[llvm] [X86] SimplifyDemandedVectorEltsForTargetNode - reduce the size of VPERMV v16f32/v16i32 nodes if the upper elements are not demanded (PR #134890)

[Phoebe Wang via llvm-commits]

================
@@ -43810,7 +43810,9 @@ bool X86TargetLowering::SimplifyDemandedVectorEltsForTargetNode(
     case X86ISD::VPERMV: {
       SmallVector<int, 16> Mask;
       SmallVector<SDValue, 2> Ops;
-      if ((VT.is256BitVector() || Subtarget.hasVLX()) &&
+      // We can always split v16i32/v16f32 AVX512 to v8i32/v8f32 AVX2 variants.
----------------
phoebewang wrote:

Can we do it for v8i64/v8f64?

https://github.com/llvm/llvm-project/pull/134890