diff options
Diffstat (limited to 'test/CodeGen/aarch64-neon-perm.c')
| -rw-r--r-- | test/CodeGen/aarch64-neon-perm.c | 324 |
1 files changed, 54 insertions, 270 deletions
diff --git a/test/CodeGen/aarch64-neon-perm.c b/test/CodeGen/aarch64-neon-perm.c index 471017a99bbd..c5d5ab18070c 100644 --- a/test/CodeGen/aarch64-neon-perm.c +++ b/test/CodeGen/aarch64-neon-perm.c @@ -888,18 +888,14 @@ poly16x8_t test_vtrn2q_p16(poly16x8_t a, poly16x8_t b) { // CHECK-LABEL: @test_vuzp_s8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.int8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.int8x8x2_t [[TMP5]], 0 @@ -912,9 +908,8 @@ int8x8x2_t test_vuzp_s8(int8x8_t a, int8x8_t b) { // CHECK-LABEL: @test_vuzp_s16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -923,9 +918,6 @@ int8x8x2_t test_vuzp_s8(int8x8_t a, int8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int16x4x2_t [[TMP7]], 0 @@ -938,9 +930,8 @@ int16x4x2_t test_vuzp_s16(int16x4_t a, int16x4_t b) { // CHECK-LABEL: @test_vuzp_s32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* @@ -949,9 +940,6 @@ int16x4x2_t test_vuzp_s16(int16x4_t a, int16x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x i32> [[VUZP1_I]], <2 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int32x2x2_t [[TMP7]], 0 @@ -964,18 +952,14 @@ int32x2x2_t test_vuzp_s32(int32x2_t a, int32x2_t b) { // CHECK-LABEL: @test_vuzp_u8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.uint8x8x2_t [[TMP5]], 0 @@ -988,9 +972,8 @@ uint8x8x2_t test_vuzp_u8(uint8x8_t a, uint8x8_t b) { // CHECK-LABEL: @test_vuzp_u16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -999,9 +982,6 @@ uint8x8x2_t test_vuzp_u8(uint8x8_t a, uint8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint16x4x2_t [[TMP7]], 0 @@ -1014,9 +994,8 @@ uint16x4x2_t test_vuzp_u16(uint16x4_t a, uint16x4_t b) { // CHECK-LABEL: @test_vuzp_u32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* @@ -1025,9 +1004,6 @@ uint16x4x2_t test_vuzp_u16(uint16x4_t a, uint16x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x i32> [[VUZP1_I]], <2 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint32x2x2_t [[TMP7]], 0 @@ -1040,9 +1016,8 @@ uint32x2x2_t test_vuzp_u32(uint32x2_t a, uint32x2_t b) { // CHECK-LABEL: @test_vuzp_f32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.float32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.float32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.float32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>* @@ -1051,9 +1026,6 @@ uint32x2x2_t test_vuzp_u32(uint32x2_t a, uint32x2_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x float> [[VUZP1_I]], <2 x float>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.float32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.float32x2x2_t [[TMP7]], 0 @@ -1066,18 +1038,14 @@ float32x2x2_t test_vuzp_f32(float32x2_t a, float32x2_t b) { // CHECK-LABEL: @test_vuzp_p8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VUZP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14> // CHECK: store <8 x i8> [[VUZP_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: store <8 x i8> [[VUZP1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.poly8x8x2_t [[TMP5]], 0 @@ -1090,9 +1058,8 @@ poly8x8x2_t test_vuzp_p8(poly8x8_t a, poly8x8_t b) { // CHECK-LABEL: @test_vuzp_p16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -1101,9 +1068,6 @@ poly8x8x2_t test_vuzp_p8(poly8x8_t a, poly8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: store <4 x i16> [[VUZP1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.poly16x4x2_t [[TMP7]], 0 @@ -1116,18 +1080,14 @@ poly16x4x2_t test_vuzp_p16(poly16x4_t a, poly16x4_t b) { // CHECK-LABEL: @test_vuzpq_s8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30> // CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31> // CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.int8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.int8x16x2_t [[TMP5]], 0 @@ -1140,9 +1100,8 @@ int8x16x2_t test_vuzpq_s8(int8x16_t a, int8x16_t b) { // CHECK-LABEL: @test_vuzpq_s16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -1151,9 +1110,6 @@ int8x16x2_t test_vuzpq_s8(int8x16_t a, int8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int16x8x2_t [[TMP7]], 0 @@ -1166,9 +1122,8 @@ int16x8x2_t test_vuzpq_s16(int16x8_t a, int16x8_t b) { // CHECK-LABEL: @test_vuzpq_s32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* @@ -1177,9 +1132,6 @@ int16x8x2_t test_vuzpq_s16(int16x8_t a, int16x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: store <4 x i32> [[VUZP1_I]], <4 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int32x4x2_t [[TMP7]], 0 @@ -1192,18 +1144,14 @@ int32x4x2_t test_vuzpq_s32(int32x4_t a, int32x4_t b) { // CHECK-LABEL: @test_vuzpq_u8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30> // CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31> // CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.uint8x16x2_t [[TMP5]], 0 @@ -1216,9 +1164,8 @@ uint8x16x2_t test_vuzpq_u8(uint8x16_t a, uint8x16_t b) { // CHECK-LABEL: @test_vuzpq_u16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -1227,9 +1174,6 @@ uint8x16x2_t test_vuzpq_u8(uint8x16_t a, uint8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint16x8x2_t [[TMP7]], 0 @@ -1242,9 +1186,8 @@ uint16x8x2_t test_vuzpq_u16(uint16x8_t a, uint16x8_t b) { // CHECK-LABEL: @test_vuzpq_u32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* @@ -1253,9 +1196,6 @@ uint16x8x2_t test_vuzpq_u16(uint16x8_t a, uint16x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: store <4 x i32> [[VUZP1_I]], <4 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint32x4x2_t [[TMP7]], 0 @@ -1268,9 +1208,8 @@ uint32x4x2_t test_vuzpq_u32(uint32x4_t a, uint32x4_t b) { // CHECK-LABEL: @test_vuzpq_f32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.float32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.float32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.float32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>* @@ -1279,9 +1218,6 @@ uint32x4x2_t test_vuzpq_u32(uint32x4_t a, uint32x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 3, i32 5, i32 7> // CHECK: store <4 x float> [[VUZP1_I]], <4 x float>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.float32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.float32x4x2_t [[TMP7]], 0 @@ -1294,18 +1230,14 @@ float32x4x2_t test_vuzpq_f32(float32x4_t a, float32x4_t b) { // CHECK-LABEL: @test_vuzpq_p8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VUZP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 8, i32 10, i32 12, i32 14, i32 16, i32 18, i32 20, i32 22, i32 24, i32 26, i32 28, i32 30> // CHECK: store <16 x i8> [[VUZP_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15, i32 17, i32 19, i32 21, i32 23, i32 25, i32 27, i32 29, i32 31> // CHECK: store <16 x i8> [[VUZP1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.poly8x16x2_t [[TMP5]], 0 @@ -1318,9 +1250,8 @@ poly8x16x2_t test_vuzpq_p8(poly8x16_t a, poly8x16_t b) { // CHECK-LABEL: @test_vuzpq_p16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -1329,9 +1260,6 @@ poly8x16x2_t test_vuzpq_p8(poly8x16_t a, poly8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VUZP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 3, i32 5, i32 7, i32 9, i32 11, i32 13, i32 15> // CHECK: store <8 x i16> [[VUZP1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.poly16x8x2_t [[TMP7]], 0 @@ -1344,18 +1272,14 @@ poly16x8x2_t test_vuzpq_p16(poly16x8_t a, poly16x8_t b) { // CHECK-LABEL: @test_vzip_s8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.int8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.int8x8x2_t [[TMP5]], 0 @@ -1368,9 +1292,8 @@ int8x8x2_t test_vzip_s8(int8x8_t a, int8x8_t b) { // CHECK-LABEL: @test_vzip_s16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -1379,9 +1302,6 @@ int8x8x2_t test_vzip_s8(int8x8_t a, int8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int16x4x2_t [[TMP7]], 0 @@ -1394,9 +1314,8 @@ int16x4x2_t test_vzip_s16(int16x4_t a, int16x4_t b) { // CHECK-LABEL: @test_vzip_s32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* @@ -1405,9 +1324,6 @@ int16x4x2_t test_vzip_s16(int16x4_t a, int16x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x i32> [[VZIP1_I]], <2 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int32x2x2_t [[TMP7]], 0 @@ -1420,18 +1336,14 @@ int32x2x2_t test_vzip_s32(int32x2_t a, int32x2_t b) { // CHECK-LABEL: @test_vzip_u8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.uint8x8x2_t [[TMP5]], 0 @@ -1444,9 +1356,8 @@ uint8x8x2_t test_vzip_u8(uint8x8_t a, uint8x8_t b) { // CHECK-LABEL: @test_vzip_u16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -1455,9 +1366,6 @@ uint8x8x2_t test_vzip_u8(uint8x8_t a, uint8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint16x4x2_t [[TMP7]], 0 @@ -1470,9 +1378,8 @@ uint16x4x2_t test_vzip_u16(uint16x4_t a, uint16x4_t b) { // CHECK-LABEL: @test_vzip_u32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* @@ -1481,9 +1388,6 @@ uint16x4x2_t test_vzip_u16(uint16x4_t a, uint16x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x i32> [[VZIP1_I]], <2 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint32x2x2_t [[TMP7]], 0 @@ -1496,9 +1400,8 @@ uint32x2x2_t