AI/rocm-systems
2
0
Форкнуть 0
Код Задачи Запросы на слияние Действия Пакеты Проекты Релизы Вики Активность
3 997 коммитов 3 Ветки 0 Теги
09bb83cf2cdcc8c7c772fe1d7c88d8dfbc54536a
Граф коммитов

3577 Коммитов

Автор SHA1 Сообщение Дата
Rahul Garg 09bb83cf2c Fix PCI Domain ID query (#1424) 
* Fix PCI Domain ID query

* Update BDF comment


[ROCm/clr commit: e2260d82a6]
 2019-10-07 14:11:52 +05:30
Evgeny Mankov 48224daaa3 [HIPIFY][SPARSE] sync with hipSPARSE 
+ Update doc and hipify-perl accordingly


[ROCm/clr commit: 201e6deaac]
 2019-10-05 15:53:04 +03:00
Evgeny Mankov 2fab0ddcc8 [HIPIFY] CUDA2HIP_Perl code clean-up and formatting 
+ Generated hipify-perl is unchanged


[ROCm/clr commit: 63af6684fa]
 2019-10-04 14:31:15 +03:00
mhbliao 85640091d5 [hip] Add the missing header <sstream>. (#1498) 
[ROCm/clr commit: 5fbb09fb39]
 2019-10-04 13:39:29 +05:30
Philip Salzmann 2dee09bd5b Fix uninitialized var in hipDeviceGetAttribute (#1497) 
This fixes the usage of an uninitialized cdattr variable in hipDeviceGetAttribute for the CUDA backend when taking the switch default, as detailed in #1317.

Note that the directed_tests/runtimeApi/device/hipGetDeviceAttribute.tst test fails for me, but it already did before applying this patch. Let's see what CI says!

[ROCm/clr commit: 11f23bba39]
 2019-10-04 13:39:19 +05:30
Rahul Garg 8d500c1b59 Add texref get APIs support (#1471) 
Added support for -
    hipTexRefGetArray
    hipTexRefGetAddressMode
    hipTexRefGetAddress



[ROCm/clr commit: d5a61736d8]
 2019-10-04 13:38:45 +05:30
Sarbojit2019 60373919d4 Removed definition of abs(), real() & imag() from hip_complex.h (#1448) 
Addresses SWDEV-201461.

[ROCm/clr commit: a7f52f8ea1]
 2019-10-04 13:38:02 +05:30
ansurya efc64a4f8a Added new Memory API's (#1399) 
Added new memory API's hipMemAllocPitch, hipMemAllocHost, hipMemsetD16, hipMemsetD16Async, hipMemsetD8Async
Modified to support all scenarios hipMemcpyParam2DAsync, hipMemcpyParam2D.


[ROCm/clr commit: 96890792b0]
 2019-10-04 13:36:31 +05:30
Maneesh Gupta e5a2ba9602 hipcc defaults to code object v3 (#1298) 
Change-Id: I272ec61090fa504d631fb54371a73dd6fd17a167

[ROCm/clr commit: c8f38a139a]
 2019-10-04 13:35:21 +05:30
Siu Chi Chan 1a7fc4e5c1 fix kernel descriptor bug with code object v3 
Change-Id: I9306b2baf36d338e36c5ab1226f74373a61a5ae0


[ROCm/clr commit: d8e09c4b70]
 2019-10-03 10:56:35 -04:00
Evgeny Mankov 7064f3a832 [HIPIFY][perl] Start to generate hipify-perl under an option -perl 
+ Get rid of --o-perl-map option for Perl output filename: the output name now is always hipify-perl
+ Option --o-perl-map-dir renamed to --o-hipify-perl-dir

Cmd example for generating hipify-perl script to the particular directory:
hipify-clang -perl --o-hipify-perl-dir=../../bin


[ROCm/clr commit: 04988dcdea]
 2019-10-03 16:30:48 +03:00
Evgeny Mankov cc2af89494 [HIPIFY] Finishing with generation of hipify-perl 
+ hipify-perl script is entirely generated by hipify-clang under an option -perl now
+ hipify-perl still has correctness gaps comparing to hipify-clang: they will be eliminated AMAP further


[ROCm/clr commit: 94aac73b3b]
 2019-10-03 16:00:52 +03:00
Evgeny Mankov 85ae269393 [HIPIFY] Return to wrapping with HIP_KERNEL_NAME(...) macro of a template instantiation kernel launch 
[REASON]

1. hip-clang is fine with the templated kernel launch, brackets are unneeded: HIP_KERNEL_NAME(...) __VA_ARGS__
2. HCC is not, thus: HIP_KERNEL_NAME(...) (__VA_ARGS__)

[TODO] Clean-up entirely kernel name wrapping when HCC is finally obsolete.

+ Update perl generation, hipify-perl, and affected tests accordingly.


