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
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
|
// RUN: %clang_cc1 -triple mips-unknown-linux -o - -emit-llvm %s | FileCheck %s -check-prefixes=ALL,O32 -enable-var-scope
// RUN: %clang_cc1 -triple mipsel-unknown-linux -o - -emit-llvm %s | FileCheck %s -check-prefixes=ALL,O32 -enable-var-scope
// RUN: %clang_cc1 -triple mips64-unknown-linux -o - -emit-llvm -target-abi n32 %s | FileCheck %s -check-prefixes=ALL,N32,NEW -enable-var-scope
// RUN: %clang_cc1 -triple mips64-unknown-linux -o - -emit-llvm -target-abi n32 %s | FileCheck %s -check-prefixes=ALL,N32,NEW -enable-var-scope
// RUN: %clang_cc1 -triple mips64-unknown-linux -o - -emit-llvm %s | FileCheck %s -check-prefixes=ALL,N64,NEW -enable-var-scope
// RUN: %clang_cc1 -triple mips64el-unknown-linux -o - -emit-llvm %s | FileCheck %s -check-prefixes=ALL,N64,NEW -enable-var-scope
#include <stdarg.h>
typedef int v4i32 __attribute__ ((__vector_size__ (16)));
int test_i32(char *fmt, ...) {
va_list va;
va_start(va, fmt);
int v = va_arg(va, int);
va_end(va);
return v;
}
// O32-LABEL: define{{.*}} i32 @test_i32(ptr{{.*}} %fmt, ...)
// N32-LABEL: define{{.*}} signext i32 @test_i32(ptr{{.*}} %fmt, ...)
// N64-LABEL: define{{.*}} signext i32 @test_i32(ptr{{.*}} %fmt, ...)
//
// O32: %va = alloca ptr, align [[$PTRALIGN:4]]
// N32: %va = alloca ptr, align [[$PTRALIGN:4]]
// N64: %va = alloca ptr, align [[$PTRALIGN:8]]
// ALL: [[V:%.*]] = alloca i32, align 4
//
// ALL: call void @llvm.va_start.p0(ptr %va)
// ALL: [[AP_CUR:%.+]] = load ptr, ptr %va, align [[$PTRALIGN]]
// O32: [[AP_NEXT:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], [[$INTPTR_T:i32]] [[$CHUNKSIZE:4]]
// NEW: [[AP_NEXT:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], [[$INTPTR_T:i32|i64]] [[$CHUNKSIZE:8]]
//
// ALL: store ptr [[AP_NEXT]], ptr %va, align [[$PTRALIGN]]
//
// O32: [[ARG:%.+]] = load i32, ptr [[AP_CUR]], align [[CHUNKALIGN:4]]
//
// N32: [[TMP:%.+]] = load i64, ptr [[AP_CUR]], align [[CHUNKALIGN:8]]
// N64: [[TMP:%.+]] = load i64, ptr [[AP_CUR]], align [[CHUNKALIGN:8]]
// NEW: [[ARG:%.+]] = trunc i64 [[TMP]] to i32
// ALL: store i32 [[ARG]], ptr [[V]], align 4
//
// ALL: call void @llvm.va_end.p0(ptr %va)
// ALL: }
long long test_i64(char *fmt, ...) {
va_list va;
va_start(va, fmt);
long long v = va_arg(va, long long);
va_end(va);
return v;
}
// ALL-LABEL: define{{.*}} i64 @test_i64(ptr{{.*}} %fmt, ...)
//
// ALL: %va = alloca ptr, align [[$PTRALIGN]]
// ALL: call void @llvm.va_start.p0(ptr %va)
// ALL: [[AP_CUR:%.+]] = load ptr, ptr %va, align [[$PTRALIGN]]
//
// i64 is 8-byte aligned, while this is within O32's stack alignment there's no
// guarantee that the offset is still 8-byte aligned after earlier reads.
