Commit Graph

Select branches

Hide Pull Requests

main

Mono Color

• a8203589b7 added example use of clusterbench_submit for running a production hibridon benchmark on IPR's alambix cluster main Guillaume Raffy 2024-12-10 18:57:54 +0100
• ccb964bb44 v0.0.16 Guillaume Raffy 2024-11-27 14:10:07 +0100
• b85f381241 v0.0.15 Guillaume Raffy 2024-11-27 13:50:38 +0100
• 263474eb5a v0.0.14 Guillaume Raffy 2024-11-27 11:02:54 +0100
• e5fb47ba64 - fixed typo HibrindoBenchmark Guillaume Raffy 2024-11-26 17:04:12 +0100
• 8d674055bc v0.0.13 Guillaume Raffy 2024-11-26 16:56:11 +0100
• e9bbc95632 v0.0.12 Guillaume Raffy 2024-11-26 16:37:10 +0100
• 866f7de27d added even more debug traces to debug a problem that only happens in ci (can't reproduce even when logged as jenkins-agent) Guillaume Raffy 2024-11-26 15:33:04 +0100
• 4b6b6ad056 added more debug traces to debug a problem that only happens in ci (can't reproduce even when logged as jenkins-agent) Guillaume Raffy 2024-11-26 15:26:51 +0100
• 06c360fa11 added debug traces to debug a problem that only happens in ci (can't reproduce even when logged as jenkins-agent) Guillaume Raffy 2024-11-26 15:18:09 +0100
• 60be658312 improved traces to debug a problem that only happens in ci (can't reproduce even when logged as jenkins-agent) Guillaume Raffy 2024-11-26 15:10:56 +0100
• 95cb65ef35 improved error message to debug a problem that only happens in ci (can't reproduce even when logged as jenkins-agent) Guillaume Raffy 2024-11-26 15:04:02 +0100
• d1aba97073 made sure the module command is available when running /ci/test_iprcluster.py Guillaume Raffy 2024-11-26 14:43:34 +0100
• c39e31a1fc fixed regression introduced in my last commit Guillaume Raffy 2024-11-26 13:42:54 +0100
• 7d019f74dd v0.0.11 Guillaume Raffy 2024-11-26 13:33:12 +0100
• 2f2ebdcf35 fixed minor styling issue Guillaume Raffy 2024-11-25 14:52:33 +0100
• 4ca2e02381 v0.0.10 Guillaume Raffy 2024-11-25 14:52:08 +0100
• 026c5f6100 v0.0.9 Guillaume Raffy 2024-11-22 16:03:29 +0100
• ea28ea0293 fixed broken commit 7976d67860 Guillaume Raffy 2024-11-22 16:00:18 +0100
• e550ba408f fixed broken commit 7976d67860 Guillaume Raffy 2024-11-22 14:25:19 +0100
• eab903a7ae reverted my last commit which was a mistake Guillaume Raffy 2024-11-22 14:24:25 +0100
• 1b30c7f4a2 fix to last commit Guillaume Raffy 2024-11-22 14:13:29 +0100
• b8bb991abd forgot this in my last commit Guillaume Raffy 2024-11-22 13:43:54 +0100
• 7976d67860 v0.0.8 Guillaume Raffy 2024-11-22 13:40:22 +0100
• e3295bf3da forgot this in my last commit Guillaume Raffy 2024-11-21 17:12:29 +0100
• 33133d6828 v0.0.7 Guillaume Raffy 2024-11-21 17:04:55 +0100
• f7d1946eb7 v0.0.6 (addendum) Guillaume Raffy 2024-11-21 16:45:22 +0100
• 9d648b4fdc v0.0.6 Guillaume Raffy 2024-11-21 08:29:45 +0100
• 7fd25890ec v0.0.5 Guillaume Raffy 2024-11-19 21:37:55 +0100
• a5b5345d39 v0.0.4 Guillaume Raffy 2024-11-19 00:09:32 +0100
• 628ac7352a v0.0.3 Guillaume Raffy 2024-11-18 23:04:29 +0100
• 862a3911f0 v0.0.2 Guillaume Raffy 2024-11-18 16:42:12 +0100
• 0dd0195d45 made tests pass when using univ-rennes vpn Guillaume Raffy 2024-11-17 11:33:12 +0100
• 6dded4234a refactored tests: - split tests into 3: - test_benchmarks.py: tests all benchmarks with the most basic resultsdb backend - test_resultsdb: tests all resultsdb backends with the most basic benchmark - test_clusterbench: tests clusterbench_submit - made tests more robust (deletes the results foder if it already exists) Guillaume Raffy 2024-11-17 11:31:48 +0100
• 3d8a0ff4ad added the following automatic parameters (parameters common to all benchmarks): - host_id (host serial number) - num_cpus - iprbench_version Guillaume Raffy 2024-11-13 17:42:01 +0100
• 59a5109ed9 the user can now output the benchmark results to a remote sql database accessed via ssh Guillaume Raffy 2024-11-13 14:57:49 +0100
• 86985d3de4 the user can now output the benchmark results to a remote sql database accessed via ssh Guillaume Raffy 2024-11-13 14:56:45 +0100
• 5337c41645 fixed stupid copy paste error that caused all the fields to have the same type Guillaume Raffy 2024-11-13 14:47:49 +0100
