Re: 21.5 build time segmentation fault

Ar an dara lá de mí Méan Fómhair, scríobh Mats Lidell: Yes, as do a few crashes in day-to-day use. My initial crashes were in part because I set gc-incremental-traversal-treshold to 10000, under the misunderstanding that it did what gc-cons-incremental-threshold does. Its default value is 400000, and with the lower value the kkcc_gc_stack grew uncontrollably large, realloc fails, and XEmacs crashed. (And then crashed again while crashing, because memory_full() calls GC recursively.) But; that shouldn’t have happened. The object that was being traversed as the realloc failed was a struct display_line that, as far as I can see, was garbage, all its values were inconsistent. Your documented crash also seems likely to be from traversal of a garbage object. I wonder is it just as simple as the redisplay structures being inconsistently marked as GC roots through the kkcc descriptions? That might fit with my never seeing the crash on this TTY-only NEWGC build. Though it should provoke the same crashes on a KKCC non-NEWGC build, which we don’t see. -- ‘As I sat looking up at the Guinness ad, I could never figure out / How your man stayed up on the surfboard after forty pints of stout’ (C. Moore)