From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:47012) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dVHIp-0003JT-9A for qemu-devel@nongnu.org; Wed, 12 Jul 2017 09:03:41 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dVHIk-0004rZ-DZ for qemu-devel@nongnu.org; Wed, 12 Jul 2017 09:03:39 -0400 Received: from mx1.redhat.com ([209.132.183.28]:18562) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1dVHIk-0004r9-7O for qemu-devel@nongnu.org; Wed, 12 Jul 2017 09:03:34 -0400 From: Markus Armbruster References: <12e6dde67e57a0ecf09fbe6103512543113ca07a.1499774331.git.alistair.francis@xilinx.com> <040c4483-73f0-7f3b-d2af-b56bc83b0e53@redhat.com> Date: Wed, 12 Jul 2017 15:03:31 +0200 In-Reply-To: (Alistair Francis's message of "Wed, 12 Jul 2017 14:27:43 +0200") Message-ID: <87zic96c4c.fsf@dusky.pond.sub.org> MIME-Version: 1.0 Content-Type: text/plain Subject: Re: [Qemu-devel] [PATCH v3 8/8] error: Add a 'error: ' prefix to error_report() List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Alistair Francis Cc: Max Reitz , philippe@mathieu-daude.net, "qemu-devel@nongnu.org Developers" Alistair Francis writes: > On Tue, Jul 11, 2017 at 7:44 PM, Max Reitz wrote: >> On 2017-07-11 14:07, Alistair Francis wrote: >>> As we don't regard error messages as a stable API the let's add a >>> 'error: ' prefix to the original error_report() messages. >>> >>> Signed-off-by: Alistair Francis >>> --- >>> >>> util/qemu-error.c | 1 + >>> 1 file changed, 1 insertion(+) >> >> This breaks quite a few qemu-iotests. > > Hmmm... Ok, I'll have to dig into that. > > I'm traveling at the moment so I won't get a chance to fix this. I'm > going to send the next version of this series, but remove this patch > and I can look at adding it again when I get back. There is already a > few things that have been pointed out that need to be fixed after this > patch set. Separating this change from the rest of the series makes sense, because it lets use get the bulk of your work in more quickly. > Do people think we should add this prefix or leave the error messages > as is? I haven't heard a definitive answer if people think this is the > right path to take. I'd leave them as is. More inertia than opposition. If you think a prefix solves a problem people have, by all means post a patch. Please explain in the commit message why the patch is useful. The current one explains only why we can change error messages, not why we *want* to change them.