AI/rocm-systems
2
0
Forkuj 0
Kod Zgłoszenia Oczekujące zmiany Actions Packages Projekty Wydania Wiki Aktywność
76 333 Commity 3 Gałęzie 0 Tagi
develop
Wykres commitów

76333 Commity

This Branch
This Branch
Wszystkie gałęzie
Autor SHA1 Wiadomość Data
Karl W Schulz 2eb05318de more refactor for profiling tests: introduce a launch_omniperf() 
utility function and use across all profiling executions

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 9f6e0a0a8b]
 2024-01-04 11:54:12 -06:00
Karl W Schulz 9f6e0a0a8b more refactor for profiling tests: introduce a launch_omniperf() 
utility function and use across all profiling executions

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz cf0efd0f28 update of profile tests to reorganize subset of output csv file 
checking into a common utility function -> check_csv_files()

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 9677666ce7]
 2024-01-04 11:54:12 -06:00
Karl W Schulz 9677666ce7 update of profile tests to reorganize subset of output csv file 
checking into a common utility function -> check_csv_files()

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz e0f6955737 update profile tests to run with 4-way parallelism 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: f0234f4571]
 2024-01-04 11:54:12 -06:00
Karl W Schulz f0234f4571 update profile tests to run with 4-way parallelism 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz 9af02e8e53 add GPU resource allocation file for parallel testing via ctest 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: a57173f504]
 2024-01-04 11:54:12 -06:00
Karl W Schulz a57173f504 add GPU resource allocation file for parallel testing via ctest 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz a47c3f11c0 expand previous profile test into multiple levaraging pytest marker 
groups; define a gpu-based RESOURCE_GROUPS to support parallel
execution on multi-gpu nodes

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 275b1080be]
 2024-01-04 11:54:12 -06:00
Karl W Schulz 275b1080be expand previous profile test into multiple levaraging pytest marker 
groups; define a gpu-based RESOURCE_GROUPS to support parallel
execution on multi-gpu nodes

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz dcfb6342d1 introduce a baseline_opts global var to house baseline profiling 
options used by all command-line tests; update --device command-line
filtering to leverage HIP_VISIBLE_DEVICES if set

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: 0f79a09a4a]
 2024-01-04 11:54:12 -06:00
Karl W Schulz 0f79a09a4a introduce a baseline_opts global var to house baseline profiling 
options used by all command-line tests; update --device command-line
filtering to leverage HIP_VISIBLE_DEVICES if set

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz 55344d586c define pytest marker groups 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: ca6d08d652]
 2024-01-04 11:54:12 -06:00
Karl W Schulz ca6d08d652 define pytest marker groups 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz cdcdc3d4ad reorg of profile mode testing functions to support parallel execution: 
* demarcate tests into eight pytest groups
* leverage new get_output_dir() function to store profile output data
  in a unique directory per test
* disable METRIC_LOGGING

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: fa587e967c]
 2024-01-04 11:54:12 -06:00
Karl W Schulz fa587e967c reorg of profile mode testing functions to support parallel execution: 
* demarcate tests into eight pytest groups
* leverage new get_output_dir() function to store profile output data
  in a unique directory per test
* disable METRIC_LOGGING

Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Karl W Schulz 6c567d532e addition of testing utility functions 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>


[ROCm/rocprofiler-compute commit: d56f10bb77]
 2024-01-04 11:54:12 -06:00
Karl W Schulz d56f10bb77 addition of testing utility functions 
Signed-off-by: Karl W Schulz <karl.schulz@amd.com>
 2024-01-04 11:54:12 -06:00
Saurabh Verma 5676c27c02 Supporting UMC counters in rocprofv2 
Change-Id: Ic96c33244d78ac5031cfd99ac2be0c12c774d1b0
 2024-01-04 12:33:08 -05:00
Saurabh Verma 492f3740b3 Supporting UMC counters in rocprofv2 
Change-Id: Ic96c33244d78ac5031cfd99ac2be0c12c774d1b0


[ROCm/rocprofiler commit: 5676c27c02]
 2024-01-04 12:33:08 -05:00
Wenkai Du 13791d7ee3 Rework barriers and adjust scope of atomics (#1019) 
[ROCm/rccl commit: abf265a911]
 2024-01-04 08:18:48 -08:00
Wenkai Du abf265a911 Rework barriers and adjust scope of atomics (#1019) 2024-01-04 08:18:48 -08:00
Jeremy Newton 42581d4172 Fix missing global symbol 
If using hsakmt as a shared library

Change-Id: I66a1849a46bd7009813d49824d0d059e8a511038
Signed-off-by: Jeremy Newton <Jeremy.Newton@amd.com>
 2024-01-04 11:14:39 -05:00
Jeremy Newton b2b4d5e7b5 Fix missing global symbol 
If using hsakmt as a shared library

Change-Id: I66a1849a46bd7009813d49824d0d059e8a511038
Signed-off-by: Jeremy Newton <Jeremy.Newton@amd.com>


[ROCm/ROCR-Runtime commit: 42581d4172]
 2024-01-04 11:14:39 -05:00
David Salinas ecc933301a Enable support for roc-obj tools on Windows 
SWDEV-301785 - roc-obj - support for Windows COFF

Change-Id: I20d53196e2d126391934351cb824b7bc926fb1ce


[ROCm/clr commit: 9fa8dcd1d1]
 2024-01-04 10:31:02 -05:00
David Salinas 9fa8dcd1d1 Enable support for roc-obj tools on Windows 
SWDEV-301785 - roc-obj - support for Windows COFF

Change-Id: I20d53196e2d126391934351cb824b7bc926fb1ce
 2024-01-04 10:31:02 -05:00
SrinivasaRao 635c7e3bf1 SWDEV-396087 - [catch2][dtest] Functional test cases for the API hipArrayGetDescriptor 
Change-Id: Ia400b7ce9a6eeb6f008cab13100fd10a86c5127f


[ROCm/hip-tests commit: 22fe5b7803]
 2024-01-04 09:44:00 -05:00
SrinivasaRao 22fe5b7803 SWDEV-396087 - [catch2][dtest] Functional test cases for the API hipArrayGetDescriptor 
Change-Id: Ia400b7ce9a6eeb6f008cab13100fd10a86c5127f
 2024-01-04 09:44:00 -05:00
SrinivasaRao 0e145c5af0 SWDEV-403479 - [catch2][dtest] Enable buffered Printf test Scenarios for Windows platform. 
Change-Id: I9c1428da5a262b81425a095dc15436621ee75ed6


[ROCm/hip-tests commit: 28eda094cc]
 2024-01-04 09:43:12 -05:00
SrinivasaRao 28eda094cc SWDEV-403479 - [catch2][dtest] Enable buffered Printf test Scenarios for Windows platform. 
Change-Id: I9c1428da5a262b81425a095dc15436621ee75ed6
 2024-01-04 09:43:12 -05:00
Ioannis Assiouras bd40edc652 SWDEV-391727 - Re-enabled Unit_hipMemRangeGetAttribute_TstCountParam 
Change-Id: I2acc1c220dc1fa1c53e739a5e8d4614022958c16


[ROCm/hip-tests commit: 9749139e50]
 2024-01-04 09:41:11 -05:00
Ioannis Assiouras 9749139e50 SWDEV-391727 - Re-enabled Unit_hipMemRangeGetAttribute_TstCountParam 
Change-Id: I2acc1c220dc1fa1c53e739a5e8d4614022958c16
 2024-01-04 09:41:11 -05:00
Ioannis Assiouras 30471a9fd9 SWDEV-438299 - Simplified the code in ExtractFatBinaryUsingCOMGR 
Change-Id: Ic15f2b71ae87ab3ca36ed5af2c816d57d09b3e70


[ROCm/clr commit: 411bccff5b]
 2024-01-04 04:01:58 -05:00
Ioannis Assiouras 411bccff5b SWDEV-438299 - Simplified the code in ExtractFatBinaryUsingCOMGR 
Change-Id: Ic15f2b71ae87ab3ca36ed5af2c816d57d09b3e70
 2024-01-04 04:01:58 -05:00
Ioannis Assiouras 1c3e482bd4 SWDEV-438531 - Fixed hipGraphExecMemcpyNodeSetParams for H2H kind 
The function erroneously returns hipErrorInvalidValue when kind
is set to hipMemcpyHostToHost

