It turns out that I did actually have an early version of the code posted on github. I've updated it to include the latest changes. I've created a couple of different branches to reflect different versions of the code as the features have been added. Unfortunately, the different branch versions have different file names (file name includes version number), so it makes tracking changes difficult. I've hopefully fixed that with the main branch, which contains the latest code that can be fitted. This should almost match the most recent files I posted up-thread (I found an error in the code where 'assign GenBankSel[0] = (rD == 0'h0) && BankDecode;' should have been 'assign GenBankSel[0] = (rD == 4'h0) && BankDecode;'. However, that change didn't seem to impact the fitting at all. Phew):Can you post your latest verilog again?
(I assume this is not anywhere in github...)
https://github.com/kgl2001/IntegraBv2-C ... 5144XL-100
This main branch still uses !nRDS instead of RnW in the PALPROM code, and I'm still using the !BBC_nRST term in the pp2a & pp2b PALPROM code (the 32k PALPROMs at banks 8 & 9).
Yes, this can be very tedious, and (in my case) there's a lot of trial and error involved.Sometimes when multi-level optimization is used designs expand in unexpected ways and you can often do a better job manually factoring out common sub-expressions. But it does get a bit tedious.
I didn't think there was, but I noticed that there was an opportunity for me to swap out one of my pp2 (32K PALPROM) types for an extra pp4 (64K PALPROM) type. However, this involved modifying some other code to free up some jumpers for selecting different switching zones for the new pp4 PALPROM. Unfortunately, I can't get this code to fit, no matter what I try. This code is available at this branch:Is there more you would like to add if you had the room?
https://github.com/kgl2001/IntegraBv2-C ... 5144XL-100
If anyone has any appetite to try and make this fit, that would be awesome. However, I suspect it's going to be difficult, and at the end of the day, if it's not possible, I won't be hugely disappointed. I'm still very happy that I've got a working version (albeit with 2 x 32k PALPROMs instead of 2 x 64k PALPROMS) in the main branch.
Statistics: Posted by KenLowe — Sat Aug 17, 2024 5:09 pm