[PATCH] tcp: Increase the initial congestion window to 10.

[PATCH] tcp: Increase the initial congestion window to 10.

[David Miller]

Signed-off-by: David S. Miller <davem@davemloft.net>
---

I've left the DCCP code to keep using RFC3390 logic, if they
wish to adopt this change in their code they can do so by
simply deleting the rfc33390_bytes_to_packets() function and
using TCP_INIT_CWND in their assignment.

 include/net/tcp.h      |   12 +++---------
 net/dccp/ccids/ccid2.c |    9 +++++++++
 net/ipv4/tcp_input.c   |    2 +-
 3 files changed, 13 insertions(+), 10 deletions(-)

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 9179111..7118668 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -196,6 +196,9 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
 /* TCP thin-stream limits */
 #define TCP_THIN_LINEAR_RETRIES 6       /* After 6 linear retries, do exp. backoff */
 
+/* TCP initial congestion window */
+#define TCP_INIT_CWND		10
+
 extern struct inet_timewait_death_row tcp_death_row;
 
 /* sysctl variables for tcp */
@@ -799,15 +802,6 @@ static inline __u32 tcp_current_ssthresh(const struct sock *sk)
 /* Use define here intentionally to get WARN_ON location shown at the caller */
 #define tcp_verify_left_out(tp)	WARN_ON(tcp_left_out(tp) > tp->packets_out)
 
-/*
- * Convert RFC 3390 larger initial window into an equivalent number of packets.
- * This is based on the numbers specified in RFC 5681, 3.1.
- */
-static inline u32 rfc3390_bytes_to_packets(const u32 smss)
-{
-	return smss <= 1095 ? 4 : (smss > 2190 ? 2 : 3);
-}
-
 extern void tcp_enter_cwr(struct sock *sk, const int set_ssthresh);
 extern __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst);
 
diff --git a/net/dccp/ccids/ccid2.c b/net/dccp/ccids/ccid2.c
index e96d5e8..fadecd2 100644
--- a/net/dccp/ccids/ccid2.c
+++ b/net/dccp/ccids/ccid2.c
@@ -583,6 +583,15 @@ done:
 	dccp_ackvec_parsed_cleanup(&hc->tx_av_chunks);
 }
 
+/*
+ * Convert RFC 3390 larger initial window into an equivalent number of packets.
+ * This is based on the numbers specified in RFC 5681, 3.1.
+ */
+static inline u32 rfc3390_bytes_to_packets(const u32 smss)
+{
+	return smss <= 1095 ? 4 : (smss > 2190 ? 2 : 3);
+}
+
 static int ccid2_hc_tx_init(struct ccid *ccid, struct sock *sk)
 {
 	struct ccid2_hc_tx_sock *hc = ccid_priv(ccid);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index eb7f82e..2f692ce 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -817,7 +817,7 @@ __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst)
 	__u32 cwnd = (dst ? dst_metric(dst, RTAX_INITCWND) : 0);
 
 	if (!cwnd)
-		cwnd = rfc3390_bytes_to_packets(tp->mss_cache);
+		cwnd = TCP_INIT_CWND;
 	return min_t(__u32, cwnd, tp->snd_cwnd_clamp);
 }
 
-- 
1.7.4

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Eric Dumazet]

Le mercredi 02 février 2011 à 17:07 -0800, David Miller a écrit :
> Signed-off-by: David S. Miller <davem@davemloft.net>
> ---
> 

Hmm, you forgot a Changelog David ;)

I thought Tom and Google guys were preparing a nice one ?

