[iyonix-support] Re: Whither !Thump?

0In message <4f24fbbb95smart1@xxxxxxxxxxxxxxxxx>
  Chris <smart1@xxxxxxxxxxxxxxxxx> wrote:

>In article <4367257.1190237026766.JavaMail.?@fh1004.dia.cp.net>,
>   c.n.l.f@xxxxxxxxxx <c.n.l.f@xxxxxxxxxx> wrote:
>> I've got 1.49 thanks. I thought that this wasn't the 32bit version
>> because it won't run on my A9. Thought I'd copied an elderly zip
>> from my Iyonix archive. Hence my search. Seems I've got the right
>> s/w but the wrong h/w! Cheers.
>
>The A9 list is the place for problems with that hardware. All I will
>say is that I have just tried Thump v. 1.49 on my A9, and it seems to
>work fine.

!Thump is 26bit, it does run on the Iyonix without Aemulor but that
does not imply that it is 32bit, it really is 26bit. It is a lot odder
than simply being 26bit. It was compiled with the 32bit flag set, it
calls these :-

$.!RunImage :
  SWI     XSharedCLibrary_LibInitAPCS_32
  SWI     XSharedCLibrary_LibInitModuleAPCS_32

Armalyser finds 171 26bit items of the less disastrous sort.

The AIF header is not 32bit flagged.

As I understand it gcc can do this unseen if one of the libraries is
26bit.

As far as the A9home goes, how successfully that runs will depend on the
compatibility settings. This aspect must be taken elsewhere.

A more relevant issue is to note a difference between OS5 and
Select/Adjust in that Select/Adjust is becoming more pro-active in
checking code standards to ensure stability. The Iyonix will only
explicitly reject a 26bit module.

None of this matters much as the author seems to have left the platform
without a forwarding address. I and others have tried to get the website
corrected. A recompile would be useful.

-- 
David Pitt.

Computing with RISC OS.
---
To alter your preferences or leave the group, 
visit http://www.freelists.org/list/iyonix-support
Other info via http://www.freelists.org/webpage/iyonix-support

Other related posts: