[TI ASM] Speed guidelines

[King Harold]

I wrote a 66 byte multiplication routine (yes, unrolled, it could be very small but it'd be slower) to multiply a by c and return the result in hl, it uses a,bc and hl.
The speed statistics are: worst case (a = 254): [309] best case (a = 1): [266]
Is that anywhere near good enough or is 266 - 309 T-states slow? (its including a ret, which takes 10 T-states, but not including the call, the pure routine would take 256 to 299 T-states)

[CoBB]

I’d say this is quite reasonable speed for such a routine.

[King Harold]

Ok, good, thanx

And how about U16 devide by U8 in an avarage of 740 T-states? is that ok?