On Mon, May 02, 2005 at 10:33:23AM +0200, Jochen Kunz wrote:
> On Sun, 1 May 2005 14:02:18 -0700
> Chuck Silvers <chuq@chuq.com> wrote:
> 
> > the panic triggered by booting from disk that was there before didn't
> > happen for me today,
> Great news! :-)

well, apparently it's not that great...  init gets an endless stream of
SIGSEGV.  I thought this was due to a busted userland, but the snapshot
I just built has the same problem.  the same kernel works when netbooted.
bizarre.  I'll try applying my previous hack and see if that still has
any effect.


> > some things still to be done include:
> > 
> >  - HIL support
> > 	do the MI/MD split of the hp300 HIL driver, or use openbsd's.
> I had a look at this some time ago. It looks like quite some work to
> make our HIL code MI. The OpenBSD code is already MI and largely
> rewitten. An import was not that hard and it worked for hp700.
> Unfortunately I never came around to convert hp300 to use this code.
> I'll take care of this again...

great, thanks.


> >  - PCX-T (PA-7200)
> > 	these machines hang as soon as virtual addressing is turned on.
> I am clueless in the VM area, all I can do is testing on a C100, J200
> and K100. (All single CPU.)

yea, I figured I would take care of that one.


> >  - EISA
> > 	does this work already?  haven't tried it.
> To be honest: Does EISA interrest? ;-)
> There is code for one of the two EISA bridges found in hp700 machines.
> IIRC it was the older chip found in some PA7000 machines.

this seems less important, but I'd think we'd like to support it eventually.
certainly other things like sysinst are more important.


> > also, we'd like to get a better idea of what hardware support is
> > actually working now.  could people try out whatever PCI cards
> > they've got an post the results?  we'd like to get an idea of
> > how many drivers need to be tweaked to work on hp700.
> I tried (e)siop(4) and tlp(4) cards with success, even some quad tlp(4)
> cards with a PCI-PCI bridge worked. An ahc(4) made problems when used in
> PCI-IO mode, IIRC it worked when the ahc(4) driver was tweaked to use
> PCI-MEM. fpa(4) did not work, but this is due to funky things fpa(4)
> does with bus_dma(9). I can try to get some more cards for testing.

I started a list at
	ftp://ftp.netbsd.org/pub/NetBSD/arch/hp700/driver-status

everyone should feel free to update that file as things are tested.

-Chuck