Harald,
>
> As for e1000 and or generic TX path changes, I don't have the time to
> review them now, sorry :( That's why I posted it to netdev, to let
> people who have an idea about the committed changes know that there is
> an issue.
>
What version of e1000 you are using?
If you are using e1000 5.7.6-k2 and haven't tried older versions, Could
you please try e1000 version 5.6.10.1-k2 (available in kernel 2.6.11)
and inform us the result?
Thanks,
Ayyappan V.
> -----Original Message-----
> From: netdev-bounce@xxxxxxxxxxx [mailto:netdev-bounce@xxxxxxxxxxx] On
> Behalf Of Harald Welte
> Sent: Sunday, April 03, 2005 10:27 PM
> To: Robert Olsson
> Cc: netdev@xxxxxxxxxxx
> Subject: Re: pktgen problem (skb refcount) in 2.6.12-rc1
>
> On Sun, Apr 03, 2005 at 09:18:30PM +0200, Robert Olsson wrote:
>
> > > I've tried to track the problem down, and I've confirmed that
skb-
> >users
> > > never goes down to 1 but instead stays at '2'.
> >
> > > The same system with the same pktgen script works fine with
2.6.11.6.
> > >
> > > I'm reporting this since it seems like it sounds like we have a
skb
> > > usage count leak somewhere :(
> >
> > Sounds like a diff could give some clues. pktgen, e1000 and TX-path
> should
> > be interesting as ev. changes in kernel config.
>
> no changes in kernel config. I've reviewed pktgen changes and
couldn't
> find something that would cause the problem. It always only
> atomic_inc'ed the ussage cound (and decrements only in error path)
which
> is perfectly fine.
>
> As for e1000 and or generic TX path changes, I don't have the time to
> review them now, sorry :( That's why I posted it to netdev, to let
> people who have an idea about the committed changes know that there is
> an issue.
>
> Cheers,
> Harald
> --
> - Harald Welte <laforge@xxxxxxxxxxxx>
> http://gnumonks.org/
>
========================================================================
==
> ==
> "Privacy in residential applications is a desirable marketing option."
> (ETSI EN 300 175-7
Ch.
> A6)
|