Subject: Re: 'program cc1 got fatal signal 11'
To: Neil A. Carson <neil@causality.com>
From: William Gallafent <William.Gallafent@comlab.ox.ac.uk>
List: port-arm32
Date: 04/27/1998 13:06:27
On Mon, 27 Apr 1998, Neil A. Carson wrote:

> William Gallafent wrote:
[that cc1 won't work ...]
> > Thinks. Was I root for that compile, and not for this one? Hmmm. I'll send
> > this anyway (forgive me if I'm right about there just being an
> > ownership/access problem somewhere...) and go and try again.
> 
> That shouldn't matter. Are you using the 1.3.1 compiler set? If so, this
> won't run properly on the bugged StrongARMs that Acorn shipped (wuthout
> telling anyone).

That'll be the problem then :( Revision K, hence the 'bugged STM^
instruction' or whatever it is message at startup :( If only I'd known
about this before the guarantee ran out I might have sent it back saying
'this chip is faulty' and demanding a newer one. Too late now though, I
suspect.

> It's not a problem with the code. If you're using SA and 1.3.1, either
> upgrade to a more modern SA, put the 710 back in, or wait for another
> compiler build that may move round the problem instructions to
> elsewhere. 

It's going to have to be the last of these I think ..... a more modern SA
(on a board from Acorn at least) is rather too expensive (getting towards
the cost of a CATS system!), and I sold the 710 :( So, will Chaltech offer
an upgrade for revision J,K SA owners whereby they remove the old chip
(for use as a keyfob, obviously) and put a shiny new unbugged one in
(They must have lots of them lying about to make CATSs with ... ?). Does
the change to the board amount to anything more than replacing the chip,
or are there changes elsewhere?

> If this isn't the problem, then ignore this posting :-)

Alas, it is. So now I'm still looking for a solution. Thanks for the
definitive answer!

Cheers,

Bill