Enhancements to M43 pins debugging

[thinkyhead]

Cleanup and optimization of #5112

• Add endstop monitoring option with E parameter
• Report PWM status of pins
• Give an extended pin report when not watching with W

[Blue-Marlin]

The pins_arduino.h files provide this info in NUM_DIGITAL_PINS and NUM_ANALOG_INPUTS

[Blue-Marlin]

endstop_monitor_flag could be checked here and would save the call of endstop_monitor() if not active.

[thinkyhead]

Thanks for the tip. I totally forgot to review that part of the code.

[Bob-the-Kuhn]

Much cleaner code

I've got a couple of bug fixes & a couple of minor changes.

All the changes are included in the attached file.
pinsDebug .zip

I couldn't get the #define PINSET(P) (defined(P) && P > -1) from the previous review to work on a pin that existed. Really strange since it uses the same method as the PIN_EXISTS macro.

When determining if a pin is connected to a PWM, the TCCRxA register is always used even if looking at TIMERxC

Lines 677-680 need to change

#define PWM_CASE(N) \
  case TIMER##N: \
    if (TCCR##N & (_BV(COM## N ##1) | _BV(COM## N ##0))) { \
      PWM_PRINT(OCR##N); \

The solution I came up with is a two part change:
Lines 677-680 now read

#define PWM_CASE(N,Z) \
  case TIMER##N##Z: \
    if (TCCR##N ##A & (_BV(COM## N ## Z ##1) | _BV(COM## N ## Z ##0))) { \
      PWM_PRINT(OCR##N ## Z); \

which means 694-724 change from
PWM_CASE(0A);
to
PWM_CASE(0,A);

In the WGM_MAKE macros the TEST function only returns 1 or 0 so the following pieces in lines 734-735 need to change from

TEST(TCCR##N##B, WGM##N##2) >> 1
TEST(TCCR##N##B, WGM##N##3) >> 1

to either

TEST(TCCR##N##B, WGM##N##2) << 2
TEST(TCCR##N##B, WGM##N##3) << 3

or

TCCR##N##B & _BV(WGM##N##2)) >> 1
TCCR##N##B & _BV(WGM##N##3)) >> 1

I prefer the latter because it's one less operation to perform and it's consistent with how the macros treat the other register.

A couple of minor changes:

• buffer[30] changed to buffer[40] per review in previous PR
• added name length to comment on line 75
  // Pin list updated from 7 OCT RCBugfix branch - max length of pin name is 24

[thinkyhead]

Really strange since it uses the same method as the PIN_EXISTS macro.

Almost. When PIN_EXISTS tries to evaluate an undefined symbol the result is:

defined(_PIN) && _PIN >= 0

…but with the proposed PINSET it ends up being:

defined() && >= 0

…which is nonsense. I hit this before, but I forgot.

[thinkyhead]

Ah, you're right about my mis-use of TEST. I forgot it simply evaluates as 1 or 0, and not as the bit.

[thinkyhead]

buffer[30] changed to buffer[40] per review in previous PR

I think 30 is fine. I forgot that justified strings already include the string length.