From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from dan.rpsys.net (5751f4a1.skybroadband.com [87.81.244.161]) by mail.openembedded.org (Postfix) with ESMTP id 0777C73E3B for ; Wed, 17 Jun 2015 21:03:35 +0000 (UTC) Received: from localhost (localhost [127.0.0.1]) by dan.rpsys.net (8.14.4/8.14.4/Debian-4.1ubuntu1) with ESMTP id t5HL3Znp014833 for ; Wed, 17 Jun 2015 22:03:35 +0100 Received: from dan.rpsys.net ([127.0.0.1]) by localhost (dan.rpsys.net [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 5uyfyDYUozgR for ; Wed, 17 Jun 2015 22:03:35 +0100 (BST) Received: from [192.168.3.10] ([192.168.3.10]) (authenticated bits=0) by dan.rpsys.net (8.14.4/8.14.4/Debian-4.1ubuntu1) with ESMTP id t5HL3NJP014825 (version=TLSv1/SSLv3 cipher=AES128-GCM-SHA256 bits=128 verify=NOT) for ; Wed, 17 Jun 2015 22:03:34 +0100 Message-ID: <1434575003.14710.38.camel@linuxfoundation.org> From: Richard Purdie To: openembedded-core Date: Wed, 17 Jun 2015 22:03:23 +0100 In-Reply-To: <1434546648.14710.25.camel@linuxfoundation.org> References: <1434546648.14710.25.camel@linuxfoundation.org> X-Mailer: Evolution 3.12.10-0ubuntu1~14.10.1 Mime-Version: 1.0 Subject: Re: [PATCH] python: Add performance patches from upstream 2.7 branch X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Patches and discussions about the oe-core layer List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 17 Jun 2015 21:03:39 -0000 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit On Wed, 2015-06-17 at 14:10 +0100, Richard Purdie wrote: > These performance patches result in about a 25% speed up of our parsing > speed with bitbake. They're already applied in 1.7 upstream and backported > from 3.X so worth applying for 2.7 too. Sadly, this was too good to be true. I'd messed up the test environment and the 25% is Ubuntu's python verses the one from OE. Ubuntu's uses PGO whereas ours does not which is likely where the speed comes from. The patches in this series do give a small speedup, but its more like 0.5-1%. Shows that PGO is likely worth investigating further though. Cheers, Richard