Subject: Re: CVSup collections for a NetBSD CVS tree
To: Jonathan Stone <jonathan@DSG.Stanford.EDU>
From: Dave <email@example.com>
Date: 04/30/1999 09:31:37
Jonathan Stone wrote:
> >Perhaps you should try the arm32 port on a processor other than the stongarm.
> >Baldness beckons ;~)
> What kind of ARM processor? Archimedes or Itsys arent exactly easy to
> come by over here. I've tried to be helpful whenever the ARM gurus ask
> about non-coherent I/D cache issues (some mips3 machines have similar,
> but far less depilating features), but its hard to do much without
> hands on the hardware.
Well, I have two Acorn RiscPC 600's. One with a strongarm, which runs up-to-date
-current with no problems(well done the team!), but I also have one with it's
original arm610 which hasn't been able to boot a -current kernel since October,
and I had to do some kludging to make one work then. The kludges no longer work.
I believe Mark Brinicombe is working on this, but I don't know if any progress has
been made. I have pinned down the problem to a call to setttb() during the
secondary bootstrap. There is no return from this call and, since it's arm
assembler(cpufuncs.S IIRC) which might as well be martian to me, I can't really
pin it down any closer than that.
The first call to setttb() during the primary bootstrap returns fine.
Make it idiot-proof and someone will make a better idiot!