[olsr-dev] debug_level=0 & standart I/O

[Andreas Tønnesen]

Nice one aaron :) It is available in Linux as well.
I'll use this call then.

- Andreas

aaron wrote:
> On Mon, 28 Mar 2005, Andreas [iso-8859-1] Tønnesen wrote:
> 
> hi!
> 
> in *BSD (dont know about linux) there is the nice daemon() syscall 
> exactly for that. It does it the "right" way.
> 
> something like
>    pid=fork()
>    switch(pid) {
> 
>      case 0: /* child */
>        blabla
>        stdin2=dup(0);
>        stdout2=dup(1);
>        stderr2=dup(2);
>        close(0);
>        close(1);
>        close(2);
>        chdir("/");
>        ...
>        connect to syslog if you want
>        blabla
> 
>      default: /parent */
>        blabla
>        wait() or exit()
>    }
> 
> 
> hope it is usefull
> 
> a.
> 
>> Hi Grégory,
>>
>> Wouldn't the right thing here be for the parent process to close the
>> handlers? Or does that not work?
>> I'll add Thomas' suggested code when I get back from vacation(tomorrow),
>> so there will be a fix for this in 0.4.9.
>>
>> - Andreas
>>
>>>
>>>         Hi,
>>>
>>>     I'm using PHP on a web server for making a page that start/stop 
>>> olsrd.
>>> When PHP execute "olsrd -i eth0 -d 0" command (for example), it waiting
>>> for the end of command... and it never happens because while the
>>> standard I/O (stdin/stdout/stderr) are not closed, php think that the
>>> process is running...
>>>
>>>     If I modified the main.c by writing "fclose(stdin); fclose(stdout);
>>> fclose(stderr)" before the setsid command (in the child code), it run
>>> ok....
>>>
>>>     So, is it a good idea to close the standard I/O after the fork 
>>> (for the
>>> child, the parent exit normaly) or is there an other solution ?
>>>
>>> Best regards,
>>>
>>> -- 
>>> Grégory Marfjan
>>>
>>>
>>>
>>> _______________________________________________
>>> olsr-dev mailing list
>>> (spam-protected)
>>> https://www.olsr.org/mailman/listinfo/olsr-dev
>>>
>>
>>
>> ---------
>> Andreas Tønnesen
>> http://www.olsr.org
>> _______________________________________________
>> olsr-dev mailing list
>> (spam-protected)
>> https://www.olsr.org/mailman/listinfo/olsr-dev
>>
>>
> 
> 
> ---
> Yesterday it worked.
> Today it is not working.
> Windows is like that.
>                        (old japanese haiku wisdom)

-- 
Andreas Tønnesen
http://www.olsr.org