This frysk-core test regularly fails for me. I'll attach a TestRunner -c FINE log.
Created attachment 1177 [details]
./TestRunner -c FINE frysk.proc.StressAttachDetachRapidlyCloningMainTask log
This stress tests seems to behave a lot better now. But occasionally it gets
into some loop and doesn't stop at all. This however seems unrepeatable with -c
A loop (with CPU usage pegged) or a hang (sitting in the event loop or waitpid)?
CPU usage pegged
Sorry, should be more precice. CPU pegged by the test program or frysk, i.e.,
is frysk sleeping?
I'm going to assume that this is from the CLASSPATH pattern matcher (which is
really really slow).
Typically, on a 1.1ghz i386 I need to specify --timeout 30 to prevent the timeout.