SWDEV-144570 - Adding extern var support for dynamically loaded modules for Texture reference.
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/hip/hip_internal.hpp#43 edit
... //depot/stg/opencl/drivers/opencl/api/hip/hip_module.cpp#42 edit
... //depot/stg/opencl/drivers/opencl/api/hip/hip_platform.cpp#43 edit
SWDEV-192384 - [HIP CQE][HIPonPAL][19.40] hipBindTexRef1DFetch, hipTextureRef2D are failed on all ASICs for both Win/Lnx
1. Correctly set the image type for textures created from arrays.
2. Allow creating any kind of image from a buffer.
ReviewBoardURL = http://ocltc.amd.com/reviews/r/18051/diff/
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/hip/hip_texture.cpp#19 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/pal/paldevice.cpp#166 edit
SWDEV-203855 - Segfault when using hipArrayCreate and hipMemcpyParam2D
1. hipArrayCreate API implementation uses a wrong parameter to check width. That parameter can be null pointer because it is used to pass the pointer back to the caller.
2. Implement hipMemcpyParam2D similar to HIP-HCC implementation. Reference: https://github.com/ROCm-Developer-Tools/HIP/blob/master/src/hip_memory.cpp
Tests:
1. PRE CHECK-IN build and test(no regression): http://ocltc:8111/viewModification.html?modId=126608&personal=true&init=1&tab=vcsModificationBuilds
2. GPU is VEGA10, OS is Windows 10, CPU is threadripper 1900x, run the test. There is not segfault or exit during hipArrayCreate and hipMemcpyParam2D function call.
ReviewBoard: http://ocltc.amd.com/reviews/r/18037/
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/hip/hip_memory.cpp#72 edit
+ Perl part of [#1458]
+ Affected functions: hipFuncSetCacheConfig, hipFuncGetAttributes
+ Implement function generateHostFunctions() in hipify-clang for that purposes
+ Update hipify-perl accordingly
+ Merge of cudaSymbolFuncCall and cudaReinterpretCastArgFuncCall matchers into a single cudaHostFuncCall matcher
+ More const std::string declarations
+ Formatting
+ Affected functions: hipFuncSetCacheConfig, hipFuncGetAttributes
+ Add a corresponding Matcher cudaReinterpretCastArgFuncCall
+ Add reinterpret_cast.cu test
TODO: Do the same for hipify-perl
cudaMemcpyToSymbol, cudaMemcpyToSymbolAsync, cudaGetSymbolSize, cudaGetSymbolAddress, cudaMemcpyFromSymbol, cudaMemcpyFromSymbolAsync
+ Perl part of [#1441]
+ Implement function generateSymbolFunctions() in hipify-clang for that purposes
+ Update hipify-perl
TODO: Eliminate dim3() issue in hipify-perl as well
cudaMemcpyToSymbol, cudaMemcpyToSymbolAsync, cudaGetSymbolSize, cudaGetSymbolAddress, cudaMemcpyFromSymbol, cudaMemcpyFromSymbolAsync
+ Add a corresponding cudaSymbolFuncCall matcher.
+ Add device_symbols.cu test for the above 6 functions, update existed.
+ Fix dim3() type cast issue, update affected tests.
TODO: Do the same in hipify-perl
SWDEV-201925 - hipArray3DCreate() not available in HIP/PAL on Windows
1. Implement hipArray3DCreate().
2. Remove the array size calculation from hipArrayCreate() as it is not used.
ReviewBoardURL = http://ocltc.amd.com/reviews/r/18005/diff/
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/hip/hip_memory.cpp#71 edit
SWDEV-201128 - [HIP] test_snli_cuda failure
Default to sync packet
Make sure GPU_NUM_MEM_DEPENDENCY is 0 for HIP
No sync packet is only used when there are mem dependency check
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/hip/hip_context.cpp#22 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocvirtual.cpp#86 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocvirtual.hpp#28 edit
+ [perl] Add generation of sub countSupportedDeviceFunctions
+ [FIX] Do not perform any replacement for supported device functions
+ [perl] Sync hipify-perl accordingly
+ Do not treat somenamespace::device_function_name as a device function
+ Fix generation of warnUnsupportedDeviceFunctions function in hipify-clang
+ Update hipify-perl based on hipify-clang -perl generation
+ Update device test math_functions.cu for hipify-perl
[Restrictions]
- hipify-perl is yet unable to handle function declarations in user namespaces
- hipify-perl is yet unable to handle using directive
* [hip][tests] add a unit test for testing hipLaunchCooperativeKernel
* use __ockl_grid_sync function
* remove already defined __ockl_grid_sync function
* use sync function for grid synchronization
+ Move generating hipify-perl and hipify-python into separate namespaces and files
+ Move getAbsoluteFilePath and getAbsoluteDirectoryPath to StringUtils
SWDEV-203438 - [HIP] AllGather RCCL test issue
The test tries to launch a kernel on two devices at once and they need to communicate with each other.
For that, it uses a custom stream for each devices.
Problem is in getNullStream we used to call syncStreams all the time
and it was syncing all the streams even the ones on different devices.
So that made the second kernel launch (on 2n dev) to wait for the first kernel to finish which
would never occur since the first one was waiting for the second one.
The fix is to not call syncStreams from getNullStream because we sync already anyway prior in general.
Affected files ...
... //depot/stg/opencl/drivers/opencl/api/hip/hip_context.cpp#21 edit
... //depot/stg/opencl/drivers/opencl/api/hip/hip_event.cpp#16 edit
... //depot/stg/opencl/drivers/opencl/api/hip/hip_internal.hpp#40 edit
... //depot/stg/opencl/drivers/opencl/api/hip/hip_memory.cpp#70 edit
... //depot/stg/opencl/drivers/opencl/api/hip/hip_module.cpp#41 edit
... //depot/stg/opencl/drivers/opencl/api/hip/hip_stream.cpp#24 edit
Evgeny Mankov
foreman
Rahul Garg
Yaxun (Sam) Liu
Sarbojit2019
mhbliao
satyanveshd
AlexBinXie
ansurya
Aryan Salmanpour