On Sun, 12 Mar 2017 18:32:04 -0700
Paul Rogers <[email protected]> wrote:

> > From: Hazel Russman <[email protected]>
> > On Sat, 11 Mar 2017 14:27:21 -0800
> > Paul Rogers <[email protected]> wrote:
> >
> > > First, allow me to ask a question.  Why do you need to go through Ch
> > > 6 again?  What is fundamentally different?  Is one 32-bit and the
> > > other 64?
> >
> > Your question puzzles me. It's a different machine. The new lfs
> > partition that I make won't have any software on it except for the
> > unpacked toolkit. Chapter 6 creates that software. How could I not go
> > through that chapter?
> >
> > The basic architecture is the same: x64_86. But the toolkit was built
> > on an Intel processor and it will be running on a Via. I hope that
> > won't cause any problems.
> >
> > From: Simon Geard <[email protected]>
> >
> > On Sat, 2017-03-11 at 14:27 -0800, Paul Rogers wrote:
> > > > First, allow me to ask a question.  Why do you need to go through
> > > > Ch 6 again?  What is fundamentally different?  Is one 32-bit and
> > > > the other 64?
> >
> > Although not explicitly saying so, I think that by "the toolkit I
> > created for my desktop build", Hazel means only Ch 5, and that Ch 6
> > will be done on the laptop.
> 
> So I'm glad I asked.  I don't understand.
> 
> I've been building (B)LFS since 4.1 for "retired" hardware--only
> appropriate--so am I.  I've been very used to setting CFLAGS & CTARGET,
> etc., for i586 & i686 architectures.  And it worked for a 7.2 system I
> built on a Coppermine (or Tualatin) Pentium-3 and installed on a little
> Via-C7 (approximately a 1GHz Pentium-3 work-alike) mini-ITX system.
> I'm only now beginning setting up my second 64-bit build--with 7.10!
> 
> What I've seen so far has led be to believe 64-bit architectures
> don't need those settings.  It worked without them for my first
> build, made on an i7-940, but installed and running in 64-bit mode on
> a Conroe Core-2 Duo.  (My process makes essentially an installable
> "distribution" from Ch-6 as-built binaries.  I don't just use a
> complete filesystem tarball.)
> 
> Now, given that both systems are 64-bit, where does gcc/glibc build
> differently for individual processor differences?  (Apart from that ABI
> parm in gmp we'd need to build a 32-bit system on a 64-bit host.  I also
> did that using that 64-bit system on the i7-940, i.e. rebuilt the same
> code proven in the 64-bit build for a 32-bit rebuild.  Right now that's
> what I'm running.)  Why can't Hazel use her Ch-6 code?  It seems to be
> working for me.
> 
> I can see it may be important what one picks when building a kernel.
> The kernel is the hardware interfacer.  When I build the initial kernel
> in Ch-8, I always pick a "lowest common denominator" I expect to be able
> to run on any of my hardware.  There's always time for kernel
> customization once it's running.
> -- 
> Paul Rogers
> [email protected]
> Rogers' Second Law: "Everything you do communicates."
> (I do not personally endorse any additions after this line. TANSTAAFL :-)
> 
> -- 
Ah, I see now. You want me to transplant the entire system, not just the 
toolkit. I suppose I could do that, then just build and install a new kernel 
more adapted to the hardware. I'd need a new xorg-server too with the 
openchrome driver. I dare say that would work, but to me it would feel a bit 
like cheating. If I wanted a ready-made system, why use LFS?
-- 

H Russman
-- 
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

Reply via email to