b847455bae9adada91a200f6536d23bfe5a97e6a
ECR #304775 - Implement profile driven wave limiter Part 2/2: OCL changes.
Profile driven wave limiter measures kernel execution time in real time and adaptively limits the number of waves per SH. This is to mitigate cache thrashing issues.
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/llvm32/include/llvm/Transforms/IPO/AMDKernelPerfHint.h#2 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm32/lib/Target/AMDIL/AMDILKernelManager.cpp#2 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm32/lib/Transforms/IPO/AMDKernelPerfHint.cpp#2 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#244 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#281 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.hpp#108 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuwavelimiter.cpp#1 add
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuwavelimiter.hpp#1 add
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/include/cal/cal.h#35 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.cpp#71 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.h#44 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#114 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/command.cpp#68 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/command.hpp#74 edit
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#228 edit
[ROCm/clr commit: 493d26db66]
Описание
No description provided
Languages
C++
67.5%
C
20.6%
Python
6.6%
CMake
3.4%
Shell
0.6%
Разное
1.1%