[ROCm/clr commit: fee2f0fa78]
 2019-10-02 16:01:07 +03:00
Evgeny Mankov 159fe8dbc3 Merge pull request #1492 from emankov/master 
[HIPIFY][#1490][fix] Populate the list of supported device atomic functions

[ROCm/clr commit: 8549cc7e1c]
 2019-10-01 14:00:07 +03:00
Evgeny Mankov 6dab0c3faa [HIPIFY][#1490][fix] Populate the list of supported device atomic functions 
+ Update hipify-perl accordingly
+ Add atomics.cu test on all atomics


[ROCm/clr commit: 2371150401]
 2019-10-01 13:57:46 +03:00
Evgeny Mankov 26bdb8c7f0 Merge pull request #1489 from emankov/master 
[HIPIFY][perl] Generate transformKernelLaunch

[ROCm/clr commit: 4a060525d5]
 2019-10-01 11:18:41 +03:00
Evgeny Mankov aaa4495aeb [HIPIFY][perl] Generate transformKernelLaunch 
+ Update hipify-perl accordingly
+ Minor refactoring


[ROCm/clr commit: 0d8d0f1ebc]
 2019-10-01 11:17:27 +03:00
satyanveshd 1c9dc1331e Reimplement hipMemGetInfo (#1447) 
Addresses SWDEV-136570. hipMemGetInfo changed to compute free memory based on information from kfd instead of relying on hc::am_tracker.


[ROCm/clr commit: 1ae4cbed4d]
 2019-10-01 12:40:36 +05:30
Maneesh Gupta 078bafd36d Bump version to 2.10 (#1483) 
Change-Id: I22ac0ae723bc537e027306dbb0f1594552fd33a4

[ROCm/clr commit: 03c1fb1a8f]
 2019-10-01 12:39:08 +05:30
Rahul Garg 7c26512e27 [ci] Enable tests on ROCm 2.8 (#1482) 
[ROCm/clr commit: 85a422916a]
 2019-10-01 12:38:47 +05:30
Sarbojit2019 3a20a14134 [dtests] Separated C macros from CPP header file (#1429) 
* Separated C macros from CPP header file

* Updated review comment


[ROCm/clr commit: 6c06fb5f0a]
 2019-10-01 12:38:32 +05:30
Evgeny Mankov 1886fdcbeb [HIPIFY][perl] Generate transformExternShared function 
+ Update hipify-perl accordingly


[ROCm/clr commit: bf5546f35e]
 2019-09-30 17:47:46 +03:00
Evgeny Mankov fa0270ae7c Merge pull request #1485 from emankov/hipify-clang 
[HIPIFY][perl] Generate simpleSubstitutions function

[ROCm/clr commit: 0e91fb36d2]
 2019-09-30 16:55:31 +03:00
Evgeny Mankov 4d2de06f2c [HIPIFY][perl] Generate simpleSubstitutions function 
+ Update hipify-perl accordingly


[ROCm/clr commit: 4860559996]
 2019-09-30 16:53:19 +03:00
Jeff Daily 280212353c hipModuleUnload should remove global variables from memtracker (#1464) 
[ROCm/clr commit: 2a53299f07]
 2019-09-30 10:41:20 +05:30
Yaxun (Sam) Liu 2ed78f20a2 Fix cast of __half for HIP-clang (#1475) 
[ROCm/clr commit: 7b3b82bd0d]
 2019-09-30 10:40:42 +05:30
satyanveshd 491270de07 Map clock64() to __builtin_readcyclecounter() (#1473) 
Fixes SWDEV-203215.

[ROCm/clr commit: ff1b23b558]
 2019-09-30 10:40:31 +05:30
Rahul Garg e410cb47bd [docs] Fix sample links in hip_profiling (#1472) 
[ROCm/clr commit: 9816122a62]
 2019-09-30 10:40:01 +05:30
eshcherb ce4c08dc1f to include hip_prof_str.h under USE_PROF_API macro (#1470) 
[ROCm/clr commit: b8139f7da4]
 2019-09-30 10:39:41 +05:30
Alex Voicu dfbb6d555c Optimise the gridDim.n * blockDim.m idiom (#1468) 
[ROCm/clr commit: 33c40ee510]
 2019-09-30 10:39:23 +05:30
AlexBinXie 933d31f374 [hip] Added gfx803,gfx1010,gfx1012 kernel binary hipModule test. (#1444) 
This is to address hipModule test seg fault in GFX1010

[ROCm/clr commit: cda068edd5]
 2019-09-30 10:38:37 +05:30
Evgeny Mankov 20496c8486 [HIPIFY] Reconcile transformation (statistics) types between hipify-clang and hipify-perl 
+ Update hipify-perl accordingly


[ROCm/clr commit: b42a476614]
 2019-09-29 17:33:37 +03:00
Evgeny Mankov de20b6c2f0 [HIPIFY][perl] Generate stat functions 
TODO: make statistics as much as similar to hipify-clang's


