Re: [patch] x64, fpu: fix possible FPU leakage in error conditions

[Suresh Siddha <suresh.b.siddha@intel.com>]

On Thu, Jul 24, 2008 at 11:59:03AM -0700, Linus Torvalds wrote:
> 
> On Thu, 24 Jul 2008, Suresh Siddha wrote:
> >
> > If we don't do init_fpu() + restore from the sane init state, process has
> > to be killed, in the paranoid failing scenario of math_state_restore()
> 
> Umm. I'm still not seeing why the right answer is not just to do "stts +
> math_used = 0".
> 
> And the specific case of math_state_restore(), that will also fix the
> problem - next time around.

Ok.

> As far as I can tell, your patch causes serious problems in case
> init_fpu() fails. Which it can do, afaik.

Only the first init_fpu() can fail (memory allocation failure). But here,
it is def not the first time.

Anyhow, your suggestion is simple and clean. Will post the patch shortly.

I have to do clear_fpu() though(stts + math_used = 0 may not be enough).

thanks,
suresh