viteno> Michael, I guess this might be of interest to your group. It is! viteno> I've just viteno> "experienced" three subsequent crashes with a HEAD xemacs using the viteno> new --kkcc switch (see attached Installation for more information). viteno> Unfortunately, the backtrace of the first crash was broken, only viteno> addresses and no code, but I've attached traces of the latter two. Hm. David's recent fix which I just checked in might help with the problem. If it fails on some specific Lisp object, as in your xemacs.bt1, could you try pobj from gdb on that object so we have a better idea what part of the invariants are failing? Many thanks for trying USE_KKCC! -- Cheers =8-} Mike Friede, Völkerverständigung und überhaupt blabla

viteno> sperber(a)informatik.uni-tuebingen.de (Michael Sperber [Mr. Preprocessor]) writes: viteno> Hi! viteno> Unfortunately, the patch hasn't helped. After having applied them and viteno> recompilation, I've started a new session in the debugger. It crashed viteno> right at the start. I've traced the gdb session. The output is viteno> attached to this message. viteno> Do you need information from higher stack levels? No, this is fine. All crash reports I've seen so far are either stack overflows (good; I know how to make them go away eventually) or GCPRO bugs. These crashes belong to the latter category: line 4360 in alloc.c is the loop marking GCPRO'd objects. It seems kkcc is merely exposing these bugs which are probably somewhere else in the code. Unless I find a way of making these reproducible for me, they'll be very hard to find, unfortunately. -- Cheers =8-} Mike Friede, Völkerverständigung und überhaupt blabla