Page 4 of 4
Re: Shifter LOAD behaviour
Posted: 28 May 2025 02:09
by Smonson
I'm not sure if it would be of use, but the verilog shifter model that I based on ijor's schematics was split out into its own repo here:
https://github.com/smonson78/st-shifter-verilog
Re: Shifter LOAD behaviour
Posted: 28 May 2025 10:02
by Badwolf
Smonson wrote: 28 May 2025 02:09
I'm not sure if it would be of use
Thanks Smonson! It probably doesn't have a direct application but it's certainly useful!
Since I'm trying to read the output of that and back-infer what it's doing to have a blow-by-blow breakdown of what it's doing is most helpful indeed.
I can also perhaps copy some of the nomenclature and the style for consistency.
Cheers,
BW
Re: Shifter LOAD behaviour
Posted: 28 May 2025 11:29
by Smonson
Badwolf wrote: 28 May 2025 10:02
Since I'm trying to read the output of that and back-infer what it's doing to have a blow-by-blow breakdown of what it's doing is most helpful indeed.
I would probably recommend that you run the simulations and then look at the waveforms in GTKWave. That allows you to see not only what's going on with the input & output pins, but also shows what's in the internal shifter registers. It's also super easy to edit the test file to create any extra scenarios that you had in mind. No need to infer!
Re: Shifter LOAD behaviour
Posted: 28 May 2025 11:41
by Badwolf
Smonson wrote: 28 May 2025 11:29
I would probably recommend that you run the simulations and then look at the waveforms in GTKWave.
Thanks, yes, I'll do that.
No need to infer!
I only meant what my CPLD is doing. It's going: "aha, this is the seventeenth LOAD instruction since VBLANK, so this must be screen address offset 32".
Cheers,
BW
Re: Shifter LOAD behaviour
Posted: 02 Jun 2025 04:56
by ijor
Smonson wrote: 28 May 2025 11:29
I would probably recommend that you run the simulations and then look at the waveforms in GTKWave.
+1. A very wise advice.
Re: Shifter LOAD behaviour
Posted: 10 Jul 2025 12:16
by Badwolf
Just a quick follow up to this as I haven't had time to work on this at all for some time but I had a head-slap moment the other day.
The sampling works perfectly. I was talking nonsense.
Actually, that's not strictly true. The sampling was all over the shop, but only when I tried to show the results!
The problem? When trying to see the output I'm isolating the Glue's VSYNC and feeding in my own one! I was resetting my counter on the original VSYNC but MMU was seeing this new generated VSYNC. Utter confusion!
So, um... Whoops!
BW