🐛 Prevent MString infinite recursion

[thinkyhead]

Followup to #24390

• Add a single typename T template for MString::append to handle unspecified types (by appending '?' to the string). This removes the infinite recursion warning for the variadic template MString::append(T, ...) and may avoid actual infinite recursion if unhandled types are passed in MString::append(T, ...).

• Add a standard handler for explicit int8_t& so this type will be handled as a small integer instead of being handled as a char or deferring to the fallback typename T.

• Suppress warnings about deprecated sprintf on native platforms.

[thisiskeithb]

The bootlooping has stopped with this PR, but now temperatures are not reported correctly over serial:

OctoPrint log:

Send: M140 S70

Recv: //action:notification Bed Heating...
Recv: ok P31 B31
Recv:  T? /0.00 B? /70.00 @:0 B@:127
Recv:  T? /0.00 B? /70.00 @:0 B@:127

Send: M104 S210

Recv: //action:notification E1 Heating...
Recv: ok P31 B31
Recv:  T? /210.00 B? /70.00 @:127 B@:127
Recv:  T? /210.00 B? /70.00 @:127 B@:127
Recv:  T? /210.00 B? /70.00 @:127 B@:127
[…]

I also can't connect to the motherboard with a BTT TFT (connected over serial) and it always says "No printer attached!".

[ellensp]

No change on eeprom issue.

stock RAMPS with EEPROM_SETTINGS enabled (and the sim)

M500

echo:Settings Stored (617 bytes; crc 59424)
Error:EEPROM datasize error.
ok

[rondlh]

The bootlooping has stopped with this PR, but now temperatures are not reported correctly over serial:
OctoPrint log:

I also can't connect to the motherboard with a BTT TFT (connected over serial) and it always says "No printer attached!".

That is because the temperature reporting is completely broken/missing after #24390 and #26037. The BTT TFT uses the temperature reporting to confirm it's connected.

[thisiskeithb]

No change on eeprom issue.

I don't see an error on an SKR V3 (with onboard EEPROM):

>>> M502
SENDING:M502
//action:notification Hardcoded Default Settings Loaded
echo:Hardcoded Default Settings Loaded
ok P31 B31

>>> M500
SENDING:M500
echo:Settings Stored (746 bytes; crc 18007)
//action:notification Settings Stored
//action:notification Settings Stored
ok P31 B31

Also no error with an SKR V2 (emulated EEPROM via flash):

>>> M502
SENDING:M502
//action:notification Hardcoded Default Settings Loaded
echo:Hardcoded Default Settings Loaded

>>> M500
SENDING:M500
echo:Settings Stored (891 bytes; crc 43788)
//action:notification Settings Stored
//action:notification Settings Stored

[RV-from-be]

@thisiskeithb
By modifying temperature.cpp:

Marlin/Marlin/src/module/temperature.cpp

Line 4184 in 4985aca

#define SFP _MIN(SERIAL_FLOAT_PRECISION, 2)

to:

    #ifdef SERIAL_FLOAT_PRECISION
      #define SFP _MIN(SERIAL_FLOAT_PRECISION, 2)
    #else
      #define SFP 2
    #endif

and:

Marlin/Marlin/src/module/temperature.cpp

Line 4188 in 4985aca

s += TS(':', p_float_t(c, SFP));

to:

    s += F(":"); s += p_float_t(c, SFP);

This should resolve the temperature reporting issue and possibly the connection with the BTT TFT.

[thisiskeithb]

 s += F(":"); s += p_float_t(c, SFP);

This should resolve the temperature reporting issue and possibly the connection with the BTT TFT.

Indeed, that does fix the temperature reporting & BTT TFT connectivity issues.