EPR #010002 - Change OpenCL version number from 1635 to 1636.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1382 edit
ECR #304775 - Update resource cache behavior
Currently, the resource cache is fixed at 64MB regardless of available video memory size. Changed the logic to use max(1/8th video memory, 64MB). This is still overrideable with the env. var. GPU_RESOURCE_CACHE_SIZE.
Improvements with changes: 18% decrease in video chat face detect time on 95w Kaveri (no change in PCMark8 score as we already achieved 30fps), 14% improvement on 19w Kaveri (this does result in an improvement in PCMark8 since the APU is slower).
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusettings.cpp#279 edit
EPR #010002 - Change OpenCL version number from 1634 to 1635.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1381 edit
EPR #010002 - Change OpenCL version number from 1633 to 1634.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1380 edit
ECR #377625 - AMDIL Function support: Calculate total private memory usage by a kernel including memory used by called functions.
This cannot be done by IPA since stack size is known only after register allocation due to potential register spill, but MachineFunctionAnalysis cannot persist after CGSCC pass with current LLVM version.
This change adds private memory usage metadata for non-kernel functions. The total private memory usage by a kernel is calculated when AMDIL is split for different kernels. BIF will contain total private memory size.
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/amdilUtils.cpp#1 add
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/amdilUtils.hpp#1 add
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Target/AMDIL/AMDILKernelManager.cpp#451 edit
... //depot/stg/opencl/drivers/opencl/compiler/llvm/lib/Target/AMDIL/AMDILKernelManager.h#51 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuprogram.cpp#175 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuprogram.hpp#54 edit
EPR #010002 - Change OpenCL version number from 1632 to 1633.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1379 edit
EPR #405753 - Fixed incorrect value of slicePitch returned from clEnqueueMapimage for 1Darray.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpumemory.cpp#111 edit
EPR #010002 - Change OpenCL version number from 1631 to 1632.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1378 edit
EPR #403493 - Block index error for CI and VI, OCL code change
Problem description: The OCL implementation requires HSA to used different block index values for CI and VI. However, the same index value is used for the same counter block in both CI and VI, which in turn caused a segmentation fault.
Root cause: HSA implementation does not know this situation before hand.
Solution: Fix to use different counter block index in CI from that in VI
Functional area: HSA perf counter implementation...
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/hsa/hsacounters.cpp#4 edit
EPR #010002 - Change OpenCL version number from 1630 to 1631.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1377 edit
EPR #397491 - only using OpenCL 1.2 onr multiple GPUs on mainline by default.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.cpp#460 edit
EPR #010002 - Change OpenCL version number from 1629 to 1630.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1376 edit
EPR #400016 - Keep the path of temp folder if the app is WIndows app
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/os/os_win32.cpp#39 edit
EPR #010002 - Change OpenCL version number from 1628 to 1629.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1375 edit
EPR #010002 - Change OpenCL version number from 1627 to 1628.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1374 edit
EPR #010002 - Change OpenCL version number from 1626 to 1627.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1373 edit
EPR #010002 - Change OpenCL version number from 1625 to 1626.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1372 edit
EPR #398128 - Windows 2015, WDDM2.0, New Residency Model
- Modify MarkUsedInCmdBuf in IOL to make sure that MakeResident is called for OpenCL (Part 2)
ReviewBoardURL = http://ocltc.amd.com/reviews/r/5684/diff/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#99 edit
ECR #304775 - Add batching to the device enqueue for possible asynchronous execution
- Increase the max device queue size to 512KB. That will allow to pass conformance tests that enqueue more jobs than the queue size.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.cpp#459 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpusched.hpp#13 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuschedcl.cpp#28 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#333 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.cpp#65 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.h#39 edit
EPR #010002 - Change OpenCL version number from 1624 to 1625.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1371 edit
