1
linux/net/ipv6
Eric Dumazet 6746960140 ipv6: RTAX_FEATURE_ALLFRAG causes inefficient TCP segment sizing
Quoting Tore Anderson from :
https://bugzilla.kernel.org/show_bug.cgi?id=42572

When RTAX_FEATURE_ALLFRAG is set on a route, the effective TCP segment
size does not take into account the size of the IPv6 Fragmentation
header that needs to be included in outbound packets, causing every
transmitted TCP segment to be fragmented across two IPv6 packets, the
latter of which will only contain 8 bytes of actual payload.

RTAX_FEATURE_ALLFRAG is typically set on a route in response to
receving a ICMPv6 Packet Too Big message indicating a Path MTU of less
than 1280 bytes. 1280 bytes is the minimum IPv6 MTU, however ICMPv6
PTBs with MTU < 1280 are still valid, in particular when an IPv6
packet is sent to an IPv4 destination through a stateless translator.
Any ICMPv4 Need To Fragment packets originated from the IPv4 part of
the path will be translated to ICMPv6 PTB which may then indicate an
MTU of less than 1280.

The Linux kernel refuses to reduce the effective MTU to anything below
1280 bytes, instead it sets it to exactly 1280 bytes, and
RTAX_FEATURE_ALLFRAG is also set. However, the TCP segment size appears
to be set to 1240 bytes (1280 Path MTU - 40 bytes of IPv6 header),
instead of 1232 (additionally taking into account the 8 bytes required
by the IPv6 Fragmentation extension header).

This in turn results in rather inefficient transmission, as every
transmitted TCP segment now is split in two fragments containing
1232+8 bytes of payload.

After this patch, all the outgoing packets that includes a
Fragmentation header all are "atomic" or "non-fragmented" fragments,
i.e., they both have Offset=0 and More Fragments=0.

With help from David S. Miller

Reported-by: Tore Anderson <tore@fud.no>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Maciej Żenczykowski <maze@google.com>
Cc: Tom Herbert <therbert@google.com>
Tested-by: Tore Anderson <tore@fud.no>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-04-27 00:03:34 -04:00
..
netfilter net: Convert all sysctl registrations to register_net_sysctl 2012-04-20 21:22:30 -04:00
addrconf_core.c net: cleanup unsigned to unsigned int 2012-04-15 12:44:40 -04:00
addrconf.c net ipv6: Convert addrconf to use register_net_sysctl 2012-04-20 21:22:29 -04:00
addrlabel.c
af_inet6.c sock: Introduce named constants for sk_reuse 2012-04-21 15:52:25 -04:00
ah6.c net: remove ipv6_addr_copy() 2011-11-22 16:43:32 -05:00
anycast.c ipv6: Remove never used function inet6_ac_check(). 2012-02-01 16:14:17 -05:00
datagram.c net: cleanup unsigned to unsigned int 2012-04-15 12:44:40 -04:00
esp6.c
exthdrs_core.c net/ipv6/exthdrs_core.c: Checkpatch cleanups 2012-04-02 04:33:46 -04:00
exthdrs.c net/ipv6/exthdrs.c: Strict PadN option checking 2012-04-12 17:36:44 -04:00
fib6_rules.c net/ipv6/fib6_rules.c: Checkpatch cleanup 2012-04-02 04:33:46 -04:00
icmp.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2012-04-10 14:30:45 -04:00
inet6_connection_sock.c tcp: bind() use stronger condition for bind_conflict 2012-04-14 15:28:55 -04:00
inet6_hashtables.c
ip6_fib.c ipv6: fix problem with expired dst cache 2012-04-13 12:58:29 -04:00
ip6_flowlabel.c net: cleanup unsigned to unsigned int 2012-04-15 12:44:40 -04:00
ip6_input.c ipv6: Add fragment reporting to ipv6_skip_exthdr(). 2011-12-03 09:35:10 -08:00
ip6_output.c ipv6: call consume_skb() in frag/reassembly 2012-04-25 20:39:46 -04:00
ip6_tunnel.c ip6_tunnel: dont drop packet but consume it 2012-04-19 14:23:55 -04:00
ip6mr.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2012-04-10 14:30:45 -04:00
ipcomp6.c
ipv6_sockglue.c net: cleanup unsigned to unsigned int 2012-04-15 12:44:40 -04:00
Kconfig
Makefile
mcast.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2012-04-10 14:30:45 -04:00
mip6.c net: remove ipv6_addr_copy() 2011-11-22 16:43:32 -05:00
ndisc.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2012-04-15 13:19:04 -04:00
netfilter.c
proc.c ipv6: fix per device IP snmp counters 2012-01-17 23:56:18 -05:00
protocol.c
raw.c ipv6: Implement IPV6_UNICAST_IF socket option. 2012-02-08 15:52:45 -05:00
reassembly.c ipv6: call consume_skb() in frag/reassembly 2012-04-25 20:39:46 -04:00
route.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2012-04-15 13:19:04 -04:00
sit.c tunnel: implement 64 bits statistics 2012-04-14 14:47:05 -04:00
syncookies.c net: remove ipv6_addr_copy() 2011-11-22 16:43:32 -05:00
sysctl_net_ipv6.c net: Delete all remaining instances of ctl_path 2012-04-20 21:22:30 -04:00
tcp_ipv6.c ipv6: RTAX_FEATURE_ALLFRAG causes inefficient TCP segment sizing 2012-04-27 00:03:34 -04:00
tunnel6.c
udp_impl.h
udp.c net: add a limit parameter to sk_add_backlog() 2012-04-23 22:28:28 -04:00
udplite.c
xfrm6_input.c
xfrm6_mode_beet.c ipsec: be careful of non existing mac headers 2012-02-23 16:50:45 -05:00
xfrm6_mode_ro.c
xfrm6_mode_transport.c
xfrm6_mode_tunnel.c ipsec: be careful of non existing mac headers 2012-02-23 16:50:45 -05:00
xfrm6_output.c xfrm6: remove unneeded NULL check in __xfrm6_output() 2012-02-01 02:52:48 -05:00
xfrm6_policy.c net: Convert all sysctl registrations to register_net_sysctl 2012-04-20 21:22:30 -04:00
xfrm6_state.c net: remove ipv6_addr_copy() 2011-11-22 16:43:32 -05:00
xfrm6_tunnel.c net: cleanup unsigned to unsigned int 2012-04-15 12:44:40 -04:00