From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephan Mueller Subject: Re: [PATCH v2 1/6] SP800-90A Deterministic Random Bit Generator Date: Thu, 20 Mar 2014 14:30:33 +0100 Message-ID: <13817639.cfCaQvh16D@myon.chronox.de> References: <2396177.vxvG2ljJL8@myon.chronox.de> <3662681.npTzbSq3ye@myon.chronox.de> <532AA307.504@ladisch.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: linux-kernel@vger.kernel.org, linux-crypto@vger.kernel.org, aquini@redhat.com, jeremy.wayne.powell@gmail.com To: Clemens Ladisch Return-path: Received: from mo4-p00-ob.smtp.rzone.de ([81.169.146.163]:52037 "EHLO mo4-p00-ob.smtp.rzone.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757256AbaCTNao convert rfc822-to-8bit (ORCPT ); Thu, 20 Mar 2014 09:30:44 -0400 In-Reply-To: <532AA307.504@ladisch.de> Sender: linux-crypto-owner@vger.kernel.org List-ID: Am Donnerstag, 20. M=E4rz 2014, 09:12:55 schrieb Clemens Ladisch: Hi Clemens, > Stephan Mueller wrote: > > This is a clean-room implementation of the DRBG defined in SP800-90= A. >=20 > Why? I guess it's for certification? As per SP800-131A, the ANSI X9.31 DRNG is sunset by the end of 2014 and= not=20 allowed to be used in FIPS 140-2 compliant environments. The kernel cry= pto API=20 implements an ANSI X9.31 DRNG in crypto/ansi_cprng.c as the only DRNG t= hat=20 complies with FIPS 140-2 at this time. Without a replacement for this ANSI X9.31 DRNG, the kernel will not hav= e an=20 =46IPS 140-2 approved DRNG any more starting from 2015. >=20 > > +static bool drbg_fips_continuous_test(struct drbg_state *drbg, > > + unsigned char *buf) > > ... > > + ret =3D memcmp(drbg->prev, buf, drbg_blocklen(drbg)); > > + ... > > + /* invert the memcmp result, because the test shall pass when the > > + * two compared values do not match */ > > + if (ret) > > + return true; > > + else > > + return false; >=20 > This looks strange. The return value of memcmp() is not really > a boolean, and the code appears not to match the comment because the > numeric value of ret is not actually inverted. How about this: Correct, the comment does not match the code as I had invered the logic= of=20 drbg_fips_continuous_test as per Rafael's comment. Yet, I did not updat= e the=20 comment. >=20 > ret =3D memcmp(...); > ... > /* the test shall pass when the compared values are not equal */ > return ret !=3D 0; I will add that change. >=20 >=20 > Regards, > Clemens Thanks Stephan --=20 | Cui bono? |