VMWare if_em breakage (was: Re: svn commit: r194865 - in head/sys: dev/e1000 modules/igb)

Jack Vogel jfvogel at gmail.com
Fri Jun 26 03:03:31 UTC 2009

The problem Robert is talking about is new, its NOT the issue with the
invalid mac address, in fact, that problem was fixed with this checkin.

It now seems to be something else, I have gotten access to a system and
installed Fusion this afternoon, I hope to repro and fix this tomorrow.

It would be of help to me if someone with Virtual Box could try the tip and
see if they see any problems.

Note that from the testing I have done there does not seem to be a problem
on real hardware, its something that is only effecting virtualized.



On Thu, Jun 25, 2009 at 7:32 PM, Daniel O'Connor <doconnor at gsoft.com.au>wrote:

> On Thu, 25 Jun 2009, Robert Watson wrote:
> > On Wed, 24 Jun 2009, Jack F Vogel wrote:
> > > Log:
> > >  Updates for both the em and igb drivers, add support
> > >  for multiqueue tx, shared code updates, new device
> > >  support, and some bug fixes.
> >
> > Since this change (and the two followups), I'm no longer able to use
> > if_em reliable in VMWare Fusion.  I get a bit of traffic, and then
> > interrupts cease firing from the (virtual) card.  If I lower the
> > interface and raise it, it will recover for a short period, and then
> > the same result.  I'm running VMWare 2.0.5, which is the latest
> > version available; if I locally back out the update changes, then it
> > works fine so it seems unlikely to be other on-going network stack
> > work.  I realize that VMWare's driver is probably not part of your
> > official test suite (not actually being a product manafactured by
> > Intel :-), but keeping VMWare working is pretty important from a
> > FreeBSD perspective!
> Not sure it was this commit but Virtual Box also has issues with em
> in -current. em complains the MAC address is invalid and refuses to
> attach.
> --
> Daniel O'Connor software and network engineer
> for Genesis Software - http://www.gsoft.com.au
> "The nice thing about standards is that there
> are so many of them to choose from."
>  -- Andrew Tanenbaum
> GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C

More information about the freebsd-current mailing list