Commit 0cd98ee
committed
AMDGPU: Test more types with minimumnum and maximumnum
We had custom lowering for the wider vectors of f16, but missing
test coverage for them. Also add more vector tests for bf16, and
split the bf16 cases into separate files so we can add globalisel
run lines.1 parent fd8b84e commit 0cd98ee
File tree
4 files changed
+45631
-990
lines changed- llvm/test/CodeGen/AMDGPU
4 files changed
+45631
-990
lines changed
0 commit comments