#2159 POT sampling is not emulated correctly

[gpz]

The POT values are currently "too stable", not only should they jitter a little bit (at least bit 0), but the values read should also produce significantly wrong (too large) values, when the sampling happens close after switch the joystick port via $dc00. This can be observed eg when reading the values without disabling the system irq (which uses $dc00 for the keyboard scanner)

[gpz]

added a test program: https://sourceforge.net/p/vice-emu/code/HEAD/tree/testprogs/SID/paddlescope/

also partial fix (lsb jitter) in r45726

[gpz]

r45727 should fix the different jitter of paddles vs 1351

[gpz]

https://sourceforge.net/p/vice-emu/code/34291/ is a similar program for the vic20, totally forgot about this

https://www.youtube.com/watch?v=0-UvoEhAEIs shows it on the real thing. Note the really odd way the sample error changes >_<

[gpz]

r45729 adds the glitchy values when switching control ports. not really accurate at all, but should do the trick for all practical purposes

[gpz]

r45731 adds dependency on display dma on vic20

i consider this done for now - for more accurate emulation we need better/more detailed tests