76333 Коммитов
Автор SHA1 Сообщение Дата
Ori Messinger b1720b42cd ROCm SMI CLI: Add --showtopoaccess Functionality
The purpose of this patch is to implement --showtopoaccess
functionality in the CLI, which shows True or False if P2P is
possible between two given GPUs.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I07d70d80ae7b484136b31d5d22780c4990029391


[ROCm/rocm_smi_lib commit: e2d9a37e5f]
2021-10-14 11:06:05 -04:00
Ori Messinger e2d9a37e5f ROCm SMI CLI: Add --showtopoaccess Functionality
The purpose of this patch is to implement --showtopoaccess
functionality in the CLI, which shows True or False if P2P is
possible between two given GPUs.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I07d70d80ae7b484136b31d5d22780c4990029391
2021-10-14 11:06:05 -04:00
Ori Messinger de16bc4552 ROCm SMI CLI: Add --showtopoaccess Functionality
The purpose of this patch is to implement --showtopoaccess
functionality in the CLI, which shows True or False if P2P is
possible between two given GPUs.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: I07d70d80ae7b484136b31d5d22780c4990029391


[ROCm/amdsmi commit: e2d9a37e5f]
2021-10-14 11:06:05 -04:00
Jason Tang 2590bfd491 SWDEV-307023 - Fix cltrace build
Change-Id: I75fd953668dc576c3132ae463e34ecb4ae2fb0c7


[ROCm/clr commit: 5c13f33da2]
2021-10-14 09:26:27 -04:00
Jason Tang 5c13f33da2 SWDEV-307023 - Fix cltrace build
Change-Id: I75fd953668dc576c3132ae463e34ecb4ae2fb0c7
2021-10-14 09:26:27 -04:00
Chauncey Hui c22c59c4a4 SWDEV-2 - Change OpenCL version number from 3377 to 3378
[ROCm/clr commit: c9f73f8d31]
2021-10-14 03:00:16 -04:00
Chauncey Hui c9f73f8d31 SWDEV-2 - Change OpenCL version number from 3377 to 3378 2021-10-14 03:00:16 -04:00
Ori Messinger a9993fb509 ROCm SMI LIB: Add rsmi_is_P2P_accessible() API
Implements rsmi_is_p2p_accessible API.
The function returns True if P2P is possible between two nodes.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: Ic7316eebcec4480175c7ad04c21a42b2e1a4c454


[ROCm/rocm_smi_lib commit: ff02042c64]
2021-10-13 22:01:33 -04:00
Ori Messinger ff02042c64 ROCm SMI LIB: Add rsmi_is_P2P_accessible() API
Implements rsmi_is_p2p_accessible API.
The function returns True if P2P is possible between two nodes.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: Ic7316eebcec4480175c7ad04c21a42b2e1a4c454
2021-10-13 22:01:33 -04:00
Ori Messinger ac01f99000 ROCm SMI LIB: Add rsmi_is_P2P_accessible() API
Implements rsmi_is_p2p_accessible API.
The function returns True if P2P is possible between two nodes.

Signed-off-by: Ori Messinger <Ori.Messinger@amd.com>
Change-Id: Ic7316eebcec4480175c7ad04c21a42b2e1a4c454


[ROCm/amdsmi commit: ff02042c64]
2021-10-13 22:01:33 -04:00
gilbertlee-amd fe4285d002 [TransferBench] Adding comment echoing to help distinguish tests (#438)
[ROCm/rccl commit: f6b7ac693e]
2021-10-13 14:56:57 -06:00
gilbertlee-amd f6b7ac693e [TransferBench] Adding comment echoing to help distinguish tests (#438) 2021-10-13 14:56:57 -06:00
kjayapra-amd 6f62f832cb SWDEV-232903 - Move hipmemset Dword optimization to ROCclr.
Change-Id: I3eae61720cbc6364f1aaac4865bfd8b6ded08097


[ROCm/clr commit: 88ed58735d]
2021-10-13 11:32:15 -04:00
kjayapra-amd 88ed58735d SWDEV-232903 - Move hipmemset Dword optimization to ROCclr.
Change-Id: I3eae61720cbc6364f1aaac4865bfd8b6ded08097
2021-10-13 11:32:15 -04:00
kjayapra-amd 7c4dfc27a7 SWDEV-232903 - Move hipmemset dword optimization into ROCclr.
Change-Id: I1af437de2771f6a48d06ce643ade6b79f551a55c


[ROCm/clr commit: 4d1195f0a7]
2021-10-13 11:32:00 -04:00
kjayapra-amd 4d1195f0a7 SWDEV-232903 - Move hipmemset dword optimization into ROCclr.
Change-Id: I1af437de2771f6a48d06ce643ade6b79f551a55c
2021-10-13 11:32:00 -04:00
pghafari 6cb3ef5d42 SWDEV-245531 - GLInterop: updating glEnv logic
Change-Id: I6103243c477547a6a2b714bbf836ce36abe84f8c


