stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Simon Horman <horms@verge.net.au>,
	"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 3.10 37/62] net: Loosen constraints for recalculating checksum in skb_segment()
Date: Mon, 13 Jan 2014 16:27:02 -0800	[thread overview]
Message-ID: <20140114002711.518466117@linuxfoundation.org> (raw)
In-Reply-To: <20140114002710.464561569@linuxfoundation.org>

3.10-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Simon Horman <horms@verge.net.au>

[ Upstream commit 1cdbcb7957cf9e5f841dbcde9b38fd18a804208b ]

This is a generic solution to resolve a specific problem that I have observed.

If the encapsulation of an skb changes then ability to offload checksums
may also change. In particular it may be necessary to perform checksumming
in software.

An example of such a case is where a non-GRE packet is received but
is to be encapsulated and transmitted as GRE.

Another example relates to my proposed support for for packets
that are non-MPLS when received but MPLS when transmitted.

The cost of this change is that the value of the csum variable may be
checked when it previously was not. In the case where the csum variable is
true this is pure overhead. In the case where the csum variable is false it
leads to software checksumming, which I believe also leads to correct
checksums in transmitted packets for the cases described above.

Further analysis:

This patch relies on the return value of can_checksum_protocol()
being correct and in turn the return value of skb_network_protocol(),
used to provide the protocol parameter of can_checksum_protocol(),
being correct. It also relies on the features passed to skb_segment()
and in turn to can_checksum_protocol() being correct.

I believe that this problem has not been observed for VLANs because it
appears that almost all drivers, the exception being xgbe, set
vlan_features such that that the checksum offload support for VLAN packets
is greater than or equal to that of non-VLAN packets.

I wonder if the code in xgbe may be an oversight and the hardware does
support checksumming of VLAN packets.  If so it may be worth updating the
vlan_features of the driver as this patch will force such checksums to be
performed in software rather than hardware.

Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 net/core/skbuff.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -2854,7 +2854,7 @@ struct sk_buff *skb_segment(struct sk_bu
 						 doffset + tnl_hlen);
 
 		if (fskb != skb_shinfo(skb)->frag_list)
-			continue;
+			goto perform_csum_check;
 
 		if (!sg) {
 			nskb->ip_summed = CHECKSUM_NONE;
@@ -2918,6 +2918,7 @@ skip_fraglist:
 		nskb->len += nskb->data_len;
 		nskb->truesize += nskb->data_len;
 
+perform_csum_check:
 		if (!csum) {
 			nskb->csum = skb_checksum(nskb, doffset,
 						  nskb->len - doffset, 0);



  parent reply	other threads:[~2014-01-14  0:27 UTC|newest]

