AI/rocm-systems
2
0
Fork 0
Code Issues Pull-Requests Actions Pakete Projekte Releases Wiki Aktivität
3.349 Commits 3 Branches 0 Tags
7ebbbd35254be59b305bbca10e8a4b472eaedf30
Commit graph

744 Commits

Autor SHA1 Nachricht Datum
Rahul Garg 131c0ea326 Fix review comments 2019-03-02 23:38:37 +05:30
Rahul Garg 77c6d5451f Fix hipMemset test for HIP/NVCC 2019-03-01 03:46:57 +05:30
Rahul Garg 026dc9457e Update hipMemset test 2019-02-28 06:54:49 +05:30
Evgeny Mankov 3a8ef9c8a2 [HIPIFY][tests] caffe2 test fix 2019-02-25 17:12:32 +03:00
Evgeny Mankov 2619f22e5c [HIPIFY][Caffe2] Initial Caffe2 support 2019-02-23 20:46:22 +03:00
Evgeny Mankov d26dc3cea7 [HIPIFY][DNN] Revise and sync with CUDNN 10 and hipDNN 
+ CUDNN_API_supported_by_HIP.md update
+ CUDA_Runtime_API_functions_supported_by_HIP.md update
+ libraryPropertyType_t is moved to Runtime
+ test update
 2019-02-13 19:40:59 +03:00
Evgeny Mankov 7ba8c440e6 [HIPIFY][RAND] revise 
+ add missing functions
+ add minimum rocRAND support
+ updated CURAND_API_supported_by_HIP.md
 2019-02-08 17:54:28 +03:00
Evgeny Mankov 385fe36f2a [HIPIFY][BLAS][tests] Add tests on hipifying to 'roc' 2019-02-07 19:25:23 +03:00
Maneesh Gupta 50a12f1ce1 Merge pull request #863 from aaronenyeshi/add-dot-instructions 
Add Dot functions as amd_mixed_dot function
 2019-01-29 06:01:01 +05:30
Aaron Enye Shi ede494bd5f Fix mixed dot for Jenkins rocm_2_0 2019-01-24 22:48:35 +00:00
Aaron Enye Shi 7e46f0639c Exclude mixed dot functions from nvcc path 2019-01-24 21:00:45 +00:00
Evgeny Mankov 4ff355357f [HIPIFY] CUDA Runtime API functions total revise (up to CUDA 10.0) 
+ for all CUDA versions starting from 7.0
+ add missing functions
+ sync with HIP
+ update CUDA_Runtime_API_functions_supported_by_HIP.md
+ update affected tests
+ fix typos
+ formatting, annotating
 2019-01-24 14:05:36 +03:00
Aaron Enye Shi f31aa946e7 Fix mixed dot for Jenkins rocm_head 2019-01-23 16:39:25 +00:00
Aaron Enye Shi 1d70f4217e Add tests for dot functions 2019-01-22 22:32:28 +00:00
Evgeny Mankov 841e056935 [HIPIFY][tests] Exclude one more test if CUDA == 7.0, which is of CUDA 7.5 2019-01-22 22:26:43 +03:00
Yaxun Sam Liu aa6abb0dcf Fix get_time in tests for nvcc 2019-01-21 14:07:50 -05:00
Yaxun Sam Liu 05b5e30187 Use chrono instead of sys/time.h in test 
sys/time.h is not available on Windows. Use C++11 <chrono> instead so that the tests
compile on Windows.
 2019-01-20 00:05:27 -05:00
Evgeny Mankov 0206fadc14 [HIPIFY] option cuda-path might be used with '--' as well as with a single dash: '-' 
+ Update README.md
 2019-01-10 18:01:33 +03:00
Evgeny Mankov fbfe005e4e [HIPIFY] Introduce CUDA installation path option '-cuda-path' 
Repeats clang's '--cuda-path' option.

[Reason]
In case of absence of any other clang's options setting '-cuda-path' allows not to specify separator '--' before clang's '--cuda-path'.

+ Tests and scripts are updated accordingly.
