1 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
2 // RUN: %clang_cc1 -no-opaque-pointers -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -mvscale-min=1 -mvscale-max=1 -fallow-half-arguments-and-returns -S -O1 -emit-llvm -o - %s -fhalf-no-semantic-interposition | FileCheck %s --check-prefix=CHECK-128
3 // RUN: %clang_cc1 -no-opaque-pointers -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -mvscale-min=4 -mvscale-max=4 -fallow-half-arguments-and-returns -S -O1 -emit-llvm -o - %s -fhalf-no-semantic-interposition | FileCheck %s --check-prefix=CHECK-512
4 
5 // REQUIRES: aarch64-registered-target
6 
7 #include <arm_sve.h>
8 
9 #define N __ARM_FEATURE_SVE_BITS
10 
11 typedef svint64_t fixed_int64_t __attribute__((arm_sve_vector_bits(N)));
12 typedef svbfloat16_t fixed_bfloat16_t __attribute__((arm_sve_vector_bits(N)));
13 typedef svbool_t fixed_bool_t __attribute__((arm_sve_vector_bits(N)));
14 
15 fixed_int64_t global_i64;
16 fixed_bfloat16_t global_bf16;
17 fixed_bool_t global_bool;
18 
19 //===----------------------------------------------------------------------===//
20 // WRITES
21 //===----------------------------------------------------------------------===//
22 
23 // CHECK-128-LABEL: @write_global_i64(
24 // CHECK-128-NEXT:  entry:
25 // CHECK-128-NEXT:    [[CASTFIXEDSVE:%.*]] = call <2 x i64> @llvm.vector.extract.v2i64.nxv2i64(<vscale x 2 x i64> [[V:%.*]], i64 0)
26 // CHECK-128-NEXT:    store <2 x i64> [[CASTFIXEDSVE]], <2 x i64>* @global_i64, align 16, !tbaa [[TBAA6:![0-9]+]]
27 // CHECK-128-NEXT:    ret void
28 //
29 // CHECK-512-LABEL: @write_global_i64(
30 // CHECK-512-NEXT:  entry:
31 // CHECK-512-NEXT:    [[CASTFIXEDSVE:%.*]] = call <8 x i64> @llvm.vector.extract.v8i64.nxv2i64(<vscale x 2 x i64> [[V:%.*]], i64 0)
32 // CHECK-512-NEXT:    store <8 x i64> [[CASTFIXEDSVE]], <8 x i64>* @global_i64, align 16, !tbaa [[TBAA6:![0-9]+]]
33 // CHECK-512-NEXT:    ret void
34 //
35 void write_global_i64(svint64_t v) { global_i64 = v; }
36 
37 // CHECK-128-LABEL: @write_global_bf16(
38 // CHECK-128-NEXT:  entry:
39 // CHECK-128-NEXT:    [[CASTFIXEDSVE:%.*]] = call <8 x bfloat> @llvm.vector.extract.v8bf16.nxv8bf16(<vscale x 8 x bfloat> [[V:%.*]], i64 0)
40 // CHECK-128-NEXT:    store <8 x bfloat> [[CASTFIXEDSVE]], <8 x bfloat>* @global_bf16, align 16, !tbaa [[TBAA6]]
41 // CHECK-128-NEXT:    ret void
42 //
43 // CHECK-512-LABEL: @write_global_bf16(
44 // CHECK-512-NEXT:  entry:
45 // CHECK-512-NEXT:    [[CASTFIXEDSVE:%.*]] = call <32 x bfloat> @llvm.vector.extract.v32bf16.nxv8bf16(<vscale x 8 x bfloat> [[V:%.*]], i64 0)
46 // CHECK-512-NEXT:    store <32 x bfloat> [[CASTFIXEDSVE]], <32 x bfloat>* @global_bf16, align 16, !tbaa [[TBAA6]]
47 // CHECK-512-NEXT:    ret void
48 //
49 void write_global_bf16(svbfloat16_t v) { global_bf16 = v; }
50 
51 // CHECK-128-LABEL: @write_global_bool(
52 // CHECK-128-NEXT:  entry:
53 // CHECK-128-NEXT:    [[TMP0:%.*]] = bitcast <vscale x 16 x i1> [[V:%.*]] to <vscale x 2 x i8>
54 // CHECK-128-NEXT:    [[CASTFIXEDSVE:%.*]] = call <2 x i8> @llvm.vector.extract.v2i8.nxv2i8(<vscale x 2 x i8> [[TMP0]], i64 0)
55 // CHECK-128-NEXT:    store <2 x i8> [[CASTFIXEDSVE]], <2 x i8>* @global_bool, align 2, !tbaa [[TBAA6:![0-9]+]]
56 // CHECK-128-NEXT:    ret void
57 //
58 // CHECK-512-LABEL: @write_global_bool(
59 // CHECK-512-NEXT:  entry:
60 // CHECK-512-NEXT:    [[TMP0:%.*]] = bitcast <vscale x 16 x i1> [[V:%.*]] to <vscale x 2 x i8>
