On Mon, Feb 01, 2010 at 11:33:49AM +0100, Rogier Krieger wrote: > On Mon, Feb 1, 2010 at 07:32, Steve Williams > <st...@williamsitconsulting.com> wrote: > > I have downloaded the current cvs code and compiled it. It exhibits the > > same problem, missing em0. > > It seems to nicely detect the hardware, just not liking its EEPROM > contents and stopping initialisation there. While you should take a > developer's word over mine, I suppose it's not surprising that > ifconfig(8) does not show the hardware. > > Seeing a few Google searches seems to indicate it's not necessarily an > OS problem. While some posts mention an Intel utility (IBAUTIL.EXE) to > configure/manage the built-in boot agent, you will probably want to > search for the correct NIC model and see which specific version/tool > you need. > > I included a link [1] to the utility a 5 minute cursory search yielded > me. Use at your own risk, since I can't really be sure it's the > correct one. >
I doubt this has todo with the boot rom. The card has a nvrom to store the macaddr etc and it seems that the access to that one is fubar-ed. There were many changes to em(4) to support newer models, some of the changes were quite intrusive and could result in failure of other cards. If possible please try some older kernels to find the commit which has caused the regression. -- :wq Claudio
