Thanks a lot. I've uploaded a patch, if you can verify it.
Ok I'll fetch the code and retest.
I've also patched another instruction that I think was wrong
timing :
P:ext JSR ext_addr
Maybe you can test it too with my newpatch to verify if it's OK
now ?
Yes no problem.
I have noticed a very small but repeatable deviation in all figures I tested in Hatari (a tiny fraction of a percent) - it may be related to MFP timing but there's a small chance it could be timing for the 'do' instruction - my test uses two nested do loops and there are 8000 iterations of the outer loop, so 8000 instances of 'do' setup time - maybe enough to cause this small difference. I will test too, just in case it's the reason.
D.