Skip to content

Commit

Permalink
Squashed 'thirdParty/alpaka/' changes from 6854077d6d..106a4975f4
Browse files Browse the repository at this point in the history 
106a4975f4 fix getFunctionAttributes for the SYCL backend
f36e1156af update CUDA version in CI
3f8456973e use inline for CUDA/HIP code when rdc is on, otherwise use static
8b9cc3c557 fix gh-pages jobA
89d5ce671c Ignore VI temporary files
4b7bd17493 Fix the device used by KernelExecutionFixture (ComputationalRadiationPhysics#2344)
2c386dc5e9 Make alpaka follow CMAKE_CUDA_RUNTIME_LIBRARY
2d652dd233 Add thread count to CPU blocks accelerators (ComputationalRadiationPhysics#2338)
dbc5ebe1e9 Fix complex math pow test (ComputationalRadiationPhysics#2336)
4995c5b22a Fix isValidWorkDivKernel to use the correct device
f571ce9197 Remove unnecessary include
a26cdbcd41 Re-enable the KernelNoTemplateGpu test
a9217fb780 Link libcudart even when libcurand is not used
9c8614143b Suppress GCC warning about casting a function to void*
ba169cdc52 Rewrite the getValidWorkDivForKernel tests
948eb757d4 Fix getValidWorkDivForKernel tests for the SYCL CPU backend
f6f94f13b5 Fix getValidWorkDivForKernel tests for the CUDA backend
f612f971a0 Reduce code duplications in matrixMulWithMdSpan (ComputationalRadiationPhysics#2326)
d1cc2e01c1 Add a matrix multiplication example using mdspan
536a183cce Add missing whitespace in enqueue log messages
81d4410eec Reduce code duplication in CUDA/HIP kernel launch
6fdec14904  add remove-restrict
5323600508 CI: improve script utils
01d123e605 fix missing C++20 STL for ICPX in the CI
d254bcd6a3 ctest: display only output of tests, which failed
c9b8c941af change documentation
b9ed742913 remove getValidWorkDiv itself
048ef8afca use getValidWorkDivForKernel in kernelfixture of tests
38805498f0 fix random strategies
4f175420f2 remove getValidWorkDiv first
7f08120428 CI_FILTER: ^linux_nvcc11.*
789344f019 ALPAKA_FN_HOST is not a type
4efdb9dc63 fix explicit instantiation issue
fe4106f88a CI_FILTER: ^linux_nvcc11.*gcc9
e6b4881b70 CI_FILTER: ^linux_nvcc11.*gcc9
e3e760ed9e make conv2dmdspan use kernelbundle
62efffe605 Add getValidWorkDivForKernel function and KernelBundle with tests
690da679bd Let the SYCL queue implement `ConceptCurrentThreadWaitFor`, `ConceptGetDev` and `ConceptQueue` (ComputationalRadiationPhysics#2314)
995c57b54b set alpaka_CXX_STANDARD in the job generator
6ad09baa38 remove nvcc11.0 and nvcc11.1 support (ComputationalRadiationPhysics#2310)
0775f7c066 clang-format and fix typo
18eeeb7b49 move complex declaration to internal namespace
3468d2f8ac add trait IsKernelTriviallyCopyable
3015eae06b update CI container to version 3.2
56c0e416bc Update Catch2 to v3.5.2 (ComputationalRadiationPhysics#2300)