 2019-01-09 20:18:36 +03:00
Maneesh Gupta 580c808400 Update hipAsynchronousStreams.cpp 
This test needs to be excluded from nvcc platform
 2019-01-07 09:26:21 +05:30
Sameer Sahasrabuddhe 8fe2a4b167 introduce test for asynchronous streams 
This test is identical to the HCC tests under
tests/Unit/OCKL/AsyncStreams, but only for synchronous operation.
 2019-01-04 05:51:50 -05:00
emankov 9abcad9407 [HIPIFY][tests] CUDA 7.0 is also supported 
+ Exclude 1 test for CUDA 7.0
+ Update README.md
 2019-01-04 01:32:30 +03:00
emankov 65155c7159 [HIPIFY][tests] Exclude CUDA 8.0 (9.0) tests if CUDA < 8.0 (9.0) 2019-01-03 18:06:49 +03:00
emankov 7fb0f4ee2a [HIPIFY][tests] Reduce cudaRegister.cu test 2019-01-03 17:17:21 +03:00
Maneesh Gupta 4d0266a495 Merge pull request #818 from yxsamliu/fix-hip-config 
Add default arguments for hipConfigureCall
 2019-01-01 10:25:25 +05:30
Maneesh Gupta cb8e28144a Merge pull request #821 from gargrahul/remove_redundancy_memset3d_test 
[dtest] Remove redundant arguments from hipMemset3D test
 2018-12-31 15:32:19 +05:30
emankov dec459efca [HIPIFY] Get rid of setting '--cuda-gpu-arch=' 
[Reasons]
+ We don't compile kernel code at least for now as HIP kernel syntax is almost equal CUDA's;
+ clang always includes PTX in its binaries, so e.g. a binary compiled with --cuda-gpu-arch= would be forwards-compatible with e.g. sm_35 GPUs.
 2018-12-29 19:48:28 +03:00
emankov 7ea586c323 [HIPIFY] Starts implicitly setting '-x cuda' by hipify-clang itself 
+ No need in setting '-x cuda' for the user anymore;
+ Testing and Readme updated accordingly.
 2018-12-29 17:04:59 +03:00
emankov 2d3f62c3e1 [HIPIFY][win] Set -std=c++14 if MSVC 
+ Due to latest Windows SDK implementation based on c++14 features, '-std+c++14' is a must;
+ Doesn't affect older versions of MSVC 2017, 2015 (the latest supported is MSVC 2015);
+ On linux still '-std=c++11' is a must;
+ Exclude '-std=' from lit config.
 2018-12-29 15:51:06 +03:00
Rahul Garg 4272119eeb Remove redundant arguments from hipMemset3D test 2018-12-20 06:51:22 +05:30
Yaxun Sam Liu ec03a8b352 Add default arguments for hipConfigureCall 2018-12-18 15:48:21 -05:00
Aaron Enye Shi 58dfeff27a Add *_rn functions back into HIP intrinsics 
Add back the round-to-nearest-even intrinsics back to HIP math
intrinsics as it caused regression.
