From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from dan.rpsys.net ([93.97.175.187]) by linuxtogo.org with esmtp (Exim 4.72) (envelope-from ) id 1UD1sG-0008A4-Hi for openembedded-core@lists.openembedded.org; Wed, 06 Mar 2013 01:06:25 +0100 Received: from localhost (dan.rpsys.net [127.0.0.1]) by dan.rpsys.net (8.14.4/8.14.4/Debian-2.1ubuntu1) with ESMTP id r25Nusga009254; Tue, 5 Mar 2013 23:57:40 GMT X-Virus-Scanned: Debian amavisd-new at dan.rpsys.net 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 i2V71NE7Q-Ct; Tue, 5 Mar 2013 23:57:40 +0000 (GMT) Received: from [192.168.3.10] (rpvlan0 [192.168.3.10]) (authenticated bits=0) by dan.rpsys.net (8.14.4/8.14.4/Debian-2.1ubuntu1) with ESMTP id r25NvWqM009276 (version=TLSv1/SSLv3 cipher=DHE-RSA-CAMELLIA256-SHA bits=256 verify=NOT); Tue, 5 Mar 2013 23:57:36 GMT Message-ID: <1362527371.11727.50.camel@ted> From: Richard Purdie To: Ross Burton Date: Tue, 05 Mar 2013 23:49:31 +0000 In-Reply-To: References: X-Mailer: Evolution 3.6.3-1 Mime-Version: 1.0 Cc: openembedded-core@lists.openembedded.org Subject: Re: [PATCH 0/4] Don't build gtk+-native X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.11 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, 06 Mar 2013 00:06:27 -0000 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit On Mon, 2013-03-04 at 12:52 +0000, Ross Burton wrote: > We're building gtk+-native (so the X stack, freetype, fontconfig, etc etc) just > to run gtk-update-icon-cache on the host when building the rootfs. > > This series adds a recipe that builds just that binary from a GTK+ tarball using > a bit of hackery. > > For a benchmark of building sato-icon-theme from nothing (no sstate or tmp): > > master takes 13:25 to run 1037 tasks. > ross/gtk takes 9:26 to run 629 tasks. > > So that's a four minute gain and 400 less tasks. Here, my core-image-sato build from populated DL_DIR time fell by about 2.5 minutes due to this changeset (46:59 to 44:17 mins) so a 5% speedup. Nice! :) It would seemingly make around six seconds on "bitbake core-image-sato -c rootfs" from sstate too (5m25 ish). Cheers, Richard