linux-kernel - Re: [BUG] local_softirq

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] [day] [month] [year] [list]

Message-ID: <20070524074534.GA21138@elte.hu>
Date:	Thu, 24 May 2007 09:45:34 +0200
From:	Ingo Molnar <mingo@...e.hu>
To:	Chuck Ebbert <cebbert@...hat.com>
Cc:	Michal Piotrowski <michal.k.k.piotrowski@...il.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	Anant Nitya <kernel@...chanda.info>,
	linux-kernel@...r.kernel.org, David Miller <davem@...emloft.net>,
	Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [BUG] local_softirq_pending storm


* Chuck Ebbert <cebbert@...hat.com> wrote:

> >  	if (need_resched() && system_state == SYSTEM_RUNNING) {
> > -		raw_local_irq_disable();
> > -		_local_bh_enable();
> > -		raw_local_irq_enable();
> > +		local_bh_enable();
> >  		__cond_resched();
> >  		local_bh_disable();
> >  		return 1;
> 
> We may have a problem with that:
> 
>  BUG: warning at kernel/softirq.c:138/local_bh_enable() (Not tainted)
>   [<c042b2ef>] local_bh_enable+0x45/0x92
>   [<c06036b7>] cond_resched_softirq+0x2c/0x42
>   [<c059d5d0>] release_sock+0x54/0xa3
>   [<c05c9428>] tcp_sendmsg+0x91b/0xa0c
>   [<c05e1bb9>] inet_sendmsg+0x3b/0x45
>   [<c059af34>] sock_aio_write+0xf9/0x105
>   [<c0476035>] do_sync_write+0xc7/0x10a
>   [<c0437265>] autoremove_wake_function+0x0/0x35
>   [<c047688e>] vfs_write+0xbc/0x154
>   [<c0476e8c>] sys_write+0x41/0x67
>   [<c0404f70>] syscall_call+0x7/0xb

hm, this place really shouldnt call cond_resched_softirq() with hardirqs 
disabled.

perhaps a buggy ->sk_backlog_rcv() handler disabled interrupts without 
restoring them?

could you enable CONFIG_PROVE_LOCKING and apply the patch below - which 
location is printed as having last disabled hardirqs?

	Ingo

--------------------->
Subject: [patch] softirqs: print out irq-trace events
From: Ingo Molnar <mingo@...e.hu>

some code is fiddling with softirqs but hardirqs are disabled, so try to 
figure out who disabled hardirqs.

Signed-off-by: Ingo Molnar <mingo@...e.hu>
---
 kernel/softirq.c |   10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

Index: linux/kernel/softirq.c
===================================================================
--- linux.orig/kernel/softirq.c
+++ linux/kernel/softirq.c
@@ -135,7 +135,15 @@ void local_bh_enable(void)
 
 	WARN_ON_ONCE(in_irq());
 #endif
-	WARN_ON_ONCE(irqs_disabled());
+	if (irqs_disabled()) {
+		static int once = 1;
+
+		if (once) {
+			once = 0;
+			print_irqtrace_events(current);
+			WARN_ON(1);
+		}
+	}
 
 #ifdef CONFIG_TRACE_IRQFLAGS
 	local_irq_save(flags);
-
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/