Return-Path: X-Original-To: apmail-openoffice-dev-archive@www.apache.org Delivered-To: apmail-openoffice-dev-archive@www.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id C32B6183D2 for ; Thu, 29 Oct 2015 14:43:03 +0000 (UTC) Received: (qmail 66717 invoked by uid 500); 29 Oct 2015 14:43:03 -0000 Delivered-To: apmail-openoffice-dev-archive@openoffice.apache.org Received: (qmail 66634 invoked by uid 500); 29 Oct 2015 14:43:03 -0000 Mailing-List: contact dev-help@openoffice.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: dev@openoffice.apache.org Delivered-To: mailing list dev@openoffice.apache.org Received: (qmail 66623 invoked by uid 99); 29 Oct 2015 14:43:03 -0000 Received: from mail-relay.apache.org (HELO mail-relay.apache.org) (140.211.11.15) by apache.org (qpsmtpd/0.29) with ESMTP; Thu, 29 Oct 2015 14:43:03 +0000 Received: from [192.168.0.103] (unknown [181.55.232.163]) by mail-relay.apache.org (ASF Mail Server at mail-relay.apache.org) with ESMTPSA id 109961A0040 for ; Thu, 29 Oct 2015 14:43:02 +0000 (UTC) Content-Type: text/plain; charset=utf-8 Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2104\)) Subject: Re: Is AOO 4.1.2 notably faster? From: Pedro Giffuni In-Reply-To: <0F40BB96-BAFB-4181-BCA0-36F45DC84202@apache.org> Date: Thu, 29 Oct 2015 09:43:00 -0500 Content-Transfer-Encoding: quoted-printable Message-Id: References: <0F40BB96-BAFB-4181-BCA0-36F45DC84202@apache.org> To: Apache OO X-Mailer: Apple Mail (2.2104) Thanks for the feedback Francis, I looked a bit into the MSVC documentation and it appears their calloc() uses malloc() internally. It is likely that there are subtle = compatibility reasons for them to avoid some tweaks that linux and other systems do. Maybe using calloc has secondary long-term effects on linux or other platforms. At least for OpenOffice, the startup time is = particularly critical so it seems good to have the OS (libc and kernel basically) be aggressive in this type of optimizations. TBH, the real reason for the change was not performance but security, and it=E2=80=99s rather curious that while I am sure other developers = spent a lot of time optimizing the startup, they had never thought of this change before. Enjoy! Pedro. > Il giorno 28/ott/2015, alle ore 20:24, Pedro Giffuni = ha scritto: >=20 > Just wondering =E2=80=A6 >=20 > Are people noticing a faster startup in AOO 4.1.2? In the BSDs we = generally > prefer to use calloc over malloc + memset when it makes sense. A = relatively > small change that I did, and was safe enough to be merged to the = release, > might have had an effect there. >=20 > It=E2=80=99s too easy to be biased when you make such changes so it = would be > interesting to get confirmation ;). >=20 > Pedro. --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscribe@openoffice.apache.org For additional commands, e-mail: dev-help@openoffice.apache.org