test_vzip_u32(uint32x2_t a, uint32x2_t b) { // CHECK-LABEL: @test_vzip_f32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.float32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.float32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.float32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>* @@ -1507,9 +1410,6 @@ uint32x2x2_t test_vzip_u32(uint32x2_t a, uint32x2_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x float> [[VZIP1_I]], <2 x float>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.float32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.float32x2x2_t [[TMP7]], 0 @@ -1522,18 +1422,14 @@ float32x2x2_t test_vzip_f32(float32x2_t a, float32x2_t b) { // CHECK-LABEL: @test_vzip_p8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VZIP_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> // CHECK: store <8 x i8> [[VZIP_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: store <8 x i8> [[VZIP1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.poly8x8x2_t [[TMP5]], 0 @@ -1546,9 +1442,8 @@ poly8x8x2_t test_vzip_p8(poly8x8_t a, poly8x8_t b) { // CHECK-LABEL: @test_vzip_p16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -1557,9 +1452,6 @@ poly8x8x2_t test_vzip_p8(poly8x8_t a, poly8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: store <4 x i16> [[VZIP1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.poly16x4x2_t [[TMP7]], 0 @@ -1572,18 +1464,14 @@ poly16x4x2_t test_vzip_p16(poly16x4_t a, poly16x4_t b) { // CHECK-LABEL: @test_vzipq_s8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> // CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31> // CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.int8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.int8x16x2_t [[TMP5]], 0 @@ -1596,9 +1484,8 @@ int8x16x2_t test_vzipq_s8(int8x16_t a, int8x16_t b) { // CHECK-LABEL: @test_vzipq_s16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -1607,9 +1494,6 @@ int8x16x2_t test_vzipq_s8(int8x16_t a, int8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int16x8x2_t [[TMP7]], 0 @@ -1622,9 +1506,8 @@ int16x8x2_t test_vzipq_s16(int16x8_t a, int16x8_t b) { // CHECK-LABEL: @test_vzipq_s32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* @@ -1633,9 +1516,6 @@ int16x8x2_t test_vzipq_s16(int16x8_t a, int16x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: store <4 x i32> [[VZIP1_I]], <4 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int32x4x2_t [[TMP7]], 0 @@ -1648,18 +1528,14 @@ int32x4x2_t test_vzipq_s32(int32x4_t a, int32x4_t b) { // CHECK-LABEL: @test_vzipq_u8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> // CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31> // CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.uint8x16x2_t [[TMP5]], 0 @@ -1672,9 +1548,8 @@ uint8x16x2_t test_vzipq_u8(uint8x16_t a, uint8x16_t b) { // CHECK-LABEL: @test_vzipq_u16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -1683,9 +1558,6 @@ uint8x16x2_t test_vzipq_u8(uint8x16_t a, uint8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint16x8x2_t [[TMP7]], 0 @@ -1698,9 +1570,8 @@ uint16x8x2_t test_vzipq_u16(uint16x8_t a, uint16x8_t b) { // CHECK-LABEL: @test_vzipq_u32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* @@ -1709,9 +1580,6 @@ uint16x8x2_t test_vzipq_u16(uint16x8_t a, uint16x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: store <4 x i32> [[VZIP1_I]], <4 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint32x4x2_t [[TMP7]], 0 @@ -1724,9 +1592,8 @@ uint32x4x2_t test_vzipq_u32(uint32x4_t a, uint32x4_t b) { // CHECK-LABEL: @test_vzipq_f32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.float32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.float32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.float32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>* @@ -1735,9 +1602,6 @@ uint32x4x2_t test_vzipq_u32(uint32x4_t a, uint32x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 2, i32 6, i32 3, i32 7> // CHECK: store <4 x float> [[VZIP1_I]], <4 x float>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.float32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.float32x4x2_t [[TMP7]], 0 @@ -1750,18 +1614,14 @@ float32x4x2_t test_vzipq_f32(float32x4_t a, float32x4_t b) { // CHECK-LABEL: @test_vzipq_p8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VZIP_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> // CHECK: store <16 x i8> [[VZIP_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31> // CHECK: store <16 x i8> [[VZIP1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.poly8x16x2_t [[TMP5]], 0 @@ -1774,9 +1634,8 @@ poly8x16x2_t