[PATCH 4/4] x86: provide a 128-bit VBROADCASTSD pseudo

[Jan Beulich]

On 20.06.2023 18:07, H.J. Lu wrote:
> On Mon, Jun 19, 2023 at 12:20 AM Jan Beulich <jbeulich@suse.com> wrote:
>>
>> On 16.06.2023 18:59, H.J. Lu wrote:
>>> On Fri, Jun 16, 2023 at 12:32 AM Jan Beulich <jbeulich@suse.com> wrote:
>>>> VBROADCASTSD not supporting 128-bit destinations in any of their AVX,
>>>> AVX2, or AVX512F incarnations is presumably because of VMOVDDUP
>>>> precisely supporting this very operation. (It is therefore different
>>>> from e.g. VPBROADCASTQ, which has no exact equivalent.) Still its
>>>> absence has led to people using VPBROADCASTQ as substitution; this could
>>>> have been avoided if such a pseudo had been supported from the very
>>>> beginning.
>>>>
>>>> Note that the pseudos try to match what the real instructions would have
>>>> used as closely as possible, i.e. VexW0 instead of VexWIG for the AVX
>>>> and AVX2 forms as well as AVX2 in the first place for the register
>>>> source form.
>>>> ---
>>>> For being the first example of us supplying such, this is partly RFC. On
>>>> top of that a question is also whether to indeed have split AVX/AVX2
>>>> templates, when in principle one (allowing for both memory and register
>>>> source) could do.
>>>>
>>>
>>> I don't think assembler should invent such instructions.
>>
>> May I ask about the "why" behind this? If such a pseudo had been there
>> from the beginning, an admittedly minor mistake like that corrected by
>> gcc commit a4df0ce78d6f likely wouldn't have been made, because no
>> special casing of V2DFmode would have been necessary in the first place.
> 
> All x86 instructions should come from the x86 SDM.

Ehem. See "clr" for an example where syntax doesn't matter (IOW I wasn't
really right in saying this is the first example). There are also various
AT&T-invented mnemonics we support (and - wrongly - even in Intel syntax).
There are further insn forms (number and/or kind of operands) which aren't
backed by the SDM.

I'm afraid I can't take this single sentence as an answer to my question
of "Why?" Even less so with not addressing at all the reason I gave why I
think we should have had such a pseudo from the beginning.

Jan