Late Letter Ballot comment on SBC-3

Elliott, Robert (Server Storage) Elliott at hp.com
Thu Jul 11 16:30:04 PDT 2013


Formatted message: <a href="http://www.t10.org/cgi-bin/ac.pl?t=r&f=r1307112_f.htm">HTML-formatted message</a>

iSCSI (RFC 3720) uses "hex notation" wording:
		The generator polynomial for this digest is given in
hex-notation
		(e.g., 0x3b stands for 0011 1011 and the polynomial is
x**5+X**4+x**3+x+1).
SPL uses the same wording as SBC for its CRC polynomials; SBC and SPL should
stay the same or change together.  SPL does not provide hex notation of the
polynomial used for SAS address hashing or the polynomial used for scrambling
(except in the sample C code).
From: owner-t10 at t10.org [mailto:owner-t10 at t10.org] On Behalf Of Ralph Weber
Sent: Thursday, 11 July, 2013 5:35 PM
To: t10 at t10.org
Subject: Late Letter Ballot comment on SBC-3
In SBC-3 r35 table 21, one finds the following definition of the DIF CRC
generator polynomial:
G(x) = x^16 + x^15 + x^11 + x^9 + x^8 + x^7 + x^5 + x^4 + x^2 + x + 1
(i.e., G(x) = 1_8BB7h)
With apologies for the lack of good superscripts in e-mail.
For reasons that will become obvious in a day or two, I have recently been
obliged to learn things I really had hoped to avoid knowing about CRC
generator polynomials. In the early going, my knee-jerk reaction to the 'G(x)
= 1_8BB7h' was something like "G(x) does not return a constant value", but I
was at a loss for a better phrasing. No more!
Remembering that x=x^1 and 1=x^0, the 1_8BB7h value is derived as follows:
Exponents 1 1111 1100 0000 0000
     of x 6 5432 1098 7654 3210
  Bitmask 1 1000 1011 1011 0111
      Hex 1   8   B    B    7
With this in mind, I request that the SBC-3 table 21 generator i.e. be
modified to read as follows:
(i.e., the bitmask representation of G(x) is 1_8BB7h)
All the best,
.Ralph
P.S. I have see some evidence that the bitmask representation is very useful
is one is building an analog engine to compute the CRC.



More information about the T10 mailing list