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Nandita Dukkipati]

Acked-by: Nandita Dukkipati <nanditad@google.com>

On Wed, Feb 2, 2011 at 5:07 PM, David Miller <davem@davemloft.net> wrote:
>
> Signed-off-by: David S. Miller <davem@davemloft.net>
> ---
>
> I've left the DCCP code to keep using RFC3390 logic, if they
> wish to adopt this change in their code they can do so by
> simply deleting the rfc33390_bytes_to_packets() function and
> using TCP_INIT_CWND in their assignment.
>
>  include/net/tcp.h      |   12 +++---------
>  net/dccp/ccids/ccid2.c |    9 +++++++++
>  net/ipv4/tcp_input.c   |    2 +-
>  3 files changed, 13 insertions(+), 10 deletions(-)
>
> diff --git a/include/net/tcp.h b/include/net/tcp.h
> index 9179111..7118668 100644
> --- a/include/net/tcp.h
> +++ b/include/net/tcp.h
> @@ -196,6 +196,9 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
>  /* TCP thin-stream limits */
>  #define TCP_THIN_LINEAR_RETRIES 6       /* After 6 linear retries, do exp. backoff */
>
> +/* TCP initial congestion window */
> +#define TCP_INIT_CWND          10
> +
>  extern struct inet_timewait_death_row tcp_death_row;
>
>  /* sysctl variables for tcp */
> @@ -799,15 +802,6 @@ static inline __u32 tcp_current_ssthresh(const struct sock *sk)
>  /* Use define here intentionally to get WARN_ON location shown at the caller */
>  #define tcp_verify_left_out(tp)        WARN_ON(tcp_left_out(tp) > tp->packets_out)
>
> -/*
> - * Convert RFC 3390 larger initial window into an equivalent number of packets.
> - * This is based on the numbers specified in RFC 5681, 3.1.
> - */
> -static inline u32 rfc3390_bytes_to_packets(const u32 smss)
> -{
> -       return smss <= 1095 ? 4 : (smss > 2190 ? 2 : 3);
> -}
> -
>  extern void tcp_enter_cwr(struct sock *sk, const int set_ssthresh);
>  extern __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst);
>
> diff --git a/net/dccp/ccids/ccid2.c b/net/dccp/ccids/ccid2.c
> index e96d5e8..fadecd2 100644
> --- a/net/dccp/ccids/ccid2.c
> +++ b/net/dccp/ccids/ccid2.c
> @@ -583,6 +583,15 @@ done:
>        dccp_ackvec_parsed_cleanup(&hc->tx_av_chunks);
>  }
>
> +/*
> + * Convert RFC 3390 larger initial window into an equivalent number of packets.
> + * This is based on the numbers specified in RFC 5681, 3.1.
> + */
> +static inline u32 rfc3390_bytes_to_packets(const u32 smss)
> +{
> +       return smss <= 1095 ? 4 : (smss > 2190 ? 2 : 3);
> +}
> +
>  static int ccid2_hc_tx_init(struct ccid *ccid, struct sock *sk)
>  {
>        struct ccid2_hc_tx_sock *hc = ccid_priv(ccid);
> diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
> index eb7f82e..2f692ce 100644
> --- a/net/ipv4/tcp_input.c
> +++ b/net/ipv4/tcp_input.c
> @@ -817,7 +817,7 @@ __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst)
>        __u32 cwnd = (dst ? dst_metric(dst, RTAX_INITCWND) : 0);
>
>        if (!cwnd)
> -               cwnd = rfc3390_bytes_to_packets(tp->mss_cache);
> +               cwnd = TCP_INIT_CWND;
>        return min_t(__u32, cwnd, tp->snd_cwnd_clamp);
>  }
>
> --
> 1.7.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Thu, 03 Feb 2011 02:53:43 +0100

> Le mercredi 02 février 2011 à 17:07 -0800, David Miller a écrit :
>> Signed-off-by: David S. Miller <davem@davemloft.net>
>> ---
>> 
> 
> Hmm, you forgot a Changelog David ;)

Maybe, or maybe not, frankly it's quite self-evident if you ask
me. :-)

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Hagen Paul Pfeifer]

On Wed, 02 Feb 2011 17:07:50 -0800 (PST), David Miller wrote:

> +/* TCP initial congestion window */
> +#define TCP_INIT_CWND		10
> +

Davem, there is _no_ research how this huge IW will behave in environments
with a small BDP. Belief it or not, but there are networks out there with a
BDP similar to ~1980. Why for heaven's sake should this work in these
environments? There are only _two_ extensive analysis one from Cherry and
and one from Ilpo. Both analysis focus on current mainstream BDP. I started
to setup a ns-2 environment but due to lack of time I am a little bit
behind schedule. Anyway, why not make this a tunable per route knob so that
it is easier to fix things, e.g. set IW back to 3.

