stray 2.4 tcp fix.

stray 2.4 tcp fix.

[davej]

Sent to davem and netdev already, and was met with silence..

diff -urpN --exclude-from=/home/davej/.exclude bk-linus/net/ipv4/tcp.c linux-2.5/net/ipv4/tcp.c
--- bk-linus/net/ipv4/tcp.c	2003-03-21 12:53:31.000000000 +0000
+++ linux-2.5/net/ipv4/tcp.c	2003-03-21 13:45:21.000000000 +0000
@@ -1189,7 +1189,8 @@ new_segment:
 
 			from += copy;
 			copied += copy;
-			seglen -= copy;
+			if ((seglen -= copy) == 0 && iovlen == 0)
+				goto out;
 
 			if (skb->len != mss_now || (flags & MSG_OOB))
 				continue;

Re: stray 2.4 tcp fix.

[Jeff Garzik]

On Mon, Mar 24, 2003 at 04:42:00PM +0000, davej@codemonkey.org.uk wrote:
> Sent to davem and netdev already, and was met with silence..
> 
> diff -urpN --exclude-from=/home/davej/.exclude bk-linus/net/ipv4/tcp.c linux-2.5/net/ipv4/tcp.c
> --- bk-linus/net/ipv4/tcp.c	2003-03-21 12:53:31.000000000 +0000
> +++ linux-2.5/net/ipv4/tcp.c	2003-03-21 13:45:21.000000000 +0000
> @@ -1189,7 +1189,8 @@ new_segment:
>  
>  			from += copy;
>  			copied += copy;
> -			seglen -= copy;
> +			if ((seglen -= copy) == 0 && iovlen == 0)
> +				goto out;

What's the purpose of combining a statement and a test?

Simply adding a test would be more clear.

	Jeff

Re: stray 2.4 tcp fix.

[David S. Miller]

On Mon, 2003-03-24 at 08:42, davej@codemonkey.org.uk wrote:
> Sent to davem and netdev already, and was met with silence..

Thanks for finding this I'll apply it, but please note:

1) I am a lossy protocol like IP, so you may need to retransmit
   a patch if I don't respond to it.

2) For some reason a lot of netdev mails aren't making their
   way to me, I believe I've been unsubscribed and I'll check
   this now.

Thanks again.

-- 
David S. Miller <davem@redhat.com>