Fix 'fpclassify: ambiguous call' compile fail in MSVC 2022 (#2426)

Similar to #2219, we see "'fpclassify': ambiguous call" error in
test_conformance\basic\test_fpmath.cpp
due to missing constexpr at
https://github.com/KhronosGroup/OpenCL-CTS/blob/9265cbb2c274/test_conformance/basic/test_fpmath.cpp#L104
This PR fixes the issue by moving utility function isnan_fp in
testHarness.h and use it.
Note this PR doesn't modify use of isnan in many tests where only
float/double values are checked.
This commit is contained in:
Wenju He
2025-08-06 00:08:04 +08:00
committed by GitHub
parent 9ca0126c54
commit e15c6eb760
12 changed files with 94 additions and 107 deletions

View File

@@ -0,0 +1,35 @@
//
// Copyright (c) 2025 The Khronos Group Inc.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
#ifndef _mathHelpers_h
#define _mathHelpers_h
#if defined(__APPLE__)
#include <OpenCL/cl_platform.h>
#else
#include <CL/cl_platform.h>
#endif
#include <cmath>
template <typename T> inline bool isnan_fp(const T &v) { return std::isnan(v); }
template <> inline bool isnan_fp<cl_half>(const cl_half &v)
{
uint16_t h_exp = (((cl_half)v) >> (CL_HALF_MANT_DIG - 1)) & 0x1F;
uint16_t h_mant = ((cl_half)v) & 0x3FF;
return (h_exp == 0x1F && h_mant != 0);
}
#endif // _mathHelpers_h