RE: e1000 issue on DQ965GF board (was 24 lost ticks with 2.6.20.10 kernel)

Yes, this was discovered through code inspection last week and I've
already made a fix (as I mentioned to Auke off-thread). It should be
submitted upstream shortly.

-----Original Message-----
From: Kok, Auke [mailto:auke-jan.h.kok@xxxxxxxxx]
Sent: Thursday, May 03, 2007 8:37 AM
To: Michel Lespinasse
Cc: Kok, Auke-jan H; Chuck Ebbert; linux-kernel@xxxxxxxxxxxxxxx; Dave
Jones; cramerj; Ronciak, John; Brandeburg, Jesse; Kirsher, Jeffrey T;
Allan, Bruce W
Subject: Re: e1000 issue on DQ965GF board (was 24 lost ticks with
2.6.20.10 kernel)

[Adding Bruce to the Cc, reply below]

Michel Lespinasse wrote:
On Wed, May 02, 2007 at 11:14:52AM -0700, Kok, Auke wrote:
I just checked and the fix I was referring to earlier didn't make it
into
2.6.21-final. You can get 2.6.21-git1 from kernel.org which has the
fix. See


http://www.kernel.org/pub/linux/kernel/v2.6/snapshots/patch-2.6.21-git1.
log

Good. So I tried that patch (well, actually only the change visible at
http://lkml.org/lkml/2007/3/6/268). I patched it into a 2.6.20.11
kernel,
using the same config file as previously. The good news is that this
fixes
my issue: there are no lost ticks anymore, and the link does negotiate
a
gigabit connection. This is a great improvement for me :)

I still seem to hit an issue if using the ethtool command, though.
when using 'ethtool -s eth0 autoneg on', the link comes up at gigabit
speed,
but a couple seconds later is comes down again, and then up at 100
megabits:

e1000: eth0: e1000_watchdog: NIC Link is Up 1000 Mbps Full Duplex
e1000: eth0: e1000_watchdog: NIC Link is Down
e1000: eth0: e1000_watchdog: NIC Link is Up 100 Mbps Full Duplex
e1000: eth0: e1000_watchdog: 10/100 speed: disabling TSO

The same thing happens if I use ifdown eth0; ifup eth0 too. Once
again,
I only observe this on my DQ965GF motherboard, the DG965RY board is
fine
(stays at gigabit speed when I issue these commands).

Is this something you could easily reproduce at Intel or would you
want me
to look into that issue on my system ?

I think we have confirmed this issue and I'll work with Bruce on seeing
where
the fix went. I might give you another patch to try.

Auke
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Relevant Pages

  • Re: e1000 issue on DQ965GF board (was 24 lost ticks with 2.6.20.10 kernel)
    ... I still seem to hit an issue if using the ethtool command, ... The same thing happens if I use ifdown eth0; ... I think we have confirmed this issue and I'll work with Bruce on seeing where the fix went. ...
    (Linux-Kernel)
  • Re: e1000 issue on DQ965GF board (was 24 lost ticks with 2.6.20.10 kernel)
    ... You can get 2.6.21-git1 from kernel.org which has the fix. ... I still seem to hit an issue if using the ethtool command, ... The same thing happens if I use ifdown eth0; ... That fixes this issue for us in our labs. ...
    (Linux-Kernel)
  • e1000 issue on DQ965GF board (was 24 lost ticks with 2.6.20.10 kernel)
    ... You can get 2.6.21-git1 from kernel.org which has the fix. ... So I tried that patch (well, actually only the change visible at ... I still seem to hit an issue if using the ethtool command, ... The same thing happens if I use ifdown eth0; ...
    (Linux-Kernel)
  • Re: Messed up my ISP/Networkmanager connection !?
    ... This overrides what ever your router is sending out as ... part of the DHCP information. ... Did as Bruce suggested. ... ]# ifdown eth0 gave me a SELinux warning in permissive mode ...
    (Fedora)
  • Re: Tracks Poll
    ... and in fact sounded much as though Bruce had written ... think if say, max is drumming is just to shitty on a particular show, ... then the show would simply not be considered, rather than trying to fix ... word word word. ...
    (rec.music.artists.springsteen)