EPR #010002 - Change OpenCL version number from 1872 to 1873.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1619 edit
EPR #010002 - Change OpenCL version number from 1871 to 1872.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1618 edit
ECR #304775 - Move AMDExportKernelNature to compiler library.
This is part of the effort to move all extra passes out of the LLVM directory and into compiler library so we can build with a pure copy of LLVM. This is far from all of the sources required to move, but I'm trying to do the migration in moderately sized pieces.
Duplicates AMDSymbolNames.h utility header to avoid having a circular dependency between LLVM and compiler library. Eventually when all extra sources are moved into compiler library, there should only be one copy left.
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/optimizer.cpp#27 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/include/AMDPasses.h#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/include/AMDSymbolName.h#1 branch
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/AMDExportKernelNature.cpp#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/llvm/include/llvm/Analysis/Passes.h#56 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm/include/llvm/InitializePasses.h#81 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Analysis/AMDExportKernelNature.cpp#16 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Analysis/Analysis.cpp#34 edit
EPR #010002 - Change OpenCL version number from 1870 to 1871.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1617 edit
ECR #304775 - back out 1179705 since it also affects AMDIL path
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusettings.cpp#322 edit
EPR #010002 - Change OpenCL version number from 1869 to 1870.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1616 edit
EPR #397491 - Disable 32 bit OCL2.0 on Win 10 for now. isWindows10OrLater() call doesn't work. So, passing a flag from GSL to runtime to disable 32 bit OCL2.0 on Win 10.
http://ocltc.amd.com/reviews/r/8234/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusettings.cpp#321 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#133 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/backend.h#8 edit
ECR #304775 - switch on "denorm" support on VI and make HSAIL default to not generating ftz modifier
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Target/HSAIL/HSAILISelDAGToDAG.cpp#48 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusettings.cpp#320 edit
EPR #419072 - [OpenCL2.0] Enable 16MB large on device queues
- Enable device queue creation up to 12MB. That should allow to run Intel SDK sample from the EPR that requires 6MB queue only.
- Currently a queue with >12.5MB size has a significant performance degradation. Thus the current max possible is 12MB. In general it's preferable to use the queue size more suitable for the task, rather than max possible.
Affected files ...
... //depot/stg/opencl/drivers/opencl/library/hsa/hsail/src/devenq/schedule.cl#10 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#115 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.hpp#38 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudefs.hpp#123 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.cpp#517 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusched.hpp#17 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#372 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.hpp#131 edit
EPR #010002 - Change OpenCL version number from 1868 to 1869.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1615 edit
EPR #010002 - Change OpenCL version number from 1867 to 1868.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1614 edit
EPR #010002 - Change OpenCL version number from 1866 to 1867.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1613 edit
ECR #304775 - Fork llvm-link into new tool called opencl-link
Most of what the patched llvm-link does has nothing to do with linking LLVM IR, and more to do with loading SPIR and specifically handling the builtin library.
Forking this into a separate tool is the fastest way to fix dependencies on large LLVM patches for the OpenCL build to work. With this in place, it should now be possible to move the various linker and SPIR conversions out of the llvm directory and into compiler library.
Ideally this would be fixed by:
1. Not always lowering the library from SPIR
2. Having a separate SPIR lowering tool for testing
3. Using function attributes and stub libraries for library FP options instead of linker flags
4. Structuring all of the SPIR conversions as passes and having a single PassManager handle all of the lowering / linking / optimization passes.
But accomplishing all of these will be more time consuming.
Branching
//depot/stg/opencl/drivers/opencl/compiler/llvm/tools/llvm-link/...
to //depot/stg/opencl/drivers/opencl/compiler/lib/linker/tools/opencl-link/...