git-subtree-dir: thirdParty/alpaka
git-subtree-split: 106a4975f48dc38cc34f6a2494a3d16774282951
  • Loading branch information
@psychocoderHPC
Third Party authored and psychocoderHPC committed Aug 7, 2024
1 parent fd24dd8 commit 8431571
Show file tree
Hide file tree
Showing 348 changed files with 14,853 additions and 6,020 deletions.
3 changes: 2 additions & 1 deletion .github/workflows/ci.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@ concurrency:
# alpaka_ACC_ANY_BT_OMP5_ENABLE : {ON, OFF}
# [ON] OMP_NUM_THREADS : {1, 2, 3, 4}
# alpaka_ACC_GPU_CUDA_ENABLE : {ON, OFF}
# [ON] ALPAKA_CI_CUDA_VERSION : {11.0, 11.1, 11.2, 11.3, 11.4, 11.5, 11.6}
# [ON] ALPAKA_CI_CUDA_VERSION : {11.2, 11.3, 11.4, 11.5, 11.6, 11.7, 11.8, 12.0, 12.1, 12.2 12.3}
# [ON] CMAKE_CUDA_COMPILER : {nvcc, [CXX==clang++]:clang++}
# alpaka_ACC_CPU_B_TBB_T_SEQ_ENABLE : {ON, OFF}
# alpaka_ACC_GPU_HIP_ENABLE : {ON, OFF}
Expand All @@ -62,6 +62,7 @@ env:
ALPAKA_CI_ONEAPI_VERSION: 2024.0
ALPAKA_CI_TBB_VERSION: 2021.10.0
ALPAKA_CI_RUN_TESTS: ON
alpaka_CXX_STANDARD: 17
alpaka_ACC_CPU_B_SEQ_T_SEQ_ENABLE: ON
alpaka_ACC_CPU_B_SEQ_T_THREADS_ENABLE: ON
alpaka_ACC_CPU_B_TBB_T_SEQ_ENABLE: ON
Expand Down
2 changes: 2 additions & 0 deletions .github/workflows/gh-pages.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,8 @@ on:
push:
branches:
- develop
env:
ALPAKA_CI_OS_NAME: "Linux"

jobs:
gh-pages:
Expand Down
1 change: 1 addition & 0 deletions .gitignore
View file
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@

# tmp files
*~
.*.swp

# netbeans project files
/nbproject/
Expand Down
6 changes: 4 additions & 2 deletions .gitlab-ci.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ variables:
# container version of the generated jobs
# should be merged with ALPAKA_GITLAB_CI_CONTAINER_VERSION
# see: script/job_generator/generate_job_yaml.py
ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION: "3.1"
ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION: "3.2"

