Re: [2.6.29-git12] BUG: scheduling while atomic: swapper/0/0x10000100

[Thomas Gleixner <tglx@linutronix.de>]

On Fri, 10 Apr 2009, Maciej Rutecki wrote:
> 2009/4/10 Thomas Gleixner <tglx@linutronix.de>:
> 
> >
> > Is this with my patch applied ?
> 
> Yes. I didn't observe any change.

Hmm. Ok, so the check in do_IRQ did not trigger. That means we need to
look for some other place.

Can you please apply the patch below and set CONFIG_FUNCTION_TRACER=y.
After boot please do as root:

# mount -t debugfs debugfs /sys/kernel/debug
# echo function >/sys/kernel/debug/tracing/current_tracer
# ech0 1 >/sys/kernel/debug/tracing/options/latency-format
# echo 1 >/sys/kernel/debug/tracing/tracing_enabled

Now wait until that scheduling while atomic bug triggers. This will
stop the tracer. Now

# cat /sys/kernel/debug/tracing/trace | bzip2 >/tmp/trace.bz2

Please upload trace.bz2 file to some place.

Thanks,

	tglx

----
diff --git a/kernel/sched.c b/kernel/sched.c
index 5724508..2e257bf 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -4912,6 +4912,7 @@ static noinline void __schedule_bug(struct task_struct *prev)
 {
 	struct pt_regs *regs = get_irq_regs();
 
+	tracing_off();
 	printk(KERN_ERR "BUG: scheduling while atomic: %s/%d/0x%08x\n",
 		prev->comm, prev->pid, preempt_count());