From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([209.51.188.92]:37535) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1h1UW9-0002pR-4J for qemu-devel@nongnu.org; Wed, 06 Mar 2019 06:15:21 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1h1UW8-00061D-4N for qemu-devel@nongnu.org; Wed, 06 Mar 2019 06:15:20 -0500 Received: from mx1.redhat.com ([209.132.183.28]:43700) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1h1UW7-00060e-P9 for qemu-devel@nongnu.org; Wed, 06 Mar 2019 06:15:20 -0500 References: <1551723614-1823-1-git-send-email-pbonzini@redhat.com> <3c67eb5e-6a7e-d884-9ce7-639584c0c63f@redhat.com> From: Paolo Bonzini Message-ID: <7e49a9e0-a532-5fe2-d17f-040f3e8fe3eb@redhat.com> Date: Wed, 6 Mar 2019 12:15:15 +0100 MIME-Version: 1.0 In-Reply-To: <3c67eb5e-6a7e-d884-9ce7-639584c0c63f@redhat.com> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PULL 00/54] Kconfig conversion, excluding ARM and MIPS List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Thomas Huth , Peter Maydell Cc: QEMU Developers , Yang Zhong On 06/03/19 12:04, Thomas Huth wrote: > On 06/03/2019 11.57, Paolo Bonzini wrote: >> On 05/03/19 10:32, Peter Maydell wrote: >>> Everything fails to build with errors like >>> make: *** No rule to make target `config-all-devices.mak', needed by >>> `subdir-aarch64-softmmu'. Stop. >>> >>> or >>> make: *** No rule to make target >>> `/home/pm215/qemu/default-configs/pci.mak', needed by >>> `aarch64-softmmu/config-devices.mak'. Stop. >>> >>> If there are special instructions for what to do with >>> build trees over the transition to kconfig, the pullreq >>> cover letter would be a good place to mention them :-) >> >> As others mentioned, this is the "usual" stale dependency issue >> requiring distclean. > > Can't we somehow add a dependency for the *.mak.d file so that they get > regenerated in this case? ... I tried to hack the Makefile, but so far I > failed... What we should do is add dummy dependencies for the .mak files, so that they do not trigger the error messages; this is how gcc -MD avoids the issue for include files. With Kconfig it should be a one-line patch, something like for fname in data.previously_included: print('%s: %s' % (argv[1], fname), file=deps) print('%s:' % fname, file=deps) but it wouldn't fix this case where the stale .mak file is preexisting. Paolo