HGN

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[John Heffner]

On Thu, Feb 3, 2011 at 9:00 AM, Hagen Paul Pfeifer <hagen@jauu.net> wrote:
>
> On Wed, 02 Feb 2011 17:07:50 -0800 (PST), David Miller wrote:
>
>> +/* TCP initial congestion window */
>> +#define TCP_INIT_CWND                10
>> +
>
> Davem, there is _no_ research how this huge IW will behave in environments
> with a small BDP. Belief it or not, but there are networks out there with a
> BDP similar to ~1980. Why for heaven's sake should this work in these
> environments? There are only _two_ extensive analysis one from Cherry and
> and one from Ilpo. Both analysis focus on current mainstream BDP. I started
> to setup a ns-2 environment but due to lack of time I am a little bit
> behind schedule. Anyway, why not make this a tunable per route knob so that
> it is easier to fix things, e.g. set IW back to 3.


There's already a per-route tunable, right (RTAX_INITCWND)?

  -John

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Hagen Paul Pfeifer]

On Thu, 3 Feb 2011 09:17:14 -0500, John Heffner wrote:

> There's already a per-route tunable, right (RTAX_INITCWND)?

Yes, __u32 cwnd = (dst ? dst_metric(dst, RTAX_INITCWND) : 0);

I was a little bit nervous ... ;)

HGN

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Hagen Paul Pfeifer]

On Thu, 03 Feb 2011 15:00:28 +0100, Hagen Paul Pfeifer wrote:

> Davem, there is _no_ research how this huge IW will behave in
environments
> with a small BDP. Belief it or not, but there are networks out there
with a
> BDP similar to ~1980. Why for heaven's sake should this work in these
> environments? There are only _two_ extensive analysis one from Cherry
and

s/Cherry/Jerry/  ;-)

Hagen

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[H.K. Jerry Chu]

On Thu, Feb 3, 2011 at 6:00 AM, Hagen Paul Pfeifer <hagen@jauu.net> wrote:
>
> On Wed, 02 Feb 2011 17:07:50 -0800 (PST), David Miller wrote:
>
>> +/* TCP initial congestion window */
>> +#define TCP_INIT_CWND                10
>> +
>
> Davem, there is _no_ research how this huge IW will behave in environments
> with a small BDP. Belief it or not, but there are networks out there with a
> BDP similar to ~1980. Why for heaven's sake should this work in these
> environments? There are only _two_ extensive analysis one from Cherry and
> and one from Ilpo. Both analysis focus on current mainstream BDP. I started

It's more than that. We have also conducted extensive, large scale
experiments covering
all BDP/RTT combinations and the result was documented in details in a
SIGCOMM/CCR paper that can be found from a summary page
http://code.google.com/speed/protocols/tcpm-IW10.html covering all our
IW10 work.

> to setup a ns-2 environment but due to lack of time I am a little bit
> behind schedule. Anyway, why not make this a tunable per route knob so that
> it is easier to fix things, e.g. set IW back to 3.

You've seemed to get this backwards. Shouldn't we set the default for the normal
cases and leave the corner cases to the initcwnd and initrwnd knobs to cover?

Jerry

>
> HGN
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Leandro Melo de Sales]

Just forwarding my reply to the list netdev and dccp lists, the text
was in HTML format and Majordomo rejected it...

