1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
|
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _
// RUN: %clang_cc1 -verify -triple powerpc64le-unknown-linux-gnu -fopenmp -x c -emit-llvm %s -o - | FileCheck %s --check-prefix=CHECK1
// RUN: %clang_cc1 -verify -triple powerpc64le-unknown-linux-gnu -fopenmp-simd -x c -emit-llvm %s -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}"
int a;
void foo(void) {
int(*b)[a];
int *(**c)[a];
#pragma omp parallel if (0)
b[0][0] = c[0][a][0][a];
}
void bar(int n, int *a) {
// expected-warning@+1 {{incompatible pointer types initializing 'int (*)[n]' with an expression of type 'int **'}}
int(*p)[n] = &a;
#pragma omp parallel if(0)
// expected-warning@+1 {{comparison of distinct pointer types ('int (*)[n]' and 'int **')}}
if (p == &a) {
}
}
// CHECK1-LABEL: define {{[^@]+}}@foo
// CHECK1-SAME: () #[[ATTR0:[0-9]+]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[B:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[C:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTBOUND_ZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1:[0-9]+]])
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr @a, align 4
// CHECK1-NEXT: [[TMP2:%.*]] = zext i32 [[TMP1]] to i64
// CHECK1-NEXT: [[TMP3:%.*]] = load i32, ptr @a, align 4
// CHECK1-NEXT: [[TMP4:%.*]] = zext i32 [[TMP3]] to i64
// CHECK1-NEXT: call void @__kmpc_serialized_parallel(ptr @[[GLOB1]], i32 [[TMP0]])
// CHECK1-NEXT: store i32 [[TMP0]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK1-NEXT: store i32 0, ptr [[DOTBOUND_ZERO_ADDR]], align 4
// CHECK1-NEXT: call void @foo.omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTBOUND_ZERO_ADDR]], i64 [[TMP2]], ptr [[B]], i64 [[TMP4]], ptr [[C]]) #[[ATTR2:[0-9]+]]
// CHECK1-NEXT: call void @__kmpc_end_serialized_parallel(ptr @[[GLOB1]], i32 [[TMP0]])
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@foo.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[VLA:%.*]], ptr noundef nonnull align 8 dereferenceable(8) [[B:%.*]], i64 noundef [[VLA1:%.*]], ptr noundef nonnull align 8 dereferenceable(8) [[C:%.*]]) #[[ATTR1:[0-9]+]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[VLA_ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[B_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[VLA_ADDR2:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[C_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK1-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK1-NEXT: store i64 [[VLA]], ptr [[VLA_ADDR]], align 8
// CHECK1-NEXT: store ptr [[B]], ptr [[B_ADDR]], align 8
// CHECK1-NEXT: store i64 [[VLA1]], ptr [[VLA_ADDR2]], align 8
// CHECK1-NEXT: store ptr [[C]], ptr [[C_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = load i64, ptr [[VLA_ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load ptr, ptr [[B_ADDR]], align 8
// CHECK1-NEXT: [[TMP2:%.*]] = load i64, ptr [[VLA_ADDR2]], align 8
// CHECK1-NEXT: [[TMP3:%.*]] = load ptr, ptr [[C_ADDR]], align 8
// CHECK1-NEXT: [[TMP4:%.*]] = load ptr, ptr [[TMP3]], align 8
// CHECK1-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP4]], i64 0
// CHECK1-NEXT: [[TMP5:%.*]] = load ptr, ptr [[ARRAYIDX]], align 8
// CHECK1-NEXT: [[TMP6:%.*]] = load i32, ptr @a, align 4
// CHECK1-NEXT: [[IDXPROM:%.*]] = sext i32 [[TMP6]] to i64
// CHECK1-NEXT: [[TMP7:%.*]] = mul nsw i64 [[IDXPROM]], [[TMP2]]
// CHECK1-NEXT: [[ARRAYIDX3:%.*]] = getelementptr inbounds ptr, ptr [[TMP5]], i64 [[TMP7]]
