From: Herbert Xu Subject: Re: [PATCH v2 1/4] Wrap crc_t10dif function all to use crypto transform framework Date: Tue, 30 Apr 2013 11:27:33 +0800 Message-ID: <20130430032733.GB20471@gondor.apana.org.au> References: <20130425132203.GA29501@gondor.apana.org.au> <1366910910.27102.126.camel@schen9-DESK> <20130426125224.GA4855@gondor.apana.org.au> <1366994657.27102.129.camel@schen9-DESK> <20130428001122.GA6973@gondor.apana.org.au> <1367268030.27102.172.camel@schen9-DESK> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "H. Peter Anvin" , "David S. Miller" , "Martin K. Petersen" , James Bottomley , Matthew Wilcox , Jim Kukunas , Keith Busch , Erdinc Ozturk , Vinodh Gopal , James Guilford , Wajdi Feghali , Jussi Kivilinna , linux-kernel , linux-crypto@vger.kernel.org, linux-scsi@vger.kernel.org To: Tim Chen Return-path: Content-Disposition: inline In-Reply-To: <1367268030.27102.172.camel@schen9-DESK> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-crypto.vger.kernel.org On Mon, Apr 29, 2013 at 01:40:30PM -0700, Tim Chen wrote: > > If I allocate the transform under the mod init instead, how can I make > sure that the fast version is already registered if I have it compiled > in? It is not clear to me how that's done looking at the libcrc32c > code. This is only an issue when everything is built-in to the kernel. In that case we could make the crc implementations register at a point earlier than device_initcall, but no earlier than subsys_initcall since that's where cryptomgr sits. Cheers, -- Email: Herbert Xu Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt