Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!mcsun!ukc!acorn!john
From: john@acorn.co.uk (John Bowler)
Newsgroups: comp.sys.acorn
Subject: Re: 32bit immediate load in ARM code
Summary: Use of NV condition code
Keywords: NV
Message-ID: <7182@acorn.co.uk>
Date: 20 May 91 18:25:48 GMT
References: <+|Q_L||@warwick.ac.uk> <1991May16.200200.24636@comp.vuw.ac.nz>
Organization: Acorn Computers Ltd, Cambridge, UK
Lines: 43

In article <1991May16.200200.24636@comp.vuw.ac.nz> gavin@comp.vuw.ac.nz (Gavin Flower) writes:
>
>Have read in several places,
> that the condition code NV
>is *NOT* forward compatible.
>
>Specifically that Acorn have
>reserved the right to reuse
>it for something else, at
>some future stage!

I was careful not to say this, only to point out that it is a very obvious
thing to want to reuse, however Alasdair Thomas's messages seems unambiguous
to me.

>To the best of my knowledge
>it is still ok with ARM3.

>****could someone at Acorn
>confirm this, also could
>they suggest what it might 
>be earmarked for?

NV works as specified on ARM2 and ARM3!  Whether and how NV is (re)used is
up to ARM Ltd (*not* Acorn Computers Ltd, although I would hope Acorn has
some influence :-).  The body of Alasdair's message seems to correspond to
the rules which we attempt to follow when generating code which will work
on all processors.  Certainly I (personally) would suggest that you attempt
to follow those rules both within hand-written ARM assembler and within
compiler generated code.

In article <7~R_RA`@warwick.ac.uk> csuwr@warwick.ac.uk (Derek Hunter) writes:
> Is there as yet any notes on a correction to the aof deficiency?

The deficiency would require extensions to AOF to allow use of a PC relative
symbol value (currently only allowed in the special context of B/BL
relocations if I remember correctly) and the insertion of bits derived from
such a value into bit fields in a word.  I don't see this being fixed on
its own - I would anticpate a more widespread review of AOF.  Obviously
since I am commenting on this I don't know about any current developments
in this area :-).

John Bowler (jbowler@acorn.co.uk)