EEPROM.length() should be static constexpr

[awgrover]

Couldn't EEPROM.length() be static constexpr? I wanted to use .length() in a class static, like:

class EEPROM_User {
  // works even if actual eeprom < 3k, i.e. wrap-around
  constexpr static int ADDR = (3169) %  EEPROM.length();
}

In Arduino IDE 1.8.12, ./hardware/arduino/avr/libraries/EEPROM/src/EEPROM.h:

uint16_t length()                    { return E2END + 1; }

And E2END is compile-time static.

I propose:

static constexpr uint16_t length()                    { return E2END + 1; }

Patch:

--- a/hardware/arduino/avr/libraries/EEPROM/src/EEPROM.h    2020-06-21 17:46:29.124231608 -0400
+++ b/hardware/arduino/avr/libraries/EEPROM/src/EEPROM.h      2020-06-21 17:56:01.608338869 -0400
@@ -124,7 +124,7 @@
     //STL and C++11 iteration capability.
     EEPtr begin()                        { return 0x00; }
     EEPtr end()                          { return length(); } //Standards requires this to be the item after the last valid entry. The returned pointer is invalid.
-    uint16_t length()                    { return E2END + 1; }
+    static constexpr uint16_t length()   { return E2END + 1; }
     
     //Functionality to 'get' and 'put' objects to and from EEPROM.
     template< typename T > T &get( int idx, T &t ){

[matthijskooijman]

Sounds good. Any idea what that means for compatibility? Can you still call EEPROM.length when you make it static, or would all code need to be changed to call EEPROMClass::length() instead?

Also, this should probably be reported at https://github.com/arduino/ArduinoCore-avr instead (or maybe this issue can be moved by the admins?).

For reference, here's the source: https://github.com/arduino/ArduinoCore-avr/blob/master/libraries/EEPROM/src/EEPROM.h

[SpenceKonde]

I have checked these changes into megaTinyCore; should work fine everywhere. Doesn't impede access like @matthijskooijman feared.