[ROCm/clr commit: 46273f86a1]
 2019-09-29 14:20:32 +03:00
Evgeny Mankov 63120a9871 [HIPIFY][doc] The latest cuDNN 7.6.4 is supported 
[ROCm/clr commit: bf2b8a6373]
 2019-09-28 20:49:56 +03:00
Evgeny Mankov f876daa083 [HIPIFY][perl] Code clean-up before continuing generation 
[ROCm/clr commit: cdd1888293]
 2019-09-28 20:44:15 +03:00
Rahul Garg ad563cef42 Merge pull request #1457 from yxsamliu/launch2 
Add new kernel launching API for hip-clang

[ROCm/clr commit: 6cfb65d11b]
 2019-09-27 12:36:26 -07:00
Evgeny Mankov a48cfd56e9 [HIPIFY] Generate more hipify-perl (continuation) 
+ Minor hipify-perl clean-up
+ Minor CUDA2HIP_Perl refactoring


[ROCm/clr commit: 0d36e13ce9]
 2019-09-27 19:10:54 +03:00
Yaxun (Sam) Liu 33b6f971fa Add new kernel launching API for hip-clang 
[ROCm/clr commit: f12b060a3b]
 2019-09-26 20:15:24 -04:00
Evgeny Mankov a5a10dc4f9 [HIPIFY] Add supported fp16 functions 
+ Update hipify-perl accordingly


[ROCm/clr commit: ff1699397b]
 2019-09-26 18:13:20 +03:00
Evgeny Mankov cea5964fd0 Merge pull request #1467 from emankov/master 
[HIPIFY][doc] Update README.md

[ROCm/clr commit: fde2733451]
 2019-09-26 17:11:16 +03:00
Evgeny Mankov 2fac98dd7f [HIPIFY][doc] Update README.md 
+ Testing, versioning


[ROCm/clr commit: e5ed366bad]
 2019-09-26 17:08:29 +03:00
Sarbojit2019 3ef2e123bb [HIP] Add tccDriver info in hipDeviceProp 
Fixes #1433.

[ROCm/clr commit: db4c0c0811]
 2019-09-26 13:53:33 +05:30
Evgeny Mankov 3ee668a9b9 [HIPIFY][#1439] Add reinterpret_cast to args of some functions 
+ Perl part of [#1458]
+ Affected functions: hipFuncSetCacheConfig, hipFuncGetAttributes
+ Implement function generateHostFunctions() in hipify-clang for that purposes
+ Update hipify-perl accordingly


[ROCm/clr commit: 1db7c1d694]
 2019-09-25 18:53:17 +03:00
Evgeny Mankov 1fa72c7dc8 [HIPIFY][perl] Get rid of obsolete warnings on cuBLAS and cuRAND 
[ROCm/clr commit: 76f2202abf]
 2019-09-25 17:37:41 +03:00
Evgeny Mankov 6fdfbb4062 [HIPIFY] HipifyAction refactoring 
+ Merge of cudaSymbolFuncCall and cudaReinterpretCastArgFuncCall matchers into a single cudaHostFuncCall matcher
+ More const std::string declarations
+ Formatting


[ROCm/clr commit: 7c62b1b2d0]
 2019-09-24 10:33:51 +03:00
Evgeny Mankov d3adb0478f [HIPIFY][#1439] Add reinterpret_cast to args of some functions 
+ Affected functions: hipFuncSetCacheConfig, hipFuncGetAttributes
+ Add a corresponding Matcher cudaReinterpretCastArgFuncCall
+ Add reinterpret_cast.cu test

TODO: Do the same for hipify-perl


[ROCm/clr commit: cf467b4550]
 2019-09-24 09:44:17 +03:00
Evgeny Mankov 7bdf8b4514 [HIPIFY][doc] Return X86 target back in LLVM's cmake 
Due to LLVM 9.0.0 release, which still needs X86 for NVPTX, LLVM trunk - doesn't.


[ROCm/clr commit: f7b3dcd6d7]
 2019-09-23 22:20:23 +03:00
Evgeny Mankov 6dfc9a99d6 [HIPIFY][perl] CUDA2HIP_Perl code cleanup 
[ROCm/clr commit: a13540b97a]
 2019-09-22 23:43:07 +03:00
Evgeny Mankov fa6a69e375 [HIPIFY][perl] generateSymbolFunctions() small refactoring 
[ROCm/clr commit: d6b0ab0fd2]
 2019-09-22 22:34:07 +03:00
Evgeny Mankov ef23cdda41 [HIPIFY][#1435][perl] Add HIP_SYMBOL wrapper to the templated Device Symbol argument of the following functions: 
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


[ROCm/clr commit: 72a658295f]
 2019-09-21 07:33:17 +03:00