------- Additional Comments From pcarlini at suse dot de 2004-10-08 15:11 ------- > Agggh! Sorry everybody. My testing was flawed: I just discovered a typo in my > config files for gcc builds. I was using --enable-_cxa_atexit instead of > --enable-__cxa_atexit. (Two underscores.)
Thanks. Now things are much more clear. > Hopefully this will resolve this issue. About this, however, I'm really unsure! If that option is meaningful only with glibc, this implies logically that the allocator is broken on anything not using glibc?!? Or we should fix it not to rely on the order of destruction?!? I'm really worried by these new (for me, I mean) problems... -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17873
