Re: [hatari-devel] DSP bug: need more explanations

[Laurent Sallafranque <laurent.sallafranque@xxxxxxx>]

Yes, it seems to make sense now, your explanation is cristal clear for me now ;)
Just have to see how to implement this in the actual code ;)

Best regards

Laurent



Le 04/01/2015 23:53, Miro Kropáček a écrit :

On Sun, Jan 4, 2015 at 11:48 PM, Laurent Sallafranque <laurent.sallafranque@xxxxxxx> wrote:

Real falcon
X1 = $10

OK, so it does work, then! If the behavior has changed with movem -> move change, that implies that:

- movem indeed somehow slowed down the AGU and it works as it should (i.e. no pipeline effect)

- when the instruction is quick enough, the pipeline effect takes place and therefore it works exactly as I described in the previous email (see the final values in r1 are always half the number of iteration)

--

MiKRO / Mystic Bytes
http://mikro.atari.org