|
|
Post by wsoft on May 11, 2024 18:48:46 GMT
The dump says it all. Look how much memory is wasted in order to access a single byte of memory.
|
|
|
|
Post by c128old on May 11, 2024 19:18:40 GMT
True, a lot of code for a poke or peek of 5 locations.
However things aren’t nearly as bad as they may seem.
Consider the use-case: the VIC bitmap can live under ff00-ff04, so a little work may pay off.
Another use-case: buffer of 256 bytes
Both options are reasonable when running z80 in bank1 because there you have no hard relying on z-page, no cpu-io-port in the way and stack typically elsewhere.
I use ff00-ff80 to store a z-page copy in a sort-of freezer code in the cpm-bios
|
|
|
|
Post by wsoft on May 11, 2024 19:36:19 GMT
Yes you can move the zero page to wherever you want I read about that. How useful that might be when programming the VIC-II- I dunno. There was only one option when selecting which bank could be used by the VIC but hey, please continue. You've sparked my interest but when it comes to the Z-80 it goes ZIP right over  . Let us also remember that the Z-80 posesses no special attributes that connect it directly to the VIC. In fact, in CP/M most routines that deal with C128 hardware are controlled by the BIOS, which uses the 85xx. The premise of your claim would assume using actual Z-80 code together with the MMU itself and that is what's so interesting about your post because only a few have gone there. |
|
. Let us also remember that the Z-80 posesses no special attributes that connect it directly to the VIC. In fact, in CP/M most routines that deal with C128 hardware are controlled by the BIOS, which uses the 85xx. The premise of your claim would assume using actual Z-80 code together with the MMU itself and that is what's so interesting about your post because only a few have gone there.