[ROCm/clr commit: 5b6b2e6d2c]
2021-10-13 09:26:52 -04:00
pghafari 5b6b2e6d2c SWDEV-245531 - GLInterop: updating glEnv logic
Change-Id: I6103243c477547a6a2b714bbf836ce36abe84f8c
2021-10-13 09:26:52 -04:00
Chauncey Hui 25b7f971e6 SWDEV-2 - Change OpenCL version number from 3376 to 3377
[ROCm/clr commit: 52e57b660b]
2021-10-13 03:00:07 -04:00
Chauncey Hui 52e57b660b SWDEV-2 - Change OpenCL version number from 3376 to 3377 2021-10-13 03:00:07 -04:00
Sean Keely 5e8d261352 Skip inital CU mask setup unless HSA_CU_MASK is defined for the GPU.
Limits CU masking application to cases where it is explicitly requested.

Change-Id: Ib65ad0ac98f86d840c0328fa15ce40c05cd4bfae
2021-10-12 20:31:56 -05:00
Sean Keely 0116378a99 Skip inital CU mask setup unless HSA_CU_MASK is defined for the GPU.
Limits CU masking application to cases where it is explicitly requested.

Change-Id: Ib65ad0ac98f86d840c0328fa15ce40c05cd4bfae


[ROCm/ROCR-Runtime commit: 5e8d261352]
2021-10-12 20:31:56 -05:00
Freddy Paul ca899ea429 Cleanup symlink to header files and folders
Due to a CPACK bug the package needs to remove header file
symlinks.  Cleanup is required for uninstall and upgrade
since each release installs to a different folder.

Change-Id: I5ec378b21e69235404781c7bce3c0203eb38eed1
2021-10-12 14:56:02 -05:00
Freddy Paul 8d550d9cbe Cleanup symlink to header files and folders
Due to a CPACK bug the package needs to remove header file
symlinks.  Cleanup is required for uninstall and upgrade
since each release installs to a different folder.

Change-Id: I5ec378b21e69235404781c7bce3c0203eb38eed1


[ROCm/ROCR-Runtime commit: ca899ea429]
2021-10-12 14:56:02 -05:00
Jason Tang e4db6ef66a SWDEV-306697 - Fix OCLGlobalOffset segfaults
If we don't create the __amd_rocclr_gwsInit kernel, we still want
to create the rest of the image related blit kernels.

Change-Id: I8bc4645f9f9116eeecbb8b22e981ac4d520f3121


[ROCm/clr commit: 55a0cf0b0c]
2021-10-12 15:13:28 -04:00
Jason Tang 55a0cf0b0c SWDEV-306697 - Fix OCLGlobalOffset segfaults
If we don't create the __amd_rocclr_gwsInit kernel, we still want
to create the rest of the image related blit kernels.

Change-Id: I8bc4645f9f9116eeecbb8b22e981ac4d520f3121
2021-10-12 15:13:28 -04:00
gilbertlee-amd ad1a620333 [TransferBench] Adding shared memory per threadblock env var. Defaulting to 1 threadblock per CU (#436)
[ROCm/rccl commit: 269f07fbc3]
2021-10-12 09:32:54 -06:00
gilbertlee-amd 269f07fbc3 [TransferBench] Adding shared memory per threadblock env var. Defaulting to 1 threadblock per CU (#436) 2021-10-12 09:32:54 -06:00
Wenkai Du b587b55c2e Add more Rome models (#434)
* Add more Rome models

* Update models and tuning

* Update tuning

[ROCm/rccl commit: 2249a1d9d3]
2021-10-12 08:23:20 -07:00
Wenkai Du 2249a1d9d3 Add more Rome models (#434)
* Add more Rome models

* Update models and tuning

* Update tuning
2021-10-12 08:23:20 -07:00
Sean Keely 19c1e92b4c Remove io_link workarounds.
KFD topology has been corrected and the defaults used by this
workaround are no longer true for all chips.

Change-Id: I0242d8077e9666ed1cf0dc3985244258ae5c0924
2021-10-11 19:15:07 -05:00
Sean Keely f19e8c43d2 Remove io_link workarounds.
KFD topology has been corrected and the defaults used by this
workaround are no longer true for all chips.

Change-Id: I0242d8077e9666ed1cf0dc3985244258ae5c0924


[ROCm/ROCR-Runtime commit: 19c1e92b4c]
2021-10-11 19:15:07 -05:00
pghafari b6ab3bbc80 SWDEV-245531 - GLInterop: exposing api for linux
Change-Id: I9aa12bf5b6463619d3d6cdaea85ffc18f50bb1ba


[ROCm/clr commit: a525605e5c]
2021-10-11 12:21:14 -04:00
pghafari a525605e5c SWDEV-245531 - GLInterop: exposing api for linux
Change-Id: I9aa12bf5b6463619d3d6cdaea85ffc18f50bb1ba
2021-10-11 12:21:14 -04:00
Xiaomeng Hou df59bfd57b Adjust the passing value for GPU agent when do max single allocation test
For APU asics, the default configuration size of video memory is
relatively small, plus the reserved region, ratio of max alloc size to
the pool size may below the expected value, so adjust it.

