Re: Request for Redhat

[Chris Wright <chrisw@sous-sol.org>]

Hi Jun,

* Jun.Kondo (jun.kondo@ctc-g.co.jp) wrote:
> CTC had the following demand;
> 
> 1. to ensure high throughput from the beginning of
> tcp connection at normal times by acquiring large
> default transmission buffer value
> 
> 2. to limit the block time of the write in order to
> prevent the timeout of upper layer applications
> even when the connection has low throughput, such
> as low rate streaming

Make sure you include network developers (netdev@vger.kernel.org Cc'd)
on networking related patches.

You'll find some helpful information regarding the email format for
submitting patches here:
  http://www.kernel.org/doc/Documentation/SubmittingPatches

Improving your Subject: and fixing the patch level for your diff come
to mind immediately.

> --- sock.h 2011-04-14 14:58:03.000000000 +0900
> +++ sock.h.mod 2011-04-21 15:30:07.000000000 +0900
> @@ -434,9 +434,12 @@
> /*
> * Compute minimal free write space needed to queue new packets.
> */
> +
> +extern int sysctl_tcp_lowat;
> +
> static inline int sk_stream_min_wspace(struct sock *sk)
> {
> - return sk->sk_wmem_queued / 2;
> + return sk->sk_wmem_queued >> sysctl_tcp_lowat;
> }

Also, this patch appears to be against an old tree (I assume the RHEL
5 tree).  To be considered upstream, you need to generate a diff against
the upstream tree.  And considering all that's changed since 2.6.18, it
would be useful to verify you still have the issue in a current upstream
Linux tree.

Alternatively, you may consider using Red Hat's Bugzilla to contact RH
engineers for help and guidance.

Good luck.

thanks,
-chris