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 4B26E7274A for ; Tue, 23 Dec 2014 10:47:21 +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 sBNAkexv010440; Tue, 23 Dec 2014 10:46:40 GMT 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 dxRgAzcfGa0d; Tue, 23 Dec 2014 10:46:40 +0000 (GMT) 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 sBNAkQTM010424 (version=TLSv1/SSLv3 cipher=AES128-GCM-SHA256 bits=128 verify=NOT); Tue, 23 Dec 2014 10:46:38 GMT Message-ID: <1419331623.1777.12.camel@linuxfoundation.org> From: Richard Purdie To: Hongxu Jia Date: Tue, 23 Dec 2014 10:47:03 +0000 In-Reply-To: References: X-Mailer: Evolution 3.12.7-0ubuntu1 Mime-Version: 1.0 Cc: paul.eggleton@linux.intel.com, openembedded-core@lists.openembedded.org Subject: Re: [PATCH 1/4] insane.bbclass: add QA check: package-missing 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: Tue, 23 Dec 2014 10:47:25 -0000 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit On Thu, 2014-12-11 at 18:27 +0800, Hongxu Jia wrote: > During packaging, check if there are any dependencies (RDEPENDS) > on packages that have ended up empty and not produced; and if so > produce an warn/error as soon as possible, instead of allowing > the build to proceed up to do_rootfs and then the package manager > reporting the package as missing. > > At the moment, we use bb.persist_data as a global database to > collect all available packages rdepends and rprovides. So in the > do_package_qa task, while package not produced, we could compute > rdepends chain, and report warn/error immediately. I liked this until I saw you were using bb.persist_data. That isn't under sstate control and will not result in deterministic builds. Have you looked at the data saved by do_packageinfo into the pkgdata structures? Can we not use the information on RDEPENDS and RPROVIDES there instead? Cheers, Richard