Hi Ben, On Monday, September 13, 2010, Ben Gardiner <bengardi...@nanometrics.ca> wrote: > Hi Ben, > > On Sun, Sep 12, 2010 at 11:45 PM, Ben Warren <biggerbadder...@gmail.com> > wrote: >> Please consider doing this differently. If you change the driver to make >> 'davinci_eth_set_mac_addr()' meet the signature of >> 'eth_device->write_hwaddr()', all the environment handling is done >> automagically. Let me know if you're confused and I'll provide more >> guidance. > > Thanks for pointing me in the right direction here. I think I get it > now; please correct me if I am wrong: board_eth_init should assign the > eth_device->write_hwaddr() function pointer then eth_initialize() will > handle calling board_eth_init and will set the MAC address from the > ethaddr env var if the MAC address in the eth_device struct is > all-zeroes. > Almost. The driver should set the function pointer, just as it does 'send()', 'init()', 'halt()', etc. This is a recent feature and the davinci driver hasn't been updated yet > Best Regards, > Ben Gardiner > > --- > Nanometrics Inc. > http://www.nanometrics.ca
Regards, Ben _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot