sigsegv?

Thu Nov 13 21:07:00 GMT 2003

On Nov 12, 2003, at 7:38 AM, Andrew Haley wrote:
>> Does this ring a bell for anybody?
>>>> #0 0x42073771 in free () from /lib/tls/libc.so.6
>> #1 0x420ebd8e in __unregister_atfork () from /lib/tls/libc.so.6
>> #2 0x42029fb8 in __cxa_finalize () from /lib/tls/libc.so.6
>> #3 0x4038ed55 in __do_global_dtors_aux ()
>> from /home/tromey/gnu/Trunk/install/lib/libgcj.so.4
>> #4 0x406563a6 in _fini () from 
>> /home/tromey/gnu/Trunk/install/lib/libgcj.so.4
>> #5 0x4000ce44 in _dl_fini () from /lib/ld-linux.so.2
>> #6 0x42029d40 in exit () from /lib/tls/libc.so.6
>> #7 0x08048bee in main (argc=1, argv=0xbfffedf4)
>> at ../../../gcc/libjava/gij.cc:163
>>>>>> This is the result of just running "gij".
>> This is with a completely fresh build as of last night. Nothing on
>> my machine has changed, it must be some new gcc bug.
>> I'm getting this too.
>> I didn't get it when I did my cast checkin a few days ago.
>> LD_ASSUME_KERNEL=2.4.17 does indeed fix the problem.
>> Jeff, how did you discover this?

Uhh, who me? ;-)
Setting LD_ASSUME_KERNEL is a reflex action whenever I get weird 
crashes. It tells glibc to use safe fallbacks for things that may be 
incompatible with recent kernels/binutils/etc.
I guess the next step to track this down would be to either use a debug 
glibc build, or try to figure out which GCC change caused it to happen. 
(BTW the problem also goes away with a more recent glibc)
Regards
Bryce