lkml.org 
[lkml]   [2005]   [Nov]   [14]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
/
DateSun, 13 Nov 2005 23:44:51 -0800
FromAndrew Morton <>
SubjectRe: start_kernel / local_irq_enable() can be very slow
Pozsar Balazs <pozsy@uhulinux.hu> wrote:
>
> I've noticed that the local_irq_enable() call in 
>  init/main.c::start_kernel() takes 0.3 .. 3.0 seconds on every boot.
>  (Measured using printk's around it.)
> 
>  I am wondering what happens during this call (which is effectively one 
>  "sti"), why does it take so much time sometimes, and why does it vary so 
>  much, why isn't it (more) deterministic.

Presumably as soon as it does the sti, the CPU takes one or more interrupts
and runs off and does something.

I wonder what?

You could do something like:

int trace_irqs;

	trace_irqs = 1;
	local_irq_enble();
	trace_irqs = 0;

then, over in handle_IRQ_event():

	if (trace_irqs)
		print_symbol("calling %s\n", (unsigned long)action->handler);

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-11-14 08:50    [from the cache]
©2003-2008