Dmitry Babokin
Dmitry Babokin
Ok, we agree that the test is correct before and after the change in term of values. I assume that the problem is in unaligned load in line 10. But...
I'm working on the templates support right now and will create a new PR shortly. This one is abandoned, I'll close it when I create a new one.
I think we need a test to specific code transformations, which happened / didn't happen.
I would say this is not a compiler issue. This is how FP precision rules are defined. `a / const1` = > `a * const2` (where `const2 = 1 /...
From compiler engineering prospective the request sounds like "use vector instructions for short vector initialization". I.e. The following code should imply vector semantic "by definition" (though it still has multiple...
In #2393 support for function templates was added. It's not complete and it's in "technical preview" state, meaning that we'd like to have some time for user feedback and evaluation...
This should be resolved now. I'm closing the issue.
The problem here is that unrolling runs too late. Specifically, there's no SROA pass to clean up after the unrolling. I've tried two approaches. 1. Run SROA after the unrolling...
PR https://github.com/ispc/ispc/pull/2393 supersedes this PR and is partially based on this one. I'm closing this one.
I'll be looking at it over the weekend, but it wouldn't hurt to clean up the commit history - there are commits that just fix build or formatting. But the...