 2018-12-18 19:31:54 +00:00
Evgeny Mankov 3d528cf491 [HIPIFY][tests][fix] Add correct CUDA version detection while testing 
+ Just do load_config for lit.site.cfg earlier in lit.cfg to get initialized cuda_version
+ Remove lit.cfg.in for CMake variables substitution
+ Return back lit.cfg
+ Exclude CUDA 9.0 tests if CUDA < 9.0
+ Report CUDA version before running tests
 2018-12-12 17:05:01 +03:00
Evgeny Mankov b955e317a7 [HIPIFY][tests][fix] Add correct CUDA version detection while testing 
+ Add lit.cfg.in for CMake variables substitution to now generated lit.cfg
+ Exclude lit.cfg
+ Exclude CUDA 9.0 tests if CUDA < 9.0
+ Report CUDA version before running tests
 2018-12-11 22:03:11 +03:00
Evgeny Mankov 09612ac03f [HIPIFY][tests] Reorganization 2018-12-04 20:47:34 +03:00
Evgeny Mankov 72d40db358 [HIPIFY][SPARSE] Add 3 more CUDA 10.0 tests 
+ lit update
+ fix typos
 2018-12-04 19:24:29 +03:00
Evgeny Mankov 116b9191f7 [HIPIFY][SPARSE] Matrix Reorderings and Format Conversion Reference 
+ cuSPARSE is supported up to CUDA 10.0
+ cuSPARSE_08 test (CUDA 10.0)
+ update CUSPARSE_API_supported_by_HIP.md
+ lit: add a rule for CUDA 10.0 tests excluding
 2018-11-30 15:33:57 +03:00
Evgeny Mankov 8a84c665e2 [HIPIFY][SPARSE] Preconditioners Reference: Tridiagonal and Pentadiagonal solves 
+ Tridiagonal Solve, Batched Tridiagonal and Pentadiagonal Solve
+ cuSPARSE_07 test
+ update CUSPARSE_API_supported_by_HIP.md
 2018-11-29 18:46:51 +03:00
Evgeny Mankov 6c085c83a2 [HIPIFY][SPARSE] Preconditioners Reference: Incomplete LU Factorization: level 0 
+ cuSPARSE_06 test
+ update CUSPARSE_API_supported_by_HIP.md
 2018-11-29 15:59:58 +03:00
Evgeny Mankov b39bd8c9a9 [HIPIFY][SPARSE] Preconditioners Reference: Incomplete Cholesky Factorization: level 0 
+ cuSPARSE_05 test
+ update CUSPARSE_API_supported_by_HIP.md
 2018-11-28 20:10:30 +03:00
Evgeny Mankov 11fedfbff6 [HIPIFY][SPARSE] Extra functions 
+ cuSPARSE_04 test
+ update CUSPARSE_API_supported_by_HIP.md
 2018-11-27 12:41:50 +03:00
Evgeny Mankov 82bbaf0b70 [HIPIFY][SPARSE] Level 3 functions 
+ cuSPARSE_03 test
+ update CUSPARSE_API_supported_by_HIP.md
 2018-11-27 11:57:25 +03:00
Evgeny Mankov c7f6266824 [HIPIFY][SPARSE] Helper and Level 1,2 functions 
+ 2 cuSPARSE tests
+ update CUSPARSE_API_supported_by_HIP.md
 2018-11-26 15:11:52 +03:00
Maneesh Gupta bcea027bf1 Merge pull request #748 from mkuron/getsymboladdress 
Implement hipGetSymbolAddress and hipGetSymbolSize
 2018-11-21 10:32:01 +05:30
emankov 30c387a811 [HIPIFY][SPARSE] Initial support 2018-11-20 17:58:05 +03:00
Michael Kuron 8610128c3e Merge branch 'master' into getsymboladdress 2018-11-20 12:03:22 +01:00
Aaron Enye Shi 0cfaa52d15 Guard rcp rounded implementation as well 
Since rcp implementations of non-default rounded versions are not correct or supported in OCML, guard them using the same macro OCML_BASIC_ROUNDED_OPERATIONS. Also update the docs and tests.
 2018-11-06 19:53:28 +00:00
Aaron Enye Shi e59c33250a Guard the OCML rounded operations instead 
Instead of commenting all these functions out, guard the functions with a macro OCML_BASIC_ROUNDED_OPERATIONS.
 2018-11-06 16:32:14 +00:00
Michael Kuron 49cc703d7a Test for hipGetSymbolSize and hipGetSymbolAddress 2018-11-06 11:39:34 +01:00
Aaron Enye Shi 2b108a2cfd Remove non-working non-default-rounded math apis 
In ROCm-Device-Libs, they have dropped the non-default-rounded versions of add, sub, mul, div, sqrt and fma. Therefore, ocml has removed the rte, rtp, rtn, and rtz counterparts. This will remove the same math APIs in HIP for _ru, _rd, _rn, and _rz.
 2018-11-05 22:34:16 +00:00