Le 10.09.2006 13:57, Ingo Molnar a écrit :
> * Andi Kleen <ak@suse.de> wrote:
> 
>>> This kernel won't boot here: it starts a GPFs loop on
>>> early boot. I attached a screenshot of the first GPF
>>> (pause_on_oops=120 helped).
>> It's lockdep's fault. This patch should fix it:
> 
> Well, it's also x86_64's fault: why does it call into a generic C 
> function (x86_64_start_kernel()) without having a full CPU state up and 
> running? i686 doesnt do it, never did.
> 
> We had frequent breakages due to this property of the x86_64 arch code 
> (many more than this single incident with lockdep), tracing and all 
> sorts of other instrumentation (including earlier versions of lockdep) 
> was hit by it again and again.
> 
> Basically, non-atomic setup of basic architecture state _is_ going to be 
> a nightmare, lockdep or not, especially if it uses common infrastructure 
> like 'current', spin_lock() or even something as simple as C functions. 
> (for example the stack-footprint tracer was once hit by this weakness of 
> the x86_64 code)
> 
>> Hackish patch to fix lockdep with PDA current
> 
> hm, this is ugly beyond words. Do you have a config i could try which 
> exhibits this problem? I'm sure there is a better solution.
> 
> 	Ingo

Andi's hackish patch helped, it runs fine now. I'll try any better patch too.

Ingo, I attached my .config to my first post. mmm... It did not reach LKML, so 
here is it again.

Thanks for your help.
-- 
laurent