Mark wrote:
> Hello,

> I came across the kernel driver code implementing some sort of IGMP snooping
> backend and as part of its functionalilty it creates a new socket address
> family, AF_IGMPSNOOP, but actually implements just a few operations for this
> type of socket:

> static struct proto_ops igmp_snoop_ops = {
>     family:       AF_IGMP_SNOOP,
>     release:      _igmp_snoop_sock_release,
>     bind:         sock_no_bind,
>     connect:      sock_no_connect,
>     socketpair:   sock_no_socketpair,
>     accept:       sock_no_accept,
>     getname:      sock_no_getname,
>     poll:         datagram_poll,
>     ioctl:        sock_no_ioctl,
>     listen:       sock_no_listen,
>     shutdown:     sock_no_shutdown,
>     setsockopt:   _igmp_snoop_setsockopt,
>     getsockopt:   sock_no_getsockopt,
>     sendmsg:      _igmp_snoop_sock_sendmsg,
>     recvmsg:      _igmp_snoop_sock_recvmsg,
>     mmap:         sock_no_mmap,
>     sendpage:     sock_no_sendpage,
> }

> However from a user space perspective a socket is created as:

> fd = socket (AF_IGMPSNOOP, SOCK_RAW, IPPROTO_IGMP)

> I'm wondering what is the rationale for adding a new family, while in fact a
> raw socket is being used to access the stack for IGMP packets? Would not it
> be the same to just create AF_INET raw socket and do the things?

> Looking forward to hearing form you !
> Thanks.

> Mark

"Les Cargill" <lcargil...@comcast.com> wrote in message

>> static struct proto_ops igmp_snoop_ops = {
>>     family:       AF_IGMP_SNOOP,
>>     release:      _igmp_snoop_sock_release,
>>     bind:         sock_no_bind,
>>     connect:      sock_no_connect,
>>     socketpair:   sock_no_socketpair,
>>     accept:       sock_no_accept,
>>     getname:      sock_no_getname,
>>     poll:         datagram_poll,
>>     ioctl:        sock_no_ioctl,
>>     listen:       sock_no_listen,
>>     shutdown:     sock_no_shutdown,
>>     setsockopt:   _igmp_snoop_setsockopt,
>>     getsockopt:   sock_no_getsockopt,
>>     sendmsg:      _igmp_snoop_sock_sendmsg,
>>     recvmsg:      _igmp_snoop_sock_recvmsg,
>>     mmap:         sock_no_mmap,
>>     sendpage:     sock_no_sendpage,
>> }

>> However from a user space perspective a socket is created as:

>> fd = socket (AF_IGMPSNOOP, SOCK_RAW, IPPROTO_IGMP)

>> I'm wondering what is the rationale for adding a new family, while in
>> fact a
>> raw socket is being used to access the stack for IGMP packets? Would not
>> it
>> be the same to just create AF_INET raw socket and do the things?

> Possibly.

> I'm pretty sure IGMP snooping is really a "bridge" thing.

> http://www.linuxfoundation.org/collaborate/workgroups/networking/brid...

> I don't know why you'd need a socket for bridge-based IGMP snooping
> at all. Bridges don't need no steenking sockets.  This being said, this
> is another one of those things...

Mark wrote:
> "Les Cargill" <lcargil...@comcast.com> wrote in message
> news:k7fdln$248$1@dont-email.me...
>>> static struct proto_ops igmp_snoop_ops = {
>>>      family:       AF_IGMP_SNOOP,
>>>      release:      _igmp_snoop_sock_release,
>>>      bind:         sock_no_bind,
>>>      connect:      sock_no_connect,
>>>      socketpair:   sock_no_socketpair,
>>>      accept:       sock_no_accept,
>>>      getname:      sock_no_getname,
>>>      poll:         datagram_poll,
>>>      ioctl:        sock_no_ioctl,
>>>      listen:       sock_no_listen,
>>>      shutdown:     sock_no_shutdown,
>>>      setsockopt:   _igmp_snoop_setsockopt,
>>>      getsockopt:   sock_no_getsockopt,
>>>      sendmsg:      _igmp_snoop_sock_sendmsg,
>>>      recvmsg:      _igmp_snoop_sock_recvmsg,
>>>      mmap:         sock_no_mmap,
>>>      sendpage:     sock_no_sendpage,
>>> }

>>> However from a user space perspective a socket is created as:

>>> fd = socket (AF_IGMPSNOOP, SOCK_RAW, IPPROTO_IGMP)

>>> I'm wondering what is the rationale for adding a new family, while in
>>> fact a
>>> raw socket is being used to access the stack for IGMP packets? Would not
>>> it
>>> be the same to just create AF_INET raw socket and do the things?

>> Possibly.

>> I'm pretty sure IGMP snooping is really a "bridge" thing.

>> http://www.linuxfoundation.org/collaborate/workgroups/networking/brid...

>> I don't know why you'd need a socket for bridge-based IGMP snooping
>> at all. Bridges don't need no steenking sockets.  This being said, this
>> is another one of those things...

> Thanks for the response.
> Another thing that comes to my mind is that perhaps the implementor of
> AF_IGMPSNOOP family wanted to have just a small subset of socket operations,
> not a full-blown socket, basically only the 'release', 'poll', setsockopt,
> sendmsg and recvmsg are implemented for this new family. May be they wanted
> to have a light-weight mechanism.

> Mark