>>> On 09.12.16 at 16:19, <jbeul...@suse.com> wrote:
>>>> On 09.12.16 at 15:56, <andrew.coop...@citrix.com> wrote:
>> It is a layering violation to remove the host_has_* part of the check
>> from the emulator.
> 
> This is from a very, very abstract perspective. The fact that
> host_and_vcpu_*() and vcpu_*() are the same in the harness
> should allow for them to be used interchangeably when only the
> harness is affected.

I should add that in a subsequent patch adding emulation of
LDDQU (an SSE3 insn) this is going to be done by slightly
extending the existing MOVDQ{U,A} etc block, i.e. using an
SSE2 insn. If I was to follow your directions, I'd have to add
separate host_has(sse2) and vcpu_has(sse3) checks. That's
just going to become insane. I have just the vcpu check in
there, and I really hope you won't force me to uglify that.

Jan


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

Reply via email to