[Bug: 21.5-b29] assertion failure while typing in minibuffer

Mike Kupfer writes: I think this is already in the tracker. Anybody who adds to the issue, please check for dupes. Note that this is really an assert in a dynarr macro, not in redisplay. Three data points: This is probably a bug in the dynarr code. Bounds-check failures in the dynarr code happen to me in a number of places, but I only once managed to discover the bare fact that dy->largest implied a zero-length array, and couldn't get past that. (It only happens when I have a deadline to meet that I can't blarney around. :-( ) *Something* is occasionally tromping on the bounds, but if it were a wild pointer I'd expect it to cause other crashes, too. It's possible that there is some kind of incorrect optimization in GCC, too, and somehow reinitialization code gets skipped because I can't see any way for this to happen. It's also possible that the reset pointers were just accidentally set that way by a wild pointer. Aidan was the last to mess with the dynarr code. I'm pretty sure I recall similar asserts *before* Aidan's patch, so don't blame him even if hg annotate does. :-) Nevertheless, unless Ben shows up, Aidan probably knows more about the dynarr code than anybody active. _______________________________________________ XEmacs-Beta mailing list XEmacs-Beta(a)xemacs.org http://lists.xemacs.org/mailman/listinfo/xemacs-beta