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/build/Makefile.api#118 edit
... //depot/stg/opencl/drivers/opencl/compiler/Makefile#60 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/Makefile#34 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/build/Makefile.common#33 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/linker.cpp#131 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/build/Makefile.complib#88 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/compliblinkerlibs#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/Makefile#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/include/AMDFixupKernelModule.h#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/include/AMDPrelink.h#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/include/AMDResolveLinker.h#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/AMDFixupKernelModule.cpp#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/AMDPrelink.cpp#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/AMDResolveLinker.cpp#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/Makefile#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/build/Makefile#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/build/Makefile.linker#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/lib/clpVectorExpansion.cpp#1 move/add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/tools/opencl-link/Android.mk#1 branch
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/tools/opencl-link/Makefile#1 branch
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/tools/opencl-link/build/Makefile#1 branch
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/tools/opencl-link/build/Makefile.opencl-link#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/linker/tools/opencl-link/opencl-link.cpp#1 add
... //depot/stg/opencl/drivers/opencl/compiler/llvm/include/llvm/AMDFixupKernelModule.h#3 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/include/llvm/AMDPrelink.h#2 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/include/llvm/AMDResolveLinker.h#7 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Linker/AMDFixupKernelModule.cpp#9 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Linker/AMDPrelink.cpp#2 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Linker/AMDResolveLinker.cpp#11 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Linker/clpVectorExpansion.cpp#3 move/delete
... //depot/stg/opencl/drivers/opencl/compiler/llvm/llvmdefs#41 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm/tools/llvm-link/llvm-link.cpp#55 edit
... //depot/stg/opencl/drivers/opencl/openclrules#87 edit
... //depot/stg/opencl/drivers/opencl/tests/hsa/bin/hsa_dist.pl#2 edit
... //depot/stg/opencl/drivers/opencl/tests/hsa/bin/test_driver.pl#18 edit
ECR #333753 - Compiler Lib: switch Bif Version to 3.1 by default for HSAIL
It is needed due to the latest AMD HSA Code Object introduction in BIF.
TODO (in separate changes):
1. Analyze the changes in sections/symbols and remove (if needed) unused anymore (in BIF31), for example, symISAMeta, check backward compatibility.
2. Move the bif versions/conversions code from libUtils to loader\Bif.
3. Refactor the bif versions/conversions code in order to get rid of copy/paste (templates?).
4. Drop aclBIFVersionCAL.
Testing: pre check-in
Reviewer: Brian Sumner, Nikolay Haustov
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/api/v0_8/acl.cpp#33 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/include/v0_8/aclEnums.h#20 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif.hpp#3 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif20.cpp#14 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif20.hpp#7 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif21.cpp#7 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif21.hpp#7 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif30.cpp#26 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif30.hpp#8 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif31.cpp#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bif31.hpp#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bifbase.cpp#53 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/loaders/bif/bifbase.hpp#23 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/bif_section_labels.hpp#22 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/v0_8/libUtils.cpp#11 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/v0_8/libUtils.h#21 edit
... //depot/stg/opencl/drivers/opencl/compiler/tools/aoc2/aoc2.cpp#75 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/binary/BIFAssumptionCheck.cpp#9 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/binary/BIFAssumptionCheck.hpp#5 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/complib/CLEnumCheck.cpp#45 edit
EPR #010002 - Change OpenCL version number from 1865 to 1866.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1612 edit
EPR #421017 - IOMMU2/SVM on CZ Win10, the bit INST_ATC of COMPUTE_PGM_HI needs to be set for device enqueue.
Affected files ...
... //depot/stg/opencl/drivers/opencl/library/hsa/hsail/src/devenq/schedule.cl#9 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusched.hpp#16 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#370 edit
EPR #424562 - Add Averaging algorithm to Wave Limiter.
1. Extract the algorithms to a sub-class of the wave limiter class.
2. Add Averaging algorithm
This averaging algorithm typically improves performance of BasemarkCL wave simulation by 8% on Tonga/Fiji than the current smooth algorithm. This change has not enable the averaging algorithm yet. Follow-up changes should be made to intelligently select which algorithm to use.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuwavelimiter.cpp#7 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuwavelimiter.hpp#4 edit
EPR #010002 - Change OpenCL version number from 1864 to 1865.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1611 edit
EPR #397491 - Back out changelist 1177450. Disable 32 bit generic address space
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#130 edit
EPR #397491 - Enable generic address space for 32 bit on Windows
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#129 edit
ECR #333753 - ORCA RT/Compiler Lib/aoc2: AMD HSA Code Object Import feature (part II) - arbitrary hidden (extra) kernargs support
Only HSAIL path is affected. It doesn't affect blit kernels.
