[clang] [llvm] [Clang][RISCV] Add Zvabd intrinsics (PR #180929)

[Craig Topper via cfe-commits]

================
@@ -0,0 +1,139 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 2
+// REQUIRES: riscv-registered-target
+// RUN: %clang_cc1 -triple riscv64 -target-feature +v -target-feature +experimental-zvabd \
+// RUN:   -disable-O0-optnone \
+// RUN:   -emit-llvm %s -o - | opt -S -passes=mem2reg | \
+// RUN:   FileCheck --check-prefix=CHECK-RV64 %s
+
+#include <riscv_vector.h>
+
+// CHECK-RV64-LABEL: define dso_local <vscale x 1 x i8> @test_vabd_vv_i8mf8
+// CHECK-RV64-SAME: (<vscale x 1 x i8> [[VS2:%.*]], <vscale x 1 x i8> [[VS1:%.*]], i64 noundef [[VL:%.*]]) #[[ATTR0:[0-9]+]] {
+// CHECK-RV64-NEXT:  entry:
+// CHECK-RV64-NEXT:    [[TMP0:%.*]] = call <vscale x 1 x i8> @llvm.riscv.vabd.nxv1i8.nxv1i8.i64(<vscale x 1 x i8> poison, <vscale x 1 x i8> [[VS2]], <vscale x 1 x i8> [[VS1]], i64 [[VL]])
+// CHECK-RV64-NEXT:    ret <vscale x 1 x i8> [[TMP0]]
+//
+vint8mf8_t test_vabd_vv_i8mf8(vint8mf8_t vs2, vint8mf8_t vs1, size_t vl) {
----------------
topperc wrote:

The absolute difference result should be an unsigned type. The difference between -128 and 127 for example is 255 not -1.

https://github.com/llvm/llvm-project/pull/180929