Skip to content
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

Jump to bottom

[Backport branch/2.3.x] Fix C++11 support of recently added tests #658

Merged
merged 3 commits into from
Nov 8, 2023
Merged

[Backport branch/2.3.x] Fix C++11 support of recently added tests #658

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
47cc55c
Fix static_assert issue in memcpy_async
ahendriksen Nov 2, 2023
e2f4ca0
Skip cp_async_bulk* tests pre-C++14
ahendriksen Nov 2, 2023
4d28095
Address review comments
miscco Nov 3, 2023
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
1 change: 1 addition & 0 deletions libcudacxx/.upstream-tests/test/cuda/barrier/cp_async_bulk_tensor_1d.pass.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
//
//===----------------------------------------------------------------------===//
//
// UNSUPPORTED: c++11
// UNSUPPORTED: libcpp-has-no-threads
// UNSUPPORTED: pre-sm-90
// UNSUPPORTED: nvrtc
Expand Down
1 change: 1 addition & 0 deletions libcudacxx/.upstream-tests/test/cuda/barrier/cp_async_bulk_tensor_2d.pass.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
//
//===----------------------------------------------------------------------===//
//
// UNSUPPORTED: c++11
// UNSUPPORTED: libcpp-has-no-threads
// UNSUPPORTED: pre-sm-90
// UNSUPPORTED: nvrtc
Expand Down
1 change: 1 addition & 0 deletions libcudacxx/.upstream-tests/test/cuda/barrier/cp_async_bulk_tensor_3d.pass.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
//
//===----------------------------------------------------------------------===//
//
// UNSUPPORTED: c++11
// UNSUPPORTED: libcpp-has-no-threads
// UNSUPPORTED: pre-sm-90
// UNSUPPORTED: nvrtc
Expand Down
1 change: 1 addition & 0 deletions libcudacxx/.upstream-tests/test/cuda/barrier/cp_async_bulk_tensor_4d.pass.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
//
//===----------------------------------------------------------------------===//
//
// UNSUPPORTED: c++11
// UNSUPPORTED: libcpp-has-no-threads
// UNSUPPORTED: pre-sm-90
// UNSUPPORTED: nvrtc
Expand Down
1 change: 1 addition & 0 deletions libcudacxx/.upstream-tests/test/cuda/barrier/cp_async_bulk_tensor_5d.pass.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
//
//===----------------------------------------------------------------------===//
//
// UNSUPPORTED: c++11
// UNSUPPORTED: libcpp-has-no-threads
// UNSUPPORTED: pre-sm-90
// UNSUPPORTED: nvrtc
Expand Down
10 changes: 10 additions & 0 deletions libcudacxx/include/cuda/std/detail/libcxx/include/__cuda/barrier.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -943,7 +943,13 @@ template <size_t _Copy_size>
inline __device__
void __cp_async_shared_global(char * __dest, const char * __src) {
// https://docs.nvidia.com/cuda/parallel-thread-execution/index.html#data-movement-and-conversion-instructions-cp-async

// If `if constexpr` is not available, this function gets instantiated even
// if is not called. Do not static_assert in that case.
#if _LIBCUDACXX_STD_VER >= 17
static_assert(_Copy_size == 4 || _Copy_size == 8 || _Copy_size == 16, "cp.async.shared.global requires a copy size of 4, 8, or 16.");
#endif // _LIBCUDACXX_STD_VER >= 17

asm volatile(
"cp.async.ca.shared.global [%0], [%1], %2, %2;"
:
Expand All @@ -970,7 +976,11 @@ void __cp_async_shared_global<16>(char * __dest, const char * __src) {
template <size_t _Alignment, typename _Group>
inline __device__
void __cp_async_shared_global_mechanism(_Group __g, char * __dest, const char * __src, _CUDA_VSTD::size_t __size) {
// If `if constexpr` is not available, this function gets instantiated even
// if is not called. Do not static_assert in that case.
#if _LIBCUDACXX_STD_VER >= 17
static_assert(4 <= _Alignment, "cp.async requires at least 4-byte alignment");
#endif // _LIBCUDACXX_STD_VER >= 17

// Maximal copy size is 16.
constexpr int __copy_size = (_Alignment > 16) ? 16 : _Alignment;
Expand Down
Loading