From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <48CE7E8D.6000301@domain.hid> Date: Mon, 15 Sep 2008 17:26:05 +0200 From: Harco Kuppens MIME-Version: 1.0 References: <48CE3096.7040502@domain.hid> <48CE6E48.6030808@domain.hid> In-Reply-To: <48CE6E48.6030808@domain.hid> Content-Type: multipart/alternative; boundary="------------010506020201080508090701" Subject: Re: [Xenomai-help] probably error in api document List-Id: Help regarding installation and common use of Xenomai List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: xenomai@xenomai.org This is a multi-part message in MIME format. --------------010506020201080508090701 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit yeah, that is what I read but the documenation for the kernel space version of rt_int_create also says that it can be called from a user space task. I think that is a mistake in the documentation. best regards Harco Kuppens Gilles Chanteperdrix wrote: > Harco Kuppens wrote: > >> Hi, >> >> I noticed in the Xenomai native API that the longer rt_create_intr : >> >> int rt_intr_create >> >> (RT_INTR *intr, const char *name, unsigned irq, rt_isr_t isr, rt_iack_t >> iack, int mode) >> >> says in its documentation : >> >> This service can be called from: >> >> * Kernel module initialization/cleanup code >> * Kernel-based task >> * *User-space task* >> >> I find it strange that it could be called from a user-space task. >> I looked in /usr/xenomai/include/native/intr.h but there it really is >> defined when __KERNEL__ is defined. >> >> So I guess this is an error in the API documention? >> > > No, there is no error: the service is implemented both in user-space and > kernel-space, and if you look at native/intr.h, you will see two > definitions of rt_intr_create: one for kernel-space and one for user-space. > > Note however that the interface in user-space is not the same as in > kernel-space, that is because in user-space, you are supposed to call > rt_intr_wait to wait for the next interrupt. > --------------010506020201080508090701 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit yeah, that is what I read but the documenation for the kernel space version of rt_int_create also says that it can be called from a user space task. I think that is a mistake in the documentation.

best regards
Harco Kuppens

Gilles Chanteperdrix wrote: 
Harco Kuppens wrote:
  
Hi,

I noticed in the Xenomai native API that  the longer rt_create_intr :

    int rt_intr_create 
<http://www.xenomai.org/documentation/trunk/html/api/group__interrupt.html#ga0a9d98556ec838008f408a56e79d304> 
(RT_INTR *intr, const char *name, unsigned irq, rt_isr_t isr, rt_iack_t 
iack, int mode)

says in its documentation :

    This service can be called from:

        * Kernel module initialization/cleanup code
        * Kernel-based task
        * *User-space task*

I find it strange that it could be called from a user-space task.
I looked in /usr/xenomai/include/native/intr.h but there it really is 
defined when __KERNEL__ is defined.

So I guess this is an error in the  API documention?
    

No, there is no error: the service is implemented both in user-space and
 kernel-space, and if you look at native/intr.h, you will see two
definitions of rt_intr_create: one for kernel-space and one for user-space.

Note however that the interface in user-space is not the same as in
kernel-space, that is because in user-space, you are supposed to call
rt_intr_wait to wait for the next interrupt.


--------------010506020201080508090701--