// O32: [[TMP1:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], i32 7
// O32: [[AP_CUR:%.+]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP1]], i32 -8)
//
// ALL: [[AP_NEXT:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], [[$INTPTR_T]] 8
// ALL: store ptr [[AP_NEXT]], ptr %va, align [[$PTRALIGN]]
//
// ALL: [[ARG:%.+]] = load i64, ptr [[AP_CUR]], align 8
//
// ALL: call void @llvm.va_end.p0(ptr %va)
// ALL: }
char *test_ptr(char *fmt, ...) {
va_list va;
va_start(va, fmt);
char *v = va_arg(va, char *);
va_end(va);
return v;
}
// ALL-LABEL: define{{.*}} ptr @test_ptr(ptr{{.*}} %fmt, ...)
//
// ALL: %va = alloca ptr, align [[$PTRALIGN]]
// ALL: [[V:%.*]] = alloca ptr, align [[$PTRALIGN]]
// ALL: call void @llvm.va_start.p0(ptr %va)
// ALL: [[AP_CUR:%.+]] = load ptr, ptr %va, align [[$PTRALIGN]]
// ALL: [[AP_NEXT:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], [[$INTPTR_T]] [[$CHUNKSIZE]]
// ALL: store ptr [[AP_NEXT]], ptr %va, align [[$PTRALIGN]]
//
// When the chunk size matches the pointer size, this is easy.
// Otherwise we need a promotion temporary.
// N32: [[TMP2:%.+]] = load i64, ptr [[AP_CUR]], align 8
// N32: [[TMP3:%.+]] = trunc i64 [[TMP2]] to i32
// N32: [[PTR:%.+]] = inttoptr i32 [[TMP3]] to ptr
// O32: [[PTR:%.+]] = load ptr, ptr [[AP_CUR]], align [[$PTRALIGN]]
// N64: [[PTR:%.+]] = load ptr, ptr [[AP_CUR]], align [[$PTRALIGN]]
// ALL: store ptr [[PTR]], ptr [[V]], align [[$PTRALIGN]]
//
// ALL: call void @llvm.va_end.p0(ptr %va)
// ALL: }
int test_v4i32(char *fmt, ...) {
va_list va;
va_start(va, fmt);
v4i32 v = va_arg(va, v4i32);
va_end(va);
return v[0];
}
// O32-LABEL: define{{.*}} i32 @test_v4i32(ptr{{.*}} %fmt, ...)
// N32-LABEL: define{{.*}} signext i32 @test_v4i32(ptr{{.*}} %fmt, ...)
// N64-LABEL: define{{.*}} signext i32 @test_v4i32(ptr{{.*}} %fmt, ...)
//
// ALL: %va = alloca ptr, align [[$PTRALIGN]]
// ALL: [[V:%.+]] = alloca <4 x i32>, align 16
// ALL: call void @llvm.va_start.p0(ptr %va)
// ALL: [[AP_CUR:%.+]] = load ptr, ptr %va, align [[$PTRALIGN]]
//
// Vectors are 16-byte aligned, however the O32 ABI has a maximum alignment of
// 8-bytes since the base of the stack is 8-byte aligned.
// O32: [[TMP1:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], i32 7
// O32: [[AP_CUR:%.+]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP1]], i32 -8)
// N32: [[TMP1:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], i32 15
// N32: [[AP_CUR:%.+]] = call ptr @llvm.ptrmask.p0.i32(ptr [[TMP1]], i32 -16)
// N64: [[TMP1:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], i32 15
// N64: [[AP_CUR:%.+]] = call ptr @llvm.ptrmask.p0.i64(ptr [[TMP1]], i64 -16)
//
// ALL: [[AP_NEXT:%.+]] = getelementptr inbounds i8, ptr [[AP_CUR]], [[$INTPTR_T]] 16
// ALL: store ptr [[AP_NEXT]], ptr %va, align [[$PTRALIGN]]
//
// O32: [[ARG:%.+]] = load <4 x i32>, ptr [[AP_CUR]], align 8
// N64: [[ARG:%.+]] = load <4 x i32>, ptr [[AP_CUR]], align 16
// N32: [[ARG:%.+]] = load <4 x i32>, ptr [[AP_CUR]], align 16
// ALL: store <4 x i32> [[ARG]], ptr [[V]], align 16
//
// ALL: call void @llvm.va_end.p0(ptr %va)
// ALL: [[VECEXT:%.+]] = extractelement <4 x i32> {{.*}}, i32 0
// ALL: ret i32 [[VECEXT]]
// ALL: }
|
