[llvm] [RISCV] Widen i1 AnyOf reductions (PR #134898)

[Luke Lau via llvm-commits]

================
@@ -103,6 +105,76 @@ bool RISCVCodeGenPrepare::visitAnd(BinaryOperator &BO) {
   return true;
 }
 
+// With EVL tail folding, an AnyOf reduction will generate an i1 vp.merge like
+// follows:
+//
+// loop:
+//   %phi = phi <vscale x 4 x i1> [ zeroinitializer, %entry ], [ %rec, %loop ]
+//   %cmp = icmp ...
+//   %rec = call <vscale x 4 x i1> @llvm.vp.merge(%cmp, i1 true, %phi, %evl)
+//   ...
+// middle:
+//   %res = call i1 @llvm.vector.reduce.or(<vscale x 4 x i1> %rec)
+//
+// However RVV doesn't have any tail undisturbed mask instructions and so we
+// need a convoluted sequence of mask instructions to lower the i1 vp.merge: see
+// llvm/test/CodeGen/RISCV/rvv/vpmerge-sdnode.ll.
+//
+// To avoid that this widens the i1 vp.merge to an i8 vp.merge, which will
+// usually be folded into a masked vor.vv.
+//
+// loop:
+//   %phi = phi <vscale x 4 x i8> [ zeroinitializer, %entry ], [ %rec, %loop ]
+//   %cmp = icmp ...
+//   %rec = call <vscale x 4 x i8> @llvm.vp.merge(%cmp, i8 true, %phi, %evl)
+//   %trunc = trunc <vscale x 4 x i8> %rec to <vscale x 4 x i1>
+//   ...
+// middle:
+//   %res = call i1 @llvm.vector.reduce.or(<vscale x 4 x i1> %rec)
+//
+// The trunc will normally be sunk outside of the loop, but even if there are
+// users inside the loop it is still profitable.
+bool RISCVCodeGenPrepare::widenVPMerge(IntrinsicInst &II) {
+  if (!II.getType()->getScalarType()->isIntegerTy(1))
----------------
lukel97 wrote:

I think this should always be profitable in any case, even if it doesn't come from say the loop vectorizer. Looks like I need to update the comment about it being folded into the vor.vv though

https://github.com/llvm/llvm-project/pull/134898