generate:
stage: generator
Expand All @@ -27,7 +27,9 @@ generate:
- apk update && apk add python3~=3.11 py3-pip
- pip3 install -r script/job_generator/requirements.txt
# it is sufficient to verify once, as the same job matrix is generated, verified and then filtered each time
- python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --verify --wave compile_only_job -o compile_only.yml
# disable verify because we know that the generator is broken: https://github.com/thombashi/allpairspy/pull/10
#- python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --verify --wave compile_only_job -o compile_only.yml
- python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --wave compile_only_job -o compile_only.yml
- python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --wave runtime_job_cpu -o runtime_cpu.yml
- python3 script/job_generator/job_generator.py ${ALPAKA_GITLAB_CI_GENERATOR_CONTAINER_VERSION} --wave runtime_job_gpu -o runtime_gpu.yml
- cat compile_only.yml
Expand Down
6 changes: 3 additions & 3 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -72,8 +72,8 @@ This library uses C++17 (or newer when available).
| OpenMP 2.0+ threads | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark:[^3] | :white_check_mark: | :white_check_mark: |
| std::thread | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
| TBB | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
| CUDA (nvcc) | :white_check_mark: <br/> (CUDA 11.0 - 12.3)[^2] | :white_check_mark: <br/> (CUDA 11.4 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 12.0 - 12.3) | :x: | :white_check_mark: <br/> (CUDA 11.0-11.2; 11.6 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 11.2, 11.6 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 11.6 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 11.7 - 12.0) | :white_check_mark: <br/> (CUDA 11.8 - 12.0) | :white_check_mark: <br/> (CUDA 12.2) | :white_check_mark: <br/> (CUDA 12.3) | :x: | :x: | :x: | :x: |
| CUDA (clang) | - | - | - | :x: | :x: | :x: | :x: | :x: | :white_check_mark: (CUDA 11.0 - 11.5) | :white_check_mark: (CUDA 11.0 - 11.5)[^1] | :white_check_mark: (CUDA 11.0 - 11.5)[^1] | :white_check_mark: (CUDA 11.0 - 11.8)[^1] | :x: | - | - |
| CUDA (nvcc) | :white_check_mark: <br/> (CUDA 11.2 - 12.5)[^2] | :white_check_mark: <br/> (CUDA 11.4 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 12.0 - 12.5) | :white_check_mark: <br/> (CUDA 12.4 - 12.5) | :white_check_mark: <br/> (11.6 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 11.2, 11.6 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 11.6 - 12.0)[^2] | :white_check_mark: <br/> (CUDA 11.7 - 12.0) | :white_check_mark: <br/> (CUDA 11.8 - 12.0) | :white_check_mark: <br/> (CUDA 12.2) | :white_check_mark: <br/> (CUDA 12.3) | :white_check_mark: <br/> (CUDA 12.4 - 12.5) | :x: | :x: | :x: |
| CUDA (clang) | - | - | - | :x: | :x: | :x: | :x: | :x: | :white_check_mark: (CUDA 11.2 - 11.5) | :white_check_mark: (CUDA 11.2 - 11.5)[^1] | :white_check_mark: (CUDA 11.2 - 11.5)[^1] | :white_check_mark: (CUDA 11.2 - 11.8)[^1] | :x: | - | - |
| [HIP](https://alpaka.readthedocs.io/en/latest/install/HIP.html) (clang) | - | - | - | :x: | :x: | :x: | :x: | :x: | :white_check_mark: (HIP 5.1 - 5.2) | :white_check_mark: (HIP 5.3 - 5.4) | :white_check_mark: (HIP 5.5 - 5.6) | :white_check_mark: (HIP 5.7 - 6.0) | :x: | - | - |
| SYCL | :x: | :x: | :x: | :x: | :x: | :x: | :x: | :x: | :x: | :x: | :x: | :x: | :white_check_mark:[^4] | :x: | :x: |

Expand All @@ -91,7 +91,7 @@ Dependencies
The **alpaka** library itself just requires header-only libraries.
However some of the accelerator back-end implementations require different boost libraries to be built.

When an accelerator back-end using *CUDA* is enabled, version *11.0* (with nvcc as CUDA compiler) or version *9.2* (with clang as CUDA compiler) of the *CUDA SDK* is the minimum requirement.
When an accelerator back-end using *CUDA* is enabled, version *11.2* (with nvcc as CUDA compiler) or version *11.2* (with clang as CUDA compiler) of the *CUDA SDK* is the minimum requirement.
*NOTE*: When using clang as a native *CUDA* compiler, the *CUDA accelerator back-end* can not be enabled together with any *OpenMP accelerator back-end* because this combination is currently unsupported.
*NOTE*: Separable compilation is disabled by default and can be enabled via the CMake flag `CMAKE_CUDA_SEPARABLE_COMPILATION`.

Expand Down
13 changes: 6 additions & 7 deletions benchmarks/babelstream/src/AlpakaStream.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -50,7 +50,7 @@ template<typename T>
void AlpakaStream<T>::init_arrays(T initA, T initB, T initC)
{
auto const workdiv = WorkDiv{arraySize / blockSize, blockSize, 1};
// auto const workdiv = alpaka::getValidWorkDiv(devAcc, arraySize);

alpaka::exec<
Acc>(queue, workdiv, InitKernel{}, std::data(d_a), std::data(d_b), std::data(d_c), initA, initB, initC);
alpaka::wait(queue);
Expand Down Expand Up @@ -78,7 +78,7 @@ template<typename T>
void AlpakaStream<T>::copy()
{
auto const workdiv = WorkDiv{arraySize / blockSize, blockSize, 1};
// auto const workdiv = alpaka::getValidWorkDiv(devAcc, arraySize);

