From: Herbert Xu Subject: Re: [RFC] [PATCH 2/7] crypto: authenc - convert to ahash Date: Mon, 20 Jul 2009 16:50:04 +0800 Message-ID: <20090720085004.GA5018@gondor.apana.org.au> References: <20090716111548.GP20288@secunet.com> <20090716111747.GR20288@secunet.com> <20090717.100334.160030992.davem@davemloft.net> <20090720062643.GA20288@secunet.com> <20090720065514.GB4026@gondor.apana.org.au> <20090720075701.GC20288@secunet.com> <20090720082116.GA4797@gondor.apana.org.au> <20090720083640.GD20288@secunet.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Miller , linux-crypto@vger.kernel.org, netdev@vger.kernel.org, Jens Axboe , Linux Kernel Mailing List To: Steffen Klassert Return-path: Received: from rhun.apana.org.au ([64.62.148.172]:42668 "EHLO arnor.apana.org.au" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1751869AbZGTIuO (ORCPT ); Mon, 20 Jul 2009 04:50:14 -0400 Content-Disposition: inline In-Reply-To: <20090720083640.GD20288@secunet.com> Sender: linux-crypto-owner@vger.kernel.org List-ID: On Mon, Jul 20, 2009 at 10:36:40AM +0200, Steffen Klassert wrote: > > For the crypto-layer it's just scatterwalk_sg_chain() that removes the > termination bit. The history behind this is that the crypto layer has used chaining way before generic chaining was invented. When the generic code was added I tried to convert crypto over to that, however, that failed because generic chaining was not ported to every architecture. So that's why the crypto layer still retains its own chaining. > For the block-layer it's blk_rq_map_integrity_sg() and two times in > blk_rq_map_sg(). I suppose we should ask Jens whether he thinks this is worth adding an interface over? Cheers, -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt