Messages in this thread | | | Date | Tue, 24 Jul 2012 22:55:07 +0200 | From | Sasha Levin <> | Subject | Re: rcu: INFO: rcu_preempt detected stalls on CPUs/tasks on v3.6 |
| |
On 07/24/2012 07:40 PM, Paul E. McKenney wrote: > The interrupt flag is zero, so interrupts are disabled. So my question > to you is "Why did do_pageattr_test() or one of the functions it called > disable interrupts for more than one hundred thousand jiffies?" > > I can't see where it is disabling interrupts, but the EFLAGS register > above does indicate that they are disabled. Or is this something useless > that is just telling us that receiving the NMI disabled interrupts?
Doesn't the stack of CPU1 show that it's currently handling an interrupt:
[ 215.026612] Call Trace: [ 215.026612] <IRQ> [ 215.026612] [<ffffffff81095d07>] native_apic_msr_eoi_write+0x27/0x30 [ 215.026612] [<ffffffff810901c4>] smp_apic_timer_interrupt+0x34/0xa0 [ 215.026612] [<ffffffff8369c4af>] apic_timer_interrupt+0x6f/0x80 [ 215.026612] <EOI>
Or am I reading that wrong?
In that case, it had nothing to do with pageattr which just happened to be on the stack at that time.
| |