ECR #333753 - RT: Bug 10185 fix
[Bug] "Failure in 1.2 conformance sub-test 'program_binary_type' of 'compiler' and sub-test 'binary_create' of 'api' when forced to run in HSAIL path".
P.S. Failure also occurs in 2.0 conformance above-stated tests.
[Problem] HSAIL text is inserted by Compiler Lib into the BIF as a symbol of .cg section, but RT extracts it as a whole section.
As a result in the extracted bits another .cg sections symbol __Loader_Map is also presented in the beginning and with trailing zeroes. This leads to cropping the whole HSAIL text while casting to string; then RT silently doesn't perform kernels' finalization cause unable to find any kernel name in the extracted hsail. Finally the binary produced based on clCreateProgramWithBinary has the different sections/symbols order from the binary produced based on clCreateProgramWithSource.
[Solution] Extract Symbol __HSAIL_main_text instead of the section aclCODEGEN contains it.
[Testing] smoke_clang, pre check-in
[Reviewers] German Andryeyev, Brian Sumner
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuprogram.cpp#172 edit
EPR #010002 - Change OpenCL version number from 1610 to 1611.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1357 edit
[ROCm/clr commit: a7b6276d0d]
EPR #010002 - Change OpenCL version number from 1610 to 1611.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1357 edit
EPR #404357 - [OCL 2.0][CPU] CPU should not report OCL 2.0 version in CLINFO
- Report OCL1.2 by default on CPU
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/cpu/cpudevice.cpp#266 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/cpu/cpusettings.cpp#26 edit
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#212 edit
[ROCm/clr commit: 75e2393a8f]
EPR #404357 - [OCL 2.0][CPU] CPU should not report OCL 2.0 version in CLINFO
- Report OCL1.2 by default on CPU
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/cpu/cpudevice.cpp#266 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/cpu/cpusettings.cpp#26 edit
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#212 edit
EPR #010002 - Change OpenCL version number from 1609 to 1610.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1356 edit
[ROCm/clr commit: 6dda08bb51]
EPR #010002 - Change OpenCL version number from 1609 to 1610.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1356 edit
ECR #304775 - Device enqueuing
- Provide scratch buffer offset for generic address space
- Use single scratch buffer for all available queues. Each queue will have a unique subbuffer in the global buffer
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.cpp#454 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.hpp#129 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusched.hpp#11 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuschedcl.cpp#24 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#329 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.hpp#120 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.cpp#63 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.h#37 edit
[ROCm/clr commit: 8fa184db38]
ECR #304775 - Device enqueuing
- Provide scratch buffer offset for generic address space
- Use single scratch buffer for all available queues. Each queue will have a unique subbuffer in the global buffer
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.cpp#454 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.hpp#129 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusched.hpp#11 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuschedcl.cpp#24 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#329 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.hpp#120 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.cpp#63 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.h#37 edit
EPR #010002 - Change OpenCL version number from 1608 to 1609.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1355 edit
[ROCm/clr commit: 7e43e18d8c]
EPR #010002 - Change OpenCL version number from 1608 to 1609.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1355 edit
EPR #010002 - Change OpenCL version number from 1607 to 1608.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1354 edit
[ROCm/clr commit: ef4b4c5b9e]
EPR #010002 - Change OpenCL version number from 1607 to 1608.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1354 edit
ECR #304775 - Add new flag GPU_IFH_MODE to allow IFH mode testing without using the inifile.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#97 edit
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#211 edit
[ROCm/clr commit: 22af57a1ec]
ECR #304775 - Add new flag GPU_IFH_MODE to allow IFH mode testing without using the inifile.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#97 edit
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#211 edit
EPR #403687 - [CQE OCL][2.0][NI][QR] WF 1.2 Conf: Failures observed in Images/ClFillImage test in Linux & Windows; Faulty CL#1057445
Issue: This issue happens because the image setting for NI is different from SI and above. It may take some time to fix NI issue for image RGB_101010.
Current user(MM team) is NOT asking RBG_101010 support for NI and previous asics.
Solution: Support image format (CL_RGB, CL_UNORM_INT_101010) on gpu only from OCL 2.0.
Code Review Request: 5494
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.cpp#107 edit
[ROCm/clr commit: 8483ca5040]
EPR #403687 - [CQE OCL][2.0][NI][QR] WF 1.2 Conf: Failures observed in Images/ClFillImage test in Linux & Windows; Faulty CL#1057445
Issue: This issue happens because the image setting for NI is different from SI and above. It may take some time to fix NI issue for image RGB_101010.
Current user(MM team) is NOT asking RBG_101010 support for NI and previous asics.
Solution: Support image format (CL_RGB, CL_UNORM_INT_101010) on gpu only from OCL 2.0.
