Граф коммитов

13474 Коммитов

Автор SHA1 Сообщение Дата
foreman af7e8beb34 P4 to Git Change 1415167 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/27 03:00:06
SWDEV-2 - Change OpenCL version number from 2429 to 2430.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2177 edit
2017-05-27 03:20:10 -04:00
foreman 8854784f49 P4 to Git Change 1414965 by gandryey@gera-w8 on 2017/05/26 15:51:24
SWDEV-79445 - OCL generic changes and code clean-up
	- use getBackendDevice() for HSA agent look-up for persistent memory allocation

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocmemory.cpp#23 edit
2017-05-26 16:08:36 -04:00
Ben Sander b2b620c12b Add isDefaultStream() accessor.
Fix code that checked for stream==nullptr after stream had been
resolved to a "true stream".
2017-05-26 13:46:48 -05:00
foreman 595192b942 P4 to Git Change 1414821 by lmoriche@lmoriche_palamida on 2017/05/26 13:04:04
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Update from git.

Affected files ...

... //depot/stg/opencl/drivers/opencl/.gitignore#1 add
... //depot/stg/opencl/drivers/opencl/CMakeLists.txt#10 edit
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/CMakeLists.txt#6 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/CMakeLists.txt#5 edit
... //depot/stg/opencl/drivers/opencl/tools/clinfo/CMakeLists.txt#1 add
2017-05-26 13:12:30 -04:00
foreman 700577fee9 P4 to Git Change 1414821 by lmoriche@lmoriche_palamida on 2017/05/26 13:04:04
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Update from git.

Affected files ...

... //depot/stg/opencl/drivers/opencl/.gitignore#1 add
... //depot/stg/opencl/drivers/opencl/CMakeLists.txt#10 edit
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/CMakeLists.txt#6 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/CMakeLists.txt#5 edit
... //depot/stg/opencl/drivers/opencl/tools/clinfo/CMakeLists.txt#1 add
2017-05-26 13:12:30 -04:00
foreman 600b7c365c P4 to Git Change 1414775 by asalmanp@asalmanp-ocl-stg on 2017/05/26 11:34:07
SWDEV-120065 - [Vega10] Multi 4k60 Hyrid VP9 Chrome Decode Causes System Hang
	- The hang happens because runtime was overwriting an incomplete IB due to incorrect result for OS event wait.
	- To fix this issue force to wait for fences if ibReuse is enabled.

	ReviewBoardURL = http://ocltc.amd.com/reviews/r/12862/

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.hpp#23 edit
2017-05-26 11:43:05 -04:00
foreman 88f304df7d P4 to Git Change 1414622 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/26 03:00:05
SWDEV-2 - Change OpenCL version number from 2428 to 2429.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2176 edit
2017-05-26 03:06:00 -04:00
Siu Chi Chan a3595d2e8c fix hip_fast_dsqrt* to call a double fp sqrt function 2017-05-25 23:15:30 -04:00
foreman 2839999356 P4 to Git Change 1414382 by cpaquot@cpaquot-ocl-lc-lnx on 2017/05/25 15:55:25
SWDEV-121486 - Create a metadata structure
	This is needed for UGL/CL interop because in UGL, they do not pass the metadata
	to KMD/KFD via amdgpu_bo_set_metadata. UGL instead will pass the SRD info
	via the mesa_in.out_driver_data.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#286 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocmemory.cpp#22 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocmemory.hpp#9 edit
2017-05-25 16:07:03 -04:00
foreman 5782f336db P4 to Git Change 1414000 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/25 03:00:04
SWDEV-2 - Change OpenCL version number from 2427 to 2428.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2175 edit
2017-05-25 03:11:46 -04:00
foreman 80b3356ed6 P4 to Git Change 1413623 by lmoriche@lmoriche_palamida on 2017/05/24 11:41:52
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Set the default triple to amdgcn-amd-amdhsa-opencl

Affected files ...

