Re: [PATCH 08/23] clocksource: drop duplicate register checking

[Daniel Walker]

On Wed, 2007-01-31 at 10:59 +0100, Ingo Molnar wrote:
> * Daniel Walker <dwalker@xxxxxxxxxx> wrote:
> 
> > This is something Thomas already dropped, [...]
> 
> (i think you forgot to Cc: Thomas here, nor is this something that 
> Thomas' change dropped.)

Yes your right .

> > [...] and I'm just sticking with that .. If you register your 
> > clocksource _twice_ your kernel will likely not work correctly (and 
> > might crash).
> 
> this is a quite bad change. John's original clocksource code protects 
> against double registry:
> 
>          if (is_registered_source(c)) {
>                  printk("register_clocksource: Cannot register %s. "
>                         "Already registered!", c->name);
>                  ret = -EBUSY;
> 
> and Thomas' change keeps that property, but doesnt printk. Your change 
> makes double registry possible, potentially crashing the kernel later 
> on! (And this isnt theoretical, double registry did happen in practice 
> when i debugged suspend problems on my SMP laptop.)

My original patch set makes the duplicate register checking better
(which I dropped) .. I'll be happy to reintroduce that part of it..

Daniel

-
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/