| Re: [hatari-devel] 2 little more questions |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/hatari-devel Archives
]
On 26/02/2012 19:09, Laurent Sallafranque wrote:
Hi,
OldCpu tests once more the address $fffffa0b.w than the newCPU.
This happens at the instruction 677 025 after the boot of hatari.
(So the first 677 024 first instructions are identical (the loops
included).
It's the first time that the 2 CPUs are doing different things.
(May it be only because of different CPU timings ?)
I think it's a cycle difference between both cpu. To be sure, it would
be better to have the traces with --trace cpu_all to see the cycle
timing of each instruction.
But it could also be a random behaviour of the MFP timer.
I haven't got the address error with the old CPU.
Any idea why it doesn't display illegal address with the old CPU ?
In hardware.txt, it's written: $FF8E09|byte |General purpose register -
does nothing |R/W (TT)
Seeing the cpu traces, something is clearly missing in the new cpu for
this range of IO registers. Even if they have no effect, they should not
give an address error.
Nicolas