File: __clang_hip_cmath.hip

package info (click to toggle)
llvm-toolchain-16 1%3A16.0.6-15~deb12u1
  • links: PTS, VCS
  • area: main
  • in suites: bookworm
  • size: 1,634,792 kB
  • sloc: cpp: 6,179,261; ansic: 1,216,205; asm: 741,319; python: 196,614; objc: 75,325; f90: 49,640; lisp: 32,396; pascal: 12,286; sh: 9,394; perl: 7,442; ml: 5,494; awk: 3,523; makefile: 2,723; javascript: 1,206; xml: 886; fortran: 581; cs: 573
file content (46 lines) | stat: -rw-r--r-- 2,281 bytes parent folder | download | duplicates (2)
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
// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
// REQUIRES: amdgpu-registered-target

// RUN: %clang_cc1 -include __clang_hip_runtime_wrapper.h      \
// RUN:   -internal-isystem %S/../../lib/Headers/cuda_wrappers \
// RUN:   -internal-isystem %S/Inputs/include \
// RUN:   -triple amdgcn-amd-amdhsa -aux-triple x86_64-unknown-unknown \
// RUN:   -target-cpu gfx906 -emit-llvm %s -fcuda-is-device -O1 -o - \
// RUN:   -D__HIPCC_RTC__ | FileCheck -check-prefix=DEFAULT %s

// Check that we end up with fast math flags set on intrinsic calls
// RUN: %clang_cc1 -include __clang_hip_runtime_wrapper.h \
// RUN:   -internal-isystem %S/../../lib/Headers/cuda_wrappers \
// RUN:   -internal-isystem %S/Inputs/include \
// RUN:   -triple amdgcn-amd-amdhsa -aux-triple x86_64-unknown-unknown \
// RUN:   -target-cpu gfx906 -emit-llvm %s -fcuda-is-device -O1 -ffinite-math-only -o - \
// RUN:   -D__HIPCC_RTC__ | FileCheck -check-prefix=FINITEONLY %s

// DEFAULT-LABEL: @test_fma_f16(
// DEFAULT-NEXT:  entry:
// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract half @__ocml_fma_f16(half noundef [[X:%.*]], half noundef [[Y:%.*]], half noundef [[Z:%.*]]) #[[ATTR6:[0-9]+]]
// DEFAULT-NEXT:    ret half [[CALL_I]]
//
// FINITEONLY-LABEL: @test_fma_f16(
// FINITEONLY-NEXT:  entry:
// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract half @__ocml_fma_f16(half noundef [[X:%.*]], half noundef [[Y:%.*]], half noundef [[Z:%.*]]) #[[ATTR6:[0-9]+]]
// FINITEONLY-NEXT:    ret half [[CALL_I]]
//
extern "C" __device__ _Float16 test_fma_f16(_Float16 x, _Float16 y,
                                            _Float16 z) {
  return fma(x, y, z);
}

// DEFAULT-LABEL: @test_pow_f16(
// DEFAULT-NEXT:  entry:
// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract half @__ocml_pown_f16(half noundef [[X:%.*]], i32 noundef [[Y:%.*]]) #[[ATTR7:[0-9]+]]
// DEFAULT-NEXT:    ret half [[CALL_I]]
//
// FINITEONLY-LABEL: @test_pow_f16(
// FINITEONLY-NEXT:  entry:
// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract half @__ocml_pown_f16(half noundef [[X:%.*]], i32 noundef [[Y:%.*]]) #[[ATTR7:[0-9]+]]
// FINITEONLY-NEXT:    ret half [[CALL_I]]
//
extern "C" __device__ _Float16 test_pow_f16(_Float16 x, int y) {
  return pow(x, y);
}