... //depot/stg/opencl/drivers/opencl/CMakeLists.txt#9 edit
2017-05-24 11:49:07 -04:00
foreman 3623f0d69e P4 to Git Change 1413623 by lmoriche@lmoriche_palamida on 2017/05/24 11:41:52
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Set the default triple to amdgcn-amd-amdhsa-opencl

Affected files ...

... //depot/stg/opencl/drivers/opencl/CMakeLists.txt#9 edit
2017-05-24 11:49:07 -04:00
Evgeny Mankov a19ecab3f2 [FIX] [HIPIFY] Add matchers for function return types.
https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/issues/73

Examples (https://github.com/thrust/thrust/blob/master/thrust/system/cuda/detail/trivial_copy.inl):

template<typename System1,
         typename System2>
cudaStream_t cuda_memcpy_stream(const thrust::cpp::execution_policy<System1> &,
                                const thrust::cuda::execution_policy<System2> &exec)

template<typename System1,
         typename System2>
cudaMemcpyKind cuda_memcpy_kind(const thrust::cuda::execution_policy<System1> &,
                                const thrust::cpp::execution_policy<System2> &)
2017-05-24 18:25:40 +03:00
foreman 5e85545c43 P4 to Git Change 1413393 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/24 03:00:04
SWDEV-2 - Change OpenCL version number from 2426 to 2427.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2174 edit
2017-05-24 03:17:11 -04:00
Ben Sander 35212632e7 Remove HIP_NUM_KERNELS_INFLIGHT. (redundant with HCC controls) 2017-05-24 01:03:28 -05:00
Ben Sander dda70ae514 Add hipHostMallocCoherent, hipHostMallocNonCoherent
Provide per-allocation control over coherent/non-coherent mem.
These overrid the default HIP_COHERENT_HOST_ALLOC setting.
2017-05-24 00:48:10 -05:00
Ben Sander d43d57d39c Remove HIP_MAX_QUEUES (replaced with HCC_MAX_QUEUES) 2017-05-23 23:48:01 -05:00
Ben Sander 92bd54d7b3 Expand test to cover copy followed by event sync 2017-05-23 23:15:45 -05:00
Ben Sander 2d5b3359c6 Use accelerator_scope for create_marker and create_blocking_marker.
As optimization when system-scope is not needed.
2017-05-23 23:15:45 -05:00
Ben Sander ca07615c37 Fix trace category for hipHostMalloc 2017-05-23 23:15:45 -05:00
foreman 2e71ba3a73 P4 to Git Change 1412538 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/23 03:00:04
SWDEV-2 - Change OpenCL version number from 2425 to 2426.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2173 edit
2017-05-23 17:24:39 -04:00
Evgeny Mankov 21d74f09b9 [FIX] [HIPIFY] Matcher for new operator is missing.
https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/issues/80

Example from CUDA 8.0.44 sample (CUDASamples\0_Simple\matrixMulDrv\matrixMulDrv.cpp):
    CUjit_option *jitOptions = new CUjit_option[jitNumOptions];
where CUjit_option is enum, should be:
    hipJitOption *jitOptions = new hipJitOption[jitNumOptions];
2017-05-23 19:45:38 +03:00
foreman 421c10fd87 P4 to Git Change 1412187 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/21 03:00:05
SWDEV-2 - Change OpenCL version number from 2424 to 2425.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2172 edit
2017-05-23 10:35:22 -04:00
Aditya Atluri 490355203b fixed erfinv build error as it is implemented in hcc
Change-Id: I27a512147c53f658a63fdf3e90f5e9cfac09ada8
2017-05-23 09:32:19 -05:00
foreman 2cea61199a P4 to Git Change 1412181 by yaxunl@yaxunl_stg_win50 on 2017/05/20 20:14:34
SWDEV-121585 - [OCL-LC-ROCm] Merge GitHub pull requests
	- Merged https://github.com/RadeonOpenCompute/ROCm-OpenCL-Runtime/pull/8

