AI/rocm-systems
2
0
Fork 0
Code Issues Pull-Requests Actions Pakete Projekte Releases Wiki Aktivität
13.687 Commits 3 Branches 0 Tags
dc34af61d77681282da4418a6efbb112337f90be
Commit graph

2422 Commits

Autor SHA1 Nachricht Datum
Assiouras, Ioannis dc34af61d7 SWDEV-543340 - Remove shared memory objects after IPC event cleanup (#745) 
This change ensures that shared memory objects (e.g., files in /dev/shm)
are unlinked once all related IPC events have been destroyed.
 2025-08-12 21:23:09 +01:00
Dittakavi, Satyanvesh c60888b0cd SWDEV-545947 - Add Implementation for hipSetValidDevices (#805) 2025-08-12 16:29:09 +05:30
Dittakavi, Satyanvesh b50d70564e SWDEV-540272 - copy params 3D struct should be populated even for memcpy 1D graph node (#702) 
SWDEV-540272 - copy params 3D struct should be populated even for 1D graph node
 2025-08-12 16:21:53 +05:30
Chaudhary, Jatin Jaikishan d7f90a3120 SWDEV-540645 - use single path for code object loading (#773) 
* use single path for code object loading
* Remove printf
* Address review comments
* Fix the declaration of ihipMallocManaged
 2025-08-12 15:15:18 +05:30
Betigeri, Sourabh 0d9428135c SWDEV-528351 - Clean up hipJitOptions (#439) 2025-08-11 21:24:44 +05:30
Betigeri, Sourabh da61c5d1da SWDEV-528351 - Cleanup template specializations and overloads for enums (#360) 
* SWDEV-528351 - Adds temlplate specialization for enums
* SWDEV-528351 - Removes template specializations for enums
 2025-08-11 21:24:24 +05:30
Chaudhary, Jatin Jaikishan ee124dda09 SWDEV-508540 - fix the max blocks per multiprocessor calculation (#737) 2025-08-11 10:17:18 +05:30
Kudchadker, Saleel 9b045922a8 SWDEV-538195 - Introduce threshold for handler submission (#723) 
- When doing device/stream sync, we can submit a handler which may
  introduce some host side delays. Use DEBUG_CLR_BATCH_CPU_SYNC_SIZE to
  batch commands for host wait. Default for HIP is 8 commands.
- Investigation is underway in ROCr but need to address this for now in
  HIP runtime.
 2025-08-06 20:34:42 -07:00
Xie, Pengda 7e777d3755 SWDEV-534394 - Kernel launch stream check (#533) 
hipErrorInvalidResourceHandle is returned if stream is not on the launching device
 2025-08-05 10:39:37 -07:00
Assiouras, Ioannis f094d16157 SWDEV-533754 - Fixed mem leak in capturing gpu packets (#365) 2025-08-04 14:58:04 +01:00
Betigeri, Sourabh 2a02d2c2f3 SWDEV-545273 - Respect HIP_LAUNCH_PARAM_BUFFER_SIZE (#770) 2025-08-03 17:32:52 -07:00
Sumner, Brian 2da49283c8 Fix undefined symbol at runtime with newer versions of LLVM (#791) 2025-08-01 08:15:02 -07:00
Patel, Jaydeepkumar c329b548eb SWDEV-541799 - Return hipSuccess from hipFreeAsync if ptr to free is … (#683) 
SWDEV-541799 - Return hipSuccess from hipFreeAsync if ptr to free is nullptr to match with hipFree.
 2025-08-01 14:41:40 +05:30
Chaudhary, Jatin Jaikishan 76d2c4135e SWDEV-543350 - restrictions on usage of hipMemcpyWithStream (#727) 
* SWDEV-543350 - Do not hipMemcpyWithStream to be called during stream capture

* update comment

---------

Co-authored-by: Jatin Chaudhary <jatchaud@amd.com>
 2025-08-01 14:41:02 +05:30
Manocha, Rahul 3f6f9d6081 SWDEV-539710 - Defer allocation of managed variable (#652) 
Co-authored-by: Rahul Manocha <rmanocha@amd.com>
 2025-07-31 08:30:23 -07:00
Xie, Jiabao(Jimbo) c88f345229 SWDEV-532824 - Fixed log_printf errors for codeql (#427) 
Co-authored-by: Jimbo Xie <jiabaxie@amd.com>
 2025-07-31 08:17:41 -07:00
Manocha, Rahul 3bbc8bd170 SWDEV-540733 - Add PCI CHIP ID device attribute (#647) 
SWDEV-540733 - Add PCI CHIP ID Device Attribute

