Re: EGCS troubles
Compile Jade with -g and see where in the coredump the signal 11 is occuring.? What does ``ldd jade'' show?? You might be mixing shared libs, that doesn't work for C++.? Could also be an exceptions problem.? Try compiling with -fnoexpcetions. [asmo...@daemon] (163) $ ldd /usr/local/bin/jade /usr/local/bin/jade: ??? libstyle.so.1 = /usr/local/lib/libstyle.so.1 (0x280b6000) ??? libspgrove.so.1 = /usr/local/lib/libspgrove.so.1 (0x282ac000) ??? libgrove.so.1 = /usr/local/lib/libgrove.so.1 (0x282ef000) ??? libsp.so.1 = /usr/local/lib/libsp.so.1 (0x282f7000) ??? libintl.so.1 = /usr/local/lib/libintl.so.1 (0x284ce000) ??? libstdc++.so.3 = /usr/lib/libstdc++.so.3 (0x284d2000) ??? libm.so.2 = /usr/lib/libm.so.2 (0x2850d000) ??? libc.so.3 = /usr/lib/libc.so.3 (0x28527000) I know for certain that libintl is gcc. Will go and do recompilations of all stuff this weekend *sigh*? ;) ? In my case (it seems all libs compiled under egcs - exactly when jade compiled): /usr/local/bin/jade: ??? libstyle.so.1 = /usr/local/lib/libstyle.so.1 (0x280b9000) ??? libspgrove.so.1 = /usr/local/lib/libspgrove.so.1 (0x282b6000) ??? libgrove.so.1 = /usr/local/lib/libgrove.so.1 (0x282f9000) ??? libsp.so.1 = /usr/local/lib/libsp.so.1 (0x28301000) ??? libstdc++.so.3 = /usr/lib/libstdc++.so.3 (0x284e4000) ??? libm.so.2 = /usr/lib/libm.so.2 (0x28521000) ??? libc.so.3 = /usr/lib/libc.so.3 (0x2853c000) ? To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message
Re: EGCS troubles
The Only way I could get Jade to work with the new compiler was with CFLAGS= -O -pipe That is not so bad. Before EGCS, we would state that -O is the only optimization that is know to always work and what we tell people to use. Mike Smith has written about this many times in Hackers and Current. -- -- David(obr...@nuxi.com -or- obr...@freebsd.org) To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message
Re: EGCS troubles
At 09:58 AM 4/9/99 +0300, Maxim Sobolev wrote: Compile Jade with -g and see where in the coredump the signal 11 is occuring. What does ``ldd jade'' show? You might be mixing shared libs, that doesn't work for C++. Could also be an exceptions problem. Try compiling with -fnoexpcetions. [asmo...@daemon] (163) $ ldd /usr/local/bin/jade /usr/local/bin/jade: libstyle.so.1 = /usr/local/lib/libstyle.so.1 (0x280b6000) libspgrove.so.1 = /usr/local/lib/libspgrove.so.1 (0x282ac000) libgrove.so.1 = /usr/local/lib/libgrove.so.1 (0x282ef000) libsp.so.1 = /usr/local/lib/libsp.so.1 (0x282f7000) libintl.so.1 = /usr/local/lib/libintl.so.1 (0x284ce000) libstdc++.so.3 = /usr/lib/libstdc++.so.3 (0x284d2000) libm.so.2 = /usr/lib/libm.so.2 (0x2850d000) libc.so.3 = /usr/lib/libc.so.3 (0x28527000) I know for certain that libintl is gcc. Will go and do recompilations of all stuff this weekend *sigh* ;) In my case (it seems all libs compiled under egcs - exactly when jade compiled): /usr/local/bin/jade: libstyle.so.1 = /usr/local/lib/libstyle.so.1 (0x280b9000) libspgrove.so.1 = /usr/local/lib/libspgrove.so.1 (0x282b6000) libgrove.so.1 = /usr/local/lib/libgrove.so.1 (0x282f9000) libsp.so.1 = /usr/local/lib/libsp.so.1 (0x28301000) libstdc++.so.3 = /usr/lib/libstdc++.so.3 (0x284e4000) libm.so.2 = /usr/lib/libm.so.2 (0x28521000) libc.so.3 = /usr/lib/libc.so.3 (0x2853c000) The Only way I could get Jade to work with the new compiler was with CFLAGS= -O -pipe The only way I have tested it is by building the Handbook -O2 -pipe built ,but Signal 11 -O2 -fno-exceptions won't build Manfred = ||man...@pacbell.net || ||Ph. (415) 681-6235|| = To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message
Re: EGCS troubles
