On 15 December 2012 17:29, Jochen Kunz <jkunz at unixag-kl.fh-kl.de> wrote:
On Sat, 15 Dec 2012 17:59:07 +0100
Holm Tiffe <holm at freibergnet.de> wrote:
NetBSD as such has dropped the VAX so far as I
know
No. Surely not!
Definitely not :)
There were a lot of VAX codegen issues with gcc 4.5, making the
NetBSD-current tree between most of netbsd-5 and netbsd-6 pretty
unusable for VAX, including the bootloader, but for netbsd-6 VAX was
switched back to gcc 4.1 and should definitely run. In fact on most
machines it should be better than netbsd-5 (some libm issues and
certain QBUS MSCP driver fixes come to mind).
and even the
last Releases aren't working anymore on a uVAXII vor example.
What brings you
to this assumption?
I think I had to
use some 5.x Release to get it booted from a DEQNA/DELQA at all.
That is the
problem. There is some strange bug in the (network) boot
loader. Maybe a GCC bug. Therefore netbooting has been broken for quite
some time. As nobody with the required skils cares it stays there.
After all NetBSD is a voluntary project. But you can use an older
version of the bootloader to boot release-6 or even -current.
The NetBSD 6 bootloader *should* be working fine, certainly the
failures I was seeing on earlier builds were no longer present.
If anyone is finding that bootloader or any other regressions in
netbsd-6, please let the port-vax at
netbsd.org list know (or just niggle
about it on here :)
Of course the biggest issue for newer versions is likely to be code
size, some form additional functionality but unfortunately too much
from more recent gcc :/
David