On Tue, 2004-07-20 at 02:04 -0500, Branden Robinson wrote: > > On Fri, Jul 16, 2004 at 04:56:08PM +0200, Josselin Mouette wrote: > > > > Program received signal SIGFPE, Arithmetic exception. > > [Switching to Thread 1078789536 (LWP 4629)] > > 0x40633310 in _mesa_test_os_sse_exception_support () > > from /usr/X11R6/lib/modules/dri/debug/i830_dri.so > > #0 0x40633310 in _mesa_test_os_sse_exception_support () > > from /usr/X11R6/lib/modules/dri/debug/i830_dri.so > > Well, I'd say the test failed. :-P
Nope. If it failed, it'd raise a SIGILL, I think. And it still wouldn't be a problem, as it's the whole point of the test to provoke a signal and catch it. The submitter correctly continued and captured a backtrace of the actual problem. The information you asked for will still be useful, and I suspect the backtrace might be more accurate with something like electric-fence. -- Earthling Michel Dänzer | Debian (powerpc), X and DRI developer Libre software enthusiast | http://svcs.affero.net/rm.php?r=daenzer
