Re: [lwip-users] TCP payload is doubled

[Kieran Mansley]

On Thu, 2009-10-22 at 08:33 +0200, Simon Goldschmidt wrote:
> That's all true, but David's and Jan's recent problem was that the
> window was changed although tcp_recved() was called right away. This
> is not an RFC-violating problem but a problem in lwIP (when and how it
> calculates the window changes).

I still think lwIP is doing the right thing in this regard.  If you call
tcp_recved() with a small value, it won't advertise that window space
right away as to do so could lead to the sender sending a small packet
to fill this little bit of extra window.  If we delay advertising the
extra window until there's a good amount (MSS, or WND / 2 with David
Empson's improvement) then the sender will always be able to send good
sized frames.

Can you explain what you think lwIP is getting wrong?

Kieran