Robert_Pluim(a)BayNetworks.COM (Robert Pluim) writes: It's definitely a bug in gcc (and apparently egcs). I played around a bit more and found that the minimal optimization options that produce the crash are: -O1 -fcse-skip-blocks -fschedule-insns A good set of options to avoid the crash seems to be: -O3 -fno-schedule-insns I started looking at the assembler and the RTL stuff generated by the different compiler-passes but got lost real quick. Maybe somebody who is a lot smarter than I am can make some sense of this. FWIW the crash occurs at the (inlined) call to SYNTAX_CODE in the macro-expanded SYNTAX() at search.c:845 Steve, how shall we proceed now? I guess quite a few people will get hit by this if we don't find a way to fix the problem... How about an entry in PROBLEMS? Has anybody any experience filing bug reports with the gcc/egcs-folks? I guess they won't be thrilled if I mail them xemacs-21.0-b50.tar.gz, right? -- Gunnar Evermann Natural Language Systems Group, University of Hamburg EMail: Gunnar.Evermann(a)nats.informatik.uni-hamburg.de