On Wed, 28 Feb 2007, Jarek Poplawski wrote:

On 27-02-2007 16:50, Ilpo Järvinen wrote:

quoted

New sysctl tcp_frto_response is added to select amongst these

...

quoted

Signed-off-by: Ilpo Järvinen <redacted>

@@ -762,15 +763,17 @@ __u32 tcp_init_cwnd(struct tcp_sock *tp,
 }
 
 /* Set slow start threshold and cwnd not falling to slow start */
-void tcp_enter_cwr(struct sock *sk)
+void tcp_enter_cwr(struct sock *sk, const int set_ssthresh)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
+	const struct inet_connection_sock *icsk = inet_csk(sk);
 
 	tp->prior_ssthresh = 0;
 	tp->bytes_acked = 0;
 	if (inet_csk(sk)->icsk_ca_state < TCP_CA_CWR) {

-	if (inet_csk(sk)->icsk_ca_state < TCP_CA_CWR) {
+	if (icsk->icsk_ca_state < TCP_CA_CWR) {

Probably something for the next "BTW".

These are going to 2.6.22, not to 2.6.21, see:
  http://marc.theaimsgroup.com/?l=linux-netdev&m=117213215924406&w=2
...or do you mean something else?

Since DaveM has already applied this, here is a patch with your correction 
alone on the top of tcp-2.6.

--
[PATCH] [TCP]: Complete icsk-to-local-variable change (in tcp_enter_cwr)

A local variable for icsk was created but this change was
missing. Spotted by Jarek Poplawski.

Signed-off-by: Ilpo Järvinen <redacted>
---

diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index d6e1776..dc221a3 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c

@@ -770,7 +770,7 @@ void tcp_enter_cwr(struct sock *sk, cons
 
 	tp->prior_ssthresh = 0;
 	tp->bytes_acked = 0;
-	if (inet_csk(sk)->icsk_ca_state < TCP_CA_CWR) {
+	if (icsk->icsk_ca_state < TCP_CA_CWR) {
 		tp->undo_marker = 0;
 		if (set_ssthresh)
 			tp->snd_ssthresh = icsk->icsk_ca_ops->ssthresh(sk);

-- 
1.4.2