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 DEE436013D for ; Thu, 26 Mar 2015 14:30:28 +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 t2QEUSff019693 for ; Thu, 26 Mar 2015 14:30:28 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 ah7HEaIdcfv7 for ; Thu, 26 Mar 2015 14:30:28 +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 t2QEUG0u019684 (version=TLSv1/SSLv3 cipher=AES128-GCM-SHA256 bits=128 verify=NOT) for ; Thu, 26 Mar 2015 14:30:27 GMT Message-ID: <1427380216.14020.100.camel@linuxfoundation.org> From: Richard Purdie To: bitbake-devel Date: Thu, 26 Mar 2015 14:30:16 +0000 X-Mailer: Evolution 3.12.10-0ubuntu1~14.10.1 Mime-Version: 1.0 Subject: Key expansion and overrides? X-BeenThere: bitbake-devel@lists.openembedded.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Patches and discussion that advance bitbake development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 26 Mar 2015 14:30:31 -0000 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Two interesting test cases: OVERIDES = "myoverride" TESTME = "myoverride" FOO = "1" FOO_${TESTME} = "2" FOO = 1 or 2? OVERIDES = "myoverride" TESTME = "myoverride" FOO = "1 2" FOO_remove_${TESTME} = "2" FOO = "1 2" or "1"? Currently expansion in overrides doesn't work. I discovered this trying to do: X_remove_pn-gcc-cross-${TARGET_ARCH} = "Y" which I do believe is a valid thing to attempt. There are two ways of fixing this I guess. One is to call expandKeys() before and after update_data() but I suspect we can still cases which won't work. The other is to move to a model where overrides are applied more dynamically and we lose the need for update_data(). There may be some performance penalty for it. I'm leaning towards trying to kill off update_data() right now... Thoughts/comments welcome. Cheers, Richard