alpaka::exec<Acc>(queue, workdiv, CopyKernel{}, std::data(d_a), std::data(d_c));
alpaka::wait(queue);
}
Expand All @@ -98,7 +98,7 @@ template<typename T>
void AlpakaStream<T>::mul()
{
auto const workdiv = WorkDiv{arraySize / blockSize, blockSize, 1};
// auto const workdiv = alpaka::getValidWorkDiv(devAcc, arraySize);

alpaka::exec<Acc>(queue, workdiv, MulKernel{}, std::data(d_b), std::data(d_c));
alpaka::wait(queue);
}
Expand All @@ -117,7 +117,7 @@ template<typename T>
void AlpakaStream<T>::add()
{
auto const workdiv = WorkDiv{arraySize / blockSize, blockSize, 1};
// auto const workdiv = alpaka::getValidWorkDiv(devAcc, arraySize);

alpaka::exec<Acc>(queue, workdiv, AddKernel{}, std::data(d_a), std::data(d_b), std::data(d_c));
alpaka::wait(queue);
}
Expand All @@ -137,7 +137,7 @@ template<typename T>
void AlpakaStream<T>::triad()
{
auto const workdiv = WorkDiv{arraySize / blockSize, blockSize, 1};
// auto const workdiv = alpaka::getValidWorkDiv(devAcc, arraySize);

alpaka::exec<Acc>(queue, workdiv, TriadKernel{}, std::data(d_a), std::data(d_b), std::data(d_c));
alpaka::wait(queue);
}
Expand All @@ -157,7 +157,7 @@ template<typename T>
void AlpakaStream<T>::nstream()
{
auto const workdiv = WorkDiv{arraySize / blockSize, blockSize, 1};
// auto const workdiv = alpaka::getValidWorkDiv(devAcc, arraySize);

alpaka::exec<Acc>(queue, workdiv, NstreamKernel{}, std::data(d_a), std::data(d_b), std::data(d_c));
alpaka::wait(queue);
}
Expand Down Expand Up @@ -197,7 +197,6 @@ template<typename T>
auto AlpakaStream<T>::dot() -> T
{
auto const workdiv = WorkDiv{dotBlockSize, blockSize, 1};
// auto const workdiv = alpaka::getValidWorkDiv(devAcc, dotBlockSize * blockSize);
alpaka::exec<Acc>(queue, workdiv, DotKernel{}, std::data(d_a), std::data(d_b), std::data(d_sum), arraySize);
alpaka::wait(queue);

Expand Down
17 changes: 16 additions & 1 deletion cmake/alpakaCommon.cmake
View file
Original file line number Diff line number Diff line change
Expand Up @@ -470,6 +470,10 @@ if(alpaka_ACC_GPU_CUDA_ENABLE)
elseif(CMAKE_CUDA_COMPILER_ID STREQUAL "NVIDIA")
message(STATUS "nvcc is used as CUDA compiler")

if(alpaka_CXX_STANDARD GREATER_EQUAL 20 AND CMAKE_VERSION VERSION_LESS "3.25.0")
message(FATAL_ERROR "CMake 3.24 and older does not support C++20 for nvcc")
endif()

# nvcc sets no linux/__linux macros on OpenPOWER linux
# nvidia bug id: 2448610
if(CMAKE_SYSTEM_NAME STREQUAL "Linux")
Expand Down Expand Up @@ -538,9 +542,20 @@ if(alpaka_ACC_GPU_CUDA_ENABLE)
endif()
endif()

