Re: status

[Jason Thorpe <thorpej%nas.nasa.gov@localhost>]

[ Sigh, yes, this is long-overdue... if only for more hours in a day... ]

On Fri, 18 Apr 1997 13:06:36 -0700 
 "Erik E. Fair" (Time Keeper) <fair%cesium.clock.org@localhost> wrote:

 > Some kind of status for the PowerPC port would be nice to see, given the
 > mega-commit that just hit the tree...

Ok ... here goes:

Some time ago, you may recall that I was planning on releasing a
cross-compiled snapshot of NetBSD/powerpc, plus toolchain diffs.
Unfortunately, that was delayed; I went to boot the snapshot floppy
image, and lo and behold, init wouldn't start ... Around the time this
happened, by Real Job(tm) schedule began to get ... well, much more
packed.  There wasn't a lot of time for scouring locore.S and pmap.c
looking for the bugs...

So, I enlisted the help of Charles Hannum ... and with the help of PSIM,
found and squished the major problems in the code.  I wish this wouldn't
have taken as long as it did, but we did manage to find them, at lesat.

In any case, it's now booting and running quite well under both PSIM
and on a Firepower LX (Charles it currently trying to get it to run
on a Power Macintosh 7300/200 - The Apple OpenFirmware implementation
is _not_ helping :-/ ).

I am planning on compiling a snapshot - however, I'm currently fighting
against the toolchain ... kernels linked with an early-March gas/binutils
snapshot don't run, due to a bug in the linker which cases the data
segment to be mapped incorrectly.  I'm currently updating to a new
(just a few days old) gas/binutils snapshot, hoping that the problem
will be magically Fixed ... if not, I'll be sending a note to Ian Taylor
about the problem.

As soon as I work out the toolchain problems, I plan on:

        - building a binary snapshot with as many executables
          as I can reasonably cross-compile (I can't easily
          build natively on the Firepower due to OpenFirmware
          not speaking to the ethernet interface on this
          particular model).

        - building two ISO 9660 floppy images, one with the boot
          program and kernel, the other with a small binary set
          to help get folks installed.

        - getting out a snapshot of the toolchain.

The next step for me is to start working on the autoconfiguration framework
for this port ... I have some ideas jotted down on this subject, but
haven't fleshed them all out yet... Once the autoconfiguration framework
is in place, I will be doing the glue for native-mode PCI on the Firepower.
This is going to depend on the "bus dma" work I've also been doing (and
making steady progress on ... geez, there are a lot of details to worry
about...)

That's status in a nutshell...

Jason R. Thorpe                                       
thorpej%nas.nasa.gov@localhost
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939