Le mardi 8 décembre 2009 17:50, vous avez écrit :
On Tuesday 2009-12-08 17:48, Jean Delvare wrote:
The case at hand is quilt's test suite. It's relatively verbose. When tests are run sequentially it's OK. If they are run in parallel then all outputs will interlace and things will become very difficult to read.
So if there is a FAIL, just rerun the testsuite with -j1.
If I'm running the test suite manually, yes, sure. But the most interesting case for me at the moment is when the test suite is run by the build service. Hmm, I guess the following in the spec file would do the trick? make -j 4 check || make check The output of the first will be unreadable, but the output of the later should be OK. That would be efficient, if not esthetically pleasant. Still I think I'll modify the test suite runner script itself to be more parallelism-compliant. It shouldn't hurt.
So I am looking for ways to run the tests in parallel while still obtaining a usable output if one of the tests fails. And I thought it might help if I could pass a flag to the tester to warn it about parallelism so that it can adjust its output. I'm not too sure how, though.
Out of curiosity, is there a shell command that would simply buffer every input until EOF and then flush everything to stdout?
The SUSE bootscripts do something like that when RUN_PARALLEL=yes in /etc/sysconfig/boot. Granted, it's not the best either, since it's just a heuristic.
Do you happen to know/remember where the code in question is? A quick look did not reveal it to me. Thanks, -- Jean Delvare Suse L3 -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org