linux/net
Gerrit Renker 86739fb96e dccp: Do not let initial option overhead shrink the MPS
This fixes a problem caused by the overlap of the connection-setup and
established-state phases of DCCP connections.

During connection setup, the client retransmits Confirm Feature-Negotiation
options until a response from the server signals that it can move from the
half-established PARTOPEN into the OPEN state, whereupon the connection is
fully established on both ends (RFC 4340, 8.1.5).

However, since the client may already send data while it is in the PARTOPEN
state, consequences arise for the Maximum Packet Size: the problem is that the
initial option overhead is much higher than for the subsequent established
phase, as it involves potentially many variable-length list-type options
(server-priority options, RFC 4340, 6.4).

Applying the standard MPS is insufficient here: especially with larger
payloads this can lead to annoying, counter-intuitive EMSGSIZE errors.

On the other hand, reducing the MPS available for the established phase by
the added initial overhead is highly wasteful and inefficient.

The solution chosen therefore is a two-phase strategy:

   If the payload length of the DataAck in PARTOPEN is too large, an Ack is sent
   to carry the options, and the feature-negotiation list is then flushed.

   This means that the server gets two Acks for one Response. If both Acks get
   lost, it is probably better to restart the connection anyway and devising yet
   another special-case does not seem worth the extra complexity.

The result is a higher utilisation of the available packet space for the data
transmission phase (established state) of a connection.

The patch (over-)estimates the initial overhead to be 32*4 bytes -- commonly
seen values were around 90 bytes for initial feature-negotiation options.

It uses sizeof(u32) to mean "aligned units of 4 bytes".
For consistency, another use of 4-byte alignment is adapted.

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-02 03:07:23 -08:00
..
9p 9p: fix sparse warning: cast adds address space 2009-02-26 23:13:32 -08:00
802 net/802: fix sparse warnings: context imbalance 2009-02-26 23:13:32 -08:00
8021q Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2009-03-01 21:35:16 -08:00
appletalk appletalk: fix warning: format not a string literal and no ... 2009-02-26 23:13:33 -08:00
atm lec: convert to net_device_ops 2009-01-21 14:02:00 -08:00
ax25 ax25: more common return path joining 2009-02-06 23:47:14 -08:00
bluetooth Bluetooth: Remove some pointless conditionals before kfree_skb() 2009-02-27 06:14:49 +01:00
bridge netlink: change nlmsg_notify() return value logic 2009-02-24 23:18:28 -08:00
can can: remove some pointless conditionals before kfree_skb() 2009-02-26 23:07:35 -08:00
core Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2009-03-01 21:35:16 -08:00
dcb DCB: fix kfree(skb) 2009-01-04 17:29:21 -08:00
dccp dccp: Do not let initial option overhead shrink the MPS 2009-03-02 03:07:23 -08:00
decnet decnet: fix sparse warnings: symbol shadows an earlier one 2009-02-26 23:13:35 -08:00
dsa net: replace uses of __constant_{endian} 2009-02-01 00:45:17 -08:00
econet net: replace uses of __constant_{endian} 2009-02-01 00:45:17 -08:00
ethernet eth: Declare an optimized compare_ether_addr_64bits() function 2008-11-23 23:24:32 -08:00
ipv4 tcp: get rid of two unnecessary u16s in TCP skb flags copying 2009-03-02 03:00:17 -08:00
ipv6 Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2009-03-01 21:35:16 -08:00
ipx net: replace uses of __constant_{endian} 2009-02-01 00:45:17 -08:00
irda net: replace uses of __constant_{endian} 2009-02-01 00:45:17 -08:00
iucv iucv: remove some pointless conditionals before kfree_skb() 2009-02-26 23:07:37 -08:00
key af_key: remove some pointless conditionals before kfree_skb() 2009-02-26 23:07:32 -08:00
lapb
llc llc: remove some pointless conditionals before kfree_skb() 2009-02-26 23:07:38 -08:00
mac80211 mac80211: Give it some time to do the TSF sync 2009-02-27 14:53:03 -05:00
netfilter Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2009-02-25 00:02:05 -08:00
netlabel netlabel: Update kernel configuration API 2008-12-31 12:54:11 -05:00
netlink netlink: remove some pointless conditionals before kfree_skb() 2009-02-26 23:07:34 -08:00
netrom netrom: convert to net_device_ops 2009-01-21 14:02:02 -08:00
packet packet: remove some pointless conditionals before kfree_skb() 2009-02-26 23:07:35 -08:00
phonet netlink: change nlmsg_notify() return value logic 2009-02-24 23:18:28 -08:00
rds rds: Fix build on powerpc. 2009-03-02 01:49:28 -08:00
rfkill net/rfkill/rfkill.c: fix unused rfkill_led_trigger() warning 2009-01-04 17:11:24 -08:00
rose rose: convert to network_device_ops 2009-01-21 14:02:04 -08:00
rxrpc RxRPC: Fix a potential NULL dereference 2009-02-06 21:50:52 -08:00
sched Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 2009-03-01 21:35:16 -08:00
sctp sctp: Inherit all socket options from parent correctly. 2009-02-16 00:03:11 -08:00
sunrpc net/sunrpc/xprtsock.c: some common code found 2009-02-06 23:48:33 -08:00
tipc net/tipc/bcast.h: use ARRAY_SIZE 2009-01-11 00:06:33 -08:00
unix unix: remove some pointless conditionals before kfree_skb() 2009-02-26 23:07:34 -08:00
wanrouter wanrouter: fix sparse warnings: context imbalance 2009-02-26 23:13:36 -08:00
wimax Merge branch 'master' of /home/davem/src/GIT/linux-2.6/ 2009-02-14 23:12:00 -08:00
wireless nl80211: Avoid AP mode BUG_ON hang with invalid lock assert 2009-02-27 14:53:04 -05:00
x25 net: replace uses of __constant_{endian} 2009-02-01 00:45:17 -08:00
xfrm Revert "xfrm: For 32/64 compatability wrt. xfrm_usersa_info" 2009-01-20 09:49:51 -08:00
compat.c net: socket infrastructure for SO_TIMESTAMPING 2009-02-15 22:43:35 -08:00
Kconfig RDS: Kconfig and Makefile 2009-02-26 23:43:35 -08:00
Makefile RDS: Kconfig and Makefile 2009-02-26 23:43:35 -08:00
nonet.c
socket.c net: socket infrastructure for SO_TIMESTAMPING 2009-02-15 22:43:35 -08:00
sysctl_net.c missing bits of net-namespace / sysctl 2008-07-27 09:45:34 -07:00
TUNABLE