another backtrace (was Re: [BRLTTY] Debugging BRLTTY crashes (Baum driver))

Jason White jasonw at ariel.its.unimelb.edu.au
Thu May 10 00:09:08 EDT 2007


On Wed, May 09, 2007 at 10:50:41PM -0400, Dave Mielke wrote:
 
> I'm fairly sure, though, that the name is brl_writeWindow(), especially since
> it does indeed contain a call to memcpy().

Yes, and that's what the line referred to in brlapi_server.c calls.
> 
> Could you please try using valgrind as well?
> 
> Something else you could try is using the -N (uyppercase) option to remove the
> complexity of BrlAPI from the picture.

I'll try valgrind next. Running under electric fence with the -N option was
easy, and it triggered the following after the braille display had been in use
for only about ten seconds. It's another call to brl_writeWindow().

(gdb) bt
#0  0x00002ba73d2be410 in memcpy () from /lib/libc.so.6
#1  0x00002ba73d9ca6e4 in ?? () from /lib/brltty/libbrlttybbm.so
#2  0x000000000040ca7f in main (argc=<value optimized out>, 
    argv=<value optimized out>) at ./main.c:2490



More information about the BRLTTY mailing list