search

mattdibi / redox-w-firmware

Firmware for nordic MCUs used in the Redox wireless keyboard
MIT License
77 stars 42 forks source link

BUGFIX: some keys could not be pressed at the same time #6

Closed rschrader closed 5 years ago

rschrader commented 5 years ago

This fixes an issue where some keys prevented other keys to be pressed when they are pressed. To fix this, the variable used to read the GPIOs is marked as volatile. This bug maybe only occurs with arm-none-eabi-gcc version 9.1.0.

rschrader commented 5 years ago

Not sure if you are interested in my sanity commits as well or just in the bugfix commit. Let me know what you think. The sanity commits just made it way easier to work in my setup and might be helpful for you as well.

Could do some more sanity stuff while working on the encryption topic if your are interested.

LucidityCrash commented 5 years ago

That might explain the issues I had when compiling on Ubuntu 19.04 ... the key rollover got really funky, both in 6kro and nkro ... I'll need to try this :) Like the change to the Makefile so you don't need to change the code to build one hand or the other.

mattdibi commented 5 years ago

Hi,

thank you for contributing. Great work!

I just tested your changes with the recent changes and everything seems to be in working order.

As I said in the comment I would ask you to update the readme with the instructions for the new scripts. Once merged I'll update the precompiled hexes to incorporate the two new PRs. :D

rschrader commented 5 years ago

Thanks!

Made the changes to the README. Also i added some location awareness to the scripts so that they can called from any location

rschrader commented 5 years ago

You should consider removing the precompiled binaries from the repository. Source repositories should not contain any binaries if possible ;). Github has a release feature for that.

mattdibi commented 5 years ago

Merged. Thank you!

Regarding the precompiled: I was thinking about that this morning. I started with the precompiled folder to follow the format used by the Mitosis repository but I think it's time to change. ;)

mattdibi commented 5 years ago

I was too eager to merge the PR and didn't spot the issue: the changes to the programming scripts broke the scripts. The HEX variable is empty and they don't write anything to the chips.

Trying to find a fix...

mattdibi commented 5 years ago

I'm sorry I reverted back the scripts to the position dependent version. I couldn't figure out how to evaluate the readlink command correctly. If you happen to fix this open a new PR.