Ethernet (Raw)

[Joel Sherrill]

gregory.menke at gsfc.nasa.gov wrote:
> 
> Joel Sherrill writes:
>  >
>  >
>  > gregory.menke at gsfc.nasa.gov wrote:
>  > >
>  > > Erwin Rol writes:
>  > >  > On Tue, 2003-03-25 at 18:12, Joel Sherrill wrote:
>  > >  >
>  > >  > > Look for tap in the networking guide and see if that helps.  I don't
>  > >  > > know if it is far enough into the stack to help.
>  > >  >
>  > >  > The TAP interface is only the receiving part ( right ? ). I hacked up
>  > >  > something like that for RTEMS 4.0 too, the problem was sending raw (or
>  > >  > almost raw , IPX) frames.
>  > >  >
>  > >
>  > > Its right above the driver itself- its a raw ethernet frame.
>  > > Depending on the if_tap function's return value, the packet is either
>  > > passed to the tcp/ip stack or discarded.
>  >
>  > Does the standard FreeBSD stack have a facility like this that was
>  > not enabled/ported?
>  >
>  > If not, then can the tap design/interface be used as a model to
>  > provide the write side?
> 
> I don't know to both questions.  if_tap operates at the very bottom
> layer of the ip stack, just after the driver submits the ethernet
> frame to it.  The sending side looks fairly involved with a
> per-interface send queue.  I would be suprised if there wasn't a hook
> in there someplace.  Do you know if SOCK_RAW is supported by the
> stack?

It looks like the code is there to support it.  I don't know if it works
though.  It might require some hook at initialization that might or
might not be happening.  For sure, there is plenty of code in
cpukit/libnetworking
which SHOULD support it.

> Gregm

-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel at OARcorp.com                 On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
Support Available                (256) 722-9985