Re: 2.6.8-rc3-mm2

[William Lee Irwin III]

* William Lee Irwin III <wli@xxxxxxxxxxxxxx> wrote:
>> Actually, what I just narrowed it down to was *only* the printk change
>> fixes it.

On Tue, Aug 10, 2004 at 11:00:51AM +0200, Ingo Molnar wrote:
> when i've seen such things on x86 it was usually some race with
> interrupts on the other CPU. Where do all the ia64 interrupts go to
> during bootup?
> the other possibility is messed up completion logic - some stuff is
> still on this CPU's kernel stack and the printk delays its
> corruption/destruction.

printk() seems to only have a few possible effects. I just tried mdelay()
for the delay effect and not much appears to have happened. I'll probably
try fiddling with schedule(), yield(), and local_irq_enable() and so on
next. Your advice is very much like what I have in mind for possibilities,
except I consider the messed up completion logic ruled out since backing
out the completion removal part of the printk() "fix"  didn't break it.


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