Re: [PATCH 2/5] x86, fpu: don't drop_fpu() in __restore_xstate_sig() if use_eager_fpu()

[Oleg Nesterov]

On 08/25, Linus Torvalds wrote:
>
> On Mon, Aug 25, 2014 at 10:09 AM, Oleg Nesterov <oleg@xxxxxxxxxx> wrote:
> >
> > I'd like to change it only because this code is the main source of the
> > nasty special case, used_math() and/or __thread_has_fpu(current) can be
> > false even if use_eager_fpu().
>
> Well, if you think it is correct (apart from missing preemption), then
> I disagree *violently* with your "clean it up by restoring things
> twice" model.

OK.

> The signal handling overhead of floating point restore is not small,
> and it's not theoretical.

Again, this only if 32bit && use_eager_fpu(), and use_eager_fpu() adds
this overhead to every context switch.

But you convinced me anyway. If this hack doesn't look "obviously safe
and correct", lets forget it.

I'll try to play with copy_from_user_in_atomic(), if nothing else just
to complete the discussion and see how the code can look in this case.

Thanks!

Oleg.

--
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/