From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Baumann Subject: Re: [PATCH v3 1/3] Eliminate recursion in setting/clearing marks in commit list Date: Sat, 14 Jan 2012 16:23:16 +0100 Message-ID: <20120114152316.GA30986@m62s10.vlinux.de> References: <1326081546-29320-1-git-send-email-pclouds@gmail.com> <1326543595-28300-1-git-send-email-pclouds@gmail.com> <1326543595-28300-2-git-send-email-pclouds@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: git@vger.kernel.org, Junio C Hamano , "Shawn O. Pearce" To: =?utf-8?B?Tmd1eeG7hW4gVGjDoWkgTmfhu41j?= Duy X-From: git-owner@vger.kernel.org Sat Jan 14 16:23:29 2012 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 1Rm5S4-0000e2-4G for gcvg-git-2@lo.gmane.org; Sat, 14 Jan 2012 16:23:28 +0100 Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755014Ab2ANPXY convert rfc822-to-quoted-printable (ORCPT ); Sat, 14 Jan 2012 10:23:24 -0500 Received: from mailout-de.gmx.net ([213.165.64.23]:34796 "HELO mailout-de.gmx.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1754851Ab2ANPXX (ORCPT ); Sat, 14 Jan 2012 10:23:23 -0500 Received: (qmail invoked by alias); 14 Jan 2012 15:23:20 -0000 Received: from m62s10.vlinux.de (EHLO m62s10.vlinux.de) [83.151.21.204] by mail.gmx.net (mp025) with SMTP; 14 Jan 2012 16:23:20 +0100 X-Authenticated: #1252284 X-Provags-ID: V01U2FsdGVkX1+3jN8g5yy3mJPjiKl1XD6ObdsgeIMXNlOfRTWBHs 3RX5u9KjTC56bO Received: by m62s10.vlinux.de (Postfix, from userid 1000) id D60EAD42AC; Sat, 14 Jan 2012 16:23:16 +0100 (CET) Content-Disposition: inline In-Reply-To: <1326543595-28300-2-git-send-email-pclouds@gmail.com> User-Agent: Mutt/1.5.20 (2009-06-14) X-Y-GMX-Trusted: 0 Sender: git-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: git@vger.kernel.org Archived-At: On Sat, Jan 14, 2012 at 07:19:53PM +0700, Nguy=E1=BB=85n Th=C3=A1i Ng=E1= =BB=8Dc Duy wrote: > Recursion in a DAG is generally a bad idea because it could be very > deep. Be defensive and avoid recursion in mark_parents_uninteresting(= ) > and clear_commit_marks(). >=20 > mark_parents_uninteresting() learns a trick from clear_commit_marks() > to avoid malloc() in (dorminant) single-parent case. ^^^^^^^^^ I think you ment dominant here.