Date: Wed, 7 Mar 2001 12:31:56 +0200 From: Ruslan Ermilov <ru@FreeBSD.ORG> To: Jonathan Lemon <jlemon@flugsvamp.com> Cc: Jonathan Lemon <jlemon@FreeBSD.ORG>, net@FreeBSD.ORG Subject: Re: Delayed checksums commit broke UDP checksum calculation Message-ID: <20010307123156.A19829@sunbay.com> In-Reply-To: <20001116091954.A19895@prism.flugsvamp.com>; from jlemon@flugsvamp.com on Thu, Nov 16, 2000 at 09:19:54AM -0600 References: <20001116120936.A45755@sunbay.com> <20001116091954.A19895@prism.flugsvamp.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Nov 16, 2000 at 09:19:54AM -0600, Jonathan Lemon wrote:
> On Thu, Nov 16, 2000 at 12:09:36PM +0200, Ruslan Ermilov wrote:
> > Hi!
> >
> > RFC768> If the computed checksum is zero, it is transmitted as all ones
> > RFC768> (the equivalent in one's complement arithmetic). An all zero
> > RFC768> transmitted checksum value means that the transmitter generated
> > RFC768> no checksum.
> >
> > This (0x0000 -> 0xFFFF) apparently got broken in udp_usrreq.c,v 1.65.
>
> Actually, it got broken in a different fashion; I had originally moved
> the above logic within in_cksum itself, and when that got taken out,
> the UDP case wasn't updated.
>
Actually, the "logic move" was implemented wrong, and then got wiped
out in revision 1.21 of sys/i386/i386/in_cksum.c.
Also, you MFC'ed this fix for i386, but did not for alpha.
(sys/alpha/alpha/in_cksum.c,v 1.5).
> > Index: ip_output.c
> > ===================================================================
> > RCS file: /home/ncvs/src/sys/netinet/ip_output.c,v
> > retrieving revision 1.116
> > diff -u -p -r1.116 ip_output.c
> > --- ip_output.c 2000/11/01 01:59:28 1.116
> > +++ ip_output.c 2000/11/16 10:05:06
> > @@ -974,6 +974,8 @@ in_delayed_cksum(struct mbuf *m)
> > ip = mtod(m, struct ip *);
> > offset = IP_VHL_HL(ip->ip_vhl) << 2 ;
> > csum = in_cksum_skip(m, ip->ip_len, offset);
> > + if (m->m_pkthdr.csum_flags & CSUM_UDP && csum == 0)
> > + csum = 0xffff; /* per RFC768 */
> > offset += m->m_pkthdr.csum_data; /* checksum offset */
> >
> > if (offset + sizeof(u_short) > m->m_len) {
>
>
> This would work. Alternatively, you could change it to:
>
> if (csum == 0)
> csum = 0xffff;
>
> So that the same logic applies to TCP packets as well. Currently, we
> can send a TCP packet with a checksum of 0, which is legal. Of possible
> interest is that Linux doesn't do this; they alwyas send a non-zero
> checksum in the TCP case, if a checksum was computed.
>
Hmm, but why would we do this for TCP? This violates RFC 793.
AFAIK, only UDP checksums are special.
Cheers,
--
Ruslan Ermilov Oracle Developer/DBA,
ru@sunbay.com Sunbay Software AG,
ru@FreeBSD.org FreeBSD committer,
+380.652.512.251 Simferopol, Ukraine
http://www.FreeBSD.org The Power To Serve
http://www.oracle.com Enabling The Information Age
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010307123156.A19829>