Pierre Labastie wrote: > So, I eventually managed to install slackware (only packages for passing the > host reqs, running jhalfs, and running the slackware package management > system. The key point (I think) is that I haven't installed the package > "gmp-5.0.5-x86_64-1", and that it does not prevent gcc to run, simply because > libgmp.so.10.0.5 is also included in elflibs (actually, it is in one of Hazel > Russmann's posts). > > And I get exactly the same error:
> libtool: link: ranlib .libs/libmpc.a > /bin/sed: can't read /usr/lib64/libgmp.la: No such file or directory > libtool: link: `/usr/lib64/libgmp.la' is not a valid libtool archive > Adding --with-gmp-include=$(pwd)gmp --with-gmp-lib=$pwd/gmp/.libs *does not > help*. > > Actually, I wonder whether it would not be cleaner to use host gmp, mpfr and > mpc completely, for building gcc pass1. Those libraries are installed anyway > with recent gcc. The only requirement would be that the headers be installed > too. > > Well, I have just read the thread "GCC build first pass: mpc build looks for > libgmp.la in the wrong place" to the end (should have done that before, I > know!). Basically, the build only fails when libmpfr.la is on the host and > libgmp.la is not. But if any is on the host, the build system uses paths on > the host. I'll try to reproduce this on Debian. It sounds like we need to add a check for libgmp on the host. Perhaps libmpfr and libmpc also. -- Bruce -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
