Closed fwyzard closed 2 years ago
Compiled with, for example
make -j`nproc` alpakatest CUDA_BASE=/usr/local/cuda-11.5.2 CUDA_ARCH=61 ROCM_BASE=/opt/rocm-5.0.2
the rsulting alpakatest
binary supports --serial
, --tbb
, --cuda
and --hip
.
(I accidentally deleted this branch, I re-opened a new PR at #357)
Support serial, TBB, CUDA and ROCm at the same time.
Split compilation by backend:
ALPAKA_..._ENABLED
macros are only defined one at a time;ALPAKA_..._PRESENT
macros to identify all backends for which support is being compiled;Add forward declaration for alpaka templates and types (thanks to Matti for the idea). Add explicit instantiation definitions and declarations to the initialisation code, and move it to the AlpakaCore "portable" library. Use new pinned host memory functionality, introduced in the latest alpaka update.
Update alpaka the
fwyzard/develop
private branch, pending integration upstream. Relevant changes include: