New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
testsuite: optional 'parallel' target uses GNU parallel to halve testing time #451
Conversation
This sounds lovely. |
You should also update the help text in the |
…ng time See the run logs below. Using 'parallel' reduces the time to run the full testsuite from 2m30s to 0m57s on my machine. All tests are run in parallel, including lib-threads. We ask parallel to preserve output order, so it should be deterministic. (Individual tests may mangle stdout and stderr in fragile ways, though.) See the Makefile comment for details. ``` rm -f _log; make clean; time make all; cp _log /tmp/trunk_log > real 2m29.947s > user 1m35.937s > sys 0m17.467s rm -f _log; make clean; time make parallel; cp _log /tmp/parallel_log > real 0m56.611s > user 2m20.861s > sys 0m21.511s diff -u /tmp/{trunk,parallel}_log | wc -l > 0 rm /tmp/{trunk,parallel}_log ```
@damiendoligez I included a Changes entry (ah, those pesky maintainer requirements...) I also included some logic to fail informatively when the tool is not present on the user's machine. The current output if you run a
For the record, there also exists a
|
testsuite: optional 'parallel' target uses GNU parallel to halve testing time
See the run logs below. Using 'parallel' reduces the time to run the
full testsuite from 2m30s to 0m57s on my machine.
All tests are run in parallel, including lib-threads.
We ask parallel to preserve output order, so it should be
deterministic. (Individual tests may mangle stdout and stderr in
fragile ways, though.) See the Makefile comment for details.