LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Yong Zhang <yong.zhang0@gmail.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Borislav Petkov <bp@alien8.de>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	x86@kernel.org, tglx <tglx@linutronix.de>,
	Ingo Molnar <mingo@elte.hu>,
	Andrew Morton <akpm@linux-foundation.org>
Subject: Re: lockdep: possible reason: unannotated irqs-off. (was: Re: Linux 2.6.38-rc4)
Date: Tue, 8 Feb 2011 20:11:09 +0800	[thread overview]
Message-ID: <20110208121108.GA4824@zhy> (raw)
In-Reply-To: <1297161712.13327.100.camel@laptop>

On Tue, Feb 08, 2011 at 11:41:52AM +0100, Peter Zijlstra wrote:
> Argh! Its an annotation nightmare that.. it didn't trigger for me when
> running that because I didn't have DEBUG_LOCKDEP=y.

Me too...

> 
> OK, let me try and come up with another way to annotate this
> del_timer_sync() muck,

Is my previous patch acceptable?

I inlined it as below(and updated based on linux-2.6.38-rc4)

---
From: Yong Zhang <yong.zhang0@gmail.com>
Subject: [PATCH 1/2] softirq: introduce loacal_bh_enable_force_wake()

If there is pending softirq, don't handle it in the caller's
context, invoke ksoftirqd directly instead.

del_timer_sync() will be the first caller.

Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Andrew Morton <akpm@linux-foundation.org>
---
 include/linux/bottom_half.h |    1 +
 kernel/softirq.c            |   21 +++++++++++++++------
 2 files changed, 16 insertions(+), 6 deletions(-)

diff --git a/include/linux/bottom_half.h b/include/linux/bottom_half.h
index 27b1bcf..665d697 100644
--- a/include/linux/bottom_half.h
+++ b/include/linux/bottom_half.h
@@ -5,5 +5,6 @@ extern void local_bh_disable(void);
 extern void _local_bh_enable(void);
 extern void local_bh_enable(void);
 extern void local_bh_enable_ip(unsigned long ip);
+extern void local_bh_enable_force_wake(void);
 
 #endif /* _LINUX_BH_H */
diff --git a/kernel/softirq.c b/kernel/softirq.c
index 68eb5ef..3c05dfa 100644
--- a/kernel/softirq.c
+++ b/kernel/softirq.c
@@ -154,9 +154,9 @@ void _local_bh_enable(void)
 
 EXPORT_SYMBOL(_local_bh_enable);
 
-static inline void _local_bh_enable_ip(unsigned long ip)
+static inline void _local_bh_enable_ip(unsigned long ip, bool force_wake)
 {
-	WARN_ON_ONCE(in_irq() || irqs_disabled());
+	WARN_ON_ONCE(in_irq() || (!force_wake && irqs_disabled()));
 #ifdef CONFIG_TRACE_IRQFLAGS
 	local_irq_disable();
 #endif
@@ -171,8 +171,12 @@ static inline void _local_bh_enable_ip(unsigned long ip)
  	 */
 	sub_preempt_count(SOFTIRQ_DISABLE_OFFSET - 1);
 
-	if (unlikely(!in_interrupt() && local_softirq_pending()))
-		do_softirq();
+	if (unlikely(!in_interrupt() && local_softirq_pending())) {
+		if (!force_wake)
+			do_softirq();
+		else
+			wakeup_softirqd();
+	}
 
 	dec_preempt_count();
 #ifdef CONFIG_TRACE_IRQFLAGS
@@ -183,16 +187,21 @@ static inline void _local_bh_enable_ip(unsigned long ip)
 
 void local_bh_enable(void)
 {
-	_local_bh_enable_ip((unsigned long)__builtin_return_address(0));
+	_local_bh_enable_ip((unsigned long)__builtin_return_address(0), false);
 }
 EXPORT_SYMBOL(local_bh_enable);
 
 void local_bh_enable_ip(unsigned long ip)
 {
-	_local_bh_enable_ip(ip);
+	_local_bh_enable_ip(ip, false);
 }
 EXPORT_SYMBOL(local_bh_enable_ip);
 
