1 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
2 // RUN: %clang_cc1 -triple thumbv8.1m.main-none-none-eabi -target-feature +mve.fp -mfloat-abi hard -fallow-half-arguments-and-returns -disable-O0-optnone -S -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
3 // RUN: %clang_cc1 -triple thumbv8.1m.main-none-none-eabi -target-feature +mve.fp -mfloat-abi hard -fallow-half-arguments-and-returns -disable-O0-optnone -DPOLYMORPHIC -S -emit-llvm -o - %s | opt -S -mem2reg | FileCheck %s
4 
5 #include <arm_mve.h>
6 
7 // CHECK-LABEL: @_Z16test_vbicq_n_s1617__simd128_int16_t(
8 // CHECK-NEXT:  entry:
9 // CHECK-NEXT:    [[TMP0:%.*]] = and <8 x i16> [[A:%.*]], <i16 11007, i16 11007, i16 11007, i16 11007, i16 11007, i16 11007, i16 11007, i16 11007>
10 // CHECK-NEXT:    ret <8 x i16> [[TMP0]]
11 //
12 int16x8_t test_vbicq_n_s16(int16x8_t a)
13 {
14 #ifdef POLYMORPHIC
15     return vbicq(a, 0xd500);
16 #else /* POLYMORPHIC */
17     return vbicq_n_s16(a, 0xd500);
18 #endif /* POLYMORPHIC */
19 }
20 
21 // CHECK-LABEL: @_Z16test_vbicq_n_u3218__simd128_uint32_t(
22 // CHECK-NEXT:  entry:
23 // CHECK-NEXT:    [[TMP0:%.*]] = and <4 x i32> [[A:%.*]], <i32 -8193, i32 -8193, i32 -8193, i32 -8193>
24 // CHECK-NEXT:    ret <4 x i32> [[TMP0]]
25 //
26 uint32x4_t test_vbicq_n_u32(uint32x4_t a)
27 {
28 #ifdef POLYMORPHIC
29     return vbicq(a, 0x2000);
30 #else /* POLYMORPHIC */
31     return vbicq_n_u32(a, 0x2000);
32 #endif /* POLYMORPHIC */
33 }
34 
35 // CHECK-LABEL: @_Z16test_vorrq_n_s3217__simd128_int32_t(
36 // CHECK-NEXT:  entry:
37 // CHECK-NEXT:    [[TMP0:%.*]] = or <4 x i32> [[A:%.*]], <i32 65536, i32 65536, i32 65536, i32 65536>
38 // CHECK-NEXT:    ret <4 x i32> [[TMP0]]
39 //
40 int32x4_t test_vorrq_n_s32(int32x4_t a)
41 {
42 #ifdef POLYMORPHIC
43     return vorrq(a, 0x10000);
44 #else /* POLYMORPHIC */
45     return vorrq_n_s32(a, 0x10000);
46 #endif /* POLYMORPHIC */
47 }
48 
49 // CHECK-LABEL: @_Z16test_vorrq_n_u1618__simd128_uint16_t(
50 // CHECK-NEXT:  entry:
51 // CHECK-NEXT:    [[TMP0:%.*]] = or <8 x i16> [[A:%.*]], <i16 -4096, i16 -4096, i16 -4096, i16 -4096, i16 -4096, i16 -4096, i16 -4096, i16 -4096>
52 // CHECK-NEXT:    ret <8 x i16> [[TMP0]]
53 //
54 uint16x8_t test_vorrq_n_u16(uint16x8_t a)
55 {
56 #ifdef POLYMORPHIC
57     return vorrq(a, 0xf000);
58 #else /* POLYMORPHIC */
59     return vorrq_n_u16(a, 0xf000);
60 #endif /* POLYMORPHIC */
61 }
62 
63 // CHECK-LABEL: @_Z16test_vcmpeqq_f1619__simd128_float16_tS_(
64 // CHECK-NEXT:  entry:
65 // CHECK-NEXT:    [[TMP0:%.*]] = fcmp oeq <8 x half> [[A:%.*]], [[B:%.*]]
66 // CHECK-NEXT:    [[TMP1:%.*]] = call i32 @llvm.arm.mve.pred.v2i.v8i1(<8 x i1> [[TMP0]])
67 // CHECK-NEXT:    [[TMP2:%.*]] = trunc i32 [[TMP1]] to i16
68 // CHECK-NEXT:    ret i16 [[TMP2]]
69 //
70 mve_pred16_t test_vcmpeqq_f16(float16x8_t a, float16x8_t b)
71 {
72 #ifdef POLYMORPHIC
73     return vcmpeqq(a, b);
74 #else /* POLYMORPHIC */
75     return vcmpeqq_f16(a, b);
76 #endif /* POLYMORPHIC */
77 }
78 
79 // CHECK-LABEL: @_Z18test_vcmpeqq_n_f1619__simd128_float16_tDh(
80 // CHECK-NEXT:  entry:
81 // CHECK-NEXT:    [[B:%.*]] = alloca half, align 2
82 // CHECK-NEXT:    [[TMP:%.*]] = alloca float, align 4
83 // CHECK-NEXT:    store float [[B_COERCE:%.*]], float* [[TMP]], align 4
84 // CHECK-NEXT:    [[TMP0:%.*]] = bitcast half* [[B]] to i8*
85 // CHECK-NEXT:    [[TMP1:%.*]] = bitcast float* [[TMP]] to i8*
86 // CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[TMP0]], i8* align 4 [[TMP1]], i32 2, i1 false)
