tcp: cubic update for net-2.6.22
Subsystems (3)
A message that is not in the archive (sent off-list or not yet received): 00e401c76d51$79a31d20$4a580e98@ncsu2cc0c3fa00
A message that is not in the archive (sent off-list or not yet received): 20070323091720.575f9162@freekitty
A message that is not in the archive (sent off-list or not yet received): [email protected]
PATCH 1/1 tcp: cubic update for net-2.6.22
The following update received from Injong updates TCP cubic to the latest version. I am running more complete tests and will have results after 4/1.
According to Injong: the new version improves on its scalability, fairness and stability. So in all properties, we confirmed it shows better performance.
NCSU results (for 2.6.18 and 2.6.20) available:
http://netsrv.csc.ncsu.edu/wiki/index.php/TCP_Testing
This version is described in a new Internet draft for CUBIC.
http://www.ietf.org/internet-drafts/draft-rhee-tcp-cubic-00.txt
--- net/ipv4/tcp_cubic.c | 8 +++++--- 1 files changed, 5 insertions(+), 3 deletions(-)
net/ipv4/tcp_cubic.c +5 -3
@@ -1,5 +1,5 @@/* -* TCP CUBIC: Binary Increase Congestion control for TCP v2.0 +* TCP CUBIC: Binary Increase Congestion control for TCP v2.1 * * This is from the implementation of CUBIC TCP in * Injong Rhee, Lisong Xu.
@@ -214,7 +214,9 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)if (ca->delay_min >0 ) {/* max increment = Smax * rtt / 0.1 */ min_cnt = (cwnd *HZ *8 )/(10 * max_increment * ca->delay_min); -if (ca->cnt < min_cnt) + +/* use concave growth when the target is above the origin */ +if (ca->cnt < min_cnt && t >= ca->bic_K) ca->cnt = min_cnt; }
@@ -400,4 +402,4 @@ module_exit(cubictcp_unregister);MODULE_AUTHOR ("Sangtae Ha, Stephen Hemminger" );MODULE_LICENSE ("GPL" );MODULE_DESCRIPTION ("CUBIC TCP" ); -MODULE_VERSION ("2.0" ); +MODULE_VERSION ("2.1" );
Signature · 1 line
1.4.4.2
From: Stephen Hemminger <[email protected]>
Date: Fri, 23 Mar 2007 13:38:57 -0700
Quoted 7 lines
The following update received from Injong updates TCP cubic to the latest version. I am running more complete tests and will have results after 4/1.
According to Injong: the new version improves on its scalability, fairness and stability. So in all properties, we confirmed it shows better performance.
Quoted 8 lines
NCSU results (for 2.6.18 and 2.6.20) available:
http://netsrv.csc.ncsu.edu/wiki/index.php/TCP_TestingThis version is described in a new Internet draft for CUBIC.
http://www.ietf.org/internet-drafts/draft-rhee-tcp-cubic-00.txtSigned-off-by Stephen Hemminger <[email protected]>
Applied, thanks Stephen.