[clang] [llvm] [AArch64] Implements FP8 SVE intrinsics for dot-product (PR #118125)

[Momchil Velikov via llvm-commits]

================
@@ -10877,24 +10923,37 @@ class sve_fp8_dot_indexed<bits<4> opc, ZPRRegOp dst_ty, Operand iop_ty, string m
   let DestructiveInstType = DestructiveOther;
   let hasSideEffects = 0;
   let mayRaiseFPException = 1;
+
+  let mayLoad  = 1;
+  let mayStore = 0;
 }
 
 // FP8 Widening Dot-Product - Indexed Group
-multiclass sve2_fp8_dot_indexed_h<string asm>{
-  def NAME : sve_fp8_dot_indexed<{0, ?, ?, ?}, ZPR16, VectorIndexH, asm> {
+multiclass sve2_fp8_dot_indexed_h<string asm, SDPatternOperator op> {
+  def NAME : sve_fp8_dot_indexed<{0, ?, ?, ?}, ZPR16, VectorIndexH32b, asm> {
     bits<3> iop;
 
     let Inst{20-19} = iop{2-1};
     let Inst{11}    = iop{0};
+
+    let mayLoad  = 1;
----------------
momchil-velikov wrote:

Done

https://github.com/llvm/llvm-project/pull/118125