test_vzipq_p8(poly8x16_t a, poly8x16_t b) { // CHECK-LABEL: @test_vzipq_p16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -1785,9 +1644,6 @@ poly8x16x2_t test_vzipq_p8(poly8x16_t a, poly8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VZIP1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> // CHECK: store <8 x i16> [[VZIP1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.poly16x8x2_t [[TMP7]], 0 @@ -1800,18 +1656,14 @@ poly16x8x2_t test_vzipq_p16(poly16x8_t a, poly16x8_t b) { // CHECK-LABEL: @test_vtrn_s8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.int8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.int8x8x2_t, %struct.int8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.int8x8x2_t [[TMP5]], 0 @@ -1824,9 +1676,8 @@ int8x8x2_t test_vtrn_s8(int8x8_t a, int8x8_t b) { // CHECK-LABEL: @test_vtrn_s16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -1835,9 +1686,6 @@ int8x8x2_t test_vtrn_s8(int8x8_t a, int8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int16x4x2_t, %struct.int16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int16x4x2_t [[TMP7]], 0 @@ -1850,9 +1698,8 @@ int16x4x2_t test_vtrn_s16(int16x4_t a, int16x4_t b) { // CHECK-LABEL: @test_vtrn_s32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.int32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* @@ -1861,9 +1708,6 @@ int16x4x2_t test_vtrn_s16(int16x4_t a, int16x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x i32> [[VTRN1_I]], <2 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int32x2x2_t, %struct.int32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int32x2x2_t [[TMP7]], 0 @@ -1876,18 +1720,14 @@ int32x2x2_t test_vtrn_s32(int32x2_t a, int32x2_t b) { // CHECK-LABEL: @test_vtrn_u8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.uint8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.uint8x8x2_t, %struct.uint8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.uint8x8x2_t [[TMP5]], 0 @@ -1900,9 +1740,8 @@ uint8x8x2_t test_vtrn_u8(uint8x8_t a, uint8x8_t b) { // CHECK-LABEL: @test_vtrn_u16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -1911,9 +1750,6 @@ uint8x8x2_t test_vtrn_u8(uint8x8_t a, uint8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint16x4x2_t, %struct.uint16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint16x4x2_t [[TMP7]], 0 @@ -1926,9 +1762,8 @@ uint16x4x2_t test_vtrn_u16(uint16x4_t a, uint16x4_t b) { // CHECK-LABEL: @test_vtrn_u32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x i32> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x i32> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x i32>* @@ -1937,9 +1772,6 @@ uint16x4x2_t test_vtrn_u16(uint16x4_t a, uint16x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x i32>, <2 x i32>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x i32> %a, <2 x i32> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x i32> [[VTRN1_I]], <2 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint32x2x2_t, %struct.uint32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint32x2x2_t [[TMP7]], 0 @@ -1952,9 +1784,8 @@ uint32x2x2_t test_vtrn_u32(uint32x2_t a, uint32x2_t b) { // CHECK-LABEL: @test_vtrn_f32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.float32x2x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.float32x2x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.float32x2x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <2 x float> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <2 x float> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <2 x float>* @@ -1963,9 +1794,6 @@ uint32x2x2_t test_vtrn_u32(uint32x2_t a, uint32x2_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <2 x float>, <2 x float>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <2 x float> %a, <2 x float> %b, <2 x i32> <i32 1, i32 3> // CHECK: store <2 x float> [[VTRN1_I]], <2 x float>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x2x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.float32x2x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.float32x2x2_t, %struct.float32x2x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.float32x2x2_t [[TMP7]], 0 @@ -1978,18 +1806,14 @@ float32x2x2_t test_vtrn_f32(float32x2_t a, float32x2_t b) { // CHECK-LABEL: @test_vtrn_p8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly8x8x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly8x8x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly8x8x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <8 x i8>* // CHECK: [[VTRN_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 0, i32 8, i32 2, i32 10, i32 4, i32 12, i32 