Ben Sander da37035a9c Improve async copy implementation. ...

- Add device-side signal waits when transitioning between command classes
(Kernel, H2D copy, D2H copy).
- Support waiting in staged memory copies as well.
- Add several chicken bits to control implementation:
    - HIP_DISABLE_ENQ_BARRIER
    - HIP_DISABLE_BIDIR_MEMCPY
    - HIP_ONESHOT_COPY_DEP
- Refactor signal pool to support efficient deallocation based on
signsequnm.
- Deallocate copy signals on eventSynchronize.
- Improve copy tests, add pingpong.


[ROCm/clr commit: 28990567fb]

2016-02-22 23:15:24 -06:00

projects/clr/hipamd

Improve async copy implementation.

2016-02-22 23:15:24 -06:00

S

描述

无详细信息

282 MiB

语言

C++ 67.5%

C 20.6%

Python 6.6%

CMake 3.4%

Shell 0.6%

其它 1.1%