• ca1f4c1a88 added autoparams `host_fqdn` and `cpu_model` Guillaume Raffy 2024-11-08 09:12:51 +0100
• 2ee1988721 the user can now choose to save the benchmark results into a sql database (sqlite at the moment) Guillaume Raffy 2024-11-05 15:39:00 +0100
• 2a91af37ff added the ability for the user to choose the resultsdb backend. Guillaume Raffy 2024-10-25 19:19:28 +0200
• 12cc0c0c8a added a mechanism to record benchmark results into a database. Guillaume Raffy 2024-10-24 18:51:24 +0200
• 25d2e489d5 fixed bug with num_cores that caused test_clusterbench to fail on hibench Guillaume Raffy 2024-10-23 16:55:17 +0200
• 98940824e4 Made benchmark `hibench` work Guillaume Raffy 2024-10-23 14:36:22 +0200
• 6000e20d6b the user of `iprbench-submit` can now choose the cluster he wants to use (previously it was hardcoded to the test cluster `DummyCluster`) Guillaume Raffy 2024-10-23 11:18:07 +0200
• fe4d66fb52 moved Singleton class to a better place Guillaume Raffy 2024-10-23 10:55:37 +0200
• 6f84732cf6 mamul1 benchmark now gets its source files from iprbench resources rather than a hardcoded directory (which was obviously not satisfactory since iprbench only worked if installed in a specfic location). Guillaume Raffy 2024-10-22 17:32:35 +0200
• 011d4eddf9 refactored iprbench to separate ipr benchmark framework from the actual benchmarks Guillaume Raffy 2024-10-22 09:16:41 +0200
• fe4a07a67e refactored all iprbench code found in `usecases/ipr/hibench` into a `iprbench` python package Guillaume Raffy 2024-10-10 18:06:09 +0200
• fb6f926cb1 improvements to hibenchonphysix: Guillaume Raffy 2024-10-09 15:16:02 +0200
• 3ca66d0063 adapted code to cope with recent change in univ rennes policy that caused github access to require the use of a proxy Guillaume Raffy 2024-10-08 17:44:22 +0200
• cf1235f62c added the option --cmake-path to allow the user to choose the cmake executable to use. Guillaume Raffy 2024-10-08 17:41:18 +0200
• 5d59650e84 added the --arch-regexp option to allow the user to restrict benchmarks on some architectures Guillaume Raffy 2024-10-08 17:27:05 +0200
• c534d7e135 improved the hosts table to ease adaptation to alambix instead of physix Guillaume Raffy 2024-10-08 17:22:26 +0200
• 350785bfee fixed regressions introduced in commit [b8c8a1b0e6] Guillaume Raffy 2024-10-08 17:12:22 +0200
• 2ba66a498d fixed bugs introduced in commit [4e0e3b60bc] Guillaume Raffy 2024-10-08 16:58:40 +0200
• f2ceeb2cdb added a mechanism to prevent starbench to hang in case the executed command fails Guillaume Raffy 2024-10-08 16:46:53 +0200
• d71bf3f67f added the --results-dir option Guillaume Raffy 2024-10-07 12:16:47 +0200
• 49aebf38a5 updated documentation regarding last commit's change (rewriting in python) Guillaume Raffy 2024-10-07 12:06:45 +0200
• 4e0e3b60bc rewrote hibench-on-physix.sh in python Guillaume Raffy 2024-10-04 18:27:21 +0200
• 46d7fd1fa7 improved showresults.py: - added the ability to show the performance per clock cycle (eg to see if avx512 has an effect on the computation) - the cpus are now sorted on x axis depending on their generation Guillaume Raffy 2024-10-04 18:24:02 +0200
• 1423090328 showresult.py now dislpays a graphic comparing the performance of hibridon on different cpus Guillaume Raffy 2024-10-01 14:25:56 +0200
• ccb0acd817 improved hibench results parser to retrieve details of configuration (eg mkl version) Guillaume Raffy 2024-09-30 17:47:15 +0200
• 1d66fc1edb added a tool to parse benchmark output files to summarize it into a table, so that it can be exploited to create graphs Guillaume Raffy 2024-09-27 23:58:14 +0200
• c4732fee87 improved usecase file hierarchy to accomodate multiple uses cases Guillaume Raffy 2024-09-27 23:53:24 +0200
• 0693b2c948 refactoring: documented mamul1 source code related to clocks to clarify the meansured times Guillaume Raffy 2024-07-01 09:30:26 +0200
• fb7608ecdd removed hardcoded path Guillaume Raffy 2024-06-30 17:16:39 +0200
• de7c1fb2dd fixed styling errors Guillaume Raffy 2024-06-30 17:15:45 +0200
