Search String: Display: Description: Sort:

Results:

References: [ +subject:/^(?:^\s*(re|sv|fwd|fw)[\[\]\d]*[:>-]+\s*)*rcv_wnd\s+\=\s+init_cwnd\*mss\s*$/: 10 ]

Total 10 documents matching your query.

1. rcv_wnd = init_cwnd*mss (score: 1)
Author: <davem@xxxxxxxxxxxxx>
Date: Wed, 27 Oct 2004 22:14:33 -0700
/archives/netdev/2004-10/msg01198.html (8,240 bytes)

2. Re: rcv_wnd = init_cwnd*mss (score: 1)
Author: ler" <davem@xxxxxxxxxxxxx>
Date: Wed, 27 Oct 2004 22:21:28 -0700
/archives/netdev/2004-10/msg01199.html (7,873 bytes)

3. RE: rcv_wnd = init_cwnd*mss (score: 1)
Author: erbert@xxxxxxxxxxxxxxxxxxx>
Date: Wed, 27 Oct 2004 23:15:48 -0700
/archives/netdev/2004-10/msg01200.html (8,964 bytes)

4. Re: rcv_wnd = init_cwnd*mss (score: 1)
Author: vem@xxxxxxxxxxxxx>
Date: Thu, 28 Oct 2004 16:56:58 -0700
/archives/netdev/2004-10/msg01274.html (8,887 bytes)

5. Re: rcv_wnd = init_cwnd*mss (score: 1)
Author: homas Graf <tgraf@xxxxxxx>
Date: Thu, 28 Oct 2004 17:55:48 -0700
/archives/netdev/2004-10/msg01291.html (8,308 bytes)

6. rcv_wnd = init_cwnd*mss (score: 1)
Author: "Meda, Prasanna" <pmeda@xxxxxxxxxx>
Date: Wed, 27 Oct 2004 22:14:33 -0700
What is the reason for checking mss with 1<<rcv_wscale? include/net/tcp.h: static inline void tcp_select_initial_window(int __space, __u32 mss, __u32 *rcv_wnd, __u32 *window_clamp, int wscale_ok, __u
/archives/netdev/2004-10/msg02696.html (8,240 bytes)

7. Re: rcv_wnd = init_cwnd*mss (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxxxxx>
Date: Wed, 27 Oct 2004 22:21:28 -0700
Because the advertised window field is 16-bits. It is interpreted as "value << rcv_wscale"
/archives/netdev/2004-10/msg02697.html (7,965 bytes)

8. RE: rcv_wnd = init_cwnd*mss (score: 1)
Author: "Meda, Prasanna" <pmeda@xxxxxxxxxx>
Date: Wed, 27 Oct 2004 23:15:48 -0700
Thanks, still it is unclear to me why are we downsizing the advertised window(rcv_wnd) to cwnd? To defeat disobeying sender, or something like below? Suppose when wscale is zero, it is now checking
/archives/netdev/2004-10/msg02698.html (8,979 bytes)

9. Re: rcv_wnd = init_cwnd*mss (score: 1)
Author: "David S. Miller" <davem@xxxxxxxxxxxxx>
Date: Thu, 28 Oct 2004 16:56:58 -0700
There is never any reason to advertise a receive window larger than the initial congestion window of the sender could ever be. Setting it properly like this also makes sure that we do receive window
/archives/netdev/2004-10/msg02772.html (8,979 bytes)

10. Re: rcv_wnd = init_cwnd*mss (score: 1)
Author: Prasanna Meda <pmeda@xxxxxxxxxx>
Date: Thu, 28 Oct 2004 17:55:48 -0700
That makes sense! But are we coping with cwnd increase on sender? Looks rcv rwnd s updated by only 1 pkt at time. Thanks, Prasanna.
/archives/netdev/2004-10/msg02789.html (8,391 bytes)


This search system is powered by Namazu