Big Fix: Array bounds crash in key handling

[Stuckya]

I was encountering intermittent crashes when my kids were using the remote. This led me to debugging the crashes and I uncovered 3 issues along the way. This was the 1st issue.

Problem:

LoadProhibited crash with EXCVADDR: 0x00000009 when pressing KEY_CHUP repeatedly, caused by array bounds violations in key handling code.

Root Cause:

• Incorrect array indexing in doShortPress() : lastTimeSent[keyCode/keypadROWS][keyCode%keypadROWS].
• Wrong keyCode calculation in keypad_processKeyStates(): row * keypadROWS + col instead of row * keypadCOLS + col.

Proposed Fix:

1. Corrected array indexing to use proper row/col derivation.
2. Fixed keyCode calculation.
3. Added bounds checking with helper functions to improve readability isKeyPressInBounds() and isKeyPressRateLimited().
4. Refactored nested conditionals with early returns, defensive programming.

[KlausMu]

Even if keypadROWS and keypadCOLS are mixed up at some place, did this really lead to a crash? Both keypadROWS and keypadCOLS are set to 5.

[Stuckya]

Even if keypadROWS and keypadCOLS are mixed up at some place, did this really lead to a crash? Both keypadROWS and keypadCOLS are set to 5.

Good point and I think you are correct, since both keypadROWS and keypadCOLS are 5, the math would work out the same either way, even if they got mixed up.

Regardless, I think this cleans the code up and prevents future regressions.

I suspect #114 was the real fix given EXCVADDR: 0x00000009 was the cause when debugging. I applied both fixes at the same time on my branch and only separated them for PR purposes. I can test these fixes independently later today.

[Stuckya]

Confirmed, #114 was the real fix and this should just be considered cleanup (improved readability) & hardening.