Re: Make TestBreakpoints deterministic

[Andrew Cagney <cagney at redhat dot com>]

Mark,

Does this actually stop the problem?  For instance, if I were to be 
running two instances of the frysk simultaneously (or any other 
operation that would leave around defunct processes), wouldn't that lead 
to similar consequences?
And under exactly what circumstance are you seeing terminated events?  
Having a wait call in tearDown, or waiting on observableProcRemovedXXX, 
does not give me warm fuzzies :-(  The latter implies that the event 
loop is running - it may have crashed.

Andrew

Mark Wielaard wrote:
> Hi,
>
> The following patch makes TestBreakpoint deterministically PASS on both
> my FC5 (x86_64) and FC6test (x86) boxes:
>
> 2006-08-21  Mark Wielaard  <mark@klomp.org>
>
>         * TestBreakpoints.java: Implements Observer.
>         (procTerminated): New boolean field.
>         (setup): Add observableProcRemoved to manager.host.
>         (tearDown): Make sure proc is really gone.
>         (update): New method.
>
> It seems that under fc6 kernels procs are kept around longer than under
> fc5. So we were seeing "about to terminate" procs under fc6 when
> restarting the eventloop, which we weren't seeing under fc5 because they
> were always already gone. You do sometimes observe this with fc5
> kernels, you will get a WARNING: No task for pid XXX from LinuxHost, but
> in those cases it seems the terminated event has already happened, or is
> never delivered to us. Under fc6 we then will get a terminated event,
> which upset this test.
>