On Sun, Mar 11, 2007 at 02:16:16PM +0100, Johnny Billquist wrote:
> The vax port is non-working since a while ago. I know that 4.99.8 is 
> working, but somewhere after that something broke.
> It hangs after saying
> "scsibus0: waiting 2 seconds for devices to settle..."
> 
> This on a 4000/90. Now, looking at some code, this appears to be some MI 
> code, which is called from the asc driver in the VAX. The same MI code 
> is used by a bunch of other platforms as well, such as the Amiga, pmax, 
> Sparc and some others (but not i386, Alpha and other "major" platforms).
> 
> So I'm curious if others with the asc driver also experiences a hang 
> now, or if it is VAX specific...?

I did a little research into this, but I still have no real idea what
exactly is causing this.

The scsibus code goes to sleep for 2 seconds, and then the system waits
in the idle loop forever. I set a breakpoint on Xhardclock, it is hit
once before scsibus_config() is called but then never again. I suspect
that there is a IPL problem, or someone clobbered the SCB vector. But
the latter seems unlikely. I still have to look into this, but I'm
currently unsure how to check the IPL from the debugger.


-- 
%SYSTEM-F-ANARCHISM, The operating system has been overthrown