On Tuesday 01 December 2009, Zach Welch wrote: > On Tue, 2009-12-01 at 01:23 -0800, David Brownell wrote: > > > > > Registers a signal handler to catch SIGSEGV in order to display the > > > > > stack where the program crashed. > > > > > > > > Is this for inside OpenOCD? If so, I'd rather just expect folk > > > > to run inside GDB. Either they're running natively and should > > > > never see SEGV ... or they should be able to fire up GDB to get > > > > this data (and likely more). > > > > > > > > > > Not everyone wants to run GDB, and not all segfaults can be predicted. > > > > Any developer who's not willing to run GDB to catch a fault ... > > take them out back behind the woodshed and "bugfix" them! ;) > > If they _can_ catch it, sure. However, you're now talking about doing > harm to the messengers. That's highly counterproductive. Asking > someone to do extra work that could have been avoided is crap attitude.
I'm actually talking hyperbole. And as for "extra work" ... it's something of a question about *who* does the work. Person seeing a problem? Motivated to resolve it. Everyone else? ... not. > > No end user should ever see segfaults, or any other flavor of > > rude/unexpected exit. If they see one that's quite a major > > bug in the code. > > I reiterate my principle point: heisenbugs. The code is not yet pretty > enough for me to feel that this handler will never be exercised in > unpredictable contexts, nor do I feel our processes have yet reached a > point where we can prevent adding such triggers in the future. And what better way to make sure we improve our processes than to stop buying band-aids to apply over spurting wounds? :) (Hyperbole again, it's late.) _______________________________________________ Openocd-development mailing list Openocd-development@lists.berlios.de https://lists.berlios.de/mailman/listinfo/openocd-development
