RF24/RF24.h + wiringPi.h = Segmentation Fault

[alandsidel]

Using wiringPi v2.60 from the "unofficial" github repo, and RF24 1.36 from here on a CanaKit RPi 3b+ running Raspbian 10. I've been successfully using the RF24 library to communicate between this RPi and several Arduino's, using C/C++ on both ends.

I want to add GPIO capability to the RPi alongside the RF24 support, so I added <wiringPi.h> before <RF24/RF24.h> in the RPi project and added some code, which resulted in a Segmentation Fault.

The call to wiringPiSetup() succeeds, but the following call to pinMode(25, INPUT); causes the SIGSEV when uncommented.

The gdb backtrace says the issue is happening in RF24 which seems odd, I would expect it to be in WiringPi. Is RF24 overriding the pinMode() call?

(gdb) bt
#0  0x76f1fb58 in bcm2835_peri_read () from /usr/local/lib/librf24.so.1
#1  0x76f1fc38 in bcm2835_peri_set_bits () from /usr/local/lib/librf24.so.1
#2  0x000133dc in main (argc=<optimized out>, argv=<optimized out>) at ./src/server.cpp:80

server.cpp:80 is the aforementioned line, pinMode(25, INPUT);

Is there an alternative way to get WiringPi playing nicely with RF24, or if not, is there a recommended GPIO library to use alongside RF24?

[TMRh20]

Are you sure you compiled correctly with wiringPi? The bcm2835_peri_read () in output would indicate you are using the bcm2835 driver.

For wiringPi you want: ./configure —driver=wiringPi make sudo make install

Then plz make sure you are using the correct pin #s for wiring Pi.

I haven’t tested it myself but don’t see why it would be a problem unless you are trying to use a pin that RF24 is using. Have you tested different pins and tested without RF24?

On Aug 2, 2020, at 7:36 AM, Allen Landsidel notifications@github.com wrote:

 Using wiringPi v2.60 from the "unofficial" github repo, and RF24 1.36 from here on a CanaKit RPi 3b+ running Raspbian 10. I've been successfully using the RF24 library to communicate between this RPi and several Arduino's, using C/C++ on both ends.

I want to add GPIO capability to the RPi alongside the RF24 support, so I added before <RF24/RF24.h> in the RPi project and added some code, which resulted in a Segmentation Fault.

The call to wiringPiSetup() succeeds, but the following call to pinMode(25, INPUT); causes the SIGSEV when uncommented.

The gdb backtrace says the issue is happening in RF24 which seems odd, I would expect it to be in WiringPi. Is RF24 overriding the pinMode() call?

(gdb) bt

0 0x76f1fb58 in bcm2835_peri_read () from /usr/local/lib/librf24.so.1

1 0x76f1fc38 in bcm2835_peri_set_bits () from /usr/local/lib/librf24.so.1

2 0x000133dc in main (argc=, argv=) at ./src/server.cpp:80

server.cpp:80 is the aforementioned line, pinMode(25, INPUT);

Is there an alternative way to get WiringPi playing nicely with RF24, or if not, is there a recommended GPIO library to use alongside RF24?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

[alandsidel]

I'm pretty sure I didn't use that driver configuration option when building RF24 since I wasn't using wiringPi at the time. I'm rebuilding it now and will let you know, thanks.

[alandsidel]

That looks like it solved the issue, thanks. The pin numbering isn't showing the correct mapping in gpio readall I think (CS0 on the board is marked CE0 in gpio) but I'm investigating that.

[TMRh20]

I think that’s normal. RPi docs seem to refer to the CSN/CS pins as CE

Unfortunate because RF24 has CSN/CS and CE for the separate control pins.

On Aug 2, 2020, at 8:50 AM, Allen Landsidel notifications@github.com wrote:

 That looks like it solved the issue, thanks. The pin numbering isn't showing the correct mapping in gpio readall I think (CS0 on the board is marked CE0 in gpio) but I'm investigating that.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.