[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [MBONED] WGLC for <draft-ietf-mboned-auto-multicast-08.txt> -UDP checksum?



> -----Original Message-----
> From: Dino Farinacci [mailto:dino at cisco.com] 
> Sent: Thursday, November 29, 2007 1:39 AM
> To: Stig Venaas
> Cc: gorry at erg.abdn.ac.uk; mboned at ietf.org
> Subject: Re: [MBONED] WGLC for 
> <draft-ietf-mboned-auto-multicast-08.txt> -UDP checksum?
> 
> >>> Right, for IPv4 you compute checksum for the IPv4 
> checksum header.  
> >>> For
> >>> IPv6 you instead have a transport checksum that include some of  
> >>> the IPv6
> >>> header fields in the computation. So, for both you have 
> to compute a
> >>> checksum, it's just two different ones.
> >>
> >> But CRCs are much stronger than the 1's complement 
> checksum in the IP
> >> and transport headers.
> >
> > Agree. I guess I should also add that the IPv4 header checksum is  
> > easier
> > to compute than the transport checksum in IPv6.
> 
> That is definitely true. It typically is over a fixed number 
> of bytes  
> (i.e. 20) which tend to always be in a high-speed FIFO.

Who knew that we would miss the IPv4 header checksum so much?

Problem is, going between different link layers, you don't always have
the protection of a link layer's CRC. Bits can get switched at
inconvenient times.

Bert


_______________________________________________
MBONED mailing list
MBONED at ietf.org
https://www1.ietf.org/mailman/listinfo/mboned