Si Owen wrote:
>I might have to look more closely to see what the LineCycleCounter value is
>throughout the loop, but it seems like a strange coincidence that the raw
>values give the correct 0xc0c0 result! Are the instruction times always
>going to be 4 t-state rounded?
Not quite that simple :) INs and OUTs are a bit awkward. First of all, INs
and OUTs take differing numbers of t-states depending on whether the port is
an internal SAM port (HMPR, etc) or an external one (Quazar etc).
For internal ports the number of t-states used depends on when the
instruction occurs. IN A,(n) and OUT (n),A takes 12 *OR* 16 t-states and IN
r,(C) and OUT (C),r takes 16 *OR* 20 t-states. The number of t-states since
some fixed point (beginning of the frame?) after the instruction seems to be
rounded up to the nearest multiple of 8. For example the timings for the
following code would be:
12/16 IN A,(STAT) ;Depends on where we've started
from!
16 IN A,(STAT)
4 NOP
12 IN A,(STAT)
16 IN A,(C)
4 NOP
20 IN A,(C)
All other versions of IN and OUT also vary by 4 t-states in the same way.
*However* :) For external ports this rounding to the nearest 8 since some
fixed point doesn't seem to occur. Also, IIRC, IN A,(n) and IN r,(C) both
take the same time for external ports (12 t-states).
Well, that should keep you going for a while :)
David Laundon.
