Re: [Xen-devel] [PATCH 2/3] x86/xsaves: fix overwriting between non-lazy/lazy xsave[sc]

[Shuai Ruan <shuai.ruan@xxxxxxxxxxxxxxx>]

On Mon, Feb 22, 2016 at 10:08:52AM -0700, Jan Beulich wrote:
> >>> On 22.02.16 at 06:35, <shuai.ruan@xxxxxxxxxxxxxxx> wrote:
> 
> First of all I wonder on what basis you collect your Cc lists on
> patches.
> 
I send the bugs-fix patch as whole. I just get the Cc lists using the
script based on the whole patchset. May be I will send the patch
seperately.
> > --- a/xen/arch/x86/i387.c
> > +++ b/xen/arch/x86/i387.c
> > @@ -118,7 +118,7 @@ static inline uint64_t vcpu_xsave_mask(const struct 
> > vcpu *v)
> >      if ( v->fpu_dirtied )
> >          return v->arch.nonlazy_xstate_used ? XSTATE_ALL : XSTATE_LAZY;
> >  
> > -    return v->arch.nonlazy_xstate_used ? XSTATE_NONLAZY : 0;
> > +    return ( cpu_has_xsaves || cpu_has_xsavec ) ? XSTATE_ALL : 
> > XSTATE_NONLAZY;
> >  }
> 
> The description lacks any mention of the performance impact,
> and what investigation was done to find ways to perhaps
> overcome this. For example, regardless of cpu_has_xsaves,
> do we really always need to _use_ XSAVES?
> 
Currently no supervisor xstates is enabled in xen or even in
guest os. Using xsaves is a little ahead (xsavec may used).  
xsavec may also cause overwriting problem like xsaves.
I will add performance impact in the description. 
I am still thinking of a better way to overcome the overhead xsave
(But have not get a better solution yet).

Thanks
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@xxxxxxxxxxxxx
> http://lists.xen.org/xen-devel
> 

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel