fix Issue 23568: Set the result type of vector comparisons to an integer vector type

[ibuclaw]

The result of vector comparisons is a single or doubleword mask of all 1s (comparison true) or all 0s (comparison false). This SIMD mask type does not have an equivalent D type, however its closest equivalent would be an integer vector of the same unit size and length.

Rationale: Currently the result of vector float comparison and equality operations is either NaNs or 0s, which requires an explicit cast to a vector int in order to use it for checking the result post operation.

For example, GDC's ternary (? :) vector intrinsic blendvector can work with the following:

blendvector(a, b, a < b);

However if the condition is a vector float type...

blendvector(f, g, cast(__vector[uint[4])(f < g));

This removes the need for such boilerplate.

[dlang-bot]

Thanks for your pull request, @ibuclaw!

Bugzilla references

Auto-close	Bugzilla	Severity	Description
✓	23568	major	Error: cannot implicitly convert expression 'a < b' of type 'float4' to 'int4'

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub run digger -- build "master + dmd#14715"