Re: mvwprintw

[Bob Rossi]

On Thu, Mar 30, 2006 at 09:38:26PM -0500, Bob Rossi wrote:
> On Thu, Mar 30, 2006 at 05:29:25PM -0500, Bob Rossi wrote:
> > On Thu, Mar 30, 2006 at 05:02:15PM -0500, Thomas Dickey wrote:
> > > On Thu, 30 Mar 2006, Bob Rossi wrote:
> > > 
> > > >I've attached both a transcript. The trace file is large. Is there a
> > > >better number I could use to give just the info needed?
> > > >The one I have is 14MB.
> > > 
> > > I'd try to cut it down by putting a call to the trace() function right 
> > > before the point where the problem appears, to turn on the trace, and
> > > then right after the relevant refresh.  14Mb sounds like you're updating
> > > the screen a lot...
> > 
> > OK, here is a smaller trace. I got to the point where the next display
> > get's corrupted. I attached with GDB, did 'call trace (8191)', followed
> > by 'c'. Then in the terminal I typed the next char, which causes CGDB to
> > write all of the lines via mvwprintw followed by a wrefresh.
> > 
> > After the display is wrong, I did 'killall -s ABRT cgdb' to end the
> > process. So it should be the minimal amount needed.
> > 
> > Hope this helps, I'm lost
> 
> Just so you know, I just downloaded the same ncurses source and built
> CGDB on a ubuntu machine. There doesn't seem to be a bug. However, if I
> ssh (from linux) into the linux machine that had the bug, it still exists.
> 
> Also, the trace log kind of makes it look like ncurses is doing the
> correct thing. What could be the problem?

At a minimum, what's the next step I can take in order to find the bug?

Bob Rossi