Re: broken re(4)
- From: Gerrit Kühn <gerrit@xxxxxxxxxxxxxxxxxxx>
- Date: Fri, 30 May 2008 14:45:10 +0200
On Fri, 30 May 2008 13:44:43 +0200 (CEST) Oliver Fromme
<olli@xxxxxxxxxxxxxxxxx> wrote about Re: broken re(4):
OF> > Talking about this controller: This is also the only board I am
OF> > using with PCI cards (and thus with a PCI riser) at all. I remember
OF> > vaguely that I had a few problems getting the controller to work in
OF> > the riser card when it put the system together. The riser has two
OF> > ports, and the controller would only work in the upper one afaicr.
OF> That rings a bell ...
OF> I remember reports of riser cards that apparently changed
OF> the timing on the PCI bus so they were only marginally
OF> compliant with the spec, or maybe not even that anymore.
Meanwhile I played through all combinations, and guess what?
I can get rid of the problems in two ways:
1. Taking out the controller card. Without the controller, everthing works
fine. The riser is still in.
2. Turn of rxcsum and txcsum. The driver can only turn them on or off
together. They're on by default, and turning them off also lets the
problems vanish.
OF> If you try to remove the controller, please also remove
OF> the riser card. It could well be that it's causing
OF> problems, especially if it's on the same PCI bus as the
OF> onboard re(4) interfaces.
Somehow the controller seems to interfer with rxc/txc.
OF> There are various kinds of riser cards (passive, active,
OF> with fixed IRQs, or with jumpers, and so on). For a
OF> related discussion see this one, for example:
OF> http://lkml.org/lkml/2007/6/24/39
OF> Typing "pci riser card jumper" in Google will give you
OF> many more pages with interesting (or frightening) stuff
OF> to read.
Thanks for the hint, now I have something to read over the weekend. ;-)
For now, I am happy that I have a working system (with controller ands
disks inside). It seems to me that the transfer rates with rxc/txc turned
off are a bit lower, but it is much more important to me that the
networking runs reliably.
cu
Gerrit
_______________________________________________
freebsd-stable@xxxxxxxxxxx mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to "freebsd-stable-unsubscribe@xxxxxxxxxxx"
- References:
- Re: broken re(4)
- From: Gerrit Kühn
- Re: broken re(4)
- From: Oliver Fromme
- Re: broken re(4)
- Prev by Date: Re: RELENG_7 amd64; memory and vm.kmem_size
- Next by Date: Re: broken re(4)
- Previous by thread: Re: broken re(4)
- Next by thread: Re: broken re(4)
- Index(es):
Relevant Pages
|
