[llvm] [InstCombine][RISCV] Convert VPIntrinsics with splat operands to splats (PR #65706)

[Luke Lau via llvm-commits]

================
@@ -0,0 +1,1240 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -S -mtriple=riscv64 -mattr=+v,+m %s -passes=vector-combine | FileCheck %s --check-prefixes=ALL,VEC-COMBINE-64
+; RUN: opt -S -mtriple=riscv64 -mattr=+v,+m %s | FileCheck %s --check-prefixes=ALL,NO-VEC-COMBINE-64
+; RUN: opt -S -mtriple=riscv32 -mattr=+v,+m %s -passes=vector-combine | FileCheck %s --check-prefixes=ALL,VEC-COMBINE-32
+; RUN: opt -S -mtriple=riscv32 -mattr=+v,+m %s | FileCheck %s --check-prefixes=ALL,NO-VEC-COMBINE-32
----------------
lukel97 wrote:

I think the RV32 and RV64 output is the same for NO-VEC-COMBINE and VEC-COMBINE, can we merge their check prefixes?

https://github.com/llvm/llvm-project/pull/65706