[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20101103224247.GB9869@mcarlson.broadcom.com>
Date: Wed, 3 Nov 2010 15:42:47 -0700
From: "Matt Carlson" <mcarlson@...adcom.com>
To: "Micha?? Miros??aw" <mirq-linux@...e.qmqm.pl>
cc: "Matthew Carlson" <mcarlson@...adcom.com>,
"netdev@...r.kernel.org" <netdev@...r.kernel.org>,
"e1000-devel@...ts.sourceforge.net"
<e1000-devel@...ts.sourceforge.net>,
"Steve Glendinning" <steve.glendinning@...c.com>,
"Greg Kroah-Hartman" <gregkh@...e.de>,
"Rasesh Mody" <rmody@...cade.com>,
"Debashis Dutt" <ddutt@...cade.com>,
"Kristoffer Glembo" <kristoffer@...sler.com>,
"linux-driver@...gic.com" <linux-driver@...gic.com>,
"linux-net-drivers@...arflare.com" <linux-net-drivers@...arflare.com>
Subject: Re: [PATCH 2/4] Ethtool: convert get_sg/set_sg calls to
hw_features flag
On Wed, Nov 03, 2010 at 03:29:10PM -0700, Micha?? Miros??aw wrote:
> On Mon, Nov 01, 2010 at 07:24:38PM -0700, Matt Carlson wrote:
> > On Fri, Oct 29, 2010 at 09:28:26PM -0700, Micha?? Miros??aw wrote:
> > > diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
> > > index 30ccbb6..b07e2d1 100644
> > > --- a/drivers/net/tg3.c
> > > +++ b/drivers/net/tg3.c
> > > @@ -11306,7 +11306,6 @@ static const struct ethtool_ops tg3_ethtool_ops = {
> > > .get_rx_csum = tg3_get_rx_csum,
> > > .set_rx_csum = tg3_set_rx_csum,
> > > .set_tx_csum = tg3_set_tx_csum,
> > > - .set_sg = ethtool_op_set_sg,
> > > .set_tso = tg3_set_tso,
> > > .self_test = tg3_self_test,
> > > .get_strings = tg3_get_strings,
> > > @@ -14681,6 +14680,7 @@ static int __devinit tg3_init_one(struct pci_dev *pdev,
> > > tp->rx_pending = TG3_DEF_RX_RING_PENDING;
> > > tp->rx_jumbo_pending = TG3_DEF_RX_JUMBO_RING_PENDING;
> > >
> > > + dev->hw_features |= NETIF_F_SG;
> > Scatter-gather should not be enabled if TG3_FLAG_BROKEN_CHECKSUMS is set. I
> > would do the following instead:
> >
> > if (!(tp->tg3_flags & TG3_FLAG_BROKEN_CHECKSUMS))
> > dev->hw_features |= NETIF_F_SG;
> >
> > TG3_FLAG_BROKEN_CHECKSUMS is set in tg3_get_invariants(), so this code
> > would need to be placed later than that function call.
>
> This bug is there now, so I'll queue this as all other hints of existent
> bugs that this patch series "uncovers".
How so? This patch would be introducing the bug. From tg3_get_invariants:
if (tp->pci_chip_rev_id == CHIPREV_ID_5700_B0)
tp->tg3_flags |= TG3_FLAG_BROKEN_CHECKSUMS;
else {
unsigned long features = NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_GRO;
tp->tg3_flags |= TG3_FLAG_RX_CHECKSUMS;
if (tp->tg3_flags3 & TG3_FLG3_5755_PLUS)
features |= NETIF_F_IPV6_CSUM;
tp->dev->features |= features;
vlan_features_add(tp->dev, features);
}
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists