constrained socket buffers patch

[coderman]

On 7/20/07, Juliusz Chroboczek <Juliusz.Chroboczek at pps.jussieu.fr> wrote:
> > this is a good idea.  16k might be even better if it worked reliably
> > (the usual default is 32 to 64k).
>
> Your information might be somewhat obsolete...
> ...
> As you may see, tor.eff.org has grown its receive window to 67 kB,
> while Lanthane has gone up to 26 kB.

it's important to note that while the constrained buffers patch will
limit the maximum size of the window, it does not do the inverse.

it is very common to have small TCP windows at start and yet still
have 64KB buffers allocated to the socket resource itself in kernel
land...  (so think of this like setting a ceiling on window, rather
than fixing it to a specific value at start)