Following the performance discussions, I would still
like to see this integrated.
By using the default value (1 second), we have the same
nofeedback timer behaviour as in 2.6.19.
The patch is the same, but I
* have reworded the configuration dialog / commit message and
comments, to make it clearer [cf. Lars' email]
* replaced multiplication by equivalent bit-shift operations
Both patches have been uploaded as 05a/05b, the former alredy
has Ian's signed-off line.
---------------------> The Patch <----------------------------
[DCCP]: Use higher RTO default for CCID3
The TFRC nofeedback timer normally expires after the maximum of 4
RTTs and twice the current send interval (RFC 3448, 4.3). On LANs
with a small RTT this can mean a high processing load and reduced
performance, since then the nofeedback timer is triggered very
frequently. As a result, the sending rate quickly converges towards
zero.
This patch provides a configuration option to set the bound for the
nofeedback timer, using as default the TCP lower bound of 1 second.
By setting the configuration option to 0, strict RFC 3448 behaviour
can be enforced for the nofeedback timer.
Signed-off-by: Gerrit Renker <gerrit@xxxxxxxxxxxxxx>
---
net/dccp/ccids/Kconfig | 24 ++++++++++++++++++++++++
net/dccp/ccids/ccid3.c | 20 +++++++++++---------
2 files changed, 35 insertions(+), 9 deletions(-)
--- a/net/dccp/ccids/Kconfig
+++ b/net/dccp/ccids/Kconfig
@@ -89,4 +89,28 @@ config IP_DCCP_CCID3_DEBUG
parameter to 0 or 1.
If in doubt, say N.
+
+config IP_DCCP_CCID3_RTO
+ int "Use higher bound for nofeedback timer"
+ default 1
+ depends on IP_DCCP_CCID3
+ ---help---
+ Use higher lower bound for nofeedback timer expiration.
+
+ The TFRC nofeedback timer normally expires after the maximum of 4
+ RTTs and twice the current send interval (RFC 3448, 4.3). On LANs
+ with a small RTT this can mean a high processing load and reduced
+ performance, since then the nofeedback timer is triggered very
+ frequently.
+
+ This option enables to set a higher lower bound for the nofeedback
+ value. Values in units of seconds can be set here.
+
+ A value of 0 disables this feature by enforcing the value specified
+ in RFC 3448. It is recommended to use the default value of 1 second,
+ as this is also the suggested lower bound for TCP (RFC 2988, 2.4).
+
+ Using a value of 2 seconds corresponds to the initial timeout (RFC
+ 3448, 4.2); higher values than this should not normally be necessary.
+
endmenu
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -245,9 +245,10 @@ static void ccid3_hc_tx_no_feedback_time
}
/*
* Schedule no feedback timer to expire in
- * max(4 * R, 2 * s/X) = max(4 * R, 2 * t_ipi)
+ * max(t_RTO, 2 * s/X) = max(t_RTO, 2 * t_ipi)
+ * See comments above regarding the value of t_RTO.
*/
- t_nfb = max(4 * hctx->ccid3hctx_rtt, 2 * hctx->ccid3hctx_t_ipi);
+ t_nfb = max(hctx->ccid3hctx_t_rto, hctx->ccid3hctx_t_ipi << 1);
break;
case TFRC_SSTATE_NO_SENT:
DCCP_BUG("Illegal %s state NO_SENT, sk=%p", dccp_role(sk), sk);
@@ -512,16 +513,17 @@ static void ccid3_hc_tx_packet_recv(stru
*/
sk->sk_write_space(sk);
- /* Update timeout interval. We use the alternative variant of
- * [RFC 3448, 3.1] which sets the upper bound of t_rto to one
- * second, as it is suggested for TCP (see RFC 2988, 2.4). */
- hctx->ccid3hctx_t_rto = max_t(u32, 4 * hctx->ccid3hctx_rtt,
- USEC_PER_SEC );
+ /* Update timeout interval for the nofeedback timer:
+ * A value of IP_DCCP_CCID3_RTO=0 lets the nofeedback timer
+ * expire as per [RFC 3448, 4]; a value of 1 second corresponds
+ * to the lower bound suggested for TCP (see RFC 2988, 2.4). */
+ hctx->ccid3hctx_t_rto = max_t(u32, hctx->ccid3hctx_rtt << 2,
+ CONFIG_IP_DCCP_CCID3_RTO * USEC_PER_SEC);
/*
* Schedule no feedback timer to expire in
- * max(4 * R, 2 * s/X) = max(4 * R, 2 * t_ipi)
+ * max(t_RTO, 2 * s/X) = max(t_RTO, 2 * t_ipi)
*/
- t_nfb = max(4 * hctx->ccid3hctx_rtt, 2 * hctx->ccid3hctx_t_ipi);
+ t_nfb = max(hctx->ccid3hctx_t_rto, hctx->ccid3hctx_t_ipi << 1);
ccid3_pr_debug("%s, sk=%p, Scheduled no feedback timer to "
"expire in %lu jiffies (%luus)\n",
