[llvm] [X86][SSE] Improve ISD::FP_TO_*INT_SAT vector lowering (PR #199416)

[Phoebe Wang via llvm-commits]

================
@@ -1945,6 +1954,11 @@ X86TargetLowering::X86TargetLowering(const X86TargetMachine &TM,
       setOperationAction(ISD::STRICT_FP_TO_UINT, VT, Custom);
     }
 
+    if (!Subtarget.hasAVX10_2()) {
+      setOperationAction(ISD::FP_TO_SINT_SAT, MVT::v16i32, Custom);
+      setOperationAction(ISD::FP_TO_UINT_SAT, MVT::v16i32, Custom);
+    }
----------------
phoebewang wrote:

ditto, hasAVX512 block.

https://github.com/llvm/llvm-project/pull/199416