Hi Ana,

Thanks for taking a look at the patch.

> Can you please explain again why a table is used for the scalar intrinsics 
> and not the other intrinisics? The other intrinsics also have long case 
> statements.

Basically because I had to start somewhere. The scalar intrinsics are
the natural place because the switch statement is already phrased as
"set a few parameters & break", we just have to delete it completely.

I definitely agree the idea is applicable to the SIMD case as well,
and I've actually done quite a bit of work in that direction; but I
wanted to make sure everyone else thought the approach was sound
before pressing on. Incremental patches and all that.

Cheers.

Tim.
_______________________________________________
cfe-commits mailing list
[email protected]
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

Reply via email to