From: YOSHIFUJI Hideaki <[EMAIL PROTECTED]>
Date: Tue, 25 Jul 2006 23:53:40 +0900 (JST)
> In article <[EMAIL PROTECTED]> (at Tue, 25 Jul 2006 10:45:51 -0400 (EDT)),
> James Morris <[EMAIL PROTECTED]> says:
>
> > The recvmsg() for raw socket seems to return random u16 value
> > from the kernel stack memory since port field is not initialized.
> > But I'm not sure this patch is correct.
> > Does raw socket return any information stored in port field?
> >
> > ---------- Start of patch ----------
> > diff -ur before/net/ipv4/raw.c after/net/ipv4/raw.c
> > --- before/net/ipv4/raw.c 2006-06-18 10:49:35.000000000 +0900
> > +++ after/net/ipv4/raw.c 2006-07-25 16:15:26.000000000 +0900
> > @@ -609,6 +609,7 @@
> > if (sin) {
> > sin->sin_family = AF_INET;
> > sin->sin_addr.s_addr = skb->nh.iph->saddr;
> > + sin->sin_port = 0;
> > memset(&sin->sin_zero, 0, sizeof(sin->sin_zero));
> > }
> > if (inet->cmsg_flags)
>
> Well, instead, should it be initalized to protocol number, shouldn't it?
Initially, this was my reaction too. But, aparently it is defined
to be zero, from TCP/IP Illustrated Volume 2, page 1055, which is
discussing rip_input() in the BSD stack:
Unlike UDP, there is no concept of a port number in raw IP, so
the sin_port field in the sockaddr_in structures is always 0.
The BSD code does not explicitly set it to zero. But it doesn't have
to because it uses "ripsrc" which is a static variable in the BSS
section of the BSD kernel image.
So this patch appears correct and I will apply it. Thanks everyone.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
