Re: [Xen-devel] [PATCH 2/2] x86/emul: Pass shadow register state to the vmfunc() hook

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

On 22/12/16 08:27, Jan Beulich wrote:
> > > > On 21.12.16 at 17:32, <andrew.cooper3@xxxxxxxxxx> wrote:
> > vmfunc can in principle modify register state, so should operate on the shadow
> > register state rather than the starting state of emulation.
> >
> > Signed-off-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
> While in principle this is fine, I'd rather see the register state
> constified for now, to demonstrate it is not being modified. I'll
> submit my two remaining follow-up patches in a minute, and
> we can then discuss which of the two to take.
The question here is how likely it is that new functionality for VMFUNC 
will be defined, which starts mutating the values.
I am not aware of anything new, so lets go with the const version for 
now (as it is one fewer parameters).  If this changes in the future, we 
can easily switch back to passing the shadow register block.
~Andrew

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