SWDEV-77584 - ORCA RT: Preparations for enabling HSAIL on OpenCL 1.2 by default. Set Blits to be always internal kernels.
[Synopsis] Option -cl-internal-kernel has not been always set for blits, which leads to difficulty/impossibility for their determination.
[Testing] pre check-in
[Reviewer] German Andryeyev
http://ocltc.amd.com/reviews/r/8799/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/device.cpp#187 edit
SWDEV-77174 - Change compiler lib def and sc lib def to be consistent.
SC interface depends on some compile options which need to be consistent between compiler library and SC.
This change will add the SC interface dependent build options to opencldefs so that they are shared between compiler lib def and sc lib def.
Both compilier lib def and sc lib def will include scinterfacedefs, which will generate consistent compile options required by sc interface based on variables defined in opencldefs.
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/complibdefs#45 edit
... //depot/stg/opencl/drivers/opencl/opencldefs#163 edit
[ROCm/clr commit: 266e0e3507]
SWDEV-77174 - Change compiler lib def and sc lib def to be consistent.
SC interface depends on some compile options which need to be consistent between compiler library and SC.
This change will add the SC interface dependent build options to opencldefs so that they are shared between compiler lib def and sc lib def.
Both compilier lib def and sc lib def will include scinterfacedefs, which will generate consistent compile options required by sc interface based on variables defined in opencldefs.
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/complibdefs#45 edit
... //depot/stg/opencl/drivers/opencl/opencldefs#163 edit
SWDEV-2 - Change OpenCL version number from 1924 to 1925.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1671 edit
[ROCm/clr commit: 5281d48d5c]
SWDEV-2 - Change OpenCL version number from 1924 to 1925.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1671 edit
SWDEV-2 - Change OpenCL version number from 1923 to 1924.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1670 edit
[ROCm/clr commit: c34973291a]
SWDEV-2 - Change OpenCL version number from 1923 to 1924.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1670 edit
SWDEV-2 - Change OpenCL version number from 1922 to 1923.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1669 edit
[ROCm/clr commit: 9d091b83a4]
SWDEV-2 - Change OpenCL version number from 1922 to 1923.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1669 edit
hsaKmtGetNodeMemoryProperties -
- Return only HSA_HEAPTYPE_SYSTEM memory for CPU only node.
- For dGPU remove redundant HSA_HEAPTYPE_FRAME_BUFFER_PRIVATE
entry.
Change-Id: I0349be39b8409a0fd64a038b8b2956191356d937
hsaKmtGetNodeMemoryProperties -
- Return only HSA_HEAPTYPE_SYSTEM memory for CPU only node.
- For dGPU remove redundant HSA_HEAPTYPE_FRAME_BUFFER_PRIVATE
entry.
Change-Id: I0349be39b8409a0fd64a038b8b2956191356d937
[ROCm/ROCR-Runtime commit: f885e551aa]
Unify fmm_get_aperture_base and fmm_get_aperture_limit into one
function. Make the return value to HSAKMT_STATUS.
Change-Id: I0b3f563ffb268947ab891f4935f61788d0af0e01
Unify fmm_get_aperture_base and fmm_get_aperture_limit into one
function. Make the return value to HSAKMT_STATUS.
Change-Id: I0b3f563ffb268947ab891f4935f61788d0af0e01
[ROCm/ROCR-Runtime commit: cb53548c89]
SWDEV-78467 - Back out changelist 1203122
was SWDEV-78467 - Enable the LF library build on Windows 32-bit
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/Makefile#3 edit
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/build/Makefile.api#123 edit
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_lqdflash_amd.cpp#4 edit
SWDEV-2 - Change OpenCL version number from 1921 to 1922.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1668 edit
[ROCm/clr commit: ba3f10d5e2]
SWDEV-2 - Change OpenCL version number from 1921 to 1922.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1668 edit
hsaKmtAllocMemory only allocates aligned address space and sets up
the scratch_physical aperture to match the allocated address space.
