On Sun, 2008-01-27 at 14:05 -0800, Ben Jackson wrote:
> On Sun, Jan 27, 2008 at 04:20:34PM +0000, Peter Clifton wrote:
> > PCB segfaults opening a layout from the command line It seems we have
> > NULL pointer dereference, (or just dereference of an uninitialised
> > pointer).
> 
> DJ added that call fairly recently (when changing snap behavior) and
> the lesstif HID can tolerate that call because there's no server round-
> trip (it's just returning a local variable that saved the shift-up-down
> state from the last event).

Good to know (and unsurprising as to the events which caused it to
break... adding core code without testing both HIDs).

Its interesting.. should the function try to make a round-trip? (IE...
is the right answer the shift key status "now", or when the last event
was?)

In general though, I remember having seen those calls, and thinking that
the lesstif approach would be more efficient. Perhaps the workaround for
the crash in this case would be to implement a similar behaviour.


It does seem strange that the core is dictating UI policy, e.g., what a
certain key should do. I guess I have to remember that PCB's core isn't
really "backend" stuff as such.. its more a shared portion which the HID
"skins" with different widgets.


-- 
Peter Clifton

Electrical Engineering Division,
Engineering Department,
University of Cambridge,
9, JJ Thomson Avenue,
Cambridge
CB3 0FA

Tel: +44 (0)7729 980173 - (No signal in the lab!)



_______________________________________________
geda-dev mailing list
[email protected]
http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev

Reply via email to