I'm cc'ing the mailing list in case anyone else has any insights. According to Kay Schenk: > Gilles Detillieux wrote: > > According to Kay Schenk: > >>I'm trying to install htDig on a Solaris 2.6 system. I get a missing > >>symbol "collect2" (I think this is the problem anyway) that your routine > >>thinks should be in libgcc.a and isn't. > >> > >>Any other ideas about where you think this might be located so I can do > >>an additional library include? ... > > It might help to have the exact error messages from the compiler/linker. ... > I am THRILLED that you got back to me--Ok, here's what's going on! > > Error line from make--cut up a bit! > > c++ -o htfuzzy -L../htlib -L../htcommon -L../db/dist -L/usr/lib > Endings.o EndingsDB.o Exact.o Fuzzy.o Metaphone.o Soundex.o > SuffixEntry.o Synonym.o htfuzzy.o Substring.o Prefix.o > ../htcommon/libcommon.a ../htlib/libht.a ../db/dist/libdb.a -lnsl > -lsocket -lgcc -shared > > --next line-- > > /usr/local/lib/gcc-lib/sparc-sun-solaris2.6/2.95.1/libgcc.a: could not > read symbols: Bad value > > collect2: ld returned 1 exit status > *** Error code 1 > > --It seems the normal "ld" spawned form gcc, normally from collect2 is > well, I don't know what it's doing. We rarely use g++ (this c++ is > really a link to that), so I don't know what's happening. > > > yes, I've been messing with the ldflags etc, but it doesn't seem to matter. > > > Done some further reading as well. > > I get this through a load with the loader without going through gcc with > the following: > > ld -o htfuzzy -dy -G -B dynamic -L../htlib -L../htcommon -L../db/dist \ > -L/usr/lib -lnsl -lsocket Endings.o EndingsDB.o Exact.o Fuzzy.o \ > Metaphone.o Soundex.o SuffixEntry.o Synonym.o htfuzzy.o Substring.o > Prefix.o\ > ../htcommon/libcommon.a ../htlib/libht.a ../db/dist/libdb.a > > --I tried to see if I could just replace your gcc load with this, but > well, I was admonished not to by one of sysadmins--which I'm not BTW.
... and later ... > Well I decided to get crazy and change all the c++ linkages to normal LD > commands for the main utilities...I believe our C linker is hinky and > has been for as long as I've been messing with it. Our sysadmins don't > have the time to do an upgrade right now. Anyway, I'm running rundig > right now and nothing has blown up so far. > > I'll give you another update as soon as I have anything to report. Well, there does seem to be something odd with your C++ installation on your system. If the c++ commmand isn't calling "ld" correctly, or if there's a problem with libgcc.a as the error message implies, then something needs to be reinstalled or reconfigured on your system. I find it odd that your "ld" command above would successfully link the htfuzzy program, though, because there's no reference to the C or C++ libraries on your command line. However, it does indeed sound like you managed to find something that work to get these programs linked. It may be that the copy of libgcc.a that c++ was telling ld to use is damaged or incompatible, and in your experimenting with ld you got it to find a library that works correctly. It might be interesting to know, for the list archives, what ld command you ended up using in your Makefiles. -- Gilles R. Detillieux E-mail: <[EMAIL PROTECTED]> Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/~grdetil Dept. Physiology, U. of Manitoba Phone: (204)789-3766 Winnipeg, MB R3E 3J7 (Canada) Fax: (204)789-3930 _______________________________________________ htdig-general mailing list <[EMAIL PROTECTED]> To unsubscribe, send a message to <[EMAIL PROTECTED]> with a subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html
