On Wed, 23 Feb 2000, Paul Kranenburg wrote:

> > Would it not be better to use the pax TMPDIR fix, and reduce the
> > size of the kernel instead? If we do come up with a better long-term
> 
> Has it been verified the reduction is sufficient?

Not empirically. However, there's not been much in the way of
sparc-specific changes on the branch since 1.4.1; I just went through
them all and there were no config file changes at all. So it's got to be
just a little bit of general expansion that caused this, and I'd think
a tweak would fix it, if this wasn't a problem in 1.4.1.

There's a chance I might not be completely !@#$%ed with work this
week-end, if not, I can look in to this myself.

> In any case, I fear it will only be a question of time before we cross
> the line again. In -current already, the INET6 libc code has topped off
> all available room in the `MD' image.

Indeed, but I think a better long term fix would be to have a single
second-stage boot loader dynamically find a place to relocate itself.
It's run in physical mode, and writing to non-existent memory doesn't
appear to produce any sort of bus fault--the bootloader appears to fail
with an illegal instruction after jumping to an area with no RAM. Given
this, one could write a test pattern into each 4 MB memory range above
the first until one reads back correctly, and then relocate there.

cjs
-- 
Curt Sampson  <cjs@cynic.net>   917 532 4208   De gustibus, aut bene aut nihil.
The most widely ported operating system in the world: http://www.netbsd.org