- [x] fix compilation errors
CUDA
BenjaminW3
All 3 comments
For my application, simply changing auto currentTaskPackage(std::shared_ptr<ITaskPkg>{nullptr}); to auto currentTaskPackage = std::shared_ptr<ITaskPkg>{nullptr}; in include/alpaka/core/ConcurrentExecPool.hpp in lines 342, 429, 550, 642 did the trick. Compilation with CUDA 10.1 and gcc-8 works for me with those changes. I don麓t know whether or not this fixes all compilation errors for all use cases/tests, but it might help.
MarkusVelten
on 15 Apr 2019
👍4
Thanks, I will try this!
BenjaminW3
on 15 Apr 2019
Works, thanks!
BenjaminW3
on 17 Apr 2019
Was this page helpful?
0 / 5 - 0 ratings
Related issues
psychocoderHPC
路
5Comments
mxmlnkn
路
5Comments
BenjaminW3
路
6Comments
ax3l
路
3Comments
BenjaminW3
路
5Comments
Most helpful comment
For my application, simply changing
auto currentTaskPackage(std::shared_ptr<ITaskPkg>{nullptr});toauto currentTaskPackage = std::shared_ptr<ITaskPkg>{nullptr};ininclude/alpaka/core/ConcurrentExecPool.hppin lines 342, 429, 550, 642 did the trick. Compilation with CUDA 10.1 and gcc-8 works for me with those changes. I don麓t know whether or not this fixes all compilation errors for all use cases/tests, but it might help.