Subject: Re: 1.5 Release documentation and Power-tec/PowerROM problems
To: Mike Pumford <mpumford@black-star.demon.co.uk>
From: Chris Gilbert <chris@buzzbee.freeserve.co.uk>
List: port-arm32
Date: 11/04/2000 14:36:55
On Sat, 04 Nov 2000, Mike Pumford wrote:
> > > Maybe a FAQ update indicating what to do if this occurs would be good.
> > > I think an initial hack would be to have a podule override which would
> > > allow an overide of the podule ID's when booting. As I already said a
> > > proper fix is to beef up the podulebus probe code so that it can tell
> > > these cards apart.
> >
> > You can pass in the boot option to force a card to be something, eg:
> > podule1=0x00000002
> > (or if I've got it the wrong way around:
> > podule1=0x00020000)
> > will force podule 1 to be an Acorn SCSI 1 card, IE it don't probe it.
>
> Wow never spotted that. Something for the FAQ perhaps?

Sounds like it :)  Don't know it if works, though, can anyone test?  (note 
that it's there in all kernels...)

> > Yep and builds are painful on RPC's.  Perhaps if I can twist my kintic
> > into doing some things it might speed up (using a RAPIDE card has speeded
> > lots of stuff up, makes CVS only take about 30 or so minutes on the main
> > src tree, perhaps I should time it)
>
> 1 day is about how long my Simtec IDE equipped RPC takes to  build the
> entire NetBSD world. The rather neat thing about this is that this build
> time has remained much the same even though the full source tree has grown.
> The only truly scary thing is that this is the amount of time that mozilla
> takes to compile as well.

I've not attempted mozilla (I've the source around here somewhere...)  as 
I've not done a recent world, so I've not got the ld.so tweak needed for it.

I have found something scary in the cpuswitch.S, and that's that it uses ffs 
to find the first queue with work, so for the most part if there's only user 
process running it was looping 12 time (eeks the pipeline!!!)  to find the 
running process that's at the head, I know not a major thing, but I suspect 
that cpu_switch is hit regularly?  Anyway I'm compiling up a kernel with this 
tweaked (I finnally got around to digging out the ffs stuff of the mailing 
list from 2 years back, the ffs version that's fixed time, does a mystical 
multiply and then lookup from table.)

Anyway once it's working it'll eventually make it into the tree I hope :)

Chris