Jeremy Huntwork wrote: > Multi-lib Debian Lenny showed that there is a hole in the current > bootstrapping method. Admittedly, it is an odd host, but it revealed a > gap.
IMHO, only odd hosts can reveal gaps :) > When we bootstrap GCC on the first pass, all we really prove is that the > *host system* is sane and is compatible with the toolchain we're > building. We really should assume that the host isn't sane. We should > only trust its compiler far enough to build Binutils and a > non-bootstrapped GCC. > I wanted to catch you here with the words "we trust the host a bit more that you say we need, because the test programs compiled by the ./configure script are compiled by the new toolchain agains the hosts's headers and libc", but you said "we should only trust its _compiler_ ...", so the catch doesn't work. > Once we have a new binutils and gcc, then we should be able to build a > usable and working Glibc, no matter what condition the host toolchain is > in. Try this starting from uclibc-based host (e.g. http://www.uclibc.org/downloads/root_fs_i386.ext2.bz2 with approprtately adjusted permissions), and you'll see that you are a bit overly optimistic. (yes, this is a case for cross-compilation, and I don't say that this should belong in LFS) > If we are able to bootstrap GCC pass2, then we know we're golden. Here I agree, even though bootstrapping has been proven to be redundant in this case. -- Alexander E. Patrakov -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
