1000
+++ working-2.5.69-bk6-stopref/kernel/module.c 2003-05-12 14:26:58.000000000 +1000
@@ -311,6 +311,7 @@ static int stopref(void *cpu)
set_cpus_allowed(current, 1UL << (unsigned long)cpu);
/* Ack: we are alive */
+ mb(); /* Theoretically the ack = 0 might not be on this CPU yet. */
atomic_inc(&stopref_thread_ack);
/* Simple state machine */
@@ -319,11 +320,13 @@ static int stopref(void *cpu)
local_irq_disable();
irqs_disabled = 1;
/* Ack: irqs disabled. */
+ mb(); /* Must read state first. */
atomic_inc(&stopref_thread_ack);
} else if (stopref_state == STOPREF_PREPARE && !prepared) {
/* Everyone is in place, hold CPU. */
preempt_disable();
prepared = 1;
+ mb(); /* Must read state first. */
atomic_inc(&stopref_thread_ack);
}
if (irqs_disabled || prepared)
@@ -333,6 +336,7 @@ static int stopref(void *cpu)
}
/* Ack: we are exiting. */
+ mb(); /* Must read state first. */
atomic_inc(&stopref_thread_ack);
if (irqs_disabled)
--
Anyone who quotes me in their sig is an idiot. -- Rusty Russell.
reply other threads:[~2003-10-09 8:51 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20031009085114.231B52C23B@lists.samba.org \
--to=rusty@rustcorp.com.au \
--cc=akpm@zip.com.au \
--cc=linux-kernel@vger.kernel.org \
--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
Powered by JetHome