-Move staging buffer locks inside the staging buffer code.
-Remove dedicated per-device completion_signal + per-device lock -
instead allocated signal from the per-stream pool. This elimintes
the lock and allows more concurrency.
-remove switch HIP_DISABLE_BIDIR_MEMCPY
[ROCm/clr commit: e64174f47a]
-Move staging buffer locks inside the staging buffer code.
-Remove dedicated per-device completion_signal + per-device lock -
instead allocated signal from the per-stream pool. This elimintes
the lock and allows more concurrency.
-remove switch HIP_DISABLE_BIDIR_MEMCPY
-Move staging buffer locks inside the staging buffer code.
-Remove dedicated per-device completion_signal + per-device lock -
instead allocated signal from the per-stream pool. This elimintes
the lock and allows more concurrency.
-remove switch HIP_DISABLE_BIDIR_MEMCPY
[ROCm/hip commit: 0af4d3623f]
-Move staging buffer locks inside the staging buffer code.
-Remove dedicated per-device completion_signal + per-device lock -
instead allocated signal from the per-stream pool. This elimintes
the lock and allows more concurrency.
-remove switch HIP_DISABLE_BIDIR_MEMCPY
SWDEV-2 - Change OpenCL version number from 2066 to 2067.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1813 edit
[ROCm/clr commit: 35181ad34b]
SWDEV-2 - Change OpenCL version number from 2066 to 2067.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1813 edit
- refactor staging buffer to operate on hsa* data structures not
hc::accelerator.
- use hsa_memory_allocate to allocate staging buffers rather than
am_alloc.
- Refactor device reset with single member function. Don't reallocate
staging buffers on reset.
- Properly track dependencies based on command type. Add new deps for
H2D and D2D rather than overloading H2D.
[ROCm/clr commit: 3b45e064f9]
- refactor staging buffer to operate on hsa* data structures not
hc::accelerator.
- use hsa_memory_allocate to allocate staging buffers rather than
am_alloc.
- Refactor device reset with single member function. Don't reallocate
staging buffers on reset.
- Properly track dependencies based on command type. Add new deps for
H2D and D2D rather than overloading H2D.
- refactor staging buffer to operate on hsa* data structures not
hc::accelerator.
- use hsa_memory_allocate to allocate staging buffers rather than
am_alloc.
- Refactor device reset with single member function. Don't reallocate
staging buffers on reset.
- Properly track dependencies based on command type. Add new deps for
H2D and D2D rather than overloading H2D.
[ROCm/hip commit: 7d500599fa]
- refactor staging buffer to operate on hsa* data structures not
hc::accelerator.
- use hsa_memory_allocate to allocate staging buffers rather than
am_alloc.
- Refactor device reset with single member function. Don't reallocate
staging buffers on reset.
- Properly track dependencies based on command type. Add new deps for
H2D and D2D rather than overloading H2D.
This option may help debug synchronization or coherency issues
involving the GPU caches. It works only on dGPUs, by changing the
cache policy of the GPUVM default aperture to "cohrent", which is
implemented as non-cached on current dGPU hardware.
Change-Id: I544ac9cc5c0cf1fa5c4e30f67aa42b3b5e44ae67
This option may help debug synchronization or coherency issues
involving the GPU caches. It works only on dGPUs, by changing the
cache policy of the GPUVM default aperture to "cohrent", which is
implemented as non-cached on current dGPU hardware.
Change-Id: I544ac9cc5c0cf1fa5c4e30f67aa42b3b5e44ae67
[ROCm/ROCR-Runtime commit: 06d391c6c9]
SWDEV-2 - Change OpenCL version number from 2065 to 2066.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1812 edit
[ROCm/clr commit: 5a093ba1e9]
SWDEV-2 - Change OpenCL version number from 2065 to 2066.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1812 edit
Still #include staging_buffer.cpp into hip_hcc.cpp.
Directed tests compile hip_hcc to static library and use the library.
[ROCm/clr commit: a1879ba59b]
Still #include staging_buffer.cpp into hip_hcc.cpp.
Directed tests compile hip_hcc to static library and use the library.
[ROCm/hip commit: 28ee7aff71]
Ben Sander
foreman
Boris Fomitchev
Boris Fomitchev
Felix Kuehling
Aditya Atluri