Re: [PATCH v2 3/3] i387: support lazy restore of FPU state

["H. Peter Anvin" <hpa@zytor.com>]

On 02/20/2012 06:18 PM, Linus Torvalds wrote:
> 
>> Although I do wonder if we should just make kernel_fpu_begin() be a
>> real function instead of inlining it. I'm not sure it makes sense to
>> inline that thing, and it might be better to export that one instead.
> 
> I do think that would be better in the long run, but for now here's an
> updated "trivial" patch to fix it.
> 

There is actually another very good reason for out-of-lining
kernel_fpu_begin/_end: it will act as a compiler barrier i we ever do
compiler-generated SSE or AVX code, which we may very well want to do.

In fact, there may even be good reason for something like:

kernel_fpu_use(void (*func)(void *), void *arg);

... where the FPU-using stuff is actively embedded in the call.

	-hpa


-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.