Actual allocation of backing memory happens in hsaKmtMapMemoryToGPU.
Change-Id: Ie709815ab9bedb3d682e096b4005fdfb5e94d3a7
hsaKmtAllocMemory only allocates aligned address space and sets up
the scratch_physical aperture to match the allocated address space.
Actual allocation of backing memory happens in hsaKmtMapMemoryToGPU.
Change-Id: Ie709815ab9bedb3d682e096b4005fdfb5e94d3a7
[ROCm/ROCR-Runtime commit: 5131ab4e64]
SWDEV-2 - Change OpenCL version number from 1920 to 1921.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1667 edit
[ROCm/clr commit: a56ada96ac]
SWDEV-2 - Change OpenCL version number from 1920 to 1921.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1667 edit
This is a hack to allow the Runtime to allocate system memory with
PreferredNode=0 on a dGPU system. We allocated it from Node 1
instead so that the node 1 GPU can map the memory. A proper fix
will be implemented together with multi-GPU support.
Change-Id: Ieb52599e5275781c04ee34405ea850bf782c523a
This is a hack to allow the Runtime to allocate system memory with
PreferredNode=0 on a dGPU system. We allocated it from Node 1
instead so that the node 1 GPU can map the memory. A proper fix
will be implemented together with multi-GPU support.
Change-Id: Ieb52599e5275781c04ee34405ea850bf782c523a
[ROCm/ROCR-Runtime commit: 590c8e522c]
SWDEV-79399 - OpenCL printf does not print correctly when the printf builtin function is called twice - clear the local printf info each time
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#305 edit
[ROCm/clr commit: 6800b9f003]
SWDEV-79399 - OpenCL printf does not print correctly when the printf builtin function is called twice - clear the local printf info each time
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#305 edit
SWDEV-2 - Change OpenCL version number from 1919 to 1920.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1666 edit
[ROCm/clr commit: 12704edd82]
SWDEV-2 - Change OpenCL version number from 1919 to 1920.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1666 edit
SWDEV-79151 - clenqueuereadImage is slow when using a pinned buffer and a row_picth!0
- Add a check if the provided rowPitch is equal to the actual transfer width. SDMA doesn't support row/slice pitches, thus runtime still has to fall back to compute in other cases
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_memobj.cpp#78 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#120 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.cpp#122 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.hpp#92 edit
[ROCm/clr commit: 69374d1cd9]
SWDEV-79151 - clenqueuereadImage is slow when using a pinned buffer and a row_picth!0
- Add a check if the provided rowPitch is equal to the actual transfer width. SDMA doesn't support row/slice pitches, thus runtime still has to fall back to compute in other cases
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_memobj.cpp#78 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#120 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.cpp#122 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.hpp#92 edit
SWDEV-79151 - clenqueuereadImage is slow when using a pinned buffer and a row_picth!0
- Add a check if the provided rowPitch is equal to the actual transfer width. SDMA doesn't support row/slice pitches, thus runtime still has to fall back to compute in other cases
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_memobj.cpp#78 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#120 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.cpp#122 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.hpp#92 edit
[ROCm/clr commit: 922e14c46d]
SWDEV-79151 - clenqueuereadImage is slow when using a pinned buffer and a row_picth!0
- Add a check if the provided rowPitch is equal to the actual transfer width. SDMA doesn't support row/slice pitches, thus runtime still has to fall back to compute in other cases
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_memobj.cpp#78 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#120 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.cpp#122 edit
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.hpp#92 edit
SWDEV-79077 - Disable >4GB large single allocation for SI because SI is using AMDIL path which doesn't not support 64bits NDRange yet.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusettings.cpp#332 edit
[ROCm/clr commit: b817bb5898]
SWDEV-79077 - Disable >4GB large single allocation for SI because SI is using AMDIL path which doesn't not support 64bits NDRange yet.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusettings.cpp#332 edit
foreman
Harish Kasiviswanathan
Felix Kuehling
Ben Goz