Commit Graph

3793 Commits

Author SHA1 Message Date
Jason Tang b38317cb3c Add COMGR_DYN_DLL to device/rocm/CMakeLists.txt
Now rocm/rocdevice.cpp also includes comgrctx.hpp, and we don't want to statically link against comgr when buidling shared libs.

Change-Id: Ic330bd860559b3e07b776c951afe6126b0f43f7d
2020-10-14 09:28:37 -04:00
Chauncey Hui a00981128a SWDEV-2 - Change OpenCL version number from 3212 to 3213 2020-10-14 03:00:02 -04:00
Vladislav Sytchenko 26d1b28b16 [PAL] Allow overriding reported asic revision
This is helpfull to do when debugging issues on lowend asics. Navi14 can be emulated as Navi10. So can Navi22 be emulated as Navi21.

Change-Id: I693ffd45a5b03657822afdc872781901bc69b65c
2020-10-13 09:36:15 -04:00
agodavar 92f1ce41dc SWDEV-254185 : hiprtc headers - handle empty headerIncludeNames
Change-Id: Ie06278c18b62cef1bdfbb8ac82728ed5667b2047
2020-10-13 09:23:43 -04:00
German Andryeyev c05de4f2f9 Add ROCr queries for HMM support
Change-Id: I2b5508bf0faf8f48dd7348d6a5202fb28de09876
2020-10-13 09:12:00 -04:00
Chauncey Hui 2a5f1ccadf SWDEV-2 - Change OpenCL version number from 3211 to 3212 2020-10-13 03:00:03 -04:00
agodavar ac72e50adc SWDEV-254185 - Added support to pass include headers to hipRTC
Change-Id: Ic7f2957b04e518c57e2fd3fc9d839de07232405e
2020-10-12 03:46:04 -04:00
Chauncey Hui 2810b8f367 SWDEV-2 - Change OpenCL version number from 3210 to 3211 2020-10-10 03:00:03 -04:00
Vladislav Sytchenko fd09a7a23c [PAL] Skip extra calls to MakeResident
With the PAL_ALWAYS_RESIDENT flag memory objects are resident at allocation time, no need to make them resident again before submit.

Also we should never evict anything with this setting, or we'll generate a VM fault.

Change-Id: Ieacc6af88ab4e09c20efd94100e148b2502e1d70
2020-10-09 14:13:32 -04:00
Chauncey Hui 099dbb7f13 SWDEV-2 - Change OpenCL version number from 3209 to 3210 2020-10-07 03:00:03 -04:00
German Andryeyev d9397590de Add option to skip AQL barrier
The change reuses HSA signals for dispatches as a wait signal.
Skipping the barrier requires to  disable L2 cache for sysmem
allocations and extra tracking for HDP access with the large bar.
ROC_BARRIER_SYNC=0 activates the new logic. Barrier sync is
still used by default.
ROC_ACTIVE_WAIT=1 enables unconditional active wait in ROCr.
The change also consolidated ROCr wait logic under single function.

