[Xenomai] xenomai-3/next: modprobe rtpacket fails with -EINVAL

[Xenomai] xenomai-3/next: modprobe rtpacket fails with -EINVAL

[Anders Blomdell]

AFAICT there are two ways to fix it, either change kernel/cobalt/rtdm/device.c:

       if (drv->device_count <= 0)
                return -EINVAL;
to:
       if (drv->device_count <= 0 ||
	   (drv->device_count == 0 && 
	    (drv->device_flags & RTDM_DEVICE_TYPE_MASK) != RTDM_PROTOCOL_DEVICE)))
                return -EINVAL;    

or change kernel/drivers/net/stack/packet/af_packet.c:

  static struct rtdm_driver packet_proto_drv = {
    .profile_info =     RTDM_PROFILE_INFO(packet,
                                        RTDM_CLASS_NETWORK,
                                        RTDM_SUBCLASS_RTNET,
                                        RTNET_RTDM_VER),
    .device_flags =     RTDM_PROTOCOL_DEVICE,
    .context_size =     sizeof(struct rtsocket),
    ...
  }

to:

  static struct rtdm_driver packet_proto_drv = {
    .profile_info =     RTDM_PROFILE_INFO(packet,
                                        RTDM_CLASS_NETWORK,
                                        RTDM_SUBCLASS_RTNET,
                                        RTNET_RTDM_VER),
    .device_flags =     RTDM_PROTOCOL_DEVICE,
    .device_count =     1,
    .context_size =     sizeof(struct rtsocket),
    ...
  }

Since the second change is what 'rtdm_driver raw_packet_proto_drv', and the code looks
prettier, I guess that the second way is the preferred one?

/Anders
-- 
Anders Blomdell                  Email: anders.blomdell@control.lth.se
Department of Automatic Control
Lund University                  Phone:    +46 46 222 4625
P.O. Box 118                     Fax:      +46 46 138118
SE-221 00 Lund, Sweden

Re: [Xenomai] xenomai-3/next: modprobe rtpacket fails with -EINVAL

[Gilles Chanteperdrix]

On Thu, Feb 05, 2015 at 12:15:22PM +0100, Anders Blomdell wrote:
> AFAICT there are two ways to fix it, either change kernel/cobalt/rtdm/device.c:
> 
>        if (drv->device_count <= 0)
>                 return -EINVAL;
> to:
>        if (drv->device_count <= 0 ||
> 	   (drv->device_count == 0 && 
> 	    (drv->device_flags & RTDM_DEVICE_TYPE_MASK) != RTDM_PROTOCOL_DEVICE)))
>                 return -EINVAL;    
> 
> or change kernel/drivers/net/stack/packet/af_packet.c:
> 
>   static struct rtdm_driver packet_proto_drv = {
>     .profile_info =     RTDM_PROFILE_INFO(packet,
>                                         RTDM_CLASS_NETWORK,
>                                         RTDM_SUBCLASS_RTNET,
>                                         RTNET_RTDM_VER),
>     .device_flags =     RTDM_PROTOCOL_DEVICE,
>     .context_size =     sizeof(struct rtsocket),
>     ...
>   }
> 
> to:
> 
>   static struct rtdm_driver packet_proto_drv = {
>     .profile_info =     RTDM_PROFILE_INFO(packet,
>                                         RTDM_CLASS_NETWORK,
>                                         RTDM_SUBCLASS_RTNET,
>                                         RTNET_RTDM_VER),
>     .device_flags =     RTDM_PROTOCOL_DEVICE,
>     .device_count =     1,
>     .context_size =     sizeof(struct rtsocket),
>     ...
>   }
> 
> Since the second change is what 'rtdm_driver raw_packet_proto_drv', and the code looks
> prettier, I guess that the second way is the preferred one?

I received a patch for this issue, I simply need to merge it. Thanks
for the reminder.

-- 
					    Gilles.

Re: [Xenomai] xenomai-3/next: modprobe rtpacket fails with -EINVAL

[Anders Blomdell]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 2015-02-05 12:50, Gilles Chanteperdrix wrote:
> I received a patch for this issue, I simply need to merge it. Thanks
> for the reminder.

OK, ".device_count =     1," works for me now. 

BTW: is it the expected behaviour that when rtpacket is not loaded, socket()
will default to __real_socket()?

/Anders


- -- 
Anders Blomdell                  Email: anders.blomdell@control.lth.se
Department of Automatic Control
Lund University                  Phone:    +46 46 222 4625
P.O. Box 118                     Fax:      +46 46 138118
SE-221 00 Lund, Sweden

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1

iQEcBAEBAgAGBQJU03c3AAoJENZYyvaDG8NcdOIH/1KOtrLaXGVOz/UR6qCTHH4w
DXiEcVaYyZfsxWFYGlaYLJX4eEyaDyUgwwzxu9btZK20C380GJvXBujNQRo/fCzT
y/lM/4mDzuWqUpdSWhEfklVM9W90pSB1nUO2FpxTtNdVYJRGrZa5MhltobrmyZKw
WXIVU9Hs47d1WevvXRbHMM1WFHh6kMcbmjYzJ596B2zAFs1qRQjQb/QaxhlCITbq
5enzxCaqXQTC1saBCcNKhg+dLviGOqUVIHJvr5PRSp+nRgTP5QKqcXbu3p0NxSf/
mHViajS7f/muPbCFFOvAytiF/bRb5qXcnqpKybpWkSihozy7vMA7GVh+Dc3BAi4=
=5r/d
-----END PGP SIGNATURE-----

Re: [Xenomai] xenomai-3/next: modprobe rtpacket fails with -EINVAL

[Gilles Chanteperdrix]

On Thu, Feb 05, 2015 at 02:59:21PM +0100, Anders Blomdell wrote:
> On 2015-02-05 12:50, Gilles Chanteperdrix wrote:
> > I received a patch for this issue, I simply need to merge it. Thanks
> > for the reminder.
> 
> OK, ".device_count =     1," works for me now. 
> 
> BTW: is it the expected behaviour that when rtpacket is not loaded, socket()
> will default to __real_socket()?

Well, it depends on the parameters you pass to socket. I guess it
only happens if you want to create a raw packet socket.

-- 
					    Gilles.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 811 bytes
Desc: not available
URL: <http://www.xenomai.org/pipermail/xenomai/attachments/20150205/03d01172/attachment.sig>