EPR #010002 - Change OpenCL version number from 1623 to 1624.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1370 edit
EPR #010002 - Change OpenCL version number from 1622 to 1623.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1369 edit
EPR #010002 - Change OpenCL version number from 1621 to 1622.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1368 edit
EPR #010002 - Change OpenCL version number from 1619 to 1620.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1366 edit
ECR #304775 - Optimization for rectangular copies(Part2). Due to HW restriction of 14bits for src and dst pitch, its advantageous to choose optimal bpp. Higher the bpp the larger the byte pitch. This indirectly helps to reduce the number of packets for buffer copy(line by line vs a single sub_win raw packet)
ReviewBoardURL = http://ocltc.amd.com/reviews/r/5605/diff/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#109 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuresource.cpp#191 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuresource.hpp#76 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.cpp#64 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLContext.h#38 edit
EPR #010002 - Change OpenCL version number from 1618 to 1619.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1365 edit
EPR #404714 - [CQE OCL][2.0][DTB]Opencl1.2 WF Conf. Math test failedon Pitcairn and Oland due to CL#1065597
- FIx for TC regression after CL#1069020. Move the lock directly to the gsl flush() calls.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#332 edit
EPR #010002 - Change OpenCL version number from 1617 to 1618.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1364 edit
EPR #404714 - [CQE OCL][2.0][DTB]Opencl1.2 WF Conf. Math test failedon Pitcairn and Oland due to CL#1065597
- Add a new MapCacheLock monitor to separate the map cache from the global lock
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.cpp#456 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.hpp#131 edit
EPR #404714 - [CQE OCL][2.0][DTB]Opencl1.2 WF Conf. Math test failedon Pitcairn and Oland due to CL#1065597
- Add VGPU lock to flush() method, because gsl flush for the same context could be called from multiple threads
- Use new scratchAlloc_ monitor for scratch reallocation
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.cpp#455 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpudevice.hpp#130 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#331 edit
EPR #010002 - Change OpenCL version number from 1616 to 1617.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1363 edit
ECR #333753 - Compiler Lib & RT: Fix for Compiler's build log printing on RT.
+ RT now asks correctly Compiler's build log by aclGetCompilerLog().
+ BuildLog is added for HSAILKernel by moving it from NullKernel class to Kernel class.
+ Compiler's Lib appendLogToCL() is fixed.
+ Usage of API's aclExtractSection/aclExtractSymbol/aclInsertSection/aclInsertSymbol in Compiler Lib itself replaced by it's inner realizations extSec/extSym/insSec/insSym due to unneded build log clearing in first case.
+ Phase info is added to build log even if CallBack function is not presented for aclCompiler.
How to verify:
set AMD_OCL_BUILD_OPTIONS_APPEND="-print-compile-phases -buildlog=stdout"
test_integer_ops integer_ctz
test_integer_ops integer_ctz cpu
Testing: make smoke_clang, selective OCL conf. tests, pre check-in
Reviewer: Brian Sumner, German Andryeyev
Review board: http://ocltc.amd.com/reviews/r/5582/
Affected files ...
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/v0_8/if_acl.cpp#46 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/brig_loader.cpp#13 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/hsail_be.cpp#31 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/v0_8/libUtils.cpp#4 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#228 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#262 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.hpp#100 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuprogram.cpp#174 edit
EPR #010002 - Change OpenCL version number from 1615 to 1616.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1362 edit
ECR #304775 - HSAIL: Direct SRD support
- Copy SRD to CB1 for image views to avoid a wait for SRD resource when image view is destroyed.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpukernel.cpp#261 edit
EPR #010002 - Change OpenCL version number from 1614 to 1615.
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1361 edit
ECR #304775 - Refactor code to do line by line copies for read\write Rect. This avoids taking the blit copy path which may be even slower.
ReviewBoardURL = http://ocltc.amd.com/reviews/r/5567/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpublit.cpp#108 edit
ECR #304775 - Correct a typo where I didnt remove the offset from the condition which made the writeRect take pinning path.
ReviewBoardURL = http://ocltc.amd.com/reviews/r/5566/diff/
Affected files ...
... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.cpp#330 edit
foreman