dbus connection abort

Andreas Volz lists at brachttal.net
Mon Mar 14 14:31:40 PDT 2011

Am Thu, 10 Mar 2011 13:27:49 +0000 schrieb Simon McVittie:

> On Thu, 10 Mar 2011 at 13:33:53 +0100, Andreas Volz wrote:
> > could you imagine that the connection is down only from
> > sending methods to dbus?
> If you sent invalid UTF-8, an invalid boolean whose value is neither
> 0 nor 1 or other malformed data, yes, this can happen. My patches on
> <https://bugs.freedesktop.org/show_bug.cgi?id=16338> diagnose this
> with a _dbus_return_if_fail rather than silently allowing it.

Hm, I'm sending all the time the same data. But I've maybe to create
more checks before I send to detect such situations. I'll maybe also try
your patches.

> > Or do I've a race condition
> Are you using threads? If so, I'm afraid you get to debug them;
> libdbus is theoretically thread-safe, and has what appears to be
> appropriate locking, but various threading problems have been
> reported.

Yes, I'm using threads. Debugging works really good for me. I could
even start both, server and client, in debug (with Anjuta IDE) and then
follow exact the connection trough Dbus-C++ and DBus (with installed
sources and dbg packages in Ubuntu) itself. So I see no problems.


Technical Blog <http://andreasvolz.wordpress.com/>

More information about the dbus mailing list