- Bug fix for peer visibility. Now contexts correctly detect when they can use SDMA for P2P vs staging buffers.
- Interface to new HCC copy_ext function.
- Improve context and peer print /debug options.
- Add comments and usage to hipPeerToPeer_simple test.
[ROCm/clr commit: 944d4eaa18]
Code simplification/cleanup:
Remove stale fallback paths that uses something besides the unpinned engine.
Remove HIP_STAGING_BUFFER env var - now is const 2, 0 no longer has
special meaning.
Change-Id: I7d24cdd1067dd0c244e87b6a83897cb135d307e7
[ROCm/clr commit: 39750da9a5]
- Enable -O3 for hipDispatchLatency.
- Use nearly-null kernel to prevent it from being optimized away.
- Formatting for hipDispatchLatency.
- Formatting for hipInfo.
[ROCm/clr commit: e8beacbfb1]
SWDEV-2 - Change OpenCL version number from 2221 to 2222.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1968 edit
[ROCm/clr commit: 4dd808cd6d]
SWDEV-102540 - Fill the device info_.boardName_ field from the info returned by ROCR's HSA_AMD_AGENT_INFO_PRODUCT_NAME query.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/roccompiler.cpp#19 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.cpp#19 edit
[ROCm/clr commit: 0dd6aae926]
SWDEV-102971 - [CQE OCL][OpenCL on PAL] "computer on low memory" issue observed folllowed by tdr while running few WF Conformance tests together
- Use custom memory allocator to avoid memory grow in PAL
- Extra clean-ups in resource alloc, including a fix for image pitch in the persistent memory
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palmemory.cpp#6 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.cpp#11 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palresource.hpp#7 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.cpp#22 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.hpp#9 edit
[ROCm/clr commit: 48590452f1]
SWDEV-2 - Change OpenCL version number from 2220 to 2221.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1967 edit
[ROCm/clr commit: fcc6d6637b]
SWDEV-94610 - Add gfx700 to the list of suported targets in HSAILProgram::linkImpl_LC. When dumping the source (-save-temps), print the options actually sent to clang as well as the options passed to OpenCL.
Affected files ...
... //depot/stg/opencl/drivers/opencl/library/build/Makefile.library#56 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/amdgpu_metadata.cpp#5 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/roccompiler.cpp#18 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocprogram.cpp#36 edit
[ROCm/clr commit: 383e97425b]
SWDEV-2 - Change OpenCL version number from 2219 to 2220.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1966 edit
[ROCm/clr commit: e9c1fb06c7]
SWDEV-2 - Change OpenCL version number from 2218 to 2219.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1965 edit
[ROCm/clr commit: f6529ae04a]
SWDEV-2 - Change OpenCL version number from 2217 to 2218.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1964 edit
[ROCm/clr commit: 34e0d956ec]
SWDEV-96479 - changing tab to space as Victor pointed out
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.cpp#18 edit
[ROCm/clr commit: c2b57553dc]
SWDEV-96479 - bug fix, exposing atomics capabilities based on GFX IP. For Hawaii the atomics is not supported hence disabled.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.cpp#17 edit
[ROCm/clr commit: d793001561]
SWDEV-102417 - Forum [205433] : Memory leak with printf statement inside kernel code
A memory leak can occur if a printf statement is inside the .cl source code but it is not used inside the __kernel code (e.g., a function inside .cl code that uses printf but never called by the __kernel). In this case compiler generates the printf metadata but printf is not used by the __kernel (i.e., the printf buffer is empty).
To fix this issue, release the transfer buffer object before returning false in PrintfDbgHSA::output function.
ReviewBoardURL = http://ocltc.amd.com/reviews/r/11394/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuprintf.cpp#42 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palprintf.cpp#3 edit
[ROCm/clr commit: 0dd7a03954]
foreman
Ben Sander
Aditya Atluri
Maneesh Gupta
Rahul Garg
pensun