# Link the CUDA Runtime library
if(CMAKE_CUDA_RUNTIME_LIBRARY STREQUAL "Shared")
target_link_libraries(alpaka INTERFACE CUDA::cudart)
else()
target_link_libraries(alpaka INTERFACE CUDA::cudart_static)
endif()

if(NOT alpaka_DISABLE_VENDOR_RNG)
# Use cuRAND random number generators
target_link_libraries(alpaka INTERFACE CUDA::cudart CUDA::curand)
if(CMAKE_CUDA_RUNTIME_LIBRARY STREQUAL "Shared")
target_link_libraries(alpaka INTERFACE CUDA::curand)
else()
target_link_libraries(alpaka INTERFACE CUDA::curand_static)
endif()
endif()
else()
message(FATAL_ERROR "Optional alpaka dependency CUDA could not be found!")
Expand Down
9 changes: 8 additions & 1 deletion docs/source/basic/cheatsheet.rst
View file
Original file line number Diff line number Diff line change
Expand Up @@ -176,15 +176,22 @@ Enqueue a memory copy from device to host
Kernel Execution
----------------
Prepare Kernel Bundle
.. code-block:: c++

HeatEquationKernel heatEqKernel;
// Arguments of KernelBundle: The kernel instance and the kernel arguments
auto const& bundeledKernel = alpaka::KernelBundle(heatEqKernel, pCurrAcc, pNextAcc, numNodesX, dx, dt);

Automatically select a valid kernel launch configuration
.. code-block:: c++

Vec<Dim, Idx> const globalThreadExtent = vectorValue;
Vec<Dim, Idx> const elementsPerThread = vectorValue;

auto autoWorkDiv = getValidWorkDiv<Acc>(
auto autoWorkDiv = getValidWorkDivForKernel<Acc>(
device,
bundeledKernel,
globalThreadExtent, elementsPerThread,
false,
GridBlockExtentSubDivRestrictions::Unrestricted);
Expand Down
1 change: 1 addition & 0 deletions example/CMakeLists.txt
View file
Original file line number Diff line number Diff line change
Expand Up @@ -26,6 +26,7 @@ add_subdirectory("helloWorld/")
add_subdirectory("helloWorldLambda/")
add_subdirectory("kernelSpecialization/")
add_subdirectory("ls/")
add_subdirectory("matrixMulWithMdspan/")
add_subdirectory("monteCarloIntegration/")
add_subdirectory("openMPSchedule/")
add_subdirectory("parallelLoopPatterns/")
Expand Down
25 changes: 10 additions & 15 deletions example/bufferCopy/src/bufferCopy.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -106,19 +106,6 @@ auto example(TAccTag const&) -> int
using Vec = alpaka::Vec<Dim, Idx>;
Vec const elementsPerThread(Vec::all(static_cast<Idx>(1)));
Vec const threadsPerGrid(Vec::all(static_cast<Idx>(10)));
using WorkDiv = alpaka::WorkDivMembers<Dim, Idx>;
WorkDiv const devWorkDiv = alpaka::getValidWorkDiv<Acc>(
devAcc,
threadsPerGrid,
elementsPerThread,
false,
alpaka::GridBlockExtentSubDivRestrictions::Unrestricted);
WorkDiv const hostWorkDiv = alpaka::getValidWorkDiv<Host>(
devHost,
threadsPerGrid,
elementsPerThread,
false,
alpaka::GridBlockExtentSubDivRestrictions::Unrestricted);

// Create host and device buffers
//
Expand Down Expand Up @@ -177,10 +164,13 @@ auto example(TAccTag const&) -> int

FillBufferKernel fillBufferKernel;

auto const& bundeledFillBufferKernel = alpaka::KernelBundle(fillBufferKernel, hostViewPlainPtrMdSpan);
auto const hostWorkDiv
= alpaka::getValidWorkDivForKernel<Host>(devHost, bundeledFillBufferKernel, threadsPerGrid, elementsPerThread);

alpaka::exec<Host>(hostQueue, hostWorkDiv, fillBufferKernel,
hostViewPlainPtrMdSpan); // 1st kernel argument


