4

After soldering a header on my Pi Zero W I used pigpio's gpiotest utility to check for problems.

I was dismayed that this produced no output and crashed the Pi. After stepping through the tests I found what had caused the problem (operator error).

I was still surprised by the output:-

Testing...
Pull up on gpio 30 failed.
Skipped non-user gpios: 
Tested user gpios: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 23 24 25 26 27 28 29 30 31 
Failed user gpios: 30

I can find no mention of gpio 30 except P5 of the Pi B Model 2. gpiotest doesn't seem to have been updated since 2104, so it is not surprising it is testing the wrong pins.

My question is - does anyone know if gpio 30 is used on current models? I had some vague idea it was used, but can find no reference.

grep Revision /proc/cpuinfo reports 

Revision    : 9000c1
Milliways
  • 62,573
  • 32
  • 113
  • 225

1 Answers1

6

That is a fault in pigpio/gpiotest and/or the hardware revision returned by the Pi Zero W.

gpiotest relies on pigpio to tell it which GPIO are safe to write.

pigpio relies on the hardware revision returned by the Pi to assemble the list of safe to write GPIO.

If pigpio does not understand the hardware revision it defaults to granting access to GPIO 0-31.

I shall change the default to the 40 pin header in pigpio (GPIO 2-27).

In hindsight that is probably a better default now.

What revision does your Pi Zero W report?

grep Revision /proc/cpuinfo

As an aside if you start the pigpio daemon with

sudo pigpiod -x0x0000000FFFFFFC

gpiotest should run as intended (assuming they are the correct permissions for the Pi Zero W).

joan
  • 71,852
  • 5
  • 76
  • 108