+void local_bh_enable_force_wake(void)
+{
+	_local_bh_enable_ip((unsigned long)__builtin_return_address(0), true);
+}
+EXPORT_SYMBOL(local_bh_enable_force_wake);
 /*
  * We restart softirq processing MAX_SOFTIRQ_RESTART times,
  * and we fall back to softirqd after that.
-- 
1.7.1

  reply	other threads:[~2011-02-08 12:11 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-02-08  0:23 Linux 2.6.38-rc4 Linus Torvalds
2011-02-08 10:17 ` lockdep: possible reason: unannotated irqs-off. (was: Re: Linux 2.6.38-rc4) Borislav Petkov
2011-02-08 10:41   ` Peter Zijlstra
2011-02-08 12:11     ` Yong Zhang [this message]
2011-02-08 12:14       ` [PATCH 2/2] timer: use local_bh_enable_force_wake() in del_timer_sync() Yong Zhang
2011-02-08 13:34       ` lockdep: possible reason: unannotated irqs-off. (was: Re: Linux 2.6.38-rc4) Yong Zhang
2011-02-08 13:48         ` Peter Zijlstra
2011-02-08 14:18           ` Peter Zijlstra
2011-02-08 15:15             ` Ingo Molnar
2011-02-08 15:51             ` [tip:core/urgent] Revert "lockdep, timer: Fix del_timer_sync() annotation" tip-bot for Peter Zijlstra
2011-02-09  1:46             ` lockdep: possible reason: unannotated irqs-off. (was: Re: Linux 2.6.38-rc4) Yong Zhang
2011-02-14 14:51             ` Yong Zhang
2011-02-14 18:53               ` Thomas Gleixner
2011-02-08 20:28 ` Heads up Linux 2.6.38-rc4 compile problems Eric W. Biederman
2011-02-08 20:44   ` Linus Torvalds
2011-02-09  9:01     ` Eric W. Biederman
2011-02-09 14:59       ` Alex Riesen
2011-02-09 16:02         ` Linus Torvalds
2011-02-13 17:39           ` Linus Torvalds
2011-02-14  2:04             ` Eric W. Biederman
2011-02-14  2:45               ` Linus Torvalds
2011-02-14  3:40                 ` Eric W. Biederman
2011-02-14  5:34                   ` Eric W. Biederman
2011-02-14 15:26                     ` Linus Torvalds
2011-02-14 15:37                     ` Eric W. Biederman
2011-02-14 16:37                       ` Linus Torvalds
2011-02-14 17:39                         ` Eric W. Biederman
2011-02-14 17:49                         ` Linus Torvalds
2011-02-14 18:08                           ` Linus Torvalds
2011-02-14 19:44                             ` Eric W. Biederman
2011-02-14 20:13                               ` Andrew Morton
2011-02-14 18:25                         ` Andi Kleen
2011-02-14 16:58                     ` Mike Snitzer
2011-02-15 14:07                       ` [Crash-utility] " Dave Anderson
2011-02-09 17:08 ` Linux 2.6.38-rc4 (test_nx: BUG) Randy Dunlap
2011-02-09 17:10   ` Arjan van de Ven
2011-02-17 19:33     ` Kees Cook
2011-02-09 17:24 ` Linux 2.6.38-rc4 (hysdn: BUG) Randy Dunlap
2011-02-09 19:44   ` Linus Torvalds
2011-02-09 21:25     ` Randy Dunlap
2011-02-09 21:57       ` David Miller
2011-02-09 22:00         ` Linus Torvalds
2011-02-09 17:26 ` Linux 2.6.38-rc4 (tty/ifx6x60: BUG) Randy Dunlap
2011-02-09 18:28   ` Alan Cox
2011-02-09 17:28 ` Linux 2.6.38-rc4 (target_core: rmmod GP fault) Randy Dunlap
2011-02-09 19:00   ` Linus Torvalds
2011-02-09 20:02     ` Nicholas A. Bellinger
2011-02-09 20:13       ` James Bottomley
2011-02-09 20:20         ` Nicholas A. Bellinger
2011-02-09 20:28           ` James Bottomley
2011-02-09 20:44             ` Nicholas A. Bellinger
2011-02-09 17:36 ` Linux 2.6.38-rc4 (other bugs) Randy Dunlap
2011-02-09 22:01   ` David Miller
2011-02-09 22:16     ` Randy Dunlap
2011-02-10  4:58     ` Linux 2.6.38-rc4 (other bugs: x25) Randy Dunlap
2011-02-10  5:48       ` David Miller
2011-02-10  6:29         ` Randy Dunlap
2011-02-10  6:35           ` David Miller
2011-02-10 19:34   ` Linux 2.6.38-rc4 (other bugs: ipmi Oops) Randy Dunlap
2011-02-10 20:03     ` Linus Torvalds
2011-02-10 20:08       ` Corey Minyard
2011-02-10 21:41       ` Randy Dunlap

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=20110208121108.GA4824@zhy \
    --to=yong.zhang0@gmail.com \
    --cc=akpm@linux-foundation.org \
    --cc=bp@alien8.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=torvalds@linux-foundation.org \
    --cc=x86@kernel.org \
    --subject='Re: lockdep: possible reason: unannotated irqs-off. (was: Re: Linux 2.6.38-rc4)' \
    /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

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).