Change-Id: I798b44d9532aa6a381a1cc19faa5a46110bf0ad6
2021-10-11 02:32:09 -04:00
Xiaomeng Hou d64a353ea0 Adjust the passing value for GPU agent when do max single allocation test
For APU asics, the default configuration size of video memory is
relatively small, plus the reserved region, ratio of max alloc size to
the pool size may below the expected value, so adjust it.

Change-Id: I798b44d9532aa6a381a1cc19faa5a46110bf0ad6


[ROCm/ROCR-Runtime commit: df59bfd57b]
2021-10-11 02:32:09 -04:00
Xiaomeng Hou 9597fe3ae5 Add gfx1035 to rocrtst.
Change-Id: I276942b8badfd5ee2914e78c6c140d80d7cf4b2d
2021-10-11 02:31:45 -04:00
Xiaomeng Hou a29a2e96e5 Add gfx1035 to rocrtst.
Change-Id: I276942b8badfd5ee2914e78c6c140d80d7cf4b2d


[ROCm/ROCR-Runtime commit: 9597fe3ae5]
2021-10-11 02:31:45 -04:00
Ammar ELWazir 023cb806f9 SWDEV-275232: Replaced the use of on_exit()
on_exit() registered exit_handler function, but in the case of OpenMP applications the registered exit func exit_handler was called after the dlclose of the library. So we removed exit_handler from roctracer as it is already handled in both rocprofiler and rocprof script.

Change-Id: I7c3d42e6ccc282e713b48b4a7faec4935e7a2600


[ROCm/roctracer commit: aa41554130]
2021-10-09 17:36:36 -04:00
Ammar ELWazir aa41554130 SWDEV-275232: Replaced the use of on_exit()
on_exit() registered exit_handler function, but in the case of OpenMP applications the registered exit func exit_handler was called after the dlclose of the library. So we removed exit_handler from roctracer as it is already handled in both rocprofiler and rocprof script.

Change-Id: I7c3d42e6ccc282e713b48b4a7faec4935e7a2600
2021-10-09 17:36:36 -04:00
Sourabh Betigeri ce4329aca7 SWDEV-292525 - Add more parentheses to fix debug build failures
Change-Id: I91bb7e1f0f40b85dd908a532a77b11c9e7406019


[ROCm/clr commit: 641b1d3968]
2021-10-09 00:04:01 +00:00
Sourabh Betigeri 641b1d3968 SWDEV-292525 - Add more parentheses to fix debug build failures
Change-Id: I91bb7e1f0f40b85dd908a532a77b11c9e7406019
2021-10-09 00:04:01 +00:00
gilbertlee-amd 227848b70f [TransferBench] Adding ability to specify suffix for numBytes (#435)
[ROCm/rccl commit: aa917c3fc8]
2021-10-08 16:36:19 -06:00
gilbertlee-amd aa917c3fc8 [TransferBench] Adding ability to specify suffix for numBytes (#435) 2021-10-08 16:36:19 -06:00
gilbertlee-amd 94c60f772d Updating licensing / attribution for documentation (#432)
[ROCm/rccl commit: a6368bac99]
2021-10-08 13:17:24 -06:00
gilbertlee-amd a6368bac99 Updating licensing / attribution for documentation (#432) 2021-10-08 13:17:24 -06:00
Kent Russell 5a8c9a54f1 kfdtest.exclude: Add KFDEvictTest.Queuetest back to gfx1030's blacklist
The shader is incompatible with gfx1030, so blacklist it there until we get
to some more ASIC-agnostic shader generation

Signed-off-by: Kent Russell <kent.russell@amd.com>
Change-Id: I71f362a80e6e8f87ecece6afa6b34a5f02bc56bf
2021-10-08 13:09:01 -04:00
Kent Russell 11cc6f4f79 kfdtest.exclude: Add KFDEvictTest.Queuetest back to gfx1030's blacklist
The shader is incompatible with gfx1030, so blacklist it there until we get
to some more ASIC-agnostic shader generation

Signed-off-by: Kent Russell <kent.russell@amd.com>
Change-Id: I71f362a80e6e8f87ecece6afa6b34a5f02bc56bf


[ROCm/ROCR-Runtime commit: 5a8c9a54f1]
2021-10-08 13:09:01 -04:00
Chauncey Hui 656850e36c SWDEV-2 - Change OpenCL version number from 3375 to 3376
[ROCm/clr commit: 6d52cc286b]
2021-10-08 03:00:05 -04:00
Chauncey Hui 6d52cc286b SWDEV-2 - Change OpenCL version number from 3375 to 3376 2021-10-08 03:00:05 -04:00