87 // CHECK-NEXT:    [[B1:%.*]] = load half, half* [[B]], align 2
88 // CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <8 x half> undef, half [[B1]], i32 0
89 // CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <8 x half> [[DOTSPLATINSERT]], <8 x half> undef, <8 x i32> zeroinitializer
90 // CHECK-NEXT:    [[TMP2:%.*]] = fcmp oeq <8 x half> [[A:%.*]], [[DOTSPLAT]]
91 // CHECK-NEXT:    [[TMP3:%.*]] = call i32 @llvm.arm.mve.pred.v2i.v8i1(<8 x i1> [[TMP2]])
92 // CHECK-NEXT:    [[TMP4:%.*]] = trunc i32 [[TMP3]] to i16
93 // CHECK-NEXT:    ret i16 [[TMP4]]
94 //
95 mve_pred16_t test_vcmpeqq_n_f16(float16x8_t a, float16_t b)
96 {
97 #ifdef POLYMORPHIC
98     return vcmpeqq(a, b);
99 #else /* POLYMORPHIC */
100     return vcmpeqq_n_f16(a, b);
101 #endif /* POLYMORPHIC */
102 }
103 
104 // CHECK-LABEL: @_Z14test_vld1q_u16PKt(
105 // CHECK-NEXT:  entry:
106 // CHECK-NEXT:    [[TMP0:%.*]] = bitcast i16* [[BASE:%.*]] to <8 x i16>*
107 // CHECK-NEXT:    [[TMP1:%.*]] = load <8 x i16>, <8 x i16>* [[TMP0]], align 2
108 // CHECK-NEXT:    ret <8 x i16> [[TMP1]]
109 //
110 uint16x8_t test_vld1q_u16(const uint16_t *base)
111 {
112 #ifdef POLYMORPHIC
113     return vld1q(base);
114 #else /* POLYMORPHIC */
115     return vld1q_u16(base);
116 #endif /* POLYMORPHIC */
117 }
118 
119 // CHECK-LABEL: @_Z16test_vst1q_p_s32Pi17__simd128_int32_tt(
120 // CHECK-NEXT:  entry:
121 // CHECK-NEXT:    [[TMP0:%.*]] = bitcast i32* [[BASE:%.*]] to <4 x i32>*
122 // CHECK-NEXT:    [[TMP1:%.*]] = zext i16 [[P:%.*]] to i32
123 // CHECK-NEXT:    [[TMP2:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP1]])
124 // CHECK-NEXT:    call void @llvm.masked.store.v4i32.p0v4i32(<4 x i32> [[VALUE:%.*]], <4 x i32>* [[TMP0]], i32 4, <4 x i1> [[TMP2]])
125 // CHECK-NEXT:    ret void
126 //
127 void test_vst1q_p_s32(int32_t *base, int32x4_t value, mve_pred16_t p)
128 {
129 #ifdef POLYMORPHIC
130     vst1q_p(base, value, p);
131 #else /* POLYMORPHIC */
132     vst1q_p_s32(base, value, p);
133 #endif /* POLYMORPHIC */
134 }
135 
136 // CHECK-LABEL: @_Z30test_vldrdq_gather_base_wb_s64P18__simd128_uint64_t(
137 // CHECK-NEXT:  entry:
138 // CHECK-NEXT:    [[TMP0:%.*]] = load <2 x i64>, <2 x i64>* [[ADDR:%.*]], align 8
139 // CHECK-NEXT:    [[TMP1:%.*]] = call { <2 x i64>, <2 x i64> } @llvm.arm.mve.vldr.gather.base.wb.v2i64.v2i64(<2 x i64> [[TMP0]], i32 576)
140 // CHECK-NEXT:    [[TMP2:%.*]] = extractvalue { <2 x i64>, <2 x i64> } [[TMP1]], 1
141 // CHECK-NEXT:    store <2 x i64> [[TMP2]], <2 x i64>* [[ADDR]], align 8
142 // CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <2 x i64>, <2 x i64> } [[TMP1]], 0
143 // CHECK-NEXT:    ret <2 x i64> [[TMP3]]
144 //
145 int64x2_t test_vldrdq_gather_base_wb_s64(uint64x2_t *addr)
146 {
147     return vldrdq_gather_base_wb_s64(addr, 0x240);
148 }
149 
150 // CHECK-LABEL: @_Z31test_vstrwq_scatter_base_wb_u32P18__simd128_uint32_tS_(
151 // CHECK-NEXT:  entry:
152 // CHECK-NEXT:    [[TMP0:%.*]] = load <4 x i32>, <4 x i32>* [[ADDR:%.*]], align 8
153 // CHECK-NEXT:    [[TMP1:%.*]] = call <4 x i32> @llvm.arm.mve.vstr.scatter.base.wb.v4i32.v4i32(<4 x i32> [[TMP0]], i32 64, <4 x i32> [[VALUE:%.*]])
154 // CHECK-NEXT:    store <4 x i32> [[TMP1]], <4 x i32>* [[ADDR]], align 8
155 // CHECK-NEXT:    ret void
156 //
157 void test_vstrwq_scatter_base_wb_u32(uint32x4_t *addr, uint32x4_t value)
158 {
159 #ifdef POLYMORPHIC
160     vstrwq_scatter_base_wb(addr, 0x40, value);
161 #else /* POLYMORPHIC */
162     vstrwq_scatter_base_wb_u32(addr, 0x40, value);
163 #endif /* POLYMORPHIC */
164 }
165