From: Andi Kleen Subject: Re: {twofish,aes}-{x86_64,i586} versus C implementations Date: Mon, 20 Aug 2007 15:06:39 +0200 Message-ID: <20070820130639.GN16680@bingen.suse.de> References: <200708200234.25620.ak@suse.de> <20070820101618.GE16680@bingen.suse.de> <20070820120605.GA13163@gondor.apana.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Andi Kleen , linux-crypto@vger.kernel.org To: Herbert Xu Return-path: Received: from cantor.suse.de ([195.135.220.2]:38951 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755223AbXHTMMn (ORCPT ); Mon, 20 Aug 2007 08:12:43 -0400 Content-Disposition: inline In-Reply-To: <20070820120605.GA13163@gondor.apana.org.au> Sender: linux-crypto-owner@vger.kernel.org List-Id: linux-crypto.vger.kernel.org > That would be the best. However, it's not hard to do a > simple probing in the kernel until modprobe(8) gets this > feature. Sounds like a big hack, and at least for aes / aes-x86_64 and twofish it's not needed. Just disable aes on x86. The only problem is the select issue with wireless. Unfortunately select CRYPTO_AES_X86_64 if X86_64 select CRYPTO_AES_I586 if X86_32 select CRYPTO_AES if !X86 produces warnings for unreferenced symbols :/ Perhaps it can be just removed for now. > > Also if one implementation is always better than the other > > then I see little reason to ever have both. > > Well it's not that useful for an assembly implementation > that works on all instances of a given architecture. I meant on x86. Sure for other architectures the C version is needed. -Andi