cr1901 / at2xt Goto Github PK

Accesses to the fields of KEYBOARD_PINS do not get optimized to a constant in certain scenarios, which results in a cascade of bitmask operations which should operate on constant values instead doing an array access to retrieve a constant value.

• Discussion (starts at linked message): https://bot.tockos.org/tockbot-moz/embedded-bot/2018-02-21/?msg=63351&page=1
• Comparison of LLVM IR: https://gist.github.com/cr1901/459a5c5d1032aac2d16b9cf62b1c959e
• Workaround: cd9731e

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

This repository currently has no open or pending branches.

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

Current AT2XT PAUSE/BREAK behavior (one or more keystrokes) on IBM 5150 is to eat the next keystroke without displaying anything. Subsequent keystrokes display as normal. Changes in lock key states are honored if such a key follows PAUSE/BREAK. However, there seem to be a few edge cases:

• CTRL+BREAK triggers scroll lock LED.
• CTRL+SCROLL also triggers scroll lock LED- supposedly the 5150 Model F keyboard version of CTRL+BREAK or BREAK alone?
• CTRL+NUM emulates pressing the PAUSE/BREAK key, but also triggers num lock LED.

Check what PAUSE, CTRL+BREAK, and BREAK alone are supposed to do. Also make sure the self.expecting_pause optimization didn't introduce this.