[llvm] [RISCV] Fold vector shift of sext/zext to widening multiply (PR #121563)

[Philip Reames via llvm-commits]

================
@@ -1417,6 +1417,9 @@ def TuneVentanaVeyron : SubtargetFeature<"ventana-veyron", "RISCVProcFamily", "V
 def TuneVXRMPipelineFlush : SubtargetFeature<"vxrm-pipeline-flush", "HasVXRMPipelineFlush",
                                              "true", "VXRM writes causes pipeline flush">;
 
+def TuneCheapVWMul : SubtargetFeature<"cheap-vwmul", "HasCheapVWMul", "true",
----------------
preames wrote:

I'm not sure the tune flag here is warranted.  I would bias towards not having it unless we know some particular core is unprofitable.  

https://github.com/llvm/llvm-project/pull/121563