--- Richard Earnshaw <rearnsha@netbsd.org> wrote:

> On Tue, 2005-01-04 at 08:21, Kamal R. Prasad wrote:
> > Hello,
> > 
> >  I found out that as soon as tlb is invalidated,
> the
> > system goes for a toss. Does the configuration
> below
> > look OK? The kernel is loaded at 0xa0200000 and
> the L1
> > page table is at 0xa01fc000.
> > Other than an incorrect base addr for the page
> table,
> > what else could cause the system to go for a toss?
> Why
> > wouldn't the system generate an exception in such
> an
> > event? Appreciate any answers.
> 
> A good way of killing your machine stone dead is to
> use vectors in page
> 0 and then forget to map the vector page in (or to
> unmap it at a later
> date).  Then on your first interrupt or page fault
> you end up in a
> deadly cycle of infinite page faults trying to
> access address zero.  The

I believe the system hangs only after the TLB is
invalidated. So it isn't a problem of interrupts
(AFAIK). I notice that the lubbock code in netbsd sets
the TLB base addr as 0xa01fc000, but isn't TLB base
addr something that points to the hardware cache?

> same applies with high vectors as well, but that's
> much less likely to
> be a problem once you've got the system up and
> running because the high
> address isn't part of 'user space'.
> 
we aren't in user-space yet. The hang occurs after
kernel is loaded, but before we can switch to the
virtual memory addressing mode.

regards
-kamal

> R.
> 



		
__________________________________ 
Do you Yahoo!? 
All your favorites on one personal page – Try My Yahoo!
http://my.yahoo.com