linux-kernel - Re: lockdep false positive? -- firewire-core transaction timer vs. scsi-core host lock

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 18 Aug 2010 17:08:37 +0800
From:	Yong Zhang <yong.zhang0@...il.com>
To:	Stefan Richter <stefanr@...6.in-berlin.de>
Cc:	Clemens Ladisch <clemens@...isch.de>,
	Peter Zijlstra <peterz@...radead.org>,
	Johannes Berg <johannes@...solutions.net>,
	Thomas Gleixner <tglx@...utronix.de>,
	linux1394-devel@...ts.sourceforge.net, linux-kernel@...r.kernel.org
Subject: Re: lockdep false positive? -- firewire-core transaction timer vs.
 scsi-core host lock

On Wed, Aug 18, 2010 at 4:09 PM, Stefan Richter
<stefanr@...6.in-berlin.de> wrote:
> Again, is the term "interrupt contexts" meant to include softIRQ contexts?

I can't think out a scene which prevent del_timer_sync() from using in
softirq. :(

Then I go back and find that comments comes before linux-2.6.12,
then in linux-2.6.12, del_timer_sync() is like this:

int del_timer_sync(struct timer_list *timer)
{
	tvec_base_t *base;
	int i, ret = 0;

	check_timer(timer);

del_again:
	ret += del_timer(timer);

	for_each_online_cpu(i) {
		base = &per_cpu(tvec_bases, i);
		if (base->running_timer == timer) {
			while (base->running_timer == timer) {
				cpu_relax();
				preempt_check_resched();
			}
			break;
		}
	}
	smp_rmb();
	if (timer_pending(timer))
		goto del_again;

	return ret;
}
EXPORT_SYMBOL(del_timer_sync);

Seems preempt_check_resched() is the proof of that comments, but
now del_timer_sync() is changed.

I don't know if there is any other restriction.

Thanks,
Yong

>
> (Nothing in firewire-core/-sbp2/-net etc. is called in hardware
> interrupts.  BTW, we actually could and maybe should change all the
> spinlocks in them from spin_lock_irq to spin_lock_bh.  Only
> firewire-ohci's local ohci->lock needs to be IRQ safe.  OTOH this could
> change if somebody comes up with a migration of the stack to threaded
> IRQ handling.)
> --
> Stefan Richter
> -=====-==-=- =--- =--=-
> http://arcgraph.de/sr/
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/