From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:42588) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1X0UKi-0007Hr-E4 for qemu-devel@nongnu.org; Fri, 27 Jun 2014 07:28:54 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1X0UKa-0008SH-IQ for qemu-devel@nongnu.org; Fri, 27 Jun 2014 07:28:44 -0400 Received: from oxygen.pond.sub.org ([144.76.244.19]:44474) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1X0UKa-0008S0-CS for qemu-devel@nongnu.org; Fri, 27 Jun 2014 07:28:36 -0400 Received: from blackfin.pond.sub.org (p5B329A43.dip0.t-ipconnect.de [91.50.154.67]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client did not present a certificate) by oxygen.pond.sub.org (Postfix) with ESMTPSA id 7404B2560B for ; Fri, 27 Jun 2014 13:28:33 +0200 (CEST) From: Markus Armbruster References: <1400966700-23165-1-git-send-email-mreitz@redhat.com> <53938249.4010504@redhat.com> <53AC93B3.9070707@redhat.com> Date: Fri, 27 Jun 2014 13:28:32 +0200 In-Reply-To: <53AC93B3.9070707@redhat.com> (Max Reitz's message of "Thu, 26 Jun 2014 23:42:11 +0200") Message-ID: <87simqy4kf.fsf@blackfin.pond.sub.org> MIME-Version: 1.0 Content-Type: text/plain Subject: Re: [Qemu-devel] [PATCH v4 0/6] iotests: Allow out-of-tree run List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Max Reitz Cc: Kevin Wolf , Fam Zheng , qemu-devel@nongnu.org, Stefan Hajnoczi Max Reitz writes: > On 07.06.2014 23:21, Max Reitz wrote: >> On 24.05.2014 23:24, Max Reitz wrote: >>> This series enables qemu-iotests to be run in a build tree outside of >>> the source tree. It also makes the tests use the command for invoking >>> the Python interpreter specified through configure instead of always >>> using "/usr/bin/env python". >> >> Ping; I do understand that this series is not urgent, but since I >> realized out-of-tree builds to be probably superior, I personally >> base all my own patches on this series, as I don't want to fiddle >> around with the iotests. Therefore, I'd be glad if someone would >> review the remaining patches so it can be merged soon. :-) > > Ping again. Because this is just convenient for development, I don't > need it in any specific release, though. I haven't found the time for a proper review, and I can't promise one right now, so I should probably keep my mouth where my money is, but here goes anyway: unless running tests is utterly trivial, tests will not be run, and avoidable mistakes happen. Case in point: I spent a non-trivial chunk of time yesterday to debug three regressions clearly visible in iotests. I did not scold the people involved in getting the regressions committed for not running these tests, because I feel strongly I can't demand tests to be run that require instructions more complex than "make WHATEVER". I don't think this is "just convenient for development". I'd say it's a must-have.