Co-authored-by: Rahul Manocha <rmanocha@amd.com>
 2025-07-30 19:43:07 -07:00
Hernandez, Gerardo 2c6ec7df45 SWDEV-534207 - fix use of uninitialized variable at hipEventRecord_common() (#771) 2025-07-23 16:49:12 +05:30
Sang, Tao b15c1657b8 SWDEV-502579 - Remove duplicated symbols (#21) 
Remove duplicated compiler symbols from hip-rt. hip-rt should use symbols such as EF_AMDGPU_MACH_AMDGCN_XXX and
ELFABIVERSION_AMDGPU_HSA_VX defined in compiler header.
 2025-07-22 10:41:01 +05:30
Dittakavi, Satyanvesh c43ee71e96 SWDEV-542905 - Update the appropriate grid size in capture hipExtModuleLaunchKernel (#740) 2025-07-21 13:57:15 +05:30
Betigeri, Sourabh 1edac1bf3a SWDEV-537486 - Change the SPIRV bundle entry id to match 4-field triple (#513) 2025-07-17 09:29:51 +05:30
Betigeri, Sourabh 0313e20706 SWDEV-540629 - Update the current device ID (#699) 2025-07-15 21:40:26 +05:30
Sang, Tao 397f303d97 SWDEV-508776 - support VGPRs validation (#274) 
Clarify some VGPRs terms description.
Fix some wrong query logics of availableVGPRs_ and
availableRegistersPerCU_ in device info.
Add hipDeviceAttributeMaxAvailableVgprsPerThread
attribute query.
Remove hardcoding of following
 info_.vgprAllocGranularity_
 info_.vgprsPerSimd_
 2025-07-09 10:46:52 -04:00
Brzak, Branislav 32e027e66c SWDEV-540609 - hipGraph capture of hipExtModuleLaunchKernel assumed grid size was a multiple of block size (#694) 2025-07-09 16:05:40 +05:30
Belton-Schure, Aidan 81238db679 SWDEV-515426 - Remove HIP_USE_RUNTIME_UNBUNDLER (#205) 
* remove HIP_USE_RUNTIME_UNBUNDLER
* clang-format
* Generic to use comgr
* Remove HIP_ALWAYS_USE_NEW_COMGR_UNBUNDLING_ACTION flag
* Removes runtime unbundling unused and debug Code
* Removes stale functions
 2025-07-08 21:45:31 +05:30
Xie, Pengda 21f6842fff SWDEV-536925 - Check for C++ compiler instead of C when building hip (#490) 
Change cmake to use CMAKE_CXX_COMPILER_ID since CMAKE_COMPILER_IS_GNUCC is deprecated
 2025-07-01 12:59:54 -07:00
Sang, Tao a7d7687b8f SWDEV-539145 - Support extended fine grained system memory pool (#603) 
* Add hipHostMalloc() new flag hipHostMallocUncached which will force to allocate pinned
host memory on extended fine grained system memory pool.
* Add hipHostAlloc() new flag hipHostAllocUncached which will force to allocate pinned
host memory on extended fine grained system memory pool.
* Add hipHostRegister() new flag hipHostRegisterUncached which will force to map
host memory onto extended fine grained system momory pool.
 2025-07-01 10:10:10 +05:30
Arandjelovic, Marko 349b83c1f6 SWDEV-512344 - Allow unmapping of only complete subbuffers (#608) 
* Do not allow partial unmapping
* Add check to prevent out of bounds unmap
* Improve code quality
* Use getOrigin instead of getOffset
 2025-06-26 15:01:33 +05:30
Kudchadker, Saleel 5d53c83806 SWDEV-535490 - Improve logging (#441) 
- Include HIP version/githash in the logs
- Add a new method to print library path
 2025-06-24 13:00:00 -07:00
Sang, Tao 7755660eeb SWDEV-533964 - Let comgr compile SPIRV to exec directly (#494) 2025-06-23 09:59:53 -04:00
Xie, Jiabao(Jimbo) 54a2e943d7 SWDEV-535184 - Fix to switch HIP_RUNTIME_API_TABLE_STEP_VERSION 9 and 10 (#448) 2025-06-02 21:24:40 +05:30