61 // CHECK-512-NEXT:    [[CASTFIXEDSVE:%.*]] = call <8 x i8> @llvm.vector.extract.v8i8.nxv2i8(<vscale x 2 x i8> [[TMP0]], i64 0)
62 // CHECK-512-NEXT:    store <8 x i8> [[CASTFIXEDSVE]], <8 x i8>* @global_bool, align 2, !tbaa [[TBAA6]]
63 // CHECK-512-NEXT:    ret void
64 //
65 void write_global_bool(svbool_t v) { global_bool = v; }
66 
67 //===----------------------------------------------------------------------===//
68 // READS
69 //===----------------------------------------------------------------------===//
70 
71 // CHECK-128-LABEL: @read_global_i64(
72 // CHECK-128-NEXT:  entry:
73 // CHECK-128-NEXT:    [[TMP0:%.*]] = load <2 x i64>, <2 x i64>* @global_i64, align 16, !tbaa [[TBAA6]]
74 // CHECK-128-NEXT:    [[CASTSCALABLESVE:%.*]] = call <vscale x 2 x i64> @llvm.vector.insert.nxv2i64.v2i64(<vscale x 2 x i64> undef, <2 x i64> [[TMP0]], i64 0)
75 // CHECK-128-NEXT:    ret <vscale x 2 x i64> [[CASTSCALABLESVE]]
76 //
77 // CHECK-512-LABEL: @read_global_i64(
78 // CHECK-512-NEXT:  entry:
79 // CHECK-512-NEXT:    [[TMP0:%.*]] = load <8 x i64>, <8 x i64>* @global_i64, align 16, !tbaa [[TBAA6]]
80 // CHECK-512-NEXT:    [[CASTSCALABLESVE:%.*]] = call <vscale x 2 x i64> @llvm.vector.insert.nxv2i64.v8i64(<vscale x 2 x i64> undef, <8 x i64> [[TMP0]], i64 0)
81 // CHECK-512-NEXT:    ret <vscale x 2 x i64> [[CASTSCALABLESVE]]
82 //
83 svint64_t read_global_i64() { return global_i64; }
84 
85 // CHECK-128-LABEL: @read_global_bf16(
86 // CHECK-128-NEXT:  entry:
87 // CHECK-128-NEXT:    [[TMP0:%.*]] = load <8 x bfloat>, <8 x bfloat>* @global_bf16, align 16, !tbaa [[TBAA6]]
88 // CHECK-128-NEXT:    [[CASTSCALABLESVE:%.*]] = call <vscale x 8 x bfloat> @llvm.vector.insert.nxv8bf16.v8bf16(<vscale x 8 x bfloat> undef, <8 x bfloat> [[TMP0]], i64 0)
89 // CHECK-128-NEXT:    ret <vscale x 8 x bfloat> [[CASTSCALABLESVE]]
90 //
91 // CHECK-512-LABEL: @read_global_bf16(
92 // CHECK-512-NEXT:  entry:
93 // CHECK-512-NEXT:    [[TMP0:%.*]] = load <32 x bfloat>, <32 x bfloat>* @global_bf16, align 16, !tbaa [[TBAA6]]
94 // CHECK-512-NEXT:    [[CASTSCALABLESVE:%.*]] = call <vscale x 8 x bfloat> @llvm.vector.insert.nxv8bf16.v32bf16(<vscale x 8 x bfloat> undef, <32 x bfloat> [[TMP0]], i64 0)
95 // CHECK-512-NEXT:    ret <vscale x 8 x bfloat> [[CASTSCALABLESVE]]
96 //
97 svbfloat16_t read_global_bf16() { return global_bf16; }
98 
99 // CHECK-128-LABEL: @read_global_bool(
100 // CHECK-128-NEXT:  entry:
101 // CHECK-128-NEXT:    [[TMP0:%.*]] = load <2 x i8>, <2 x i8>* @global_bool, align 2, !tbaa [[TBAA6]]
102 // CHECK-128-NEXT:    [[CASTFIXEDSVE:%.*]] = call <vscale x 2 x i8> @llvm.vector.insert.nxv2i8.v2i8(<vscale x 2 x i8> undef, <2 x i8> [[TMP0]], i64 0)
103 // CHECK-128-NEXT:    [[TMP1:%.*]] = bitcast <vscale x 2 x i8> [[CASTFIXEDSVE]] to <vscale x 16 x i1>
104 // CHECK-128-NEXT:    ret <vscale x 16 x i1> [[TMP1]]
105 //
106 // CHECK-512-LABEL: @read_global_bool(
107 // CHECK-512-NEXT:  entry:
108 // CHECK-512-NEXT:    [[TMP0:%.*]] = load <8 x i8>, <8 x i8>* @global_bool, align 2, !tbaa [[TBAA6]]
109 // CHECK-512-NEXT:    [[CASTFIXEDSVE:%.*]] = call <vscale x 2 x i8> @llvm.vector.insert.nxv2i8.v8i8(<vscale x 2 x i8> undef, <8 x i8> [[TMP0]], i64 0)
110 // CHECK-512-NEXT:    [[TMP1:%.*]] = bitcast <vscale x 2 x i8> [[CASTFIXEDSVE]] to <vscale x 16 x i1>
111 // CHECK-512-NEXT:    ret <vscale x 16 x i1> [[TMP1]]
112 //
113 svbool_t read_global_bool() { return global_bool; }
114