6, i32 14> // CHECK: store <8 x i8> [[VTRN_I]], <8 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <8 x i8>, <8 x i8>* [[TMP1]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i8> %a, <8 x i8> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: store <8 x i8> [[VTRN1_I]], <8 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.poly8x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.poly8x8x2_t, %struct.poly8x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.poly8x8x2_t [[TMP5]], 0 @@ -2002,9 +1826,8 @@ poly8x8x2_t test_vtrn_p8(poly8x8_t a, poly8x8_t b) { // CHECK-LABEL: @test_vtrn_p16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly16x4x2_t, align 8 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly16x4x2_t, align 8 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly16x4x2_t, align 8 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i16> %a to <8 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i16> %b to <8 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i16>* @@ -2013,9 +1836,6 @@ poly8x8x2_t test_vtrn_p8(poly8x8_t a, poly8x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i16>, <4 x i16>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i16> %a, <4 x i16> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: store <4 x i16> [[VTRN1_I]], <4 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.poly16x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 16, i32 8, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL_I]], align 8 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.poly16x4x2_t, %struct.poly16x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.poly16x4x2_t [[TMP7]], 0 @@ -2028,18 +1848,14 @@ poly16x4x2_t test_vtrn_p16(poly16x4_t a, poly16x4_t b) { // CHECK-LABEL: @test_vtrnq_s8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30> // CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31> // CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.int8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.int8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.int8x16x2_t, %struct.int8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.int8x16x2_t [[TMP5]], 0 @@ -2052,9 +1868,8 @@ int8x16x2_t test_vtrnq_s8(int8x16_t a, int8x16_t b) { // CHECK-LABEL: @test_vtrnq_s16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -2063,9 +1878,6 @@ int8x16x2_t test_vtrnq_s8(int8x16_t a, int8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: store <8 x i16> [[VTRN1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int16x8x2_t, %struct.int16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int16x8x2_t [[TMP7]], 0 @@ -2078,9 +1890,8 @@ int16x8x2_t test_vtrnq_s16(int16x8_t a, int16x8_t b) { // CHECK-LABEL: @test_vtrnq_s32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.int32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.int32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.int32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* @@ -2089,9 +1900,6 @@ int16x8x2_t test_vtrnq_s16(int16x8_t a, int16x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: store <4 x i32> [[VTRN1_I]], <4 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.int32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.int32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.int32x4x2_t, %struct.int32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.int32x4x2_t [[TMP7]], 0 @@ -2104,18 +1912,14 @@ int32x4x2_t test_vtrnq_s32(int32x4_t a, int32x4_t b) { // CHECK-LABEL: @test_vtrnq_u8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30> // CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31> // CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.uint8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.uint8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.uint8x16x2_t, %struct.uint8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.uint8x16x2_t [[TMP5]], 0 @@ -2128,9 +1932,8 @@ uint8x16x2_t test_vtrnq_u8(uint8x16_t a, uint8x16_t b) { // CHECK-LABEL: @test_vtrnq_u16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -2139,9 +1942,6 @@ uint8x16x2_t test_vtrnq_u8(uint8x16_t a, uint8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: store <8 x i16> [[VTRN1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint16x8x2_t, %struct.uint16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint16x8x2_t [[TMP7]], 0 @@ -2154,9 +1954,8 @@ uint16x8x2_t test_vtrnq_u16(uint16x8_t a, uint16x8_t b) { // CHECK-LABEL: @test_vtrnq_u32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.uint32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.uint32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.uint32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x