Problems with 5.x [ boot-loader related, NOT hardware related ]

[Matt Emmerton]

> On 2005-Jun-30 18:18:16 -0400, Matt Emmerton <matt at gsicomp.on.ca> wrote:
> >> More information required (as almost always):
> >> - type and model of maxhine?
> >
> >It's home-made.  It doesn't have a type or model.  FWIW, it's ran FreeBSD
> >fine since 3.2.
>
> Presumably you bought the motherboard.  What brand is it?

DFI  686IPK.  A piece of crap, really, but it worked fine for -stable all
the way through 3.x and 4.x.

> >> - describe in detail your buildworld problems (and any other problems),
> >> specifically any error messages you might get.
> >
> >Basically, it konks out immediately while trying to rebuild 'make'
because
> >of missing system headers.  Probably a caveat with trying to build 5.x on
a
> >4.x machine but I didn't see anything in special instructions (UPDATING)
or
> >on the web about this.
>
> Can you please cut-and-paste the exact commandline you use to buildworld,
> the contents of /etc/make.conf and the tail of the make output (back to
> the last few command echos).

I don't have the info.  I'd need to reinstall 4.10 on it, cvsup to -stable
and try again.

Frankly, I'm not concerned why attempting to buildworld 5.x sources on a 4.x
machine screws up -- I'm more interested why 5.x won't even install.  The
"build from source" was just a last-ditch attempt at 3am this morning.

Is there anyone who actually _worked_ on the BTX and boot loader changes in
5.x willing to comment on my particular problems getting any flavour of 5.x
to install on this machine?

> >> Have you ruled out hardware problems? By running memtest86 for example?
> >> (Yes, I know that 4.10 installs ok. That proves almost nothing.)
> >
> >The system has run fine for 3+ years - the hardware is fine.
>
> Hardware can fail from old age.  Anything with moving parts (fans and
> disk drives) as well as electrolytic capacitors have a finite lifetime.

That doesn't explain why 4.10 was working fine on it yesterday and a few
hours later 5.x doesn't

If the hardware was broken, I'd be intelligent enough to realize that and
wouldn't waste anyone's time here.  I've been tracking -stable for 7 years
and have seen my share of broken hardware to know what those problems look
like, so let's just focus on the problem at hand, okay?

--
Matt