Hi, news from the gcc front ;-) After another 'funny' incident (In the end of stage3 I found I forgot to compile libc_g with new stack size so one stage of gcc was compiled with 24 KB stack and I could start again since comparation between stages was not correct -- probably due to non-crashing-but-still-here bug due to a low stack). OK. Now, this morning I've seen stage-to-stage gcc comparation succeed and finally libstdc++ compilation begin. However, my happiness was not long -- see attached output log. (config.log has no sense to attach since the last test succeed, there's return code 0 and then nothing). I'm not expert in this configure/m4 stuff, please, could someone take a look what is going next to this test to be able to duplicate this crash? It seems it's something serious, on my falcon, when I've tried --disable-bootstrap and 'make' reached this same test (-ffunction-sections), machine just hangs and on aranym I have this Bus error or aranym crash or hang too -- so it seems to something really nasty. My configure flags were: --disable-multilib --disable-libstdcxxpch --disable-nls --enable-lnaguages=c,c++. -- MiKRO / Mystic Bytes http://mikro.atari.org
Attachment:
gcc.log
Description: Binary data