---------- Forwarded message ----------
From: Leandro Melo de Sales <leandroal@gmail.com>
Date: Thu, Feb 3, 2011 at 2:48 PM
Subject: Re: [PATCH] tcp: Increase the initial congestion window to 10.
To: David Miller <davem@davemloft.net>
Cc: netdev@vger.kernel.org, dccp@vger.kernel.org, therbert@google.com,
Angelo Perkusich <Angelo.Perkusich@gmail.com>, Hyggo Almeida
<hyggo.almeida@gmail.com>


On Wed, Feb 2, 2011 at 10:07 PM, David Miller <davem@davemloft.net> wrote:
>
> Signed-off-by: David S. Miller <davem@davemloft.net>
> ---
>
> I've left the DCCP code to keep using RFC3390 logic, if they
> wish to adopt this change in their code they can do so by
> simply deleting the rfc33390_bytes_to_packets() function and
> using TCP_INIT_CWND in their assignment.
>
>  include/net/tcp.h      |   12 +++---------
>  net/dccp/ccids/ccid2.c |    9 +++++++++
>  net/ipv4/tcp_input.c   |    2 +-
>  3 files changed, 13 insertions(+), 10 deletions(-)
>
> diff --git a/include/net/tcp.h b/include/net/tcp.h
> index 9179111..7118668 100644
> --- a/include/net/tcp.h
> +++ b/include/net/tcp.h
> @@ -196,6 +196,9 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
>  /* TCP thin-stream limits */
>  #define TCP_THIN_LINEAR_RETRIES 6       /* After 6 linear retries, do exp. backoff */
>
> +/* TCP initial congestion window */
> +#define TCP_INIT_CWND          10
> +
>  extern struct inet_timewait_death_row tcp_death_row;
>
>  /* sysctl variables for tcp */
> @@ -799,15 +802,6 @@ static inline __u32 tcp_current_ssthresh(const struct sock *sk)
>  /* Use define here intentionally to get WARN_ON location shown at the caller */
>  #define tcp_verify_left_out(tp)        WARN_ON(tcp_left_out(tp) > tp->packets_out)
>
> -/*
> - * Convert RFC 3390 larger initial window into an equivalent number of packets.
> - * This is based on the numbers specified in RFC 5681, 3.1.
> - */
> -static inline u32 rfc3390_bytes_to_packets(const u32 smss)
> -{
> -       return smss <= 1095 ? 4 : (smss > 2190 ? 2 : 3);
> -}
> -
>  extern void tcp_enter_cwr(struct sock *sk, const int set_ssthresh);
>  extern __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst);
>
> diff --git a/net/dccp/ccids/ccid2.c b/net/dccp/ccids/ccid2.c
> index e96d5e8..fadecd2 100644
> --- a/net/dccp/ccids/ccid2.c
> +++ b/net/dccp/ccids/ccid2.c
> @@ -583,6 +583,15 @@ done:
>        dccp_ackvec_parsed_cleanup(&hc->tx_av_chunks);
>  }
>
> +/*
> + * Convert RFC 3390 larger initial window into an equivalent number of packets.
> + * This is based on the numbers specified in RFC 5681, 3.1.
> + */
> +static inline u32 rfc3390_bytes_to_packets(const u32 smss)
> +{
> +       return smss <= 1095 ? 4 : (smss > 2190 ? 2 : 3);
> +}
> +
>  static int ccid2_hc_tx_init(struct ccid *ccid, struct sock *sk)
>  {
>        struct ccid2_hc_tx_sock *hc = ccid_priv(ccid);
> diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
> index eb7f82e..2f692ce 100644
> --- a/net/ipv4/tcp_input.c
> +++ b/net/ipv4/tcp_input.c
> @@ -817,7 +817,7 @@ __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst)
>        __u32 cwnd = (dst ? dst_metric(dst, RTAX_INITCWND) : 0);
>
>        if (!cwnd)
> -               cwnd = rfc3390_bytes_to_packets(tp->mss_cache);
> +               cwnd = TCP_INIT_CWND;
>        return min_t(__u32, cwnd, tp->snd_cwnd_clamp);
>  }
>
> --
> 1.7.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe dccp" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

