LUMIERA.clone/tests/test.conf

LOGSUPPRESS='^\(\*\*[0-9]*\*\* \)\?[0-9]\{10,\}[:!] \(TRACE\|INFO\|NOTICE\|WARNING\|ERR\|TODO\|PLANNED\|FIXME\|DEPRECATED\|UNIMPLEMENTED\|RESOURCE_ANNOUNCE\|RESOURCE_ENTER\|RESOURCE_STATE\|RESOURCE_LEAVE\):'

LIMIT_CPU=10
LIMIT_VSZ=$((8 * 1014 * 1024))
update for nobug-201002.1 2010-02-03 10:38:12 +01:00			`LOGSUPPRESS='^\(\\[0-9]\\* \)\?[0-9]\{10,\}[:!] \(TRACE\\|INFO\\|NOTICE\\|WARNING\\|ERR\\|TODO\\|PLANNED\\|FIXME\\|DEPRECATED\\|UNIMPLEMENTED\\|RESOURCE_ANNOUNCE\\|RESOURCE_ENTER\\|RESOURCE_STATE\\|RESOURCE_LEAVE\):'`
Increase valgrind time limits for tests and fix test.sh glitch On my slow laptop some tests fail because of timeouts under valgrind, going to 55/60 seconds should be sane. Fixing an undiscovered bug ignoring timeout configs in test.sh. 2010-07-16 21:47:08 +02:00
Testsuite: change resource limits to work on Debian/Jessie this is strange; we need to crank up the VSize to 6 GiB to get the 5 massively multithreaded tests to run. I checked those tests with valgrind and can not see any such huge allocation 2015-08-14 23:05:21 +02:00			`LIMIT_CPU=10`
Testsuite: increase virtual memory limit Set ulimit -v setting to 8 GiB (setting is given in kbyte) Otherwise it is not possible to start 100 Threads. This is surprising, because the actual memory usage of the tests in question are minuscule and also TOP does not show any significant memory peak when running the test. 2023-10-04 22:42:37 +02:00			`LIMIT_VSZ=$((8 * 1014 * 1024))`