// CHECK1-NEXT: [[ARRAYIDX4:%.*]] = getelementptr inbounds ptr, ptr [[ARRAYIDX3]], i64 0
// CHECK1-NEXT: [[TMP8:%.*]] = load ptr, ptr [[ARRAYIDX4]], align 8
// CHECK1-NEXT: [[TMP9:%.*]] = load i32, ptr @a, align 4
// CHECK1-NEXT: [[IDXPROM5:%.*]] = sext i32 [[TMP9]] to i64
// CHECK1-NEXT: [[ARRAYIDX6:%.*]] = getelementptr inbounds i32, ptr [[TMP8]], i64 [[IDXPROM5]]
// CHECK1-NEXT: [[TMP10:%.*]] = load i32, ptr [[ARRAYIDX6]], align 4
// CHECK1-NEXT: [[TMP11:%.*]] = load ptr, ptr [[TMP1]], align 8
// CHECK1-NEXT: [[TMP12:%.*]] = mul nsw i64 0, [[TMP0]]
// CHECK1-NEXT: [[ARRAYIDX7:%.*]] = getelementptr inbounds i32, ptr [[TMP11]], i64 [[TMP12]]
// CHECK1-NEXT: [[ARRAYIDX8:%.*]] = getelementptr inbounds i32, ptr [[ARRAYIDX7]], i64 0
// CHECK1-NEXT: store i32 [[TMP10]], ptr [[ARRAYIDX8]], align 4
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@bar
// CHECK1-SAME: (i32 noundef signext [[N:%.*]], ptr noundef [[A:%.*]]) #[[ATTR0]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[N_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[A_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[P:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTTHREADID_TEMP_:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[DOTBOUND_ZERO_ADDR:%.*]] = alloca i32, align 4
// CHECK1-NEXT: [[TMP0:%.*]] = call i32 @__kmpc_global_thread_num(ptr @[[GLOB1]])
// CHECK1-NEXT: store i32 [[N]], ptr [[N_ADDR]], align 4
// CHECK1-NEXT: store ptr [[A]], ptr [[A_ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load i32, ptr [[N_ADDR]], align 4
// CHECK1-NEXT: [[TMP2:%.*]] = zext i32 [[TMP1]] to i64
// CHECK1-NEXT: store ptr [[A_ADDR]], ptr [[P]], align 8
// CHECK1-NEXT: call void @__kmpc_serialized_parallel(ptr @[[GLOB1]], i32 [[TMP0]])
// CHECK1-NEXT: store i32 [[TMP0]], ptr [[DOTTHREADID_TEMP_]], align 4
// CHECK1-NEXT: store i32 0, ptr [[DOTBOUND_ZERO_ADDR]], align 4
// CHECK1-NEXT: call void @bar.omp_outlined(ptr [[DOTTHREADID_TEMP_]], ptr [[DOTBOUND_ZERO_ADDR]], i64 [[TMP2]], ptr [[P]], ptr [[A_ADDR]]) #[[ATTR2]]
// CHECK1-NEXT: call void @__kmpc_end_serialized_parallel(ptr @[[GLOB1]], i32 [[TMP0]])
// CHECK1-NEXT: ret void
//
//
// CHECK1-LABEL: define {{[^@]+}}@bar.omp_outlined
// CHECK1-SAME: (ptr noalias noundef [[DOTGLOBAL_TID_:%.*]], ptr noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[VLA:%.*]], ptr noundef nonnull align 8 dereferenceable(8) [[P:%.*]], ptr noundef nonnull align 8 dereferenceable(8) [[A:%.*]]) #[[ATTR1]] {
// CHECK1-NEXT: entry:
// CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[VLA_ADDR:%.*]] = alloca i64, align 8
// CHECK1-NEXT: [[P_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: [[A_ADDR:%.*]] = alloca ptr, align 8
// CHECK1-NEXT: store ptr [[DOTGLOBAL_TID_]], ptr [[DOTGLOBAL_TID__ADDR]], align 8
// CHECK1-NEXT: store ptr [[DOTBOUND_TID_]], ptr [[DOTBOUND_TID__ADDR]], align 8
// CHECK1-NEXT: store i64 [[VLA]], ptr [[VLA_ADDR]], align 8
// CHECK1-NEXT: store ptr [[P]], ptr [[P_ADDR]], align 8
// CHECK1-NEXT: store ptr [[A]], ptr [[A_ADDR]], align 8
// CHECK1-NEXT: [[TMP0:%.*]] = load i64, ptr [[VLA_ADDR]], align 8
// CHECK1-NEXT: [[TMP1:%.*]] = load ptr, ptr [[P_ADDR]], align 8
// CHECK1-NEXT: [[TMP2:%.*]] = load ptr, ptr [[A_ADDR]], align 8
// CHECK1-NEXT: [[TMP3:%.*]] = load ptr, ptr [[TMP1]], align 8
// CHECK1-NEXT: [[CMP:%.*]] = icmp eq ptr [[TMP3]], [[TMP2]]
// CHECK1-NEXT: br i1 [[CMP]], label [[IF_THEN:%.*]], label [[IF_END:%.*]]
// CHECK1: if.then:
// CHECK1-NEXT: br label [[IF_END]]
// CHECK1: if.end:
// CHECK1-NEXT: ret void
//
|