Thread overview: 64+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-01-14  0:26 [PATCH 3.10 00/62] 3.10.27-stable review Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 01/62] irqchip: renesas-irqc: Fix irqc_probe error handling Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 02/62] clocksource: em_sti: Set cpu_possible_mask to fix SMP broadcast Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 03/62] gpio-rcar: R-Car GPIO IRQ share interrupt Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 04/62] HID: Revert "Revert "HID: Fix logitech-dj: missing Unifying device issue"" Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 05/62] IPv6: Fixed support for blackhole and prohibit routes Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 06/62] net: do not pretend FRAGLIST support Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 07/62] rds: prevent BUG_ON triggered on congestion update to loopback Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 08/62] macvtap: Do not double-count received packets Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 09/62] macvtap: update file current position Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 10/62] tun: " Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 11/62] macvtap: signal truncated packets Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 12/62] virtio: delete napi structures from netdev before releasing memory Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 13/62] packet: fix send path when running with proto == 0 Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 14/62] ipv6: dont count addrconf generated routes against gc limit Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 15/62] net: drop_monitor: fix the value of maxattr Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 16/62] net: unix: allow set_peek_off to fail Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 17/62] tg3: Initialize REG_BASE_ADDR at PCI config offset 120 to 0 Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 18/62] netvsc: dont flush peers notifying work during setting mtu Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 19/62] ipv6: fix illegal mac_header comparison on 32bit Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 20/62] net: unix: allow bind to fail on mutex lock Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 22/62] net: inet_diag: zero out uninitialized idiag_{src,dst} fields Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 23/62] drivers/net/hamradio: Integer overflow in hdlcdrv_ioctl() Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 25/62] net: fec: fix potential use after free Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 26/62] ipv6: always set the new created dsts from in ip6_rt_copy Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 27/62] rds: prevent dereference of a NULL device Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 28/62] net: rose: restore old recvmsg behavior Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 29/62] vlan: Fix header ops passthru when doing TX VLAN offload Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 30/62] virtio_net: fix error handling for mergeable buffers Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 31/62] virtio-net: make all RX paths handle errors consistently Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 32/62] virtio_net: dont leak memory or block when too many frags Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 33/62] virtio-net: fix refill races during restore Greg Kroah-Hartman
2014-01-14  0:26 ` [PATCH 3.10 34/62] net: llc: fix use after free in llc_ui_recvmsg Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 35/62] netpoll: Fix missing TXQ unlock and and OOPS Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 36/62] bridge: use spin_lock_bh() in br_multicast_set_hash_max Greg Kroah-Hartman
2014-01-14  0:27 ` Greg Kroah-Hartman [this message]
2014-01-14  0:27 ` [PATCH 3.10 38/62] ARM: fix footbridge clockevent device Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 39/62] ARM: fix "bad mode in ... handler" message for undefined instructions Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 40/62] ARM: dts: exynos5250: Fix MDMA0 clock number Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 41/62] ARM: shmobile: kzm9g: Fix coherent DMA mask Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 42/62] ARM: shmobile: armadillo: " Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 43/62] ARM: shmobile: mackerel: " Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 45/62] parisc: Ensure full cache coherency for kmap/kunmap Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 46/62] ahci: add PCI ID for Marvell 88SE9170 SATA controller Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 47/62] clk: clk-divider: fix divisor > 255 bug Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 48/62] clk: samsung: exynos4: Correct SRC_MFC register Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 49/62] clk: samsung: exynos5250: Add CLK_IGNORE_UNUSED flag for the sysreg clock Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 50/62] clk: exynos5250: fix sysmmu_mfc{l,r} gate clocks Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 51/62] mfd: rtsx_pcr: Disable interrupts before cancelling delayed works Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 52/62] ACPI / TPM: fix memory leak when walking ACPI namespace Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 53/62] ACPI / Battery: Add a _BIX quirk for NEC LZ750/LS Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 54/62] mac80211: move "bufferable MMPDU" check to fix AP mode scan Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 55/62] intel_pstate: Add X86_FEATURE_APERFMPERF to cpu match parameters Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 56/62] SCSI: sd: Reduce buffer size for vpd request Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 57/62] netfilter: nf_nat: fix access to uninitialized buffer in IRC NAT helper Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 58/62] x86, fpu, amd: Clear exceptions in AMD FXSAVE workaround Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 59/62] sched: Fix race on toggling cfs_bandwidth_used Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 60/62] sched: Fix cfs_bandwidth misuse of hrtimer_expires_remaining Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 61/62] sched: Fix hrtimer_cancel()/rq->lock deadlock Greg Kroah-Hartman
2014-01-14  0:27 ` [PATCH 3.10 62/62] sched: Guarantee new group-entities always have weight Greg Kroah-Hartman
2014-01-14  3:02 ` [PATCH 3.10 00/62] 3.10.27-stable review Guenter Roeck
2014-01-14 23:12   ` Greg Kroah-Hartman
2014-01-14 19:30 ` Shuah Khan
2014-01-14 23:12   ` Greg Kroah-Hartman

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=20140114002711.518466117@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=davem@davemloft.net \
    --cc=horms@verge.net.au \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@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).