// Copy host to device Buffer
//
// A copy operation of one buffer into
Expand Down Expand Up @@ -213,10 +203,15 @@ auto example(TAccTag const&) -> int
auto deviceBufferMdSpan2 = alpaka::experimental::getMdSpan(deviceBuffer2);

TestBufferKernel testBufferKernel;
auto const& bundeledTestBufferKernel = alpaka::KernelBundle(testBufferKernel, deviceBufferMdSpan1);

// Let alpaka calculate good block and grid sizes given our full problem extent
auto const devWorkDiv
= alpaka::getValidWorkDivForKernel<Acc>(devAcc, bundeledTestBufferKernel, threadsPerGrid, elementsPerThread);

alpaka::exec<Acc>(devQueue, devWorkDiv, testBufferKernel, deviceBufferMdSpan1);
alpaka::exec<Acc>(devQueue, devWorkDiv, testBufferKernel, deviceBufferMdSpan2);


// Print device Buffer
//
// Because we really like to flood our
Expand Down
15 changes: 8 additions & 7 deletions example/complex/src/complex.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -55,15 +55,16 @@ auto example(TAccTag const&) -> int
// Define the work division
Idx const threadsPerGrid = 1u;
Idx const elementsPerThread = 1u;
auto const workDiv = alpaka::getValidWorkDiv<Acc>(
devAcc,
threadsPerGrid,
elementsPerThread,
false,
alpaka::GridBlockExtentSubDivRestrictions::Unrestricted);

ComplexKernel complexKernel;

auto const& bundeledKernel = alpaka::KernelBundle(complexKernel);
// Let alpaka calculate good block and grid sizes given our full problem extent
auto const workDiv
= alpaka::getValidWorkDivForKernel<Acc>(devAcc, bundeledKernel, threadsPerGrid, elementsPerThread);

// Run the kernel
alpaka::exec<Acc>(queue, workDiv, ComplexKernel{});
alpaka::exec<Acc>(queue, workDiv, complexKernel);
alpaka::wait(queue);

// Usage of alpaka::Complex<T> on the host side is the same as inside kernels, except math functions are not
Expand Down
19 changes: 11 additions & 8 deletions example/conv2DWithMdspan/src/conv2DWithMdspan.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -129,14 +129,6 @@ auto example(TAccTag const&) -> int
//
auto outputDeviceMemory = alpaka::allocBuf<DataType, Idx>(devAcc, extent);

// Let alpaka calculate good block and grid sizes given our full problem extent.
alpaka::WorkDivMembers<Dim, Idx> const workDiv(alpaka::getValidWorkDiv<DevAcc>(
devAcc,
extent,
Vec::ones(),
false,
alpaka::GridBlockExtentSubDivRestrictions::Unrestricted));

// Prepare convolution filter at host
//
std::vector<DataType> const filter = {0.11, 0.12, 0.13, 0.14, 0.15, 0.21, 0.22, 0.23, 0.24, 0.25, 0.31, 0.32, 0.33,
Expand All @@ -155,6 +147,17 @@ auto example(TAccTag const&) -> int
// Construct kernel object
ConvolutionKernelMdspan2D convolutionKernel2D;

// Make a bundle
auto const& bundeledKernel = alpaka::KernelBundle(
convolutionKernel2D,
alpaka::experimental::getMdSpan(bufInputAcc),
alpaka::experimental::getMdSpan(outputDeviceMemory),
alpaka::experimental::getMdSpan(bufFilterAcc));

// Let alpaka calculate good block and grid sizes given our full problem extent.
auto const workDiv = alpaka::getValidWorkDivForKernel<DevAcc>(devAcc, bundeledKernel, extent, Vec::ones());


// Run the kernel, pass 3 arrays as 2D mdspans
alpaka::exec<DevAcc>(
queueAcc,
Expand Down
Loading

0 comments on commit 8431571

Please sign in to comment.