netdev - Re: [PATCH 2/2] tcp: Fix tcp_prequeue() to get correct rto

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Sat, 2 May 2009 13:20:02 +0300 (EEST)
From:	"Ilpo Järvinen" <ilpo.jarvinen@...sinki.fi>
To:	Satoru SATOH <satoru.satoh@...il.com>
cc:	Netdev <netdev@...r.kernel.org>
Subject: Re: [PATCH 2/2] tcp: Fix tcp_prequeue() to get correct rto_min value

On Sat, 2 May 2009, Satoru SATOH wrote:

> This patch depends on the previous patch, makes tcp_prequeue() uses the
> correct tcp rto_min value returns from tcp_rto_min().
> 
> Signed-off-by: Satoru SATOH <satoru.satoh@...il.com>
> 
> ---
>  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 284cc68..a54d139 100644
> --- a/include/net/tcp.h
> +++ b/include/net/tcp.h
> @@ -920,7 +920,7 @@ static inline int tcp_prequeue(struct sock *sk, struct sk_buff *skb)
>  			wake_up_interruptible(sk->sk_sleep);
>  			if (!inet_csk_ack_scheduled(sk))
>  				inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK,
> -						          (3 * TCP_RTO_MIN) / 4,
> +						          (3 * tcp_rto_min(sk)) / 4,
>  							  TCP_RTO_MAX);
>  		}
>  		return 1;
> 

I'm not fully sure we really want this... I wonder did you run into some 
real problem with the TCP_RTO_MIN based version as it's not setting RTO 
timer but delayed ACK timer...? The change could have nasty effects to 
interactivity of a flow if somebody is hit by nagle delay and the peer
has tuned their minimum RTO. Not that I in anyway believe it should be 
that useful to set that minimum RTO because of frto that is nowadays
enabled.

In any case, I see very little point in putting them into a separate 
patches due to non-complex nature of the actual change (with complex 
changes it might help though).

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