[llvm] [X86] Try Folding icmp of v8i32 -> fcmp of v8f32 on AVX (PR #82290)

[via llvm-commits]

================
@@ -23299,6 +23299,110 @@ static SDValue LowerVSETCC(SDValue Op, const X86Subtarget &Subtarget,
     }
   }
 
+  // We get bad codegen for v8i32 compares on avx targets (without avx2) so if
+  // possible convert to a v8f32 compare.
+  if (VTOp0.getVectorElementType() == MVT::i32 && VTOp0 == MVT::v8i32 &&
+      Subtarget.hasAVX() && !Subtarget.hasAVX2()) {
+    std::optional<KnownBits> KnownOps[2];
+    // Check if an op is known to be in a certain range.
+    auto OpInRange = [&DAG, Op, &KnownOps](unsigned OpNo, bool CmpLT,
+                                           const APInt Bound) {
+      if (!KnownOps[OpNo].has_value())
+        KnownOps[OpNo] = DAG.computeKnownBits(Op.getOperand(OpNo));
----------------
goldsteinn wrote:

We can get away with only doing `Op1` if its a small constants (which is the most likely path).

https://github.com/llvm/llvm-project/pull/82290