Ensure that MultiplyAddEstimate exits for cases where GT_MUL is unsupported (#109842)

Author: tannergooding

src/coreclr/jit/hwintrinsicxarch.cpp

@@ -3233,6 +3233,38 @@ GenTree* Compiler::impSpecialIntrinsic(NamedIntrinsic        intrinsic,
                 break;
             }
 
+            if ((simdSize == 32) && !varTypeIsFloating(simdBaseType) &&
+                !compOpportunisticallyDependsOn(InstructionSet_AVX2))
+            {
+                // We can't deal with TYP_SIMD32 for integral types if the compiler doesn't support AVX2
+                break;
+            }
+
+            assert(simdSize != 64 || IsBaselineVector512IsaSupportedDebugOnly());
+
+            if (varTypeIsLong(simdBaseType))
+            {
+                if (TARGET_POINTER_SIZE == 4)
+                {
+                    // TODO-XARCH-CQ: 32bit support
+                    break;
+                }
+
+                if ((simdSize == 32) && compOpportunisticallyDependsOn(InstructionSet_AVX2))
+                {
+                    // Emulate NI_AVX512DQ_VL_MultiplyLow with AVX2 for SIMD32
+                }
+                else if ((simdSize == 16) && compOpportunisticallyDependsOn(InstructionSet_SSE41))
+                {
+                    // Emulate NI_AVX512DQ_VL_MultiplyLow with SSE41 for SIMD16
+                }
+                else
+                {
+                    // Software fallback
+                    break;
+                }
+            }
+
             op3 = impSIMDPopStack();
             op2 = impSIMDPopStack();
             op1 = impSIMDPopStack();