David and others,
    I don't think this change will make sense for DCCP, once IW10 is
for the cases of short-lived flows. DCCP is used on scenarios for
multimedia flows, which are, in general, long-lived flows. So, IMO the
way how we calculate IW for DCCP is appropriate, at least considering
a quick answer. In fact, nowadays we need better congestion control
algorithms for DCCP, and in this sense we are working on adapt TCP
Cubic for DCCP as a CCID, and also we started some work to make DCCP
support TCP pluggable mechanism , which will allow us to use TCP
congestion control algorithms in DCCP.
    I've been following all the discussions about IW10 in the TCPM
discussion list and also been reading all materials in this context
(papers and ppts) provided by Tom Herbert and Nandita Dukkipati (along
with other researchers from Networking Research Lab at North Carolina
State University), but although the results is really great (at least
for scenarios they've experimented and regardless to all opinions
expressed at TCPM), DCCP should calculate its IW as specified in
RFC3390 rather than set it to IW10. For the moment, I understand that
for DCCP, we have to discuss more about this in a separated thread.

Leandro.

--
Leandro Melo de Sales
Professor at Institute of Computing at Federal University of Alagoas, Brazil
PhD candidate in Computer Science
Pervasive and Embedded Computing Laboratory, UFCG

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Hagen Paul Pfeifer]

* H.K. Jerry Chu | 2011-02-03 13:17:19 [-0800]:

>> to setup a ns-2 environment but due to lack of time I am a little bit
>> behind schedule. Anyway, why not make this a tunable per route knob so that
>> it is easier to fix things, e.g. set IW back to 3.
>
>You've seemed to get this backwards. Shouldn't we set the default for the normal
>cases and leave the corner cases to the initcwnd and initrwnd knobs to cover?

Yes Jerry, I phrase this a little bit abstruse. Apply David patch and use per
route options to reduce the IW if some curious behavior appear.

It seems the time is come to increase the IW. ;-)

Hagen

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Alexander Zimmermann]

[-- Attachment #1: Type: text/plain, Size: 1822 bytes --]

Hi David,

Am 03.02.2011 um 02:07 schrieb David Miller:

> 
> Signed-off-by: David S. Miller <davem@davemloft.net>
> ---
> 
> I've left the DCCP code to keep using RFC3390 logic, if they
> wish to adopt this change in their code they can do so by
> simply deleting the rfc33390_bytes_to_packets() function and
> using TCP_INIT_CWND in their assignment.
> 
> include/net/tcp.h      |   12 +++---------
> net/dccp/ccids/ccid2.c |    9 +++++++++
> net/ipv4/tcp_input.c   |    2 +-
> 3 files changed, 13 insertions(+), 10 deletions(-)
> 
> diff --git a/include/net/tcp.h b/include/net/tcp.h
> index 9179111..7118668 100644
> --- a/include/net/tcp.h
> +++ b/include/net/tcp.h
> @@ -196,6 +196,9 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
> /* TCP thin-stream limits */
> #define TCP_THIN_LINEAR_RETRIES 6       /* After 6 linear retries, do exp. backoff */
> 

Could you add a reference to draft?

