Optimization comparison: 3.3, 3.4, mainline, tree-ssa

[Geert Bosch]

On Apr 19, 2004, at 12:52, Kaveh R. Ghazi wrote:
>> It seems that it would make sense to have a more generic inline
>> expansion of these well-known functions at the tree level.
>>    -Geert
>
> ...but using trees can be tricky.  It's not hard to model e.g. asin at
> the tree level using atan2(x, sqrt(1.0 - x * x)); however this is only
> a win if we know the entire expression transforms into inline assembly
> code.  So we don't want to do this transformation on platforms that
> don't also have inline intrinsics for atan2 and sqrt.
>
> Therefore, I think it makes sense to do this in the md file where we
> know whether we have the other functions expanded.

Yes, this definitely will be target-dependent. The advantage of
having the compiler itself do the transformation is that we can
decide wether or not the expansion will be a good idea, based
on optimization flags, estimated execution count, cost etc.

Maybe a way to specify possible tree transformations in the .md
file would be the best approach.

   -Geert