File: nvptx_device_math_sincos.cpp

package info (click to toggle)
swiftlang 6.0.3-2
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 2,519,992 kB
  • sloc: cpp: 9,107,863; ansic: 2,040,022; asm: 1,135,751; python: 296,500; objc: 82,456; f90: 60,502; lisp: 34,951; pascal: 19,946; sh: 18,133; perl: 7,482; ml: 4,937; javascript: 4,117; makefile: 3,840; awk: 3,535; xml: 914; fortran: 619; cs: 573; ruby: 573
file content (63 lines) | stat: -rw-r--r-- 1,944 bytes parent folder | download | duplicates (12)
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
// REQUIRES: nvptx-registered-target
// RUN: %clang_cc1 -internal-isystem %S/Inputs/include -fopenmp -triple powerpc64le-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm-bc %s -o %t-ppc-host.bc
// RUN: %clang_cc1 -internal-isystem %S/../../lib/Headers/openmp_wrappers         -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -fopenmp -triple nvptx64-nvidia-cuda -aux-triple powerpc64le-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-target-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s
// RUN: %clang_cc1 -internal-isystem %S/../../lib/Headers/openmp_wrappers -DCMATH -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -fopenmp -triple nvptx64-nvidia-cuda -aux-triple powerpc64le-unknown-unknown -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm %s -fopenmp-is-target-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s

#ifdef CMATH
#include <cmath>
#else
#include <math.h>
#endif

// 4 calls to sincos(f), all translated to __nv_sincos calls:

// CHECK-NOT: _Z.sincos
// CHECK: call void @__nv_sincos(double
// CHECK-NOT: _Z.sincos
// CHECK: call void @__nv_sincosf(float
// CHECK-NOT: _Z.sincos
// CHECK: call void @__nv_sincos(double
// CHECK-NOT: _Z.sincos
// CHECK: call void @__nv_sincosf(float
// CHECK-NOT: _Z.sincos

// single precision wrapper
inline void sincos(float x, float* __restrict__ sin, float* __restrict__ cos)
{
  sincosf(x, sin, cos);
}

template<typename T>
void test_sincos(T x)
{
  T res_sin, res_cos;

  #pragma omp target map(from: res_sin, res_cos)
  {
    sincos(x, &res_sin, &res_cos);
  }

}

int main(int argc, char **argv)
{

#if !defined(C_ONLY)
  test_sincos<double>(0.0);
  test_sincos<float>(0.0);
#endif

  #pragma omp target
  {
    double s, c;
    sincos(0, &s, &c);
  }

  #pragma omp target
  {
    float s, c;
    sincosf(0.f, &s, &c);
  }

  return 0;
}