Update clang builtins to the new GVL intrinsics
This commit updates all the existing clang builtins to match the expectations of the LLVM IR as of llvm!8 This change also adds a generated test and improved the test generator. Another important detail is the addition of a load/store intrinsic for 2xf64, otherwise we can't test the widening/narrowing conversions. See #12
Loading
Please sign in to comment