Change-Id: I6bd1be30aa88258da1b1f9de319ef5a45852afd8
2020-10-06 08:37:12 -04:00
Sarbojit Sarkar 4a025e1a87 [perf]hipMalloc performance optimization
Change-Id: I6e8a918cc1c4cafad197b09e10755cd180e11ead
2020-10-06 03:19:41 -04:00
Sarbojit Sarkar 8ab8fac173 SWDEV-253548 : clean up gfx-arch macros
Change-Id: I8deb2ea44f556260bb78d24f68b04b0c730ed4d8
2020-10-06 03:17:09 -04:00
Chauncey Hui c7f03dae75 SWDEV-2 - Change OpenCL version number from 3208 to 3209 2020-10-06 03:00:02 -04:00
German Andryeyev df63547906 Add support for CPU device in advise
Change-Id: I7250e0183580c14cd3d6050ef85f9ce26e36f4a8
2020-10-05 12:52:46 -04:00
Tao Sang bc7291e85c Fix empty cpu agent on raven system
Change-Id: I30f5e65367613152ce96b80b13e1c9f2a28da807
2020-10-05 10:54:57 -04:00
Chauncey Hui 65b88ba364 SWDEV-2 - Change OpenCL version number from 3207 to 3208 2020-10-03 03:00:03 -04:00
Vladislav Sytchenko f26bb1d76d [PAL] Add entry for unannounced device
Change-Id: I323f240068ee1eff6981e69b7a2837f06157e44b
2020-10-02 13:52:25 -04:00
Vladislav Sytchenko 11bdf0005f [PAL] Detect if resizable bar is enabled
Change-Id: I54b2e878d1cf7235a848a4b243f878538e226702
2020-10-02 12:25:00 -04:00
Chauncey Hui 189ec6f794 SWDEV-2 - Change OpenCL version number from 3206 to 3207 2020-10-02 03:00:03 -04:00
German Andryeyev af8426b0e4 Revert "Reduce the default size of the signal pool"
This reverts commit e68d671a51.

Reason for revert: a regression

Change-Id: I78180ba011f45af9a4cce110b14f379aa10f7d3a
2020-10-01 09:56:05 -04:00
Chauncey Hui d8404bfee9 SWDEV-2 - Change OpenCL version number from 3205 to 3206 2020-10-01 03:00:04 -04:00
Vladislav Sytchenko 1183455755 Fix Windows build
Change-Id: Icdc723843541a5c13503670b5f972b78db0cccbf
2020-09-30 14:40:07 -04:00
Jatin Chaudhary d0f009a0ee Adding and populating available registers per cu value
Change-Id: Ide012038a9d74b20166ce1e69a9f2c0cbdce0d73
2020-09-28 05:30:23 -04:00
Chauncey Hui b26a283118 SWDEV-2 - Change OpenCL version number from 3204 to 3205 2020-09-26 03:00:04 -04:00
kjayapra-amd 4071cf1a16 SWDEV-241896 - Remove extern support in ROCclr/HIP since compiler added Texture Reference support.
Change-Id: I2cf27e13a53860fe11e9245c9cbdcb8bd88a7872
2020-09-25 10:20:16 -04:00
Jason Tang 40b432f992 Avoid using sdma engine if there's none
Change-Id: Id8d373683f3e1667ab461b4c5b8b490377b64cc0
2020-09-25 09:48:10 -04:00
Aryan Salmanpour 2e199bd492 Fix a crash when printf used in a kernel launched on a stream with custom CU mask
SWDEV-249719 - root cause: queues with custom CU mask are not inserted
into queuePool_ (i.e., queue of reusable HSA queues) of ROC device class
causing a crash when creating hostcall buffers for printf

Change-Id: Ieee7005d9a5a30b3113394ce23ee65927126d0d6
2020-09-25 09:25:19 -04:00
Chauncey Hui 0cef35c402 SWDEV-2 - Change OpenCL version number from 3203 to 3204 2020-09-24 03:00:02 -04:00
kjayapra-amd 18352d189b SWDEV-253063 - Code changes to make Image Buffer Workaround only for targets gfx 10.1
Change-Id: I17044a1c0775f427b9ba712eb3fd5ab21ed88b0e
2020-09-23 11:07:15 -04:00
Chauncey Hui 2f98dbedca SWDEV-2 - Change OpenCL version number from 3202 to 3203 2020-09-22 03:00:03 -04:00
Vladislav Sytchenko e1b916260a Fix Windows legacy LLVM build
The os.hpp header gets added to the include path of legacy llvm via the compiler lib. Having "windows.h" included causes a lot conflicts with LLVM headers, as they forward declare many Windows types. Best to not include it here.

