Paul Rogers schreef:
I built LFS 8.0 systemd 32 bits succesfully on my desktop (AMD64
processor).
I copied the system to a laptop (32 bits Intel processor) with rsync. It
boots all right there but it refuses to compile new sources.
Upward compatibility does not mean backward compatibility!
You have different CPU architectures, which means cross-compiling
throughout. Even "compatible" Intel CPUs may not be entirely
compatible, e.g. Pentium 3 vs Pentium 4 vs Core2. That whole process is
very delicate. In general, without due care, the RoT is "within
families" and "upward only", and maybe not even the last.
If you go back in the archives for the last couple or three months
you'll find other threads on this kind of issue--one mine,
crosscompiling for a 32-bit i686 box, with a 32-bit LFS running on an
8-core i7 box (for obvious reasons). Ken's right, GMP is first and
foremost among the stumbling blocks, especially where gcc is involved,
but there are others, e.g. I'm not sure how to tell Clang to
cross-compile, audio-video drivers, MPEG decoders, etc.
It is clear where the problem lies. Thanks.
Pierre Labastie gave me some tips for a solution: recompile gmp on your
desktop renaming configfsf.guess to config.guess and configfsf.sub to
config.sub (maybe check permissions so that both are executable), and
transfer it to the laptop.
Then possibly recompile gmp on the laptop, to get some optimization
adapted to the laptop processor.
I am going to try them. If you have others, please mail me.
Hans.
--
http://lists.linuxfromscratch.org/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
Do not top post on this list.
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?
http://en.wikipedia.org/wiki/Posting_style
