[PATCH] D32344: InstCombine/AMDGPU: Fix constant folding of llvm.amdgcn.{icmp, fcmp}

[Nicolai Hähnle via Phabricator via llvm-commits]

nhaehnle added inline comments.


================
Comment at: lib/Transforms/InstCombine/InstCombineCalls.cpp:3435-3437
+        if (CCmp->isNullValue())
+          return replaceInstUsesWith(
+              *II, ConstantExpr::getSExt(CCmp, II->getType()));
----------------
arsenm wrote:
> Multiline, so braces
Fixing this before I commit.


================
Comment at: test/Transforms/InstCombine/amdgcn-intrinsics.ll:1538
   ret i64 %result
 }
----------------
arsenm wrote:
> arsenm wrote:
> > Check that attributes 4 contains convergent?
> This is really missing nounwind?
Yes. It's an attribute of the call-site, not of the function. read_register itself is nounwind readonly, but doesn't have the convergent attribute.


https://reviews.llvm.org/D32344