Change-Id: I60c44a8d28660368f1a4a95741e1053ef3528fa1
2020-09-21 11:52:01 -04:00
Chauncey Hui 7fd56c6f13 SWDEV-2 - Change OpenCL version number from 3201 to 3202 2020-09-19 03:00:04 -04:00
kjayapra-amd 7462e39954 SWDEV-252542 - Fixing Win Compilation on SWDEV-241902.
Change-Id: If76f79002b265dccf6da4acef1ff9372d8b0a2ff
2020-09-18 12:11:56 -04:00
Satyanvesh Dittakavi 90c694988b set free mem to zero avoding underflow
Change-Id: I964b41695fb41712624146c5959a35b37eacf52d
2020-09-18 11:56:57 -04:00
Tao Sang 77c9226d87 Resolve macro conflict in opencl build
Change-Id: I752dff96652a55e4fbf719e84ee7d3ec49f5f20a
2020-09-16 17:57:24 -04:00
Chauncey Hui 0f422622ca SWDEV-2 - Change OpenCL version number from 3200 to 3201 2020-09-16 03:00:03 -04:00
kjayapra-amd a66c56d641 SWDEV-241902 - Changes to pass file descriptor and offset to load code object.
Change-Id: I0243cccdeaa533b2a56fde42f12d5424c3b63a3b
2020-09-15 07:54:24 -04:00
Chauncey Hui de6e02f98b SWDEV-2 - Change OpenCL version number from 3199 to 3200 2020-09-15 03:00:03 -04:00
kjayapra-amd 11b723a605 SWDEV-251941 - Adding gfx1030,gfx1031 support to mainline
Change-Id: I752dff86651a55f4fbf719e84ee7d3ec49f5f207
2020-09-14 09:07:44 -04:00
Chauncey Hui a35de40c13 SWDEV-2 - Change OpenCL version number from 3198 to 3199 2020-09-13 03:00:03 -04:00
Vladislav Sytchenko be0d98b433 Fix typo
Change-Id: I571e84cab9258dfd5e94b70865505f452ed89016
2020-09-12 20:16:27 -04:00
Vladislav Sytchenko 29efce7520 Add missing include
Change-Id: I21683ddb981581a0456ad0aa79843a44422a6ab0
2020-09-12 20:16:09 -04:00
Chauncey Hui 33138c7a52 SWDEV-2 - Change OpenCL version number from 3197 to 3198 2020-09-12 03:00:02 -04:00
Jason Tang 740f1ba0aa SWDEV-252095 - Fix progvar_prog_scope_misc test Aborted (core dumped)
The progvar_prog_scope_misc test expects CL_DEVICE_MAX_MEM_ALLOC_SIZE to be multiple of integer.

Change-Id: Ic0690f5ab06ffa3feafb718beabbcf83ab53978c
2020-09-11 12:36:39 -04:00
Chauncey Hui 51d0c6bbef SWDEV-2 - Change OpenCL version number from 3196 to 3197 2020-09-11 03:00:03 -04:00
Sarbojit Sarkar 677208f2de Fix for memory leak detected in hipModule*load [SWDEV:250740]
Change-Id: I8c4930b64764620a896079b14b27f6f87fbad292
2020-09-10 10:50:46 -04:00
Chauncey Hui e01cdb5530 SWDEV-2 - Change OpenCL version number from 3195 to 3196 2020-09-10 03:00:03 -04:00
Laurent Morichetti 5d4b6f74d3 Use std::atomic
Replace amd::Atomic with std::atomic. Remove make_atomic uses by
converting the variable to std::atomic and making sure the memory
order is relaxed when synchronizes-with is not needed.

Delete utils/atomic.hpp.

Change-Id: I0b36db8d604a8510ac6e36b32885fd16a1b8ccfa
2020-09-09 14:55:29 -04:00
Jason Tang cef70aa8d3 Export tiny_printf in PAL build
Change-Id: I6166bb3a4d177f03c60e14116c41b9b6da6e6ed2
2020-09-09 10:55:35 -04:00