Skip to content

[CUDA] Remove obsolete GPU-side __constexpr_* wrappers. #139164

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
May 12, 2025
Merged

[CUDA] Remove obsolete GPU-side __constexpr_* wrappers. #139164

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
50 changes: 0 additions & 50 deletions clang/lib/Headers/cuda_wrappers/cmath
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,56 +39,6 @@
__attribute__((device)) long double logb(long double);
__attribute__((device)) long double scalbn(long double, int);

namespace std {

// For __constexpr_fmin/fmax we only need device-side overloads before c++14
// where they are not constexpr.
#if _LIBCPP_STD_VER < 14

__attribute__((device))
inline _LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX14 float __constexpr_fmax(float __x, float __y) _NOEXCEPT {
return __builtin_fmaxf(__x, __y);
}

__attribute__((device))
inline _LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX14 double __constexpr_fmax(double __x, double __y) _NOEXCEPT {
return __builtin_fmax(__x, __y);
}

__attribute__((device))
inline _LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX14 long double
__constexpr_fmax(long double __x, long double __y) _NOEXCEPT {
return __builtin_fmaxl(__x, __y);
}

template <class _Tp, class _Up, __enable_if_t<is_arithmetic<_Tp>::value && is_arithmetic<_Up>::value, int> = 0>
__attribute__((device))
_LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX14 typename __promote<_Tp, _Up>::type
__constexpr_fmax(_Tp __x, _Up __y) _NOEXCEPT {
using __result_type = typename __promote<_Tp, _Up>::type;
return std::__constexpr_fmax(static_cast<__result_type>(__x), static_cast<__result_type>(__y));
}
#endif // _LIBCPP_STD_VER < 14

// For logb/scalbn templates we must always provide device overloads because
// libc++ implementation uses __builtin_XXX which gets translated into a libcall
// which we can't handle on GPU. We need to forward those to CUDA-provided
// implementations.

template <class _Tp>
__attribute__((device))
_LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX14 _Tp __constexpr_logb(_Tp __x) {
return ::logb(__x);
}

template <class _Tp>
__attribute__((device))
_LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX20 _Tp __constexpr_scalbn(_Tp __x, int __exp) {
return ::scalbn(__x, __exp);
}

} // namespace std//

#endif // _LIBCPP_STD_VER

#endif // include guard