On Sun, May 31, 2015 at 10:16:57PM -0400, Sonic wrote: > On Sun, May 31, 2015 at 3:37 PM, Sonic <sonicsm...@gmail.com> wrote: > > On Sun, May 31, 2015 at 3:49 AM, Paul de Weerd <we...@weirdnet.nl> wrote: > >> I can confirm my i217-LM still works with the latest snapshot (typing > >> this e-mail over it). > > > > I'm currently upgrading -current but as I've seen no changes to > > if_em.c I'm skeptical of it working. > > I'm glad it's working for you Paul but it isn't working here with > -current. Works in FreeBSD and Fedora. > > However as it was reported earlier in the thread by Paulo to be > working in 5.6 but not in 5.7 I reverted the following files back to > their 5.6 release states (keeping the rest of the kernel at -current): > if_em.c > if_em.h > if_em_hw.c > if_em_hw.h > if_em_soc.c > > Then rebuilt the kernel and the i217-LM now works. > It appears that somewhere between 5.6 and now one or more of the > updates to those if_em* files made a change some systems don't like. > > Thanks, > > Chris >
Are you using vlans or anything that would raise the mtu? There aren't many revisions between 5.6 and 5.7 for em and none look all that suspicous for i217. Can you find which revision introduced the regression?
