Re: [PATCH 4/6] kvm, emulator: Add initial three-byte insns support

[Gleb Natapov <gleb@redhat.com>]

On Tue, Oct 29, 2013 at 11:04:57AM +0100, Borislav Petkov wrote:
> On Tue, Oct 29, 2013 at 11:50:43AM +0200, Gleb Natapov wrote:
> > There are two three opcode tables, so third_opcode_byte is ambiguous.
> 
> Actually there's also 0F_3A and there are also other prefixes besides f0
> and f1. Oh, and those tables are not completely full so I can imagine
> more stuff coming in later....
> 
> I know what you're thinking by now, btw :-)
> 
:)

> > What about pfx_0f_38_f0 and pfx_0f_38_f1?
> 
> Yeah, those make it much more explicit.
> 
> I wanted to keep the "three_byte" in the name in there somewhere,
> though, so that it is clear that we're dealing with three byte opcodes
> instead of requiring the onlooking innocent person to know the opcodes.
> 
> How about:
> 
> three_byte_0f_38_f0
> three_byte_0f_38_f1
> three_byte_0f_3a_50
> ...
> 
> Last one is an example only.
> 
Looks OK to me.

> Btw, we might want to reconsider that whole tabular representation when
> more stuff needs to be added...
> 
Of course. When tables will start to show their limitation we can always
change to something else.

--
			Gleb.