Code Review Request: 5494
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/platform/memory.cpp#107 edit
EPR #010002 - Change OpenCL version number from 1606 to 1607.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1353 edit
[ROCm/clr commit: b758009d47]
EPR #010002 - Change OpenCL version number from 1606 to 1607.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1353 edit
EPR #010002 - Change OpenCL version number from 1605 to 1606.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1352 edit
[ROCm/clr commit: ba4d7b82b6]
EPR #010002 - Change OpenCL version number from 1605 to 1606.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1352 edit
ECR #333753 - RS compilation path.
RS compilation will be in 2 stages, first generates BRIG and the next is done via HSA Finalize API (that involves a load step).
Existing code in compiler/lib has a bug in that when the final output expected is HSAIL_BINARY, compilation should stop with invoking the llvm compiler (and the built-in assembler), not go all the way to ISA.
Tests: precheckin, hsa smoke
hsa/tests/RS/ test harness will be changed in a separate changelist.
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/v0_8/if_acl.cpp#44 edit
... //depot/stg/opencl/drivers/opencl/compiler/tools/aoc2/aoc2.cpp#57 edit
[ROCm/clr commit: 587b069737]
ECR #333753 - RS compilation path.
RS compilation will be in 2 stages, first generates BRIG and the next is done via HSA Finalize API (that involves a load step).
Existing code in compiler/lib has a bug in that when the final output expected is HSAIL_BINARY, compilation should stop with invoking the llvm compiler (and the built-in assembler), not go all the way to ISA.
Tests: precheckin, hsa smoke
hsa/tests/RS/ test harness will be changed in a separate changelist.
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/v0_8/if_acl.cpp#44 edit
... //depot/stg/opencl/drivers/opencl/compiler/tools/aoc2/aoc2.cpp#57 edit
EPR #403199 - Reduced the max number of command queues
"multi_device_context" fails with CF mode because there is not enough buffer for IB on Linux
http://ocltc.amd.com/reviews/r/5454/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#210 edit
[ROCm/clr commit: c90818dd9a]
EPR #403199 - Reduced the max number of command queues
"multi_device_context" fails with CF mode because there is not enough buffer for IB on Linux
http://ocltc.amd.com/reviews/r/5454/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/flags.hpp#210 edit
EPR #403896 - [OpenGL] - Use GFXIP instead of asic family name
This is the first step to using a GFXIP version instead of the asic family names wherever possible in our driver. It is only being used for a few checks right now, and there will be more changes after this to use it in more places.
- Added GFXIP so that we can reference IP version instead of checking asic family names. This is being done so it's more clear about which version of gfx ip a given asic is, since marketing often uses a different family name than engineering (ex. Hawaii is a VI part from marketing standpoint). This also reflects the HW team's move towards using the gfxip version instead of a family name. Another benefit is it will clean up our asic family checks quite a bit, and they should be a bit more forward looking as well.
- Added some asserts for when gfxip 9 comes along we know the places we need to validate. It hasn't been determined yet if gfxip 9 will use the same hwl, so if it doesn't, these asserts will be removed then.
- Removed all references of Kauai, Tiran, and Maui from our driver since these asics were all cancelled at some point previously.
- Changed Distributed Tessellation enablement criteria to be gfxip 8+ && numSE > 1 instead of checking asic names.
ReviewBoardURL = http://ocltc.amd.com/reviews/r/5450/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#96 edit
[ROCm/clr commit: 11dccb06ae]
EPR #403896 - [OpenGL] - Use GFXIP instead of asic family name
This is the first step to using a GFXIP version instead of the asic family names wherever possible in our driver. It is only being used for a few checks right now, and there will be more changes after this to use it in more places.
- Added GFXIP so that we can reference IP version instead of checking asic family names. This is being done so it's more clear about which version of gfx ip a given asic is, since marketing often uses a different family name than engineering (ex. Hawaii is a VI part from marketing standpoint). This also reflects the HW team's move towards using the gfxip version instead of a family name. Another benefit is it will clean up our asic family checks quite a bit, and they should be a bit more forward looking as well.
- Added some asserts for when gfxip 9 comes along we know the places we need to validate. It hasn't been determined yet if gfxip 9 will use the same hwl, so if it doesn't, these asserts will be removed then.
- Removed all references of Kauai, Tiran, and Maui from our driver since these asics were all cancelled at some point previously.
- Changed Distributed Tessellation enablement criteria to be gfxip 8+ && numSE > 1 instead of checking asic names.
ReviewBoardURL = http://ocltc.amd.com/reviews/r/5450/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#96 edit
EPR #010002 - Change OpenCL version number from 1604 to 1605.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1351 edit
[ROCm/clr commit: 5ac6db1c98]
EPR #010002 - Change OpenCL version number from 1604 to 1605.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1351 edit
EPR #010002 - Change OpenCL version number from 1603 to 1604.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1350 edit
[ROCm/clr commit: 430eb21c68]
EPR #010002 - Change OpenCL version number from 1603 to 1604.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1350 edit
ECR #304775 - Allow disabling of the CPU device when CPU_MAX_COMPUTE_UNITS is 0.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/cpu/cpudevice.cpp#265 edit
[ROCm/clr commit: b9d45ad9c5]
foreman