[llvm] [RISCV][VLOPT] Add support for checkUsers when UserMI is a Single-Width Integer Reduction (PR #120345)

[Craig Topper via llvm-commits]

================
@@ -963,46 +1031,58 @@ bool RISCVVLOptimizer::checkUsers(const MachineOperand *&CommonVL,
       break;
     }
 
-    const MCInstrDesc &Desc = UserMI.getDesc();
-    if (!RISCVII::hasVLOp(Desc.TSFlags) || !RISCVII::hasSEWOp(Desc.TSFlags)) {
-      LLVM_DEBUG(dbgs() << "    Abort due to lack of VL or SEW, assume that"
-                           " use VLMAX\n");
+    auto VLOp = getVLForUser(UserOp);
+    if (!VLOp) {
       CanReduceVL = false;
       break;
     }
 
-    unsigned VLOpNum = RISCVII::getVLOpNum(Desc);
-    const MachineOperand &VLOp = UserMI.getOperand(VLOpNum);
-
-    // Looking for an immediate or a register VL that isn't X0.
-    assert((!VLOp.isReg() || VLOp.getReg() != RISCV::X0) &&
-           "Did not expect X0 VL");
-
     // Use the largest VL among all the users. If we cannot determine this
     // statically, then we cannot optimize the VL.
-    if (!CommonVL || RISCV::isVLKnownLE(*CommonVL, VLOp)) {
-      CommonVL = &VLOp;
+    if (!CommonVL || RISCV::isVLKnownLE(*CommonVL, *VLOp)) {
+      CommonVL.emplace(*VLOp);
----------------
topperc wrote:

Why do we need to use `emplace` here? Does `CommonVl = *VLOp` not work?

https://github.com/llvm/llvm-project/pull/120345