From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:33973) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1ZjBQ3-0000pa-GD for qemu-devel@nongnu.org; Mon, 05 Oct 2015 15:27:32 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1ZjBQ2-0002K3-Gv for qemu-devel@nongnu.org; Mon, 05 Oct 2015 15:27:31 -0400 Sender: Paolo Bonzini References: <9B0F7F7B0E031F4A99FC8A50E22E3AA360FC1AFD@EU-MBX-01.mgc.mentorg.com> <9B0F7F7B0E031F4A99FC8A50E22E3AA360FC1B8E@EU-MBX-01.mgc.mentorg.com> <561103DF.5000704@redhat.com> <9B0F7F7B0E031F4A99FC8A50E22E3AA360FC2658@EU-MBX-01.mgc.mentorg.com> From: Paolo Bonzini Message-ID: <5612CF1D.6050400@redhat.com> Date: Mon, 5 Oct 2015 21:27:25 +0200 MIME-Version: 1.0 In-Reply-To: <9B0F7F7B0E031F4A99FC8A50E22E3AA360FC2658@EU-MBX-01.mgc.mentorg.com> Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Subject: Re: [Qemu-devel] [Qemu-discuss] TCP options ipv4 and ipv6 have no effect List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: "Sair, Umair" , Peter Maydell Cc: QEMU Developers , "qemu-discuss@nongnu.org" On 05/10/2015 20:03, Sair, Umair wrote: >> The first if handles the "default to N" case, the second handles "default to Y", the (absent) else case handles "default to PF_UNSPEC". > > Can you please elaborate it. Also I am not understanding the reason for inverting the values of addr->has_ipv* in second if condition. If a value is absent, you can either default it to Y (then the value is !addr->has_xyz || addr->xyz) or default it to N (then the value is addr->has_xyz && addr->xyz). The desired logic is: - with one absent value and one "on", the absent value should be "off" - with one absent value and one "off", the absent value should be "on" - with two absent values, the absent values can be both left absent The patch works like this: >> static void inet_addr_to_opts(QemuOpts *opts, const InetSocketAddress *addr) { >> - bool ipv4 = addr->ipv4 || !addr->has_ipv4; >> - bool ipv6 = addr->ipv6 || !addr->has_ipv6; >> + bool ipv4 = addr->has_ipv4 && addr->ipv4; >> + bool ipv6 = addr->has_ipv6 && addr->ipv6; >> >> - if (!ipv4 || !ipv6) { >> + if (ipv4 || ipv6) { >> qemu_opt_set_bool(opts, "ipv4", ipv4, &error_abort); >> qemu_opt_set_bool(opts, "ipv6", ipv6, &error_abort); This handles the case where the absent value should be "off": at least one value is present _and_ true. >> + } else if (addr->has_ipv4 || addr->has_ipv6) { >> + qemu_opt_set_bool(opts, "ipv4", !addr->has_ipv4, &error_abort); >> + qemu_opt_set_bool(opts, "ipv6", !addr->has_ipv6, &error_abort); This handles the case where the absent value should be "on". We know that whichever the present value is, it is false; therefore !addr->has_xyz || addr->xyz simplifies to just !addr->has_xyz. > I believe that the fix for the issue under discussion will be > committed to qemu repo very soon, Did you test the patch, and did it work for you? If so, it is customary to reply with a line like "Tested by: Sair, Umair ". > so I'll like to add one more thing > which requires to be fixed along with it. In 'tcp_chr_accept' > function of qemu-char.c, the data type of saddr should be > sockaddr_in6 so that it works with both IPv6 and IPv4 on Windows > (works for linux without it because of accept4 and works with this > solution as well!). Can you send a patch for it? Thanks! Paolo