Change-Id: I6810b9f04f3218e517fd2f96410d1375e6ac6ff0


[ROCm/clr commit: af86c1b8a2]
 2024-01-04 03:47:39 -05:00
Ioannis Assiouras af86c1b8a2 SWDEV-438531 - Fixed hipGraphExecMemcpyNodeSetParams for H2H kind 
The function erroneously returns hipErrorInvalidValue when kind
is set to hipMemcpyHostToHost

Change-Id: I6810b9f04f3218e517fd2f96410d1375e6ac6ff0
 2024-01-04 03:47:39 -05:00
Satyanvesh Dittakavi d1901f2017 SWDEV-434846 - Limit the gpu single allocation percentage for all MI300 versions 
Change-Id: I33dea3eaab249ce3f9a624d38267489f99cd530c


[ROCm/clr commit: 755eb2962c]
 2024-01-03 23:47:44 -05:00
Satyanvesh Dittakavi 755eb2962c SWDEV-434846 - Limit the gpu single allocation percentage for all MI300 versions 
Change-Id: I33dea3eaab249ce3f9a624d38267489f99cd530c
 2024-01-03 23:47:44 -05:00
Ziyue Yang e3d45f9de4 Improve MSCCL algorithms (#1023) 
[ROCm/rccl commit: 0a53077c9c]
 2024-01-03 14:51:34 -08:00
Ziyue Yang 0a53077c9c Improve MSCCL algorithms (#1023) 2024-01-03 14:51:34 -08:00
German Andryeyev b8e7a858a5 SWDEV-311271 - Switch to sorted map for pool allocations 
Sorted map can work much faster for many allocations and a low reuse frequency

Change-Id: I6dba29ebc8bfacdf34307149b6a2b194890b2932


[ROCm/clr commit: a1fffac595]
 2024-01-03 17:36:05 -05:00
German Andryeyev a1fffac595 SWDEV-311271 - Switch to sorted map for pool allocations 
Sorted map can work much faster for many allocations and a low reuse frequency

Change-Id: I6dba29ebc8bfacdf34307149b6a2b194890b2932
 2024-01-03 17:36:05 -05:00
shadi f318f0f7e6 SWDEV-388256 - correct API name 
Signed-off-by: shadi <shadi.dashmiz@amd.com>
Change-Id: Icbf79613268fe0737a8033abfbcc9f33d785bc12


[ROCm/clr commit: a071c4d2fc]
 2024-01-03 15:06:41 -05:00
shadi a071c4d2fc SWDEV-388256 - correct API name 
Signed-off-by: shadi <shadi.dashmiz@amd.com>
Change-Id: Icbf79613268fe0737a8033abfbcc9f33d785bc12
 2024-01-03 15:06:41 -05:00
Ammar ELWazir d20003ed92 Fixing ROCTracer to match versioning changes in HIP Runtime 
Change-Id: Ibde2419314824f12acb4cf51dc675cb14a51851c


[ROCm/roctracer commit: aa3c07c257]
 2024-01-03 12:29:24 -06:00
Ammar ELWazir aa3c07c257 Fixing ROCTracer to match versioning changes in HIP Runtime 
Change-Id: Ibde2419314824f12acb4cf51dc675cb14a51851c
 2024-01-03 12:29:24 -06:00
Ammar ELWazir a59d56f9bc Fixing ROCProfilerV2 to match versioning changes in HIP Runtime 
Change-Id: Ifb0f98356acebcb7cf5372340b36178de000405d
 2024-01-03 12:28:08 -06:00
Ammar ELWazir 80329e8364 Fixing ROCProfilerV2 to match versioning changes in HIP Runtime 
Change-Id: Ifb0f98356acebcb7cf5372340b36178de000405d


[ROCm/rocprofiler commit: a59d56f9bc]
 2024-01-03 12:28:08 -06:00
Julia Jiang 96aa10d337 SWDEV-410522 - Fix in HIP API header for incorrect/missing API descriptions 
Change-Id: Ia16a555133c7e57ef61efacc489387e9e0dd248f


[ROCm/hip commit: f83941ea9a]
 2024-01-03 12:53:11 -05:00
Julia Jiang f83941ea9a SWDEV-410522 - Fix in HIP API header for incorrect/missing API descriptions 
Change-Id: Ia16a555133c7e57ef61efacc489387e9e0dd248f
 2024-01-03 12:53:11 -05:00