I admit, this will be an embarrassing question, the both process need to know when to send the message (process A) and when to receive it (process B), I forget the classic UNIX's stuff called synchronisation tool like semaphore. I really forget this magic ;). But anyway, thanks for reading self-answered question.
Ardhan
--- ajhwb@xxxxxxxx wrote:
From: "Ardhan Madras" <ajhwb@xxxxxxxx>
To: "Glynn Clements" <glynn@xxxxxxxxxxxxxxxxxx>
Cc: <linux-c-programming@xxxxxxxxxxxxxxx>
Subject: Re: POSIX Message Queues
Date: Thu, 2 Sep 2010 02:51:56 -0700
Hi all,
I use message queue to exchange data between 2 unrelated process, as shown below
Program A:
while (1) {
mq_receive(); /* wait and receive request */
/* process request */
mq_send(); /* reply the request result */
}
Program B:
mq_send(); /* send request */
mq_receive(); /* process result */
How it's possible to process A that call mq_send() receive it's *own* send data using mq_receive(), so process B that intended to read the message (mq_receive()) could not read it?, the problem is *sometime* happen, at most program A reply what program B requested. Did I miss something here?
Ardhan
--- glynn@xxxxxxxxxxxxxxxxxx wrote:
From: Glynn Clements <glynn@xxxxxxxxxxxxxxxxxx>
To: <ajhwb@xxxxxxxx>
Cc: <linux-c-programming@xxxxxxxxxxxxxxx>
Subject: Re: POSIX Message Queues
Date: Tue, 15 Jun 2010 16:33:25 +0100
Ardhan Madras wrote:
> It was mq_open(), sorry my bad ;p
>
> I did umask(0001) before creating the queue, it's works now. 'Program
> A' that requested O_RDWR flag to the queue can send and receive
> message as 'Program B' did. I also try to demonstrating this problem
> with another machine (Linux of course), as result i can O_RDWR'ed
> program A queue flag without have to call umask() first. Is this
> behavior can be set through system configuration?
A process' umask is inherited from its parent. For interactive
sessions, you can set it in your ~/.profile or similar. For daemons
started by init, you can set it in the rc script which starts the
daemon.
> fchmod() also works here, sorry, but what do you mean with 'not
> portable'?, do you mean fchmod() intended to be used only for file, in
> other words "not every system support fchmod() with message queue
> descriptor" right?
Yes.
--
Glynn Clements <glynn@xxxxxxxxxxxxxxxxxx>
_____________________________________________________________
Listen to KNAC, Hit the Home page and Tune In Live! ---> http://www.knac.com
--
To unsubscribe from this list: send the line "unsubscribe linux-c-programming" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
_____________________________________________________________
Listen to KNAC, Hit the Home page and Tune In Live! ---> http://www.knac.com
--
To unsubscribe from this list: send the line "unsubscribe linux-c-programming" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
