Description of problem: # gdb --args /usr/bin/gcj-dbtool <apparently whatever> .... Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 46912496220496 (LWP 12548)] 0x0000003167903909 in _dl_addr () from /lib64/libc.so.6 (gdb) bt #0 0x0000003167903909 in _dl_addr () from /lib64/libc.so.6 #1 0x00000034021234fb in _Jv_RegisterLibForGc () from /usr/lib64/libgcj.so.7rh #2 0x00000034021183ac in _Jv_RegisterClasses () from /usr/lib64/libgcj.so.7rh #3 0x000000000040271e in _init () #4 0x0000000000000003 in ?? () #5 0x0000000000404a20 in ?? () #6 0x0000003167820a3e in __libc_start_main () from /lib64/libc.so.6 #7 0x0000000000402b49 in ?? () #8 0x00007fffa3494cd8 in ?? () #9 0x0000000000000000 in ?? () This program is used by by /usr/bin/rebuild-gcj-db which, in turn, shows up in various %post scripts and those clearly fail. I do not know if the problem is truly limited to x86_64. Version-Release number of selected component (if applicable): gcc-4.1.1-12 libgcj-4.1.1-12 How reproducible: always
The problem re-appeared after the latest round of updates which did NOT touch gcc or libgcj. I guess that changes in glibc will be the most likely trigger here. A backtrace from gdb with libgcj-4.1.1-13 looks exactly like the above; only addresses are slightly different.
*** This bug has been marked as a duplicate of 200635 ***