From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:46986) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1dXS3U-000498-6y for qemu-devel@nongnu.org; Tue, 18 Jul 2017 08:56:49 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1dXS3Q-0003TC-64 for qemu-devel@nongnu.org; Tue, 18 Jul 2017 08:56:48 -0400 Received: from mx1.redhat.com ([209.132.183.28]:36374) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1dXS3P-0003Sv-Vu for qemu-devel@nongnu.org; Tue, 18 Jul 2017 08:56:44 -0400 Date: Tue, 18 Jul 2017 13:56:35 +0100 From: "Daniel P. Berrange" Message-ID: <20170718125635.GI11927@redhat.com> Reply-To: "Daniel P. Berrange" References: <20170407143847.GM2138@work-vm> <1573e08f-9805-eaf4-3fe1-c1172b6ef6e4@amsat.org> <1c55bc6d-7ef7-dcf5-81e8-22712f575691@amsat.org> <6e4fb953-17b4-a11b-4f98-8ddfd5ff12f8@redhat.com> <39dc6b92-dd56-fe33-cd85-92c03d69f133@redhat.com> <20170717164641.GA31820@work-vm> <20170718072351.GB11927@redhat.com> <503be976-144b-d468-04b2-12dd99947bf1@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <503be976-144b-d468-04b2-12dd99947bf1@redhat.com> Subject: Re: [Qemu-devel] Hacks for building on gcc 7 / Fedora 26 List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eric Blake Cc: "Dr. David Alan Gilbert" , Gerd Hoffmann , Philippe =?utf-8?Q?Mathieu-Daud=C3=A9?= , qemu-devel@nongnu.org On Tue, Jul 18, 2017 at 07:35:58AM -0500, Eric Blake wrote: > On 07/18/2017 02:23 AM, Daniel P. Berrange wrote: > >> How would it not help? Are we using gcc 7 on windows builds? Adding > >> the assert is enough to shut up new gcc; old gcc was already silent; and > >> if mingw is still on old gcc, it doesn't matter whether assert() is > >> marked noreturn for what this patch is doing. > > > > Mingw isn't using a fork of GCC anymore, its all mainline. Thus Fedora's > > mingw gcc packages track native gcc packages. IOW i686-w64-mingw32-gcc > > is already on version 7.1.0 in Fedora > > So I guess that means running 'make docker-test-mingw@fedora' on Fedora > 26 will find out if we need further tweaks? Trying it now... > > Nope, didn't get very far :( > > $ make docker-test-mingw@fedora > BUILD fedora > make[1]: Entering directory '/home/eblake/qemu' > ARCHIVE qemu.tgz > usage: git archive [] [...] > ... > make[1]: *** [/home/eblake/qemu/tests/docker/Makefile.include:35: > docker-src.2017-07-18-07.35.20.4101] Error 129 > make[1]: Leaving directory '/home/eblake/qemu' > make: *** [/home/eblake/qemu/tests/docker/Makefile.include:156: > docker-run-test-mingw@fedora] Error 2 > > Am I not doing it right? That works fine for me - perhaps V=1 will tell you what's wrong with the git archive command. In any case, despite having gcc 7.1.0, I don't see the error messages in question when running a mingw build. It could be that the new problems GCC 7 reports are only triggered when combined with a suitable C library like GCC, to get the symbols annotated to enable fortify source to check them. Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|