Pure LFS - e2fs trouble

Greg Schafer gschafer at zip.com.au
Mon Feb 17 13:44:11 PST 2003


On Mon, Feb 17, 2003 at 09:51:17PM +0100, Jules Bj?rn Colding wrote:
> Ok, after a bit of googling I now know that rdtsc is the "Read-Time
> Stamp Counter". This instruction is only available on Pentiums and
> higher. The problem is that I am compiling for a i486 in a fast dual
> Athlon.
> 
> So, something from my build machine has sneaked over into my "pure" lfs
> build.
> 
> I know that this probably has something to do with my build and will
> continue to look into this. I am wide open for suggestions though..

You need to go read the "crosscompiling-x86.txt" hint. Ignore everything in
that hint apart from the kernel module uname hack. Do not fiddle with --host
or --target. Do not fiddle with -march or -mcpu. If you follow these
instructions everything will "just work".

Be warned though, glibc's (even 2.2.5) have a bug when compiled by gcc-3.2.x
for i386 and i486 which result in a broken ldconfig. The "ld.so.cache" will
be 64 bytes in size and anything you add to /etc/ld.so.conf will not get
recognised. This has come up on the lists more than once, the last time only
a few days ago. Search the archives for more info.

Lastly, this has nothing to do with the pure LFS hint. You didn't follow the
hint to the letter because you fiddled with optimisaton despite the big fat
warning not to do so. This is a generic problem and your post should have
gone to lfs-support. Could you please be a bit more careful about
apportioning blame in your posts?

Greg
-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list