Re: BBR/BBS 65C02 instruction cycle counts

From: smf <smf_at_null.net>
Date: Wed, 29 Jun 2016 21:09:26 +0100
Message-ID: <5e1c18a4-06b1-6734-c79f-33ee3a729a03@null.net>
It's only going to be 6 cycles, if crossing a page takes an extra cycle. 
The 6502 branch instructions do, but that is likely for backwards 
compatibility.

It's much more likely to always be 5. Rockwell say it's 5 on one page 
and imply it's 5/6/7 on another, so one page is definitely wrong as they 
disagree & the 5/6/7 doesn't make sense. If you go for 4/5/6 then you're 
assuming that both pages are wrong. Which is possible, but much less 
likely than them putting an ** on one page when it wasn't meant to be there.

Unless you get hold of some hardware to verify it, or some software that 
is sensitive to timing that has been developed on real hardware then 
it's actually not that important.

More important is the 65ce02, where the instructions take less clock 
cycles as it's the basis for the commodore 65. I don't think commodore 
used 65c02 in anything.

On 29/06/2016 12:59, Kajtár Zsolt wrote:
> the page crossing branch should be at least 6 cycles. I vote for 
> 4/5/6, unless
> there's an extra dummy cycle stuffed in for which I don't see a reason yet.
>


       Message was sent through the cbm-hackers mailing list
Received on 2016-06-29 21:00:02

Archive generated by hypermail 2.2.0.