Re: [PATCH v2 1/2] x86/fpu: Combine fpu_ctxt and xsave_area in arch_vcpu

["Alejandro Vallejo" <alejandro.vallejo@cloud.com>]

On Mon Aug 12, 2024 at 4:16 PM BST, Jan Beulich wrote:
> On 08.08.2024 15:41, Alejandro Vallejo wrote:
> > --- a/xen/arch/x86/domctl.c
> > +++ b/xen/arch/x86/domctl.c
> > @@ -1344,7 +1344,10 @@ void arch_get_info_guest(struct vcpu *v, vcpu_guest_context_u c)
> >  #define c(fld) (c.nat->fld)
> >  #endif
> >  
> > -    memcpy(&c.nat->fpu_ctxt, v->arch.fpu_ctxt, sizeof(c.nat->fpu_ctxt));
> > +    memcpy(&c.nat->fpu_ctxt, &v->arch.xsave_area->fpu_sse,
> > +           sizeof(c.nat->fpu_ctxt));
> > +    BUILD_BUG_ON(sizeof(c.nat->fpu_ctxt) != sizeof(fpusse_t));
>
> While it may seem unlikely that it would change going forward, I think
> that such build-time checks should make no implications at all. I.e.
> here the right side ought to be sizeof(v->arch.xsave_area->fpu_sse)
> even if that's longer.

Sounds sensible.

>
> Personally I also think that BUILD_BUG_ON(), just like BUG_ON(), would
> better live ahead of the construct they're for.
>
> Same again in at least one more place.
>
> Jan

Ack, sure.

Cheers,
Alejandro