Luzynski, Sebastian 109795f6c9 SWDEV-508967 - Fix error code mismatch for hipBindTexture API (#423) 
Signed-off-by: Sebastian Luzynski <Sebastian.Luzynski@amd.com>
 2025-05-29 13:13:12 +05:30
Betigeri, Sourabh c0c3aa3637 SWDEV-508963 - Change hipMemcpy subset of methods to use const qualifier for src (#139) (#418) 
Change-Id: Ia5b68bacff634d3d9e5757f515bac50887895bc3
 2025-05-29 13:07:51 +05:30
Manocha, Rahul 7177f1c280 SWDEV-460098 - Match hipModuleLaunchKernel and hipModuleLaunchCooperativeKernel with cuda (#414) 
Co-authored-by: Rahul Manocha <rmanocha@amd.com>
 2025-05-29 13:04:54 +05:30
Arandjelovic, Marko df06b2b692 SWDEV-531009 - Revert SWDEV-525653 - Make hipGetDeviceProperties and hipChooseDevice use th… (#381) 
This reverts commit b006380ff6.
 2025-05-29 13:02:45 +05:30
Li, Todd tiantuo d96016bcc5 SWDEV-514950 - fix hipMallocManaged & hipMemAdvise when capturing stream (#413) 2025-05-28 13:48:48 +05:30
Li, Todd tiantuo 24cdff6cce SWDEV-516511 - fix hipCreateSurfaceObject & hipDestroySurfaceObject when capturing stream (#412) 2025-05-28 13:46:17 +05:30
Li, Todd tiantuo c7f3ae6d2f SWDEV-504307 SWDEV-504308 - fix hipDeviceSetCacheConfig and hipDeviceSetSharedMemConfig when capturing stream (#411) 2025-05-28 13:44:24 +05:30
Dittakavi, Satyanvesh 93c6ffdec4 SWDEV-534857 - Fix the hip build failure on Windows (#409) 2025-05-28 13:41:31 +05:30
Manocha, Rahul b5e9bc55cb SWDEV-475482 - [6.4 Preview] Match hipTexObjectCreate with cuda (#408) 
Co-authored-by: Rahul Manocha <rmanocha@amd.com>
 2025-05-28 13:21:45 +05:30
Betigeri, Sourabh 988c42a6e1 SWDEV-508969 - Fixes error code mismatch for hipBindTexture2D API (#406) 2025-05-28 13:17:15 +05:30
Patel, Jaydeepkumar b751614cd8 SWDEV-508982 - Handle hipMemPoolCreate, hipMemPoolDestory & hipDeviceSetMemPool during stream capture. (#398) 2025-05-27 14:48:21 +05:30
Brzak, Branislav ee2d4a0358 SWDEV-518962 - Match hipCtxGetApiVersion signature with Cuda (#327) 2025-05-26 20:28:22 +05:30
Belton-Schure, Aidan 7aae33951b SWDEV-508966 - hipGetTextureAlignmentOffset return hipErrorInvalidTexture (#389) 2025-05-22 23:19:45 +05:30
Belton-Schure, Aidan 51dd60a64f SWDEV-508977 - return error in global capture mode for hipEventQuery (#388) 2025-05-22 23:17:57 +05:30
Stojiljkovic, Vladana a0e946c2cc Match hipStreamAddCallback capture behavior with nvidia (#386) 
Co-authored-by: Anusha GodavarthySurya <Anusha.GodavarthySurya@amd.com>
 2025-05-22 20:54:04 +05:30
Stojiljkovic, Vladana 63c4020e49 Match hipLaunchCooperativeKernelMultiDevice behavior with nvidia (#385) 
Co-authored-by: Anusha GodavarthySurya <Anusha.GodavarthySurya@amd.com>
 2025-05-22 20:52:37 +05:30
Stojiljkovic, Vladana 2ddfe40de0 Allow hipMemAdvise capturing only in relaxed mode (#384) 
Co-authored-by: Anusha GodavarthySurya <Anusha.GodavarthySurya@amd.com>
 2025-05-22 20:52:00 +05:30
Stojiljkovic, Vladana 69112331ac Allow hipMallocManaged capturing only in relaxed mode (#383) 
Co-authored-by: Anusha GodavarthySurya <Anusha.GodavarthySurya@amd.com>
 2025-05-22 20:51:30 +05:30
Arandjelovic, Marko 9ef5602453 Match cuda behaviour for VMM APIs (#382) 2025-05-22 20:49:08 +05:30