From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff King Subject: Re: [PATCH] Disable OpenSSL SHA1 implementation by default Date: Mon, 22 Feb 2010 06:23:26 -0500 Message-ID: <20100222112326.GA21929@coredump.intra.peff.net> References: <20100222110814.GA3247@progeny.tock> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: git@vger.kernel.org, Nicolas Pitre , Robert Shearman , Ben Walton To: Jonathan Nieder X-From: git-owner@vger.kernel.org Mon Feb 22 13:20:11 2010 Return-path: Envelope-to: gcvg-git-2@lo.gmane.org Received: from vger.kernel.org ([209.132.180.67]) by lo.gmane.org with esmtp (Exim 4.69) (envelope-from ) id 1NjWNx-0007lq-4Q for gcvg-git-2@lo.gmane.org; Mon, 22 Feb 2010 12:23:33 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752139Ab0BVLX2 convert rfc822-to-quoted-printable (ORCPT ); Mon, 22 Feb 2010 06:23:28 -0500 Received: from peff.net ([208.65.91.99]:49374 "EHLO peff.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751773Ab0BVLX1 (ORCPT ); Mon, 22 Feb 2010 06:23:27 -0500 Received: (qmail 26184 invoked by uid 107); 22 Feb 2010 11:23:41 -0000 Received: from coredump.intra.peff.net (HELO coredump.intra.peff.net) (10.0.0.2) by peff.net (qpsmtpd/0.40) with (AES128-SHA encrypted) SMTP; Mon, 22 Feb 2010 06:23:41 -0500 Received: by coredump.intra.peff.net (sSMTP sendmail emulation); Mon, 22 Feb 2010 06:23:26 -0500 Content-Disposition: inline In-Reply-To: <20100222110814.GA3247@progeny.tock> Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Mon, Feb 22, 2010 at 05:08:14AM -0600, Jonathan Nieder wrote: > The OpenSSL SHA-1 routine is about as fast as block-sha1, but linking > to libcrypto slows down the startup of git commands by an appreciable > amount. Use the BLK_SHA1 implementation by default instead. What is your definition of "about as fast"? I benchmarked up to a 20% slow-down a while back: http://article.gmane.org/gmane.comp.version-control.git/126995 Now my complaint then was specifically about removing openssl sha1 support entirely, and I have no problem with setting OPENSSL_SHA1 in my build options, but it does make sense to me that the default should be whatever is fastest for most people. And that means benchmarking blk_sha1 versus the libcrypto linking slow-down on several machines to get actual numbers. Unfortunately, I think we may end up with an apples-to-oranges comparison, as sha1-heavy tasks are affected one way, and scripted many-git-invocation tasks the other way. > Typed =E2=80=9Cmake NO_OPENSSL=3D1=E2=80=9D for the umpteenth time to= day, but this time > I thought I should something about it. echo 'NO_OPENSSL=3D1' >config.mak ? > -ifndef NO_OPENSSL > +ifdef OPENSSL_TLS > + BASIC_CFLAGS +=3D -DOPENSSL_TLS > + USE_OPENSSL =3D Yes > +endif Doesn't this flip the default for using TLS on imap-send? -Peff