From: Sridhar Samudrala <sri@us.ibm.com>
To: "Brandeburg, Jesse" <jesse.brandeburg@intel.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>,
"Kirsher, Jeffrey T" <jeffrey.t.kirsher@intel.com>,
netdev <netdev@vger.kernel.org>
Subject: Re: [PATCH net-next-2.6] IFix IPv6 GSO type checks in Intel ethernet drivers
Date: Tue, 19 Jan 2010 14:36:17 -0800 [thread overview]
Message-ID: <1263940577.1635.29.camel@w-sridhar.beaverton.ibm.com> (raw)
In-Reply-To: <alpine.WNT.2.00.1001191306450.5776@jbrandeb-desk1.amr.corp.intel.com>
On Tue, 2010-01-19 at 13:11 -0800, Brandeburg, Jesse wrote:
>
> On Tue, 19 Jan 2010, Sridhar Samudrala wrote:
>
> > Found this problem when testing IPv6 from a KVM guest to a remote
> > host via e1000e device on the host.
> > The following patch fixes the check for IPv6 GSO packet in Intel
> > ethernet drivers to use skb_is_gso_v6(). SKB_GSO_DODGY is also set
> > when packets are forwarded from a guest.
> >
> > Signed-off-by: Sridhar Samudrala <sri@us.ibm.com
>
> Looks fine, thanks! The current code in net-next and 2.6.32 both have
> exactly the same condition in skb_is_gso_v6, so I'm not sure that this
> patch alone will fix any issues, FYI. If this patch is part of another
> set or dependent upon another for the actual change in behavior mentioned
> in the comment, I think it should be noted or sent in a series.
The bug is in comparing the flag SKB_GSO_TCPV6 directly. I replaced the
compare with the call to skb_is_gso_v6.
This is not a regression and the bug is present even in 2.6.32 and
earlier kernels. So it would be good if this patch can go into 2.6.32
and also -stable kernels.
Thanks
Sridhar
> Ack-ed by: Jesse Brandeburg <jesse.brandeburg@intel.com>
>
> >
> > diff --git a/drivers/net/e1000e/netdev.c b/drivers/net/e1000e/netdev.c
> > index 3d57ca5..2353e95 100644
> > --- a/drivers/net/e1000e/netdev.c
> > +++ b/drivers/net/e1000e/netdev.c
> > @@ -3771,7 +3771,7 @@ static int e1000_tso(struct e1000_adapter *adapter,
> > 0, IPPROTO_TCP, 0);
> > cmd_length = E1000_TXD_CMD_IP;
> > ipcse = skb_transport_offset(skb) - 1;
> > - } else if (skb_shinfo(skb)->gso_type == SKB_GSO_TCPV6) {
> > + } else if (skb_is_gso_v6(skb)) {
> > ipv6_hdr(skb)->payload_len = 0;
> > tcp_hdr(skb)->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
> > &ipv6_hdr(skb)->daddr,
> > diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
> > index d967949..d4bbdf0 100644
> > --- a/drivers/net/igb/igb_main.c
> > +++ b/drivers/net/igb/igb_main.c
> > @@ -3422,7 +3422,7 @@ static inline int igb_tso_adv(struct igb_ring *tx_ring,
> > iph->daddr, 0,
> > IPPROTO_TCP,
> > 0);
> > - } else if (skb_shinfo(skb)->gso_type == SKB_GSO_TCPV6) {
> > + } else if (skb_is_gso_v6(skb)) {
> > ipv6_hdr(skb)->payload_len = 0;
> > tcp_hdr(skb)->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
> > &ipv6_hdr(skb)->daddr,
> > diff --git a/drivers/net/igbvf/netdev.c b/drivers/net/igbvf/netdev.c
> > index a6c3920..ce33671 100644
> > --- a/drivers/net/igbvf/netdev.c
> > +++ b/drivers/net/igbvf/netdev.c
> > @@ -1963,7 +1963,7 @@ static int igbvf_tso(struct igbvf_adapter *adapter,
> > iph->daddr, 0,
> > IPPROTO_TCP,
> > 0);
> > - } else if (skb_shinfo(skb)->gso_type == SKB_GSO_TCPV6) {
> > + } else if (skb_is_gso_v6(skb)) {
> > ipv6_hdr(skb)->payload_len = 0;
> > tcp_hdr(skb)->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
> > &ipv6_hdr(skb)->daddr,
> > diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
> > index 81971ed..b2c0025 100644
> > --- a/drivers/net/ixgbe/ixgbe_main.c
> > +++ b/drivers/net/ixgbe/ixgbe_main.c
> > @@ -5113,7 +5113,7 @@ static int ixgbe_tso(struct ixgbe_adapter *adapter,
> > iph->daddr, 0,
> > IPPROTO_TCP,
> > 0);
> > - } else if (skb_shinfo(skb)->gso_type == SKB_GSO_TCPV6) {
> > + } else if (skb_is_gso_v6(skb)) {
> > ipv6_hdr(skb)->payload_len = 0;
> > tcp_hdr(skb)->check =
> > ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
> > diff --git a/drivers/net/ixgbevf/ixgbevf_main.c b/drivers/net/ixgbevf/ixgbevf_main.c
> > index 39544af..652b6f4 100644
> > --- a/drivers/net/ixgbevf/ixgbevf_main.c
> > +++ b/drivers/net/ixgbevf/ixgbevf_main.c
> > @@ -2756,7 +2756,7 @@ static int ixgbevf_tso(struct ixgbevf_adapter *adapter,
> > IPPROTO_TCP,
> > 0);
> > adapter->hw_tso_ctxt++;
> > - } else if (skb_shinfo(skb)->gso_type == SKB_GSO_TCPV6) {
> > + } else if (skb_is_gso_v6(skb)) {
> > ipv6_hdr(skb)->payload_len = 0;
> > tcp_hdr(skb)->check =
> > ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
> >
> >
> >
next prev parent reply other threads:[~2010-01-19 22:36 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-01-19 21:02 [PATCH net-next-2.6] IFix IPv6 GSO type checks in Intel ethernet drivers Sridhar Samudrala
2010-01-19 21:11 ` Brandeburg, Jesse
2010-01-19 22:14 ` David Miller
2010-01-19 23:15 ` Jeff Kirsher
2010-01-19 22:36 ` Sridhar Samudrala [this message]
2010-01-19 22:39 ` Herbert Xu
2010-01-20 0:44 ` Equal cost multipath for IPv6? Templin, Fred L
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1263940577.1635.29.camel@w-sridhar.beaverton.ibm.com \
--to=sri@us.ibm.com \
--cc=herbert@gondor.apana.org.au \
--cc=jeffrey.t.kirsher@intel.com \
--cc=jesse.brandeburg@intel.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).