To use offline by aoc2:
aoc2 -hsacodeobject=<importing_code_object_filename> -numhiddenkernargs=<num> -cl-std=CL2.0 -march=hsail(-64) -mdevice=Bonaire <source_cl_filename>
To use online by setting env:
AMD_DEBUG_HSA_NUM_HIDDEN_KERNARGS=<num>
where num >= 0. If num == 0, then no additional arguments will be added on RT for every kernel. The default value is unchanged and equal to 6 for now.
Misc:
+ get rid of PRE & POST defines in Compiler Lib, as they started to conflict with ugl\gl\gs\hwl\ headers with the same defines.
+ minor copy/paste eliminations & typo fixes
+ ocltst complib tests update
Testing: pre check-in, manually based on ocl sdk MatrixMultiplication
Reviewers: Brian Sumner, German Andryeyev, Nikolay Haustov, Artem Tamazov
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/v0_8/if_acl.cpp#72 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/hsail_be.cpp#49 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/metadata.cpp#8 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/include/v0_8/aclDefs.h#5 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/include/v0_8/aclEnums.h#19 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/include/v0_8/aclStructs.h#17 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/bif_section_labels.hpp#21 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/v0_8/libUtils.cpp#10 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/v0_8/libUtils.h#20 edit
... //depot/stg/opencl/drivers/opencl/compiler/tools/aoc2/aoc2.cpp#74 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/device.cpp#181 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#249 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#291 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.hpp#113 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuprogram.cpp#199 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#369 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/hsa/hsaprogram.cpp#38 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/hsa_foundation/hsakernel.cpp#8 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/hsa_foundation/hsakernel.hpp#5 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/hsa_foundation/hsaprogram.cpp#19 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/hsa_foundation/hsavirtual.cpp#43 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/complib/CLAssumptionCheck.cpp#43 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/complib/CLEnumCheck.cpp#44 edit
EPR #010002 - Change OpenCL version number from 1863 to 1864.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1610 edit
EPR #010002 - Change OpenCL version number from 1862 to 1863.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1609 edit
EPR #010002 - Change OpenCL version number from 1861 to 1862.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1608 edit
EPR #010002 - Change OpenCL version number from 1860 to 1861.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1607 edit
EPR #010002 - Change OpenCL version number from 1859 to 1860.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1606 edit
EPR #010002 - Change OpenCL version number from 1858 to 1859.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1605 edit
EPR #419362 - Forum [170348]: problem with printf for OpenCL 2.0 kernel build - added a set of missing symbols to flex including space, added a extra backslash to escape sequences that could not be handled by flex and also to colon which is defined in flex as a delimiter!
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/MDParser/AMDILMDParser.l#3 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/MDParser/AMDILMDParser.output#2 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/MDParser/lex.yy.cpp#4 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Target/HSAIL/HSAILProducePrintfMetadata.cpp#6 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Transforms/Scalar/AMDPrintfRuntimeBinding.cpp#10 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#290 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/compiler/OCLPrintSpecialChars.cpp#1 add
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/compiler/OCLPrintSpecialChars.h#1 add
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/compiler/TestList.cpp#41 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/compiler/build/Makefile.compiler#44 edit
EPR #010002 - Change OpenCL version number from 1857 to 1858.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1604 edit
EPR #010002 - Change OpenCL version number from 1856 to 1857.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1603 edit
ECR #333753 - cl-denorms-are-zero runtime change - comment out "singleFpDenorm_ = true" for now.
So AMD_GPU_FORCE_SINGLE_FP_DENORM is required to enable this, until a few issues are resolved.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusettings.cpp#316 edit
foreman