Date: Sat, 26 May 2007 18:52:17 +0300 From: Stefan Lambrev <stefan.lambrev@sun-fish.com> To: Jack Vogel <jfvogel@gmail.com> Cc: freebsd-net@freebsd.org Subject: Re: em0 - bge0 failed to work at 1000baseTX Message-ID: <465857B1.1080104@sun-fish.com> In-Reply-To: <2a41acea0705111501h41170824qddbf8fee9d9dacc7@mail.gmail.com> References: <462E3B4A.5030307@sun-fish.com> <2a41acea0704241103r59a1fa8di7e7747e191eea787@mail.gmail.com> <462F0CBF.6020507@sun-fish.com> <2a41acea0704250936l46ed31a3w930ac8fba04df810@mail.gmail.com> <2a41acea0705041349o7bce9e1eh63ef1a036f80d1b6@mail.gmail.com> <463F2318.4070702@sun-fish.com> <2a41acea0705111501h41170824qddbf8fee9d9dacc7@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi Jack, Jack Vogel wrote: > On 5/7/07, Stefan Lambrev <stefan.lambrev@sun-fish.com> wrote: >> Hi again, >> >> Jack Vogel wrote: >> >> > >> > The new driver I just checked into CURRENT has this fix, I hope its >> > going to be in the May snapshot, Stefan, can you try CURRENT to >> > see if it solves your problem? >> > >> > Jack >> I have problems with loading if_em as module on 7.0-current from today: >> >> link_elf: symbol e1000_init_function_pointers_82575 undefined >> >> so I compiled in the kernel, and it works, but nothing changed about my >> problem - it still refuses to stay on 1000baseTX for more then 2 seconds >> and auto-negotiate to 100baseTX. >> >> em0: <Intel(R) PRO/1000 Network Connection Version - 6.5.0> port >> 0x2000-0x201f mem 0xf0500000-0xf051ffff,0xf0524000-0xf0524fff irq 19 at >> device 25.0 on pci0 >> em0: Ethernet address: 00:0f:fe:4e:78:16 >> em0: [FILTER] >> >> I'm ready to test any new patches :) > > My patch made assumptions about what is causing your problem, > and that may be incorrect. Our test group does not have that Broadcom > adapter but they are going to get one. > > They had a suggestion that you can try, go into your BIOS, and > then the Advanced Menu, I believe under that you will find settings > for the NorthBridge, in that submenu there is settings for HECI, > disable that, it is the interface to the management processor. > > Then boot up and test again, if it now autonegs correctly then > it is the management getting in your way, on the other hand if > you still see it happen then we're back to square one :) > > Good luck, > > Jack I finally succeed with building latest current, found another Gige Intel card and make more tests: Both servers are running freebsd 7.0 new server with dual-port em network card (7.0-CURRENT #9: Thu May 24): em1: <Intel(R) PRO/1000 Network Connection Version - 6.5.2> port 0x2000-0x201f mem 0xb8c00000-0xb8c1ffff,0xb8400000-0xb87fffff irq 19 at device 0.1 on pci4 em1: Ethernet address: 00:15:17:1b:08:ed em1: [FILTER] problematic server with verbose boot (FreeBSD 7.0-CURRENT #8: Sat May 26) : em0: <Intel(R) PRO/1000 Network Connection Version - 6.5.2> port 0x2000-0x201f mem 0xf0500000-0xf051ffff,0xf0524000-0xf0524fff irq 19 at device 25.0 on pci0 em0: Reserved 0x20000 bytes for rid 0x10 type 3 at 0xf0500000 em0: attempting to allocate 1 MSI vectors (1 supported) msi: routing MSI IRQ 256 to vector 49 em0: using IRQ 256 for MSI em0: Reserved 0x1000 bytes for rid 0x14 type 3 at 0xf0524000 em0: bpf attached em0: Ethernet address: 00:0f:fe:4e:78:16 em0: [FILTER] Just after boot I see: em0: Link is up 1000 Mbps Full Duplex em0: link state changed to UP em0: Link is Down em0: link state changed to DOWN em0: Link is up 100 Mbps Full Duplex em0: link state changed to UP So the problem is not with bge0, but em0. -- Best Wishes, Stefan Lambrev ICQ# 24134177
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?465857B1.1080104>