Closed GoogleCodeExporter closed 8 years ago
To get around the problem I've changed my implementation of
SigarProcessManager.kill to have:
sigar.kill(pid, Sigar.getSigNum( "KILL" ) );
That will result in a signal of -1 being sent on Windows (which is OK) and on
other systems that don't support SIGKILL (which will probably result in EINVAL).
Original comment by jim.tal...@groupgti.com
on 24 Oct 2011 at 1:31
I am guessing you have correctly configured sigar and included the appropriate
runtime library from sigar-dist?
Its strange that this is a problem inside sigar, since Sigar should support
windows also.
However, I can investigate a bit more when I have time, and commit your changes
to trunk. I don´t see a problem with it at first glance.
Original comment by shervin.asgari@gmail.com
on 26 Oct 2011 at 11:01
Yes I have the correct runtime library loaded (otherwise it wouldn't do
anything).
Reading the source code shows that the problem is there - I guess sigar could
argue that it's a question of how it is used rather than a problem with sigar
itself, but as it stands it makes the kill(long pid, String signame) method
unusable in a cross-platform manner.
Original comment by jim.tal...@groupgti.com
on 8 Nov 2011 at 10:28
The error
sigarProcessManager(org.artofsolving.jodconverter.process.ProcessManagerTest):
kill failed happens to me when I try do a officeManager.stop() without doing a
officeManager.start() first.
Original comment by SergioCa...@gmail.com
on 9 Nov 2011 at 3:52
Added changes to trunk
Original comment by shervin.asgari@gmail.com
on 2 Dec 2011 at 11:45
Original issue reported on code.google.com by
jim.tal...@groupgti.com
on 24 Oct 2011 at 1:15