From mboxrd@z Thu Jan 1 00:00:00 1970 From: Grant Edwards To: Jonathan Larmour Cc: Hugo Tyson , ecos-discuss@sources.redhat.com Subject: Re: [ECOS] DHCP shuts down Ethernet device? Date: Mon, 23 Apr 2001 13:35:00 -0000 Message-id: <20010423153657.A2947@visi.com> References: <20010423123926.A2035@visi.com> <3AE4822A.24EC454B@redhat.com> X-SW-Source: 2001-04/msg00340.html On Mon, Apr 23, 2001 at 08:27:38PM +0100, Jonathan Larmour wrote: > > It shuts down everything, deliberately. > > > > > I think I'm going to have to either comment out the above call, > > > or make my Ethernet driver's _stop() method into a noop... > > > > Either of those would do. > > I'm wondering: isn't the problem just that configuring the > interface down calls the _stop() method? Why not just not do > that? The common eth driver already just has eth_drv_recv() > return immediately if the interface is down. Of course it > depends where any non-TCP/IP stack is plugging in... In my case, The non-IP stack hooks into the Ethernet driver at the same level that the TCP/IP stack does. I decided to change my _stop() method so that it doesn't shut off the interface, it just sets a flag so that packets normally passed to the TCP/IP stack are discarded (this should result in the behvior expected by the TCP/IP stack). > It probably shouldn't be the default behaviour admittedly, > otherwise we're just doing a bunch of interrupt processing for > no reason in the most common case. Right. -- Grant Edwards grante@visi.com