Hi Segher,
on 2022/9/22 22:05, Segher Boessenkool wrote:
> Hi!
>
> On Thu, Sep 22, 2022 at 10:28:23AM +0800, Kewen.Lin wrote:
>> on 2022/9/22 05:56, Segher Boessenkool wrote:
>>> On Fri, Jun 24, 2022 at 10:02:19AM +0800, HAO CHEN GUI wrote:
>>> In the other direction I am worried that the unspecs will degrade
>>> performance (relative to smin/smax) when -ffast-math *is* active (and
>>> this new builtin code and pattern doesn't blow up).
>>
>> For fmin/fmax it would be fine, since they are transformed to {MAX,MIN}
>> EXPR in middle end, and yes, it can degrade for the bifs, although IMHO
>> the previous expansion to smin/smax contradicts with the bif names (users
>> expect to map them to xs{min,max}dp than others).
>
> But builtins *never* say to generate any particular instruction. They
> say to generate code that implements certain functionality. For many
> builtins this does of course boil down to specific instructions, but
> even then it could be optimised away completely or replace with
> something more specific if things can be folded or such.
ah, your explanation refreshed my mind, thanks! Previously I thought the
bifs with specific mnemonic as part of their names should be used to generate
specific instructions, it's to save users' efforts using inline-asm, if
we want them to represent the generic functionality (not bind with specific),
we can use some generic names instead. As your explanation, binding at
fast-math isn't needed, then I think Haochen's patch v7 with gimple folding
can avoid the concern on degradation at fast-math (still smax/smin), nice. :)
BR,
Kewen