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/CMakeLists.txt#4 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/roccompiler.cpp#33 edit
2017-05-23 10:30:44 -04:00
pensun 2523f8a492 fix GGL helper header file, reorder for C++17
Change-Id: I3d9ddfe670bf7e3e8e7bd85e52cc61f48c19c213
2017-05-22 08:52:43 -05:00
Evgeny Mankov 51b7420359 [HIPIFY] [FIX] [HIPIFY] Matcher for pointer to enum var declaration is missing.
https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/issues/79

Example from CUDA 8.0.44 sample (CUDASamples\0_Simple\matrixMulDrv\matrixMulDrv.cpp):
    CUjit_option *jitOptions = new CUjit_option[jitNumOptions];
where CUjit_option is enum, should be:
    hipJitOption *jitOptions = new hipJitOption[jitNumOptions];

[TODO]
1. new CUjit_option -> new hipJitOption.
   Matcher for new operator is missing:
https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/issues/79
2. Merge matchers cudaEnumDecl and cudaEnumVarPtr.
2017-05-19 17:39:09 +03:00
Evgeny Mankov ae9f14ef9c [HIP] [HIPIFY] [FIX] cuModuleLoadDataEx -> hipModuleLoadDataEx
https://github.com/GPUOpen-ProfessionalCompute-Tools/HIP/issues/81

1.	Do not use JIT options on HCC path, call hipModuleLoadData instead.
2.	NVCC path is unchanged, to call cuModuleLoadDataEx with all options.
3.	Get rid of manual hipification, based on #ifdef #else for NVCC/HIP.
4.	Update documentation accordingly.
2017-05-19 17:22:14 +03:00
foreman 738528f6f3 P4 to Git Change 1411633 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/19 03:00:04
SWDEV-2 - Change OpenCL version number from 2423 to 2424.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2171 edit
2017-05-19 03:07:36 -04:00
foreman 6d7fb53d4a P4 to Git Change 1411361 by gandryey@gera-lnx-rcf-lc on 2017/05/18 14:22:45
SWDEV-121873 - [ROCm CQE][Vega10][QR] Observed Aborted (core dumped) while running clinfo due to CL#1410373
	- Fix device handle default initialization

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/pro/prodevice.hpp#2 edit
2017-05-18 14:47:03 -04:00
Aditya Atluri 0433a2e608 fixed vector type issues by reverting to old code, changed __half2 to map to vector types in llvm
Change-Id: I7317408c25e8c1a0c02a346042c9137e160c8bbd
2017-05-18 10:51:07 -05:00
foreman fabcd11161 P4 to Git Change 1411057 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/18 03:00:04
SWDEV-2 - Change OpenCL version number from 2422 to 2423.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2170 edit
2017-05-18 03:04:34 -04:00
foreman 842f311464 P4 to Git Change 1410993 by lmoriche@lmoriche_opencl_dev2 on 2017/05/17 22:25:50
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Make it possible to use cl2.hpp

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/khronos/headers/opencl2.0/CL/cl2.hpp#2 edit
... //depot/stg/opencl/drivers/opencl/tools/clinfo/clinfo.cpp#3 edit
2017-05-17 22:31:48 -04:00
foreman 757432840f P4 to Git Change 1410861 by gandryey@gera-lnx-rcf-lc on 2017/05/17 17:03:55
SWDEV-120180 - [amdgpu-pro] OpenCL support for SSG
	Correct define name

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_lqdflash_amd.cpp#20 edit
2017-05-17 18:52:41 -04:00
foreman 94759e212e P4 to Git Change 1410787 by gandryey@gera-w8 on 2017/05/17 15:15:22
SWDEV-120180 - [amdgpu-pro] OpenCL support for SSG
	- Protect amdgpu-pro headers with the define for easier open source integration

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.cpp#52 edit
... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocmemory.cpp#21 edit
2017-05-17 15:49:02 -04:00
foreman 928d2bcf66 P4 to Git Change 1410684 by gandryey@gera-w8 on 2017/05/17 13:00:21
SWDEV-120180 - [amdgpu-pro] OpenCL support for SSG
	Fix open source build

Affected files ...

... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_lqdflash_amd.cpp#19 edit
2017-05-17 13:09:07 -04:00
foreman 83937abeae P4 to Git Change 1410638 by lmoriche@lmoriche_palamida on 2017/05/17 11:34:07
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Add FindROCR.cmake, and use find_package() to locate the ROCR

Affected files ...

... //depot/stg/opencl/drivers/opencl/cmake/modules/FindROCR.cmake#2 edit
2017-05-17 11:50:44 -04:00
foreman 8f8d05040f P4 to Git Change 1410629 by lmoriche@lmoriche_palamida on 2017/05/17 11:27:06
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Add FindROCR.cmake, and use find_package() to locate the ROCR

Affected files ...

... //depot/stg/opencl/drivers/opencl/CMakeLists.txt#8 edit
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/CMakeLists.txt#5 edit
... //depot/stg/opencl/drivers/opencl/cmake/modules/FindROCR.cmake#1 add
... //depot/stg/opencl/drivers/opencl/runtime/CMakeLists.txt#3 edit
2017-05-17 11:40:24 -04:00
foreman 85b83e06cf P4 to Git Change 1410629 by lmoriche@lmoriche_palamida on 2017/05/17 11:27:06
SWDEV-102733 - [OCL-LC-ROCm] Cmake build Write CMakeLists.txt to enable building with and without the DK environment
	- Add FindROCR.cmake, and use find_package() to locate the ROCR

Affected files ...

... //depot/stg/opencl/drivers/opencl/CMakeLists.txt#8 edit
... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/CMakeLists.txt#5 edit
... //depot/stg/opencl/drivers/opencl/cmake/modules/FindROCR.cmake#1 add
... //depot/stg/opencl/drivers/opencl/runtime/CMakeLists.txt#3 edit
2017-05-17 11:40:24 -04:00
foreman c84fa0811c P4 to Git Change 1410625 by gandryey@gera-w8 on 2017/05/17 11:20:57
SWDEV-79445 - OCL generic changes and code clean-up
	Remove unused wait() call

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocmemory.cpp#20 edit
2017-05-17 11:29:38 -04:00
Ben Sander ee37a31799 Return precise address for hipHostGetDevicePointer. 2017-05-17 07:36:06 -05:00
foreman 540e578066 P4 to Git Change 1410518 by johtaylo@johtaylo-jtincrementor-increment on 2017/05/17 03:00:04
SWDEV-2 - Change OpenCL version number from 2421 to 2422.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#2169 edit
2017-05-17 07:14:00 -04:00
Aditya Atluri 93fa174900 changed vector types to make sure it generate proper llvm vector types
Change-Id: I6c4616dae137dc4eac35e5827dc5b7f3251e0247
2017-05-16 21:35:40 -05:00
Ben Sander 848d77ebb5 Make hipMultiThreadStreams1 test a little harsher.
Fail faster if synchronization rules are violated.
Run vectorAddRevers to read last elements of array first - if the
vector add kernel starts before preceding copy finishes we
will read stale data and flag the error.

Increase default array sizes, so synchronization errors more easily
exposed.
2017-05-16 19:04:25 -05:00
Ben Sander fad9104b5f Fix HIP_TRACE_API so kernel launch only printed when requested. 2017-05-16 19:04:25 -05:00
Ben Sander 8bc6ee5932 Add initial HIP_SYNC_NULL_STREAM=0 mode.
This eliminates host-synchronization for null stream.  Instead, the
null-stream uses GPU-side events to wait for other streams.
Default is OFF pending additional testing.

Add enhanced null-stream test.

Also refine HIP_TRACE_API.
2017-05-16 19:04:25 -05:00
Ben Sander 5d2072aba1 Update tests README 2017-05-16 19:04:25 -05:00
Ben Sander a5a12942b2 Doc update - split hip_debugging.md into separate file. 2017-05-16 19:04:25 -05:00
Ben Sander 61c8633113 split debugging into separate .md file 2017-05-16 19:04:25 -05:00
Ben Sander 7e7ba5027f Add HIP_TRACE_API=4. Only display memory allocation/free apis. 2017-05-16 19:04:25 -05:00