> +/* TCP initial congestion window */
> +#define TCP_INIT_CWND		10
> +
> extern struct inet_timewait_death_row tcp_death_row;
> 
> /* sysctl variables for tcp */
> @@ -799,15 +802,6 @@ static inline __u32 tcp_current_ssthresh(const struct sock *sk)
> /* Use define here intentionally to get WARN_ON location shown at the caller */
> #define tcp_verify_left_out(tp)	WARN_ON(tcp_left_out(tp) > tp->packets_out)
> 
> -/*
> - * Convert RFC 3390 larger initial window into an equivalent number of packets.
> - * This is based on the numbers specified in RFC 5681, 3.1.
> - */
> -static inline u32 rfc3390_bytes_to_packets(const u32 smss)
> -{
> -	return smss <= 1095 ? 4 : (smss > 2190 ? 2 : 3);
> -}
> -
> extern void tcp_enter_cwr(struct sock *sk, const int set_ssthresh);
> extern __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst);
> 



[-- Attachment #2: Signierter Teil der Nachricht --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Gerrit Renker]

| I've left the DCCP code to keep using RFC3390 logic, if they
| wish to adopt this change in their code they can do so by
| simply deleting the rfc33390_bytes_to_packets() function and
| using TCP_INIT_CWND in their assignment.
| 
|  include/net/tcp.h      |   12 +++---------
|  net/dccp/ccids/ccid2.c |    9 +++++++++
Thank you for keeping the compatility, this seems an excellent way forward.

The change only affects CCID-2, TCP-like congestion control. This still has
some unresolved issues (related to cwnd) in its implementation.

Once these have been resolved, we can return to tracking the state of the 
art in TCP, for the moment your patch works best.

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[Gerrit Renker]

Leandro,
| >  include/net/tcp.h      |   12 +++---------
| >  net/dccp/ccids/ccid2.c |    9 +++++++++
|
David's change only affects CCID-2, CCID-3/4 remain unaffected,
they use rfc3390_initial_rate() (which does almost the same).
 
|     I don't think this change will make sense for DCCP, once IW10 is
| for the cases of short-lived flows. DCCP is used on scenarios for
| multimedia flows, which are, in general, long-lived flows. So, IMO the
| way how we calculate IW for DCCP is appropriate, at least considering
| a quick answer. 
I agree with you, and also vote to eventually bring CCID-2 on a par with
the state of art in TCP. Before we can do this, the CCID-2 implementation
needs some work (as per separate thread with Samuel Jero).

| In fact, nowadays we need better congestion control algorithms for DCCP,
| and in this sense we are working on adapt TCP Cubic for DCCP as a CCID,
| and also we started some work to make DCCP support TCP pluggable
| mechanism , which will allow us to use TCP congestion control
| algorithms in DCCP.
This is very good news and am very much looking forward to the new approach.


| DCCP should calculate its IW as specified in RFC3390 rather than set
| it to IW10. For the moment, I understand that for DCCP, we have to 
| discuss more about this in a separated thread.
Agree that we should keep an eye on this - perhaps too early to turn it into a
should. Networking speeds are growing, and many of the simulations were based
on TCP Reno which stems from the 10 Mbit ethernet era.

Re: [PATCH] tcp: Increase the initial congestion window to 10.

[David Miller]

From: Alexander Zimmermann <alexander.zimmermann@comsys.rwth-aachen.de>
Date: Fri, 04 Feb 2011 09:38:34 +0100

> Could you add a reference to draft?
> 
>> +/* TCP initial congestion window */
>> +#define TCP_INIT_CWND		10
>> +
>> extern struct inet_timewait_death_row tcp_death_row;

Sure, done.

--------------------
tcp: Add reference to initial CWND ietf draft.

Suggested by Alexander Zimmermann

Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/tcp.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/include/net/tcp.h b/include/net/tcp.h
index 7118668..adfe6db 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -196,7 +196,7 @@ extern void tcp_time_wait(struct sock *sk, int state, int timeo);
 /* TCP thin-stream limits */
 #define TCP_THIN_LINEAR_RETRIES 6       /* After 6 linear retries, do exp. backoff */
 
-/* TCP initial congestion window */
+/* TCP initial congestion window as per draft-hkchu-tcpm-initcwnd-01 */
 #define TCP_INIT_CWND		10
 
 extern struct inet_timewait_death_row tcp_death_row;
-- 
1.7.4