Re: [PATCH 3/3] x86: ignore high register select bit(s) in 32- and 16-bit modes

["H.J. Lu" <hjl dot tools at gmail dot com>]

On Tue, Nov 14, 2017 at 5:42 AM, Jan Beulich <JBeulich@suse.com> wrote:
>>>> On 14.11.17 at 13:43, <hjl.tools@gmail.com> wrote:
>> On Mon, Nov 13, 2017 at 11:32 PM, Jan Beulich <JBeulich@suse.com> wrote:
>>> +[      ]*[a-f0-9]+:    62 f2 7d 0f 90 0c 00    vpgatherdd \(%eax,%xmm0(,1)?\),%xmm1\{%k7\}
>>> +[      ]*[a-f0-9]+:    62 d2 7d 0f 90 0c 00    vpgatherdd \(%eax,%xmm0(,1)?\),%xmm1\{%k7\}
>>> +[      ]*[a-f0-9]+:    62 f2 7d 07 90 0c 00    vpgatherdd \(%eax,%xmm0(,1)?\),%xmm1\{%k7\}
>
> I assume this is what ...
>
>> Where are there "(,1)?"?  Under what conditions will
>> assembler/disassembler generate
>> different outputs?
>
> ... you refer to here? I've added those because for the purpose of
> the test it is not relevant whether the disassembler produces them,
> just like I like to do in other cases where parts of the expected output
> are optional to be present. I consider it bad practice not to do so,
> and I've been bitten more than once by others not being careful to
> permit such optional pieces (or often the other way around - marking
> them as required when really they are optional, and once one goes
> and cleans up after others, one then also has to touch all these badly
> written test cases). Granted I could have gone _farther_ with this,
> by e.g. also permitting blanks after the commas (which the
> disassembler really should emit for readability).
>

Please drop those optional characters if disassembler doesn't generate
them.  When we make changes to disassembler,  we need to verify/know
the impact of the changes.

-- 
H.J.