From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:49222) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aVykv-0000uC-JJ for qemu-devel@nongnu.org; Wed, 17 Feb 2016 04:50:46 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1aVyks-0008LA-Ch for qemu-devel@nongnu.org; Wed, 17 Feb 2016 04:50:45 -0500 Received: from indium.canonical.com ([91.189.90.7]:46196) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1aVykr-0008L0-OV for qemu-devel@nongnu.org; Wed, 17 Feb 2016 04:50:41 -0500 Received: from loganberry.canonical.com ([91.189.90.37]) by indium.canonical.com with esmtp (Exim 4.76 #1 (Debian)) id 1aVykq-0006eP-6p for ; Wed, 17 Feb 2016 09:50:40 +0000 Received: from loganberry.canonical.com (localhost [127.0.0.1]) by loganberry.canonical.com (Postfix) with ESMTP id 0CFC42E80D4 for ; Wed, 17 Feb 2016 09:50:40 +0000 (UTC) MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Date: Wed, 17 Feb 2016 09:42:37 -0000 From: Liang Chen Sender: bounces@canonical.com Message-Id: <20160217094237.24220.45325.malonedeb@chaenomeles.canonical.com> Errors-To: bounces@canonical.com Subject: [Qemu-devel] [Bug 1546445] [NEW] support vhost user without specifying vhostforce Reply-To: Bug 1546445 <1546445@bugs.launchpad.net> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Public bug reported: [Impact] * vhost-user falls back to virtio-net which causes performance lose without specifying the vhostforce option. But it should be the default behavior for vhost-user, since guests using PMD doesn't support msi-x. [Test Case] create a vhost-user virtio backend without specifying the vhostforce opti= on, i.e. -netdev type=3Dvhost-user,id=3Dmynet1,chardev=3D start the VM vhost-user is not enabled [Regression Potential] * none vhost user nic doesn't support non msi guests(like pxe stage) by default. Vhost user nic can't fall back to qemu like normal vhost net nic does. So w= e should enable it for non msi guests. The problem has been fix in qemu upstream - http://git.qemu.org/?p=3Dqemu.git;a=3Dcommitdiff;h=3D24f938a682d934b133863e= b421aac33592f7a09e. And the patch needs to be backported to 1:2.2+dfsg-5expubuntu9.8 . ** Affects: qemu Importance: Undecided Assignee: Liang Chen (cbjchen) Status: New ** Changed in: qemu Assignee: (unassigned) =3D> Liang Chen (cbjchen) -- = You received this bug notification because you are a member of qemu- devel-ml, which is subscribed to QEMU. https://bugs.launchpad.net/bugs/1546445 Title: support vhost user without specifying vhostforce Status in QEMU: New Bug description: [Impact] * vhost-user falls back to virtio-net which causes performance lose without specifying the vhostforce option. But it should be the default behavior for vhost-user, since guests using PMD doesn't support msi-x. [Test Case] create a vhost-user virtio backend without specifying the vhostforce op= tion, i.e. -netdev type=3Dvhost-user,id=3Dmynet1,chardev=3D start the VM vhost-user is not enabled [Regression Potential] * none vhost user nic doesn't support non msi guests(like pxe stage) by default. Vhost user nic can't fall back to qemu like normal vhost net nic does. So= we should enable it for non msi guests. The problem has been fix in qemu upstream - http://git.qemu.org/?p=3Dqemu.git;a=3Dcommitdiff;h=3D24f938a682d934b13386= 3eb421aac33592f7a09e. And the patch needs to be backported to 1:2.2+dfsg-5expubuntu9.8 . To manage notifications about this bug go to: https://bugs.launchpad.net/qemu/+bug/1546445/+subscriptions