i32> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x i32> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x i32>* @@ -2165,9 +1964,6 @@ uint16x8x2_t test_vtrnq_u16(uint16x8_t a, uint16x8_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x i32>, <4 x i32>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: store <4 x i32> [[VTRN1_I]], <4 x i32>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.uint32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.uint32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.uint32x4x2_t, %struct.uint32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.uint32x4x2_t [[TMP7]], 0 @@ -2180,9 +1976,8 @@ uint32x4x2_t test_vtrnq_u32(uint32x4_t a, uint32x4_t b) { // CHECK-LABEL: @test_vtrnq_f32( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.float32x4x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.float32x4x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.float32x4x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <4 x float> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <4 x float> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <4 x float>* @@ -2191,9 +1986,6 @@ uint32x4x2_t test_vtrnq_u32(uint32x4_t a, uint32x4_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <4 x float>, <4 x float>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <4 x float> %a, <4 x float> %b, <4 x i32> <i32 1, i32 5, i32 3, i32 7> // CHECK: store <4 x float> [[VTRN1_I]], <4 x float>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.float32x4x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.float32x4x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.float32x4x2_t, %struct.float32x4x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.float32x4x2_t [[TMP7]], 0 @@ -2206,18 +1998,14 @@ float32x4x2_t test_vtrnq_f32(float32x4_t a, float32x4_t b) { // CHECK-LABEL: @test_vtrnq_p8( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly8x16x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly8x16x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly8x16x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast i8* [[TMP0]] to <16 x i8>* // CHECK: [[VTRN_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30> // CHECK: store <16 x i8> [[VTRN_I]], <16 x i8>* [[TMP1]] // CHECK: [[TMP2:%.*]] = getelementptr inbounds <16 x i8>, <16 x i8>* [[TMP1]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31> // CHECK: store <16 x i8> [[VTRN1_I]], <16 x i8>* [[TMP2]] -// CHECK: [[TMP3:%.*]] = bitcast %struct.poly8x16x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP4:%.*]] = bitcast %struct.poly8x16x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP3]], i8* [[TMP4]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP5:%.*]] = load %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP6:%.*]] = getelementptr inbounds %struct.poly8x16x2_t, %struct.poly8x16x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP7:%.*]] = extractvalue %struct.poly8x16x2_t [[TMP5]], 0 @@ -2230,9 +2018,8 @@ poly8x16x2_t test_vtrnq_p8(poly8x16_t a, poly8x16_t b) { // CHECK-LABEL: @test_vtrnq_p16( // CHECK: [[RETVAL_I:%.*]] = alloca %struct.poly16x8x2_t, align 16 -// CHECK: [[__RET_I:%.*]] = alloca %struct.poly16x8x2_t, align 16 // CHECK: [[RETVAL:%.*]] = alloca %struct.poly16x8x2_t, align 16 -// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET_I]] to i8* +// CHECK: [[TMP0:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL_I]] to i8* // CHECK: [[TMP1:%.*]] = bitcast <8 x i16> %a to <16 x i8> // CHECK: [[TMP2:%.*]] = bitcast <8 x i16> %b to <16 x i8> // CHECK: [[TMP3:%.*]] = bitcast i8* [[TMP0]] to <8 x i16>* @@ -2241,9 +2028,6 @@ poly8x16x2_t test_vtrnq_p8(poly8x16_t a, poly8x16_t b) { // CHECK: [[TMP4:%.*]] = getelementptr inbounds <8 x i16>, <8 x i16>* [[TMP3]], i32 1 // CHECK: [[VTRN1_I:%.*]] = shufflevector <8 x i16> %a, <8 x i16> %b, <8 x i32> <i32 1, i32 9, i32 3, i32 11, i32 5, i32 13, i32 7, i32 15> // CHECK: store <8 x i16> [[VTRN1_I]], <8 x i16>* [[TMP4]] -// CHECK: [[TMP5:%.*]] = bitcast %struct.poly16x8x2_t* [[RETVAL_I]] to i8* -// CHECK: [[TMP6:%.*]] = bitcast %struct.poly16x8x2_t* [[__RET_I]] to i8* -// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[TMP5]], i8* [[TMP6]], i64 32, i32 16, i1 false) #2 // CHECK: [[TMP7:%.*]] = load %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL_I]], align 16 // CHECK: [[TMP8:%.*]] = getelementptr inbounds %struct.poly16x8x2_t, %struct.poly16x8x2_t* [[RETVAL]], i32 0, i32 0 // CHECK: [[TMP9:%.*]] = extractvalue %struct.poly16x8x2_t [[TMP7]], 0 |