<br><div class="gmail_quote">2012/7/3 Oliver Schonefeld <span dir="ltr"><<a href="mailto:schonefeld@ids-mannheim.de" target="_blank">schonefeld@ids-mannheim.de</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi,<br></blockquote><div><br>Hi Oliver,<br>Â </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
the last commit to upssched.c (Merge Coverity branch, r3555) broke<br>
upssched with commands that have a second argument.<br>
If a command with a second argument is used, "enc" in line 697 is not an<br>
empty buffer and  snprintfcat() just appends some stuff, i.e. the<br>
following command<br>
 CANCEL onbatt online<br>
becomes<br>
 onlineCANCEL "onbatt" "online"<br>
and is then sent to the pipe-fd.<br>
<br>
The attached patch replaces snprintfcat() with snprintf() and fixes the<br>
problem.<br></blockquote></div><br>thanks a lot for your report.<br>the issue and solution are both confirmed.<br><br>I've applied your patch to the trunk (r3670), and I'm now considering a 2.6.5 release for soon.<br>
<br>note that there are efforts underway to have unit tests, along with integrating the regression test suite, run with our buildbots:<br><a href="http://buildbot.networkupstools.org/">http://buildbot.networkupstools.org/</a><br>
<br>this would provide a decent CI (continuous integration) and prevent us from future regression.<br><br>cheers,<br clear="all">Arnaud<br><br>