RE: [PATCH 1/4] x86: rename vec_encoding and vex_encoding_*

["Cui, Lili" <lili.cui@intel.com>]

> On 18.02.2024 06:59, Cui, Lili wrote:
> >> Even with just VEX these weren't limited to vector insns. With APX
> >> the set of non-vector ones covered has greatly increased. Drop the
> >> vec_ prefix. Also drop the vex_ ones off of the enumerators, as they weren't
> appropriate anyway:
> >> Should have been vec_ then, too.
> >>
> >> --- a/gas/config/tc-i386.c
> >> +++ b/gas/config/tc-i386.c
> >> @@ -445,16 +445,16 @@ struct _i386_insn
> >>      /* Disable instruction size optimization.  */
> >>      bool no_optimize;
> >>
> >> -    /* How to encode vector instructions.  */
> >> +    /* How to encode instructions.  */
> >>      enum
> >>        {
> >> -	vex_encoding_default = 0,
> >> -	vex_encoding_vex,
> >> -	vex_encoding_vex3,
> >> -	vex_encoding_evex,
> >> -	vex_encoding_evex512,
> >> -	vex_encoding_error
> >> -      } vec_encoding;
> >> +	encoding_default = 0,
> >> +	encoding_vex,
> >> +	encoding_vex3,
> >> +	encoding_evex,
> >> +	encoding_evex512,
> >
> > "encoding_evex512" seems redundant, duplicating with encoding_evex, do
> you know the reason for keeping it?
> 
> Well, did you look at how it's used, or at the commit that introduced it?
> 
> > Could you add some comments for " encoding_evex512", thanks.
> 
> If any comments are to be added here, then for every one of the enumerators.
> 

Well, you used evex512 to mark out "sae" and "zmm" for optimization. Maybe it's because you're so familiar with this area, but it's special to me. Of course, you are free not to annotate it.

Lili.