Thanks a lot Hussam!
I could not get it working using the /sys/class/gpio/gpioxx/edge. Maybe
/sys/class/gpioxx/value would work? But by modifying the sample device tree
overlay from www.hipstercircuits.com and /dev/input/event1. I was able to
get it going using the following code.
#include <stdio.h>
#include <stddef.h>
#include <time.h>
#include <iostream>
#include <fcntl.h>
#include <unistd.h>
#include <sys/types.h>
#include <string.h>
#include <termios.h>
#include <signal.h>
#include <stdlib.h>
using namespace std;
volatile int counter = 0;
void IRQHandler(int sig){
counter ++;
cout << counter;
}
int main(void){
struct sigaction saio;
int fd1;
fd1 = open("/dev/input/event1", O_RDWR|O_NONBLOCK);
saio.sa_handler = IRQHandler;
//saio.sa_mask = 0;
saio.sa_flags = 0;
saio.sa_restorer = NULL;
sigaction(SIGIO,&saio,NULL);
fcntl(fd1, F_SETOWN, getpid());
fcntl(fd1, F_SETFL, FASYNC);
while(1){
}
}
On Sun, Feb 23, 2014 at 8:57 PM, Hussam Hertani <halhe...@gmail.com> wrote:
> Dorian, here's a little example. First open the file and get a file
> descriptor, then use FCNTL with SETOWN command to send SIGIO to current
> process. Use fcntl to set the FASYNC flag then populate the signal
> structure. This code was not tested but hopefully should give you an idea
> as to what to do...
>
>
> int counter = 0;
>
> int main(){
> struct sigaction saio;
> int fd ;
> fd = open("/sys/class/blahblah/edge",O_RDWR | O_NONBLOCK); // open file
> descriptor for the edge
>
> /* install the signal handler before making the device asynchronous */
> saio.sa_handler = signal_handler_IO;
> saio.sa_mask = 0;
> saio.sa_flags = 0;
> saio.sa_restorer = NULL;
> sigaction(SIGIO,&saio,NULL);
>
> /* allow the process to receive SIGIO */
> fcntl(fd, F_SETOWN, getpid());
> /* Make the file descriptor asynchronous (the manual page says only
> O_APPEND and O_NONBLOCK, will work with F_SETFL...) */
> fcntl(fd, F_SETFL, FASYNC);
>
>
> }
>
> void signal_handler_IO (int status)
> {
> counter++;
> }
>
> Linux purists generally don't like using signals,,,too many can cause
> many conflicts/race conditions that are hard to debug.
> http://ajaxxx.livejournal.com/62378.html
>
> most of the above code was taken from this link that has code for SIGIO
> asynchronous serial port access:
> http://www.tldp.org/HOWTO/pdf/Serial-Programming-HOWTO.pdf
>
> For more on signals in general
> http://beej.us/guide/bgipc/
> http://ph7spot.com/musings/introduction-to-unix-signals-and-system-calls
> http://www.linuxprogrammingblog.com/all-about-linux-signals?page=1
>
> For more on fcntl:
> http://linux.die.net/man/2/fcntl
>
> There's also real-time signals. I haven't heard of those before today.
> But they seem to have less race conditions issues. I'll have to read up on
> them. Unfortunately My Linux Signal Programming fu is very limited. I
> recommend that you get yourself a good text reference on Linux System
> Programming as well. It will help.
>
> Regards,
> Hussam
>
>
> On Sun, Feb 23, 2014 at 7:35 PM, Dorian Levy <dorianal...@gmail.com>wrote:
>
>> Something like SIGIO is what I was trying to do. I was
>> trying signal(SIGINT, IRQHandler), but something is not
>> right. How do I attach it, and why do Linux enthusiasts not recommend this?
>>
>>
>>
>> On Sun, Feb 23, 2014 at 1:40 PM, <halhe...@gmail.com> wrote:
>>
>>> Dorian,
>>>
>>> I haven't tried any of these myself but I see no reason why they
>>> shouldn't work.
>>>
>>> - you could use a fork to create a child process. In this child
>>> process use the poll()/select() calls to detect a change in the edge
>>> file.
>>> When this happens a variable could be incremented and piped back to the
>>> parent process.
>>> - You could use threads...i.e. a thread calls poll()/select() to
>>> detect a change in the edge file and then updates a counts
>>> variables.....with threads all threads share the variables so this will
>>> be
>>> easier. but could be more convoluted.....
>>> - You could try monitoring the edge file with GDK/GTK libs
>>> monitoring I/O calls
>>> https://developer.gnome.org/gtk-tutorial/2.90/x1770.html. Qt has a
>>> QIODevice class that does the same thing.
>>> - You could also attach the SIGIO signal to the file descriptor for
>>> the edge file and write a signal handler that increments a
>>> variable...though Linux purists do not recommend this
>>>
>>> Hussam
>>>
>>>
>>>
>>> On Saturday, February 22, 2014 1:41:02 PM UTC-5, Dorian Levy wrote:
>>>>
>>>> I want to use a hall effect motor encoder to determine the number of
>>>> revolutions of a wheel. Using the BB-KEYS I am able to set
>>>> up /dev/inputs/event1 that captures the event and records the number of
>>>> interrupts in /proc/interrupts. I believe setting
>>>> /sys/class/gpio/gpioxx/edge will do pretty much the same. I haven't been
>>>> able to figure out how to count them in a C/C++ code. Is it bes to set up
>>>> an interrupt and count or try to read the /proc/interrupt file? Polling is
>>>> not an option because there are two wheel that need to be counted. I would
>>>> prefer to have and ISR that counts each time it is triggered, but I have
>>>> not been able to figure out how to set it up. Any help would be greatly
>>>> appreciated since I have spent the last few days trying to research this
>>>> without much luck.
>>>>
>>>> Thanks,
>>>> Dorian
>>>>
>>> --
>>> For more options, visit http://beagleboard.org/discuss
>>> ---
>>> You received this message because you are subscribed to a topic in the
>>> Google Groups "BeagleBoard" group.
>>> To unsubscribe from this topic, visit
>>> https://groups.google.com/d/topic/beagleboard/RhSCj3GgAj8/unsubscribe.
>>> To unsubscribe from this group and all its topics, send an email to
>>> beagleboard+unsubscr...@googlegroups.com.
>>> For more options, visit https://groups.google.com/groups/opt_out.
>>>
>>
>> --
>> For more options, visit http://beagleboard.org/discuss
>> ---
>> You received this message because you are subscribed to a topic in the
>> Google Groups "BeagleBoard" group.
>> To unsubscribe from this topic, visit
>> https://groups.google.com/d/topic/beagleboard/RhSCj3GgAj8/unsubscribe.
>> To unsubscribe from this group and all its topics, send an email to
>> beagleboard+unsubscr...@googlegroups.com.
>> For more options, visit https://groups.google.com/groups/opt_out.
>>
>
> --
> For more options, visit http://beagleboard.org/discuss
> ---
> You received this message because you are subscribed to a topic in the
> Google Groups "BeagleBoard" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/beagleboard/RhSCj3GgAj8/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> beagleboard+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
--
For more options, visit http://beagleboard.org/discuss
---
You received this message because you are subscribed to the Google Groups
"BeagleBoard" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to beagleboard+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