• 249ef1f3e7 refactored by isolated all core functions and classes into a core.py source file Guillaume Raffy 2024-06-30 16:20:34 +0200
• 68cb7169c7 fixed dangerous-default-value / W0102 potential problem Guillaume Raffy 2024-06-30 16:08:10 +0200
• c05ff89d29 added a complete standalone usage example (mamul1: mutiplication of matrices) in the form of a unit test Guillaume Raffy 2024-06-30 15:53:03 +0200
• 3dc0d12307 decoupled starbench_cmake_app from git repos information, so that starbench_cmake_app can now be used with any source code provider, not only from git repositories (eg an existing directory tree) Guillaume Raffy 2024-06-28 15:07:37 +0200
• b8c8a1b0e6 turned starbench into an installable package Guillaume Raffy 2024-06-21 14:48:00 +0200
• a43eb68db5 fixed all styling warnings and comments, and documented the code Guillaume Raffy 2024-06-21 08:50:36 +0200
• dc897e9225 fixed styling errors detected by pylint Guillaume Raffy 2024-06-20 11:51:53 +0200
• f87e564528 reorganized files and documented Guillaume Raffy 2022-06-10 08:39:22 +0200
• 165da43619
  Create README.md g-raffy 2022-06-09 15:59:52 +0200
• 2c349315cd the job script is now saved to a location that should avoid it to be written by multiple jobs at the same time, thus avoiding race conditions issues. Guillaume Raffy 2022-06-09 11:16:36 +0200
• 5bc51c78ca fixed bug that caused starbench.py source code to be included twice (and one of them causing syntax errors) Guillaume Raffy 2022-06-09 11:15:00 +0200
• 43ae0a3068 cleanup and documented Guillaume Raffy 2022-06-09 09:09:32 +0200
• 733fda5517 added mechanism to download benchmark results on work.global on success Guillaume Raffy 2022-06-09 08:58:04 +0200
• 75c4b98be0 improved name of job Guillaume Raffy 2022-06-09 08:56:47 +0200
• 58dbfc9be6 added ifort build to hibridon's benchmark Guillaume Raffy 2022-06-09 08:54:50 +0200
• f2b8d6cdb4 now the hibridon benchmark runs one job for each machine type on physix Guillaume Raffy 2022-06-07 19:01:30 +0200
• 936dfa793a made changes needed to get hibridon benchmark running on physix (ipr's cluster): Guillaume Raffy 2022-06-07 14:52:56 +0200
• 6715cd1714 made starbench compatible with python 3.5 + : Guillaume Raffy 2022-06-07 12:44:25 +0200
• fab373f3c1 made `hibench.job` hibridon independent so that: - `hibench.job` could be reused as is by another project - it gives more flexibility to lauch-perf-jobs.sh Guillaume Raffy 2022-06-03 16:10:47 +0200
• 176db9f719 now that `hibench.py` has been made generic (it has no hibridon related code anymore), I renamed it as `starbench.py` Guillaume Raffy 2022-06-03 14:13:32 +0200
• 959fba13b3 the test command is no longer hardcoded and can be chosen via command line arguments Guillaume Raffy 2022-06-02 17:52:38 +0200
• 226871547f removed compiler loop as the used compiler can now be defined by the user via cmake defines Guillaume Raffy 2022-06-02 17:34:16 +0200
• 431f728793 the cmake options are no longer hardcoded Guillaume Raffy 2022-06-02 15:59:27 +0200
• ac639b3a08 fixed bug that coused the checkout to fait because it was performed from a non-git repository Guillaume Raffy 2022-06-02 15:58:00 +0200
• 97a04d2831 it's now possible to choose aversion of the code to test Guillaume Raffy 2022-06-02 14:34:14 +0200
• 05ec8a5181 made `hibench.py` (better name than `starbench.py`) parse options so that it is more generic. Guillaume Raffy 2022-06-02 13:01:16 +0200
• c65cd60a54 measure_hibridon_perf now takes the name of the test as a parameter to make it more generic Guillaume Raffy 2022-06-01 16:17:10 +0200
• ddcb9f1175 now the number of threads per run is restricted to what the user asks Guillaume Raffy 2022-06-01 12:06:58 +0200
• 0ecddcf8f8 minor flake fix Guillaume Raffy 2022-06-01 12:06:10 +0200
• d45d53c3ac fixed bugs discovered after running from scratch Guillaume Raffy 2022-06-01 12:05:28 +0200
• e643be2435 cleanup by removing debug traces and replacing them with useful messages Guillaume Raffy 2022-05-31 19:03:00 +0200
• 9b12ce7cd2 added a way to redirect stdout and stderr to files to avoid pollution on the console Guillaume Raffy 2022-05-31 18:55:16 +0200