SORT BY:

LIST ORDER
THREAD
AUTHOR
SUBJECT


SEARCH

IPS HOME


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

    RE: iSCSI CRC: A CRC-checking example



    Joe and Bob,
    
    I doubt that any implementation today will be doing CRC computation in a bit
    serial circuit. The computation is generally done either byte parallel or,
    increasingly, multi-byte parallel. The order in which the bits of a byte are
    assigned to polynomial coefficients makes no difference to byte parallel CRC
    implementation complexity. The only time complexity would be effected by bit
    order is when one was doing a serial shift register implementation of the
    CRC generator as some of the earliest Ethernet implementations did. The
    technical reasons which applied to Ethernet in chosing the CRC bit checking
    order, burst protection and serial implementation complexity, don't apply to
    iSCSI.
    
    The only reason for choosing one bit order over another for iSCSI CRC
    checking is the limits of human comprehension. Which bit order is most
    likely to be gotten right by the humans implementing the standard?
    
     Ethernet 
              + many people are already accustomed to its bit order
              - most significant byte first, least significant bit first can
    cause confusion
     Most significant everything first 
              + more internally consistent than Ethernet
              - different from Ethernet
    
    The most important thing is that we specify whichever one we chose as
    clearly as possible.
    
    Regards,
    Pat 
                 
    
    -----Original Message-----
    From: Joe Gervais [mailto:joeg@alacritech.com]
    Sent: Thursday, August 23, 2001 7:21 PM
    To: Robert Snively; Black_David@emc.com; sanjay_goyal@ivivity.com;
    ips@ece.cmu.edu
    Subject: RE: iSCSI CRC: A CRC-checking example
    
    
    Bob,
    
    If the data was going on token ring or some other media, sure, who cares
    about an Ethernet approach, but the data is generally being sent and
    received over Ethernet, and there is advantage at least in some
    implementations to be in Ethernet bit/byte order.
    
    Additionally, as David said earlier today, if it isn't broke, let's not fix
    it. Can we get consensus and move on?
    
    Joe
    -----Original Message-----
    From: owner-ips@ece.cmu.edu [mailto:owner-ips@ece.cmu.edu]On Behalf Of
    Robert Snively
    
    ... snip ...
    So I would have expected that we would avoid any choice that
    makes that less convenient.  The Ethernet approach to those
    bytes would strike me as being the second-least convenient
    approach.  Clearly, throwing hardware at this could correct
    for that inconvenience, but it seems to me that any
    optimizations we make should focus on what we want to do, not
    what Ethernet did.
    
    Bob
    
    


Home

Last updated: Tue Sep 04 01:03:54 2001
6315 messages in chronological order