Граф коммитов

82 Коммитов

Автор SHA1 Сообщение Дата
Vladislav Sytchenko 8f0a226660 _aligned_malloc() on Windows first takes size, then alignment, which is the opposite of how the similar function behaves on Linux. Memory allocated by it also has to be freed using _aligned_free(), unlike Linux where we can use regular free().
Edit aligned_alloc() macro and add a aligned_free() one to align with the above behaviour.
2019-10-17 18:58:32 -04:00
Vladislav Sytchenko c747b77ac1 hipMemset2D and hipMemset3D tests should be passing by default. 2019-10-16 11:02:38 -04:00
Vladislav Sytchenko f5af263aba In the hipMemset2D and hipMemset3D tests synchronize with the default stream after performing an async memset. 2019-10-15 17:15:49 -04:00
Vladislav Sytchenko 346bfa90d6 Update indentation in the hipMemset3D test. Replace all tabs with four spaces. 2019-10-15 15:29:14 -04:00
Vladislav Sytchenko 00425bdf3d Add async subtest to hipMemSet3D 2019-10-15 14:24:04 -04:00
Vladislav Sytchenko 3db2ecc52b hipMemset2D test should pass only if both async and sync subtests pass. 2019-10-15 14:20:14 -04:00
ansurya 87834500b9 Fix for directed tests failure (#1511)
directed_tests/runtimeApi/module/hipLaunchCooperativeKernel.tst - Disabling test temporarily until driver support is available.
directed_tests/runtimeApi/memory/hipArray.tst - Disabling test temporarily to reimplement it correctly.
2019-10-10 19:58:41 +05:30
ansurya ba9c6e13e4 Added new Memory API's (#1399)
Added new memory API's hipMemAllocPitch, hipMemAllocHost, hipMemsetD16, hipMemsetD16Async, hipMemsetD8Async
Modified to support all scenarios hipMemcpyParam2DAsync, hipMemcpyParam2D.
2019-10-04 13:36:31 +05:30
AlexBinXie 5ed1f3e2c8 [hip]Skip test when hipHostMallocCoherent is not supported by implementation (#1380) 2019-09-16 08:31:43 +00:00
Rahul Garg 44422065a1 [dtests] refactor windows specific changes (#1313)
* [dtests] refactor windows specific changes

* Refactor hipMemoryAllocateCoherentDriver - PR- 1309

* Fix missing z in _putenv_s

* Revert "Fix missing z in _putenv_s"

This reverts commit 099a1b20a5c75c5f122d57c0ad2bca01745cdc9c.

* Refactor changes from PR 1299

* Update hipEnvVarDriver.cpp
2019-08-16 02:13:00 +00:00
amd-lthakur 9b31d26237 [dtests] Fix build issues with hipMemoryAllocateCoherentDriver.cpp on windows (#1309)
Compilation error being observed due to popen(), pclose() and setenv() linux calls on windows. Replaced with appropriate calls on windows.
2019-08-09 11:53:16 +00:00
amd-lthakur 9abae7114c [dtests] Fix build issue with hipMemcpy_simple.cpp on windows (#1306)
Compilation error being observed on windows due to aligned_alloc() call. Mapped the call to _aligned_malloc() for windows.
2019-08-09 11:52:46 +00:00
Aaron Enye Shi 0de4caa085 Fix hipMemcpy-size test running out of Host Mem (#1224)
* Fix hipMemcpy-size test running out of Host Mem

The hipMemcpy-size uses a maxElem calculated from the total GPU mem /8. Then it will allocate 4 times that amount of host memory. This tests begins failing when there is not enough host memory, such as on systems with 32GB GPU mem, and 16GB RAM. This fixes the test if not enough host memory is available on the system.

* Add windows support to hipMemcpy-size fix

* avoid linking extra libs for windows

* HIPMemcpy-size Remove freeCPU including swap
2019-07-24 07:49:20 +00:00
Maneesh Gupta 9f2d1453fb [hit] Rename RUN -> TEST & RUN_NAMED -> TEST_NAMED
Change-Id: I75e24f15129973cee15fc9dac65d678bd2172074
2019-05-09 09:59:18 +05:30
Rahul Garg 2bc2c46d4d Add hipMallocManaged default functional support (#1036)
* Add hipMallocManaged default functional support

