Re: [PATCH] x86: always supply .cpuid() handler to x86_emulate()

[Wei Liu <wei.liu2@citrix.com>]

On Fri, Nov 11, 2016 at 09:21:00AM -0700, Jan Beulich wrote:
> >>> On 11.11.16 at 14:54, <wei.liu2@citrix.com> wrote:
> > On Thu, Nov 10, 2016 at 05:30:25AM -0700, Jan Beulich wrote:
> >> With us incremementally adding proper CPUID checks to x86_emulate()
> >> (see commit de05bd965a ["x86emul: correct {,F}CMOV and F{,U}COMI{,P}
> >> emulation"]) it is no longer appropriate to invoke the function with
> >> that hook being NULL, as long as respective instructions may get used
> >> in that case.
> >> 
> >> Signed-off-by: Jan Beulich <jbeulich@suse.com>
> >> ---
> >> I've noticed the problem first with a not yet submitted patch checking
> >> cx8 (PV 32-bit kernels occasionally use cmpxchg8b for page table
> >> accesses), and I think we don't strictly need the change here for 4.8,
> >> but then again it may be better to handle this properly right away.
> > 
> > I'm fine with having this if we have some free cycles. If it is reviewed
> > today, feel free to commit it. Otherwise I will try to sneak it in after
> > it is reviewed.
> 
> In the hope that you don't mind, I've translated this to R-a-b.

That's absolutely fine.

> 
> Jan
> 

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