On Wednesday, 07 January 2009 at 20:02, Tobias Mummert wrote: > * Brendan Cully [2009-01-07 03:25 CET]: > > > Er. That wasn't supposed to fix it, it was just supposed to improve > > the backtrace. Possibly the make clean alone did the trick -- you > > could make clean and make and see if it continues to work. If it > > breaks, we have a bug that will be no fun to track down. > > Sorry to say this, but a "make clean && make" build results also in the > same segmentation fault.
Does it crash if you make CFLAGS="-ggdb -O2"? If so, does the back trace look any better?
