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
| // RUN: %clang_cc1 -x cl -O1 -emit-llvm %s -o - -triple x86_64-linux-gnu | FileCheck %s -check-prefix=OPT
// RUN: %clang_cc1 -x cl -O0 -emit-llvm %s -o - -triple x86_64-linux-gnu | FileCheck %s -check-prefix=NOOPT
// OpenCL essentially reduces all shift amounts to the last word-size
// bits before evaluating. Test this both for variables and constants
// evaluated in the front-end.
// OPT: @gtest1 = local_unnamed_addr constant i64 2147483648
__constant const unsigned long gtest1 = 1UL << 31;
// NOOPT: @negativeShift32
int negativeShift32(int a,int b) {
// NOOPT: %array0 = alloca [256 x i8]
char array0[((int)1)<<40];
// NOOPT: %array1 = alloca [256 x i8]
char array1[((int)1)<<(-24)];
// NOOPT: ret i32 65536
return ((int)1)<<(-16);
}
//OPT: @positiveShift32
int positiveShift32(int a,int b) {
//OPT: [[M32:%.+]] = and i32 %b, 31
//OPT-NEXT: [[C32:%.+]] = shl i32 %a, [[M32]]
int c = a<<b;
int d = ((int)1)<<33;
//OPT-NEXT: [[E32:%.+]] = add nsw i32 [[C32]], 2
int e = c + d;
//OPT-NEXT: ret i32 [[E32]]
return e;
}
//OPT: @positiveShift64
long positiveShift64(long a,long b) {
//OPT: [[M64:%.+]] = and i64 %b, 63
//OPT-NEXT: [[C64:%.+]] = ashr i64 %a, [[M64]]
long c = a>>b;
long d = ((long)8)>>65;
//OPT-NEXT: [[E64:%.+]] = add nsw i64 [[C64]], 4
long e = c + d;
//OPT-NEXT: ret i64 [[E64]]
return e;
}
typedef __attribute__((ext_vector_type(4))) int int4;
//OPT: @vectorVectorTest
int4 vectorVectorTest(int4 a,int4 b) {
//OPT: [[VM:%.+]] = and <4 x i32> %b, <i32 31, i32 31, i32 31, i32 31>
//OPT-NEXT: [[VC:%.+]] = shl <4 x i32> %a, [[VM]]
int4 c = a << b;
//OPT-NEXT: [[VF:%.+]] = add <4 x i32> [[VC]], <i32 2, i32 4, i32 16, i32 8>
int4 d = {1, 1, 1, 1};
int4 e = {33, 34, -28, -29};
int4 f = c + (d << e);
//OPT-NEXT: ret <4 x i32> [[VF]]
return f;
}
//NOOPT-LABEL: @vectorScalarTest
int4 vectorScalarTest(int4 a,int b) {
//NOOPT: [[SP0:%.+]] = insertelement <4 x i32> undef
//NOOPT: [[SP1:%.+]] = shufflevector <4 x i32> [[SP0]], <4 x i32> undef, <4 x i32> zeroinitializer
//NOOPT: [[VSM:%.+]] = and <4 x i32> [[SP1]], <i32 31, i32 31, i32 31, i32 31>
//NOOPT: [[VSC:%.+]] = shl <4 x i32> [[VSS:%.+]], [[VSM]]
int4 c = a << b;
//NOOPT: [[VSF:%.+]] = shl <4 x i32> [[VSC1:%.+]], <i32 2, i32 2, i32 2, i32 2>
//NOOPT: [[VSA:%.+]] = add <4 x i32> [[VSC2:%.+]], [[VSF]]
int4 d = {1, 1, 1, 1};
int4 f = c + (d << 34);
//NOOPT: ret <4 x i32>
return f;
}
|