Pio Update

[hoffman373]

Updated pio assembly program to:

• honor host interruptions while device is trying to send
• allow sending to occur without interrupts
• send flags bytes to indicate a send success as well as a send failure
• updated host send detection logic
• added comments
• had to swap clock and data pin order to ease reading from both pins Updated main application to:
• isolated PS/2 code into a pstransceiver file to share between keyboard and mouse
• treat multi-byte messages as single unit for resends

[hoffman373]

I have tested the keyboard support, but I have not tested the mouse. If you guys are interested, take a look at the branch and see how it works and what you think about it.

[No0ne]

Thanks for the PR, had a short look into it: changing the pins has to be done on the HV side since I used a fixed PCB for that, see https://raw.githubusercontent.com/No0ne/ps2x2pico/main/hw1.jpg I will try to fix the mouse support and merge it afterwards, thank you!

[hoffman373]

I am going to make a pcb board with ps/2 headers and the level shifter all on one pcb.

I've used osh park to make PCBs that are nice for hobby work.

On Sat, Jan 28, 2023, 12:45 PM No0ne @.***> wrote:

Thanks for the PR, had a short look into it: changing the pins has to be done on the HV side since I used a fixed PCB for that, see https://raw.githubusercontent.com/No0ne/ps2x2pico/main/hw1.jpg I will try to fix the mouse support and merge it afterwards, thank you!

— Reply to this email directly, view it on GitHub https://github.com/No0ne/ps2x2pico/pull/6#issuecomment-1407449209, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACSDYMQUFXOFVH5BZDQ7BJDWUVLK5ANCNFSM6AAAAAAUJYBGMY . You are receiving this because you authored the thread.Message ID: @.***>

[hoffman373]

I tried to find a way to not change the pin orders. The 32 instruction limit drove me nuts.

On Sat, Jan 28, 2023, 1:00 PM Dustin Hoffman @.***> wrote:

I am going to make a pcb board with ps/2 headers and the level shifter all on one pcb.

I've used osh park to make PCBs that are nice for hobby work.

On Sat, Jan 28, 2023, 12:45 PM No0ne @.***> wrote:

Thanks for the PR, had a short look into it: changing the pins has to be done on the HV side since I used a fixed PCB for that, see https://raw.githubusercontent.com/No0ne/ps2x2pico/main/hw1.jpg I will try to fix the mouse support and merge it afterwards, thank you!

— Reply to this email directly, view it on GitHub https://github.com/No0ne/ps2x2pico/pull/6#issuecomment-1407449209, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACSDYMQUFXOFVH5BZDQ7BJDWUVLK5ANCNFSM6AAAAAAUJYBGMY . You are receiving this because you authored the thread.Message ID: @.***>

[No0ne]

True, the instructions limit is hard, thats why I did the receive start interrupt. Changing pins is OK, just leave the LV pins connected the current GPIOs and change the HV label and the software pin.

[hoffman373]

OK I follow you. I changed the README to reflect the HV being swapped.

[No0ne]

Can you tell me at wich lines in the PIO programm you took advantage of the swapped data/clock pins?

[hoffman373]

Yes. Let me write you back tomorrow. Sorry for the delay.

[hoffman373]

OK, I it has been a few months since I have worked on this. But this is what I remember. On line 17 "wait 1 pin 1" waits for clock to go high by indexing off of the value set on line 81 with the C function call "sm_config_set_in_pins(&c, dat);"

On line 32 we read data from from the same pin set on line 81.

If we were to flip the order of the pins, then line 17 would become "wait 1 pin 0", which is totally fine, however on line 32 we would be reading from the clock pin instead of the data pin.

Without using wait the program needs more instructions in order to check the clock for high on line 17

[No0ne]

Thanks, understood!