Ok,
but now I am compiling a module to write for an I/O port (code below)
With the previous version of next branch I can compile it without any
error, insert it check that it is working fine by reading te I/O port from
an user space app, but with this new version, when compiling the module, I
get:
CC [M] /mnt/hgfs/apps-3.x/10-portAccessSyncKernel/writeportK.o
In file included from include/xenomai/rtdm/driver.h:43:0,
from
/mnt/hgfs/apps-3.x/10-portAccessSyncKernel/writeportK.c:8:
include/xenomai/cobalt/kernel/init.h:24:33: fatal error:
cobalt/uapi/corectl.h: No such file or directory
#include <cobalt/uapi/corectl.h>
^
compilation terminated.
The makefile is just:
obj-m += writeportK.o
$(eval EXTRA_CFLAGS := $(shell xeno-config --kcflags))
all:
make -C /lib/modules/$(shell uname -r)/build M=$(PWD) modules
And the code of the module:
#include <linux/kernel.h>
#include <linux/module.h>
#include <rtdm/driver.h> //defines: rtdm_task_t, rtdm_sem_t, ... (also
includes rtdm.h)
#define TASK_PRIO 20 //99 is Highest RT priority, 0 is Lowest
#define us 1000 //1000 ns = 1 us (micro sec)
#define ms 1000000ll //1000000 ns = 1 ms
#define TASK_PERIOD 500*ms //0.5 secs period
#define COUNT 0x80
#define DATA 0x81
rtdm_task_t td_pw; //RTDM task descriptor
rtdm_sem_t semA; //RTDM semaphore descriptor
void writeport (void *arg) {
static unsigned char count=0;
int err;
for (;;) {
err = rtdm_task_wait_period(); //deschedule until next period
if (err<0) {
rtdm_printk("writeportK: rtdm_task_wait_period, error %d\n", err);
break; //avoid kernel corruption when removing module
//trying to execute code of killed thread
}
outb (count, COUNT); //outb takes aprox. 1 Micro sec.
outb (count%4, DATA); //outb takes aprox. 1 Micro sec.
rtdm_printk("writeportK: count = %d\n", count); //check with dmesg
if (++count == 100) count=0;
}
}
int init_module(void) {
int err;
err = rtdm_task_init (&td_pw, "writeportK", writeport, NULL, TASK_PRIO,
TASK_PERIOD);
if (err < 0) rtdm_printk("writeportK: rtdm_task_init, error %d\n", err);
else {
err = rtdm_task_set_period(&td_pw, TASK_PERIOD); //set period
if (err < 0) rtdm_printk("writeportK: rtdm_task_set_period %d\n", err);
else rtdm_printk("writeportK successfully loaded\n");
}
return err;
}
void cleanup_module(void) {
rtdm_task_destroy(&td_pw);
rtdm_printk("writeportK unloaded\n");
}
MODULE_LICENSE("GPL"); //To have access to Xenomai symbols
On 19 March 2015 at 18:29, Philippe Gerum <[email protected]> wrote:
> On 03/19/2015 06:23 PM, Philippe Gerum wrote:
> > On 03/17/2015 06:19 PM, Helder Daniel wrote:
> >> ok
> >>
> >> On 17 March 2015 at 17:10, Philippe Gerum <[email protected]
> >> <mailto:[email protected]>> wrote:
> >>
> >> On 03/17/2015 06:06 PM, Helder Daniel wrote:
> >> > I am sorry I just found now.
> >> > It still happens with Mercury (with the configuration described
> in the
> >> > previous email)
> >> >
> >> > When I compiled another app and try to run it I got the same
> error:
> >> >
> >> > root@debian:~/09-portAccessSyncTwoAppsC++# ./writeport
> >> > sysregd:
> create_directory_recursive("/var/run/xenomai/anon/system"):
> >> > Transport endpoint is not connected
> >> > sysregd:
> create_directory_recursive("/var/run/xenomai/anon/system"):
> >> > Transport endpoint is not connected
> >> > sysregd:
> create_directory_recursive("/var/run/xenomai/anon/system"):
> >> > Transport endpoint is not connected
> >> > 3"012.188| WARNING: [main] cannot connect to registry daemon
> >> > 3"012.188| BUG: [main] initialization failed, EAGAIN
> >>
> >> Ok. This one is different from the ENOMEM issue, which is
> >> Cobalt-specific. The problem with mounting the anon/system root is a
> >> FUSE issue, which still needs a work around it seems.
> >>
> >
> > The ENOMEM issue is now fixed in -next. I'll tackle the C++ parsing
> > problem with libtrank next.
> >
>
> This commit in -next fixes the link stage for the readport/writeport test.
>
>
> http://git.xenomai.org/xenomai-3.git/commit/?h=next&id=f537649c1b69a4f7b6d8f22ef344faf3414fe079
>
> --
> Philippe.
>
--
Helder Daniel
UALG - FCT
DEEI
http://w3.ualg.pt/~hdaniel
_______________________________________________
Xenomai mailing list
[email protected]
http://www.xenomai.org/mailman/listinfo/xenomai
