Cross compiling for i486 - too late now?

davidh davidh-lfs at
Thu Apr 22 18:36:32 PDT 2004

On Thu, 22 Apr 2004 19:24:09 +0100, mrnobby wrote
> "davidh" <davidh-lfs at> wrote in message
> news:20040421121750.M74570 at
> > Hi everyone,
> >
> > I'm building a LFS (my second one), but this time I'm cross-compiling
> > for an i486.  Unfortunately I didn't read the cross compiling hint
> > until now - I'm already up to compiling GCC (pass 2; near the end of
> > chapter 5)
> >
> > My question is this:  Am I too late to start using the uname hack now?
> >
> > Since I've only built the toolchain, can I now install the uname hack
> > and have gcc build me i486 binaries from now on?
> >
> > If I have to start over again, then so be it... but thats a waste of
> > around 4 hours build time...
> Hi,
> If you've been reading, you'll know that I am building LFS 5.0
> on an i386/DX-20. Native. If not, check out
> You'll see that the first
> package, Binutils, took 3.25 hours just to un-pack. Four hours is no
> time at all!

Truly amazing :)  I still have a 386dx16 lying around which I used for an 
internet gateway 4 years ago!

The problem is that I have a lot of other things going on as well, and I 
don't really have enough time to sit around and wait :)  So if it's possible 
to use the idle CPU power of my P4-2.4 instead of waiting for the slower 
computer it would be worth it.

> My recommendation would 
> be to forget the cross compiling and build LFS 5.0 straight on to 
> your 486. Even if it was a lowly 25 MHz version, it would probably 
> only take 3 weeks at most. You'll also be guaranteed 100% i486 code 
> as well ;-)

The target host is actually a VIA C3 processor at 533Mhz (one of those Mini-
ITX boards) so I don't think the build will take *that* long... Problem is 
that I can't find a Linux distro that actually runs on it - the processor 
tells the OS it can run i686 code, but that is not true since an i686 kernel 
crashes it - you have to compile it for i486 to run successfully.


More information about the lfs-support mailing list