ipv6: Should use consistent conditional judgement for ip6 fragment between __ip6_appe...
authorZheng Li <james.z.li@ericsson.com>
Wed, 28 Dec 2016 15:23:46 +0000 (23:23 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 7 Aug 2017 01:59:44 +0000 (18:59 -0700)
commitaeb230318ddbfc4b06f0d4befd755090ec0c4d6f
treeee1fa9eb0910bd2c63318cb66229e89b20590b13
parentea703cb0140bdeed8b5576816d3cbe6cf09c44d5
ipv6: Should use consistent conditional judgement for ip6 fragment between __ip6_append_data and ip6_finish_output

[ Upstream commit e4c5e13aa45c23692e4acf56f0b3533f328199b2 ]

There is an inconsistent conditional judgement between __ip6_append_data
and ip6_finish_output functions, the variable length in __ip6_append_data
just include the length of application's payload and udp6 header, don't
include the length of ipv6 header, but in ip6_finish_output use
(skb->len > ip6_skb_dst_mtu(skb)) as judgement, and skb->len include the
length of ipv6 header.

That causes some particular application's udp6 payloads whose length are
between (MTU - IPv6 Header) and MTU were fragmented by ip6_fragment even
though the rst->dev support UFO feature.

Add the length of ipv6 header to length in __ip6_append_data to keep
consistent conditional judgement as ip6_finish_output for ip6 fragment.

Signed-off-by: Zheng Li <james.z.li@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <alexander.levin@verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
net/ipv6/ip6_output.c