[PATCH] D19426: [AArch64] Use the reciprocal estimation machinery

[James Molloy via llvm-commits]

Hi Evandro,

> It seems to me that your objection is not so much against this patch as against the machinery in the DAGCombiner.
> 
> 

Perhaps. The important thing is how we decide whether to perform this optimisation or not. A hook function, as we normally use, has the ability to be extended to add more logic. A subtarget feature does not. I wouldn’t like to see optimization decisions being switched on or off by subtarget features.

Cheers,

James

> On 27 Apr 2016, at 17:26, Evandro Menezes <e.menezes at samsung.com> wrote:
> 
> evandro added a comment.
> 
> James,
> 
> It seems to me that your objection is not so much against this patch as against the machinery in the DAGCombiner.
> 
> I understand your points, even adding that the series takes the pressure from the unit(s) that perform division and square root and puts it unto the unit(s) that perform multiplication.
> 
> I do intend to investigate this issue further, but I think that it's incremental, if not tangential, to what this patch proposes.
> 
> Finally, I'm interested in understanding better what you think would be more appropriate than using features.
> 
> Thank you.
> 
> 
> Repository:
>  rL LLVM
> 
> http://reviews.llvm.org/D19426
> 
> 
>