From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:33969) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1e7iEe-00021k-JO for qemu-devel@nongnu.org; Thu, 26 Oct 2017 09:30:13 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1e7iEb-00023D-Td for qemu-devel@nongnu.org; Thu, 26 Oct 2017 09:30:12 -0400 Received: from mail-qk0-x234.google.com ([2607:f8b0:400d:c09::234]:46518) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1e7iEb-000225-PC for qemu-devel@nongnu.org; Thu, 26 Oct 2017 09:30:09 -0400 Received: by mail-qk0-x234.google.com with SMTP id k123so4235511qke.3 for ; Thu, 26 Oct 2017 06:30:09 -0700 (PDT) Sender: =?UTF-8?Q?Philippe_Mathieu=2DDaud=C3=A9?= References: <20171025123056.3165-1-richard.henderson@linaro.org> <20171026070608.GC5847@redhat.com> <20171026100645.GA24581@redhat.com> From: =?UTF-8?Q?Philippe_Mathieu-Daud=c3=a9?= Message-ID: <8774f2a0-274f-b358-c316-d3484f15b2cb@amsat.org> Date: Thu, 26 Oct 2017 10:29:59 -0300 MIME-Version: 1.0 In-Reply-To: <20171026100645.GA24581@redhat.com> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 8bit Subject: Re: [Qemu-devel] [PULL 00/11] Disassembler patches List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Daniel P. Berrange" , Peter Maydell Cc: Richard Henderson , QEMU Developers On 10/26/2017 07:06 AM, Daniel P. Berrange wrote: > Do we really expect make/configure todo the right thing when going > backwards in time ? "Yes"? Ideally at least :) > I've always assumed that if you go back in time > when you need to do a 'git clean -f -x d' and re-run configure from> scratch. I certainly don't do that and would rather not have to think about it. I also don't think about calling it when bisecting, however looking at the man page, this is somehow suggested in the example: ยท Automatically bisect with temporary modifications (hot-fix): [...] # undo the tweak to allow clean flipping to the next commit git reset --hard This could be a warning displayed via a post-checkout hook, for this particular merge hash, or if you go way too back in time... > Certainly in the past various makefile changes in QEMU would> break, or silently not correctly recompile stuff when going backwards > in time. Travis is already taking too long, but we could add a such test (merge/build/checkout backward/configure/quick build), not sure if we gain much.