At 09:37 AM 4/9/99 -0700, David O'Brien wrote: The Only way I could get Jade to work with the new compiler was with CFLAGS= -O -pipe That is not so bad. Before EGCS, we would state that -O is the only optimization that is know to always work and what we tell people to use. Mike Smith has written about this many times in Hackers and Current. mysql322 from ports is another one, if you try and compile it with the stock -O -pipe it builds up till almost the end but when it gets to sql.yacc.cc the machine just hangs there and finally dies (no input or output). I have to go into the debugger and reboot. But if you add -fno-exceptions it builds fine.It's taken a couple of days but I've weeded out all of the programs that depended on old gcc libs and rebuilt them. Manfred = ||man...@pacbell.net || ||Ph. (415) 681-6235|| = To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message
Re: EGCS troubles
eebsd.dsl -t sgml handbook.sgml *** Signal 11 Interesting that when it was compiled using pgcc port it worked just fine. Then it would probably do fine using the EGCS Port. But, the port != /usr/src. The library setup of EGCS in /usr/src is very different than the Port -- -- David(obr...@nuxi.com -or- obr...@freebsd.org) To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message
RE: EGCS troubles
On 08-Apr-99 Maxim Sobolev wrote: Hi, While this problem might not belongs to this list however it is difficult in this case to distinguish exactly. Something broken in new egcs c++ compiler - jade doesn't work with following symptoms (checked on two machines). Apr 7 20:26:04 daemon /kernel: pid 61135 (jade), uid 0: exited on signal 11 (core dumped) Verified... David, might this be due to the compiler itself of the libraries? Any way of making certain? --- Jeroen Ruigrok van der Werven http://www.freebsdzine.org asmodai(at)wxs.nlThe idea does not replace the work... Network/Security Specialist http://home.wxs.nl/~asmodai *BSD: Powered by Knowledge Know-how http://www.freebsd.org To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message
Re: EGCS troubles
Something broken in new egcs c++ compiler - jade doesn't work with following symptoms (checked on two machines). Apr 7 20:26:04 daemon /kernel: pid 61135 (jade), uid 0: exited on signal 11 (core dumped) Verified... David, might this be due to the compiler itself of the libraries? Any way of making certain? Compile Jade with -g and see where in the coredump the signal 11 is occuring. What does ``ldd jade'' show? You might be mixing shared libs, that doesn't work for C++. Could also be an exceptions problem. Try compiling with -fnoexpcetions. -- -- David(obr...@nuxi.com -or- obr...@freebsd.org) To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message
Re: EGCS troubles
On 08-Apr-99 David O'Brien wrote: Something broken in new egcs c++ compiler - jade doesn't work with following symptoms (checked on two machines). Apr 7 20:26:04 daemon /kernel: pid 61135 (jade), uid 0: exited on signal 11 (core dumped) Verified... David, might this be due to the compiler itself of the libraries? Any way of making certain? Compile Jade with -g and see where in the coredump the signal 11 is occuring. What does ``ldd jade'' show? You might be mixing shared libs, that doesn't work for C++. Could also be an exceptions problem. Try compiling with -fnoexpcetions. [asmo...@daemon] (163) $ ldd /usr/local/bin/jade /usr/local/bin/jade: libstyle.so.1 = /usr/local/lib/libstyle.so.1 (0x280b6000) libspgrove.so.1 = /usr/local/lib/libspgrove.so.1 (0x282ac000) libgrove.so.1 = /usr/local/lib/libgrove.so.1 (0x282ef000) libsp.so.1 = /usr/local/lib/libsp.so.1 (0x282f7000) libintl.so.1 = /usr/local/lib/libintl.so.1 (0x284ce000) libstdc++.so.3 = /usr/lib/libstdc++.so.3 (0x284d2000) libm.so.2 = /usr/lib/libm.so.2 (0x2850d000) libc.so.3 = /usr/lib/libc.so.3 (0x28527000) I know for certain that libintl is gcc. Will go and do recompilations of all stuff this weekend *sigh* ;) --- Jeroen Ruigrok van der Werven http://www.freebsdzine.org asmodai(at)wxs.nlThe idea does not replace the work... Network/Security Specialist http://home.wxs.nl/~asmodai *BSD: Powered by Knowledge Know-how http://www.freebsd.org To Unsubscribe: send mail to majord...@freebsd.org with unsubscribe freebsd-current in the body of the message