1 ; RUN: llc -mtriple=arm-none-linux-gnu -mattr=+neon,+i8mm -float-abi=hard < %s -o -| FileCheck %s
3 define <4 x i32> @smmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) {
5 ; CHECK-LABEL: smmla.v4i32.v16i8
6 ; CHECK: vsmmla.s8 q0, q1, q2
7 %vmmla1.i = tail call <4 x i32> @llvm.arm.neon.smmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) #3
8 ret <4 x i32> %vmmla1.i
11 define <4 x i32> @ummla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) {
13 ; CHECK-LABEL: ummla.v4i32.v16i8
14 ; CHECK: vummla.u8 q0, q1, q2
15 %vmmla1.i = tail call <4 x i32> @llvm.arm.neon.ummla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) #3
16 ret <4 x i32> %vmmla1.i
19 define <4 x i32> @usmmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) {
21 ; CHECK-LABEL: usmmla.v4i32.v16i8
22 ; CHECK: vusmmla.s8 q0, q1, q2
23 %vusmmla1.i = tail call <4 x i32> @llvm.arm.neon.usmmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) #3
24 ret <4 x i32> %vusmmla1.i
27 define <2 x i32> @usdot.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) {
29 ; CHECK-LABEL: usdot.v2i32.v8i8
30 ; CHECK: vusdot.s8 d0, d1, d2
31 %vusdot1.i = tail call <2 x i32> @llvm.arm.neon.usdot.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) #3
32 ret <2 x i32> %vusdot1.i
35 define <2 x i32> @usdot_lane.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) {
37 ; CHECK-LABEL: usdot_lane.v2i32.v8i8
38 ; CHECK: vusdot.s8 d0, d1, d2[0]
39 %0 = bitcast <8 x i8> %b to <2 x i32>
40 %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <2 x i32> zeroinitializer
41 %1 = bitcast <2 x i32> %shuffle to <8 x i8>
42 %vusdot1.i = tail call <2 x i32> @llvm.arm.neon.usdot.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %1) #3
43 ret <2 x i32> %vusdot1.i
46 define <2 x i32> @sudot_lane.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) {
48 ; CHECK-LABEL: sudot_lane.v2i32.v8i8
49 ; CHECK: vsudot.u8 d0, d1, d2[0]
50 %0 = bitcast <8 x i8> %b to <2 x i32>
51 %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <2 x i32> zeroinitializer
52 %1 = bitcast <2 x i32> %shuffle to <8 x i8>
53 %vusdot1.i = tail call <2 x i32> @llvm.arm.neon.usdot.v2i32.v8i8(<2 x i32> %r, <8 x i8> %1, <8 x i8> %a) #3
54 ret <2 x i32> %vusdot1.i
57 define <4 x i32> @usdotq_lane.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <8 x i8> %b) {
59 ; CHECK-LABEL: usdotq_lane.v4i32.v16i8
60 ; CHECK: vusdot.s8 q0, q1, d4[0]
61 %0 = bitcast <8 x i8> %b to <2 x i32>
62 %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <4 x i32> zeroinitializer
63 %1 = bitcast <4 x i32> %shuffle to <16 x i8>
64 %vusdot1.i = tail call <4 x i32> @llvm.arm.neon.usdot.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %1) #3
65 ret <4 x i32> %vusdot1.i
68 define <4 x i32> @sudotq_lane.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <8 x i8> %b) {
70 ; CHECK-LABEL: sudotq_lane.v4i32.v16i8
71 ; CHECK: vsudot.u8 q0, q1, d4[0]
72 %0 = bitcast <8 x i8> %b to <2 x i32>
73 %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <4 x i32> zeroinitializer
74 %1 = bitcast <4 x i32> %shuffle to <16 x i8>
75 %vusdot1.i = tail call <4 x i32> @llvm.arm.neon.usdot.v4i32.v16i8(<4 x i32> %r, <16 x i8> %1, <16 x i8> %a) #3
76 ret <4 x i32> %vusdot1.i
79 declare <4 x i32> @llvm.arm.neon.smmla.v4i32.v16i8(<4 x i32>, <16 x i8>, <16 x i8>) #2
80 declare <4 x i32> @llvm.arm.neon.ummla.v4i32.v16i8(<4 x i32>, <16 x i8>, <16 x i8>) #2
81 declare <4 x i32> @llvm.arm.neon.usmmla.v4i32.v16i8(<4 x i32>, <16 x i8>, <16 x i8>) #2
82 declare <2 x i32> @llvm.arm.neon.usdot.v2i32.v8i8(<2 x i32>, <8 x i8>, <8 x i8>) #2
83 declare <4 x i32> @llvm.arm.neon.usdot.v4i32.v16i8(<4 x i32>, <16 x i8>, <16 x i8>) #2