* Fix build error

* Add dtest
2019-04-24 16:50:03 +05:30
Rahul Garg e5e7651a4a Fix D2DAsync test 2019-04-18 07:35:06 +05:30
Maneesh Gupta 51f29b9cee Update hipMemset.cpp
Address build issues on nvcc path.
2019-03-05 12:11:11 +05:30
Wen-Heng (Jack) Chung bfde8a7fab Add direct test for hipMemsetD32 and hipMemsetD32Async 2019-03-04 17:20:32 +00:00
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
Rahul Garg 4272119eeb Remove redundant arguments from hipMemset3D test 2018-12-20 06:51:22 +05:30
Maneesh Gupta 52e320f396 Replace hipLaunchKernel -> hipLaunchKernelGGL
Change-Id: I4d99009e1199811d417becf1e1b934ec4d4e30be
2018-10-17 14:32:25 +05:30
saleelk 8bc73424d1 Fix hipHostRegister mem leak and copy kind in hipMemcpyAsync (#587) 2018-07-23 11:04:14 +05:30
Rahul Garg 18944113ed Cleaned up hipMemset3D test 2018-07-06 14:54:08 +05:30
Rahul Garg da302c3e93 Added hipMemset3D 2018-05-07 10:24:30 +05:30
Rahul Garg 3cfb9c0d40 Added hipMemset2DAsync support 2018-04-17 18:27:27 +05:30
Maneesh Gupta 1ba06f63c4 Apply .clangformat to all repo source files
Change-Id: I7e79c6058f0303f9a98911e3b7dd2e8596079344
2018-03-12 11:29:03 +05:30
Phaneendr-kumar Lanka 7cceebd9ed [nvccTests] Resubmit hipMemcpyDtoD & inline_asm_vadd 2017-12-18 14:46:19 +05:30
Phaneendr-kumar Lanka eea7d495c7 [nvccWarnings] Fix warnings seen with dtests on nvcc path 2017-12-14 14:10:37 +05:30
Alex Voicu 32e11e7dc6 Revert "Revert adoption of CUDA indexing in general - this can only work with later versions of the compiler, just like module based dispatch, and thus must be guarded against usage in earlier (e.g. 1.6) versions."
This reverts commit d2fd1f5
2017-11-29 21:49:10 +00:00
Alex Voicu fbaf729f88 Revert "Revert adoption of CUDA indexing in general - this can only work with later versions of the compiler, just like module based dispatch, and thus must be guarded against usage in earlier (e.g. 1.6) versions."
This reverts commit d2fd1f5
2017-11-29 21:36:29 +00:00
Alex Voicu c5f2b22d0d Re-sync with upstream. 2017-11-20 15:34:50 +00:00
Ben Sander 5a7a28ad29 Merge branch 'feature_natural_indexing' of https://github.com/AlexVlx/HIP 2017-11-19 15:25:17 -06:00
Ben Sander e0c3f684ae Temporarily disable P2P on nvidia (fails on dual GPU) 2017-11-19 15:21:37 -06:00
Alex Voicu cffd0e14eb This implements the trivial change needed to move back from the hip{Something}_{x, y, z} macros to the natural CUDA syntax of Something.{x, y, z}. This is contained in lines 384-404 in hip_runtime.h. All of the other changes have to do with changing unit tests to use this syntax. The macros are retained for backwards compatibility. 2017-11-19 01:54:12 +00:00
Alex Voicu bd78459690 Update new tests so as to make them work with new variadic based launch mechanisms. 2017-11-10 00:14:34 +00:00
Alex Voicu 3d248927e4 Merge remote-tracking branch 'origin/master' into feature_use_module_based_dispatch_instead_of_pfe
# Conflicts:
#	tests/src/runtimeApi/stream/hipStreamSync2.cpp
2017-11-08 10:26:30 +00:00
Alex Voicu d8e323d4b5 Clean up trailing whitespace so as to reduce noise in #246. 2017-11-08 00:08:55 +00:00
Phaneendr-kumar Lanka 9ee330451a Modified tests as per review 2017-11-06 15:13:36 +05:30
Phaneendr-kumar Lanka 093279b10b More tests for memory,stream & error APIs 2017-11-06 11:31:52 +05:30
Alex Voicu c2482d1255 This switches HIP from its currently convoluted macro + pfe based dispatch mechanism to a more natural one partially based on the existing module API. The basic idea is that HCC will always correctly emit __global__ functions: as empty-bodied stubs, on host, and as kernels, on device. It then becomes trivial to obtain the mangled name on host, at dispatch, from the function's address, and then to use the mangled name to retrieve the kernel. This should address all problems stemming from serialisation, dubious mismatches due to the manufactured functor, macro-isms et al. It also immediately enables support for generalised globals as a consequence of that being available in the module API. Finally, it will make debug much easier, since the actual names of the __global__ functions will automatically be used in traces etc. One detail is that due to how dispatch works now (hipLaunchKernel and hipLaunchKernelGGL are themselves variadic function templates which deduce the function type of the callee), in certain cases it may be necesssary to insert explicit casts to ensure that the variadic argument list selects a viable overload - this can be observed in some unit tests. Eventually we may be able to remove this limitation, but for now it does not appear terribly onerous. The code is not extremely HIPpie, nor is it fully optimised, but rather is intended as a starting point for the HIP team to make its own. 2017-11-01 15:09:59 +00:00
Ben Sander ca1230300a Clean up test to address review feedback. 2017-10-25 16:08:16 -05:00
Ben Sander 8e04b33c95 Test cleanup. 2017-10-07 07:54:10 -07:00
Rahul Garg 2f617bac6d Disable failing test p2p_copy_coherency 2017-08-02 14:47:13 +05:30
Ben Sander 7912e61502 Clean up old USE_* and RELEASE.md notes. 2017-06-23 18:05:30 -05:00
Rahul Garg a9449533a1 Update P2P test for using memcpy and kernel tests
Change-Id: Ib0f8fc9425e6e85fd11d7d02395c52bc713dcb37
2017-06-12 11:53:25 +05:30
Maneesh Gupta 1c93d8592e hipMemcpy-size test: reduce max size to make it work correctly on nvcc path
Change-Id: I9ce9f5a9e141ffd8ddf961269010b33358e02771
2017-06-07 15:25:54 +05:30
Maneesh Gupta 1efb6ce994 p2p_copy_coherency test: gracefully handle single gpu case
Change-Id: I216663f67ef58c673136332635dab8b57079b909
2017-06-07 15:23:37 +05:30
Ben Sander 8ce6d17983 Update tests.
Fix some NVCC issues.
Add hipStreamSync2, record_event tests.
2017-06-04 20:18:37 -05:00