[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Network died completely in 2.6.29 - also on m68k :)



2009/4/16 Kolbjørn Barmen <linux-m68k@kolla.no>:
> After getting my ebuilds going, I compiled and installed 2.6.29 first on
> aranym, where it works just fine, and then on one of my A1200s. On the
> A1200 it seemed to work fine, for a while that is, suddenly the network
> stopped up - completely. Nothing in dmesg, but rmmod/modprobe apne does
> not make any difference, and I cant even ping 127.0.0.1 - so I guess this
> is the infamous 2.6.29 bug that was fixed in 2.6.29.1, and that is
> described here http://lkml.org/lkml/2009/3/27/421
>
> So, I tryed to pull out a m68k-v2.6.29.1 from git, but appearantly there is no
> such thing. So¸ it is probably to add the "Disable GRO on legacy netif_rx
> path" patch (in posting above) to the v2.6.29 branch, no?

Usually I do not track stable myself, that's a distributor's issue ;-)
But I gave it a try, and did:

    git remote add stable
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6-stable.git
    git remote update
    git merge v2.6.29.1

(git merge stable/linux-2.6.29.y should have worked, too)
and pushed the result to m68k-v2.6.29. However, I don't know if I
should keep on doing
this for each stable release, as you can easily do it yourself,
without having to wait for me.

Furthermore, it complicates the process of extracting patches for e.g.
the Debian kernel,
unless I would rebase the m68k-v2.6.29 branch on top of the latest
stable version.
But I prefer not to rebase long-lived branches like the stable branch.

What do people think?

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


Reply to: