From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:56889) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bozxp-0007XS-HW for qemu-devel@nongnu.org; Tue, 27 Sep 2016 17:30:58 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1bozxl-0007AN-CX for qemu-devel@nongnu.org; Tue, 27 Sep 2016 17:30:56 -0400 Received: from mx6-phx2.redhat.com ([209.132.183.39]:44529) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1bozxl-00079D-3F for qemu-devel@nongnu.org; Tue, 27 Sep 2016 17:30:53 -0400 Date: Tue, 27 Sep 2016 17:30:51 -0400 (EDT) From: Paolo Bonzini Message-ID: <1072277924.3196177.1475011851599.JavaMail.zimbra@redhat.com> In-Reply-To: <20160927140138.GD26148@potion> References: <20160922210432.18680-1-rkrcmar@redhat.com> <20160922210432.18680-6-rkrcmar@redhat.com> <20160927140138.GD26148@potion> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH 5/5] intel_iommu: do not allow EIM without KVM support List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Radim =?utf-8?B?S3LEjW3DocWZ?= Cc: qemu-devel@nongnu.org, Peter Xu , Igor Mammedov , Richard Henderson , Eduardo Habkost , "Michael S. Tsirkin" ----- Original Message ----- > From: "Radim Kr=C4=8Dm=C3=A1=C5=99" > To: "Paolo Bonzini" > Cc: qemu-devel@nongnu.org, "Peter Xu" , "Igor Mammedov= " , "Richard Henderson" > , "Eduardo Habkost" , "Michael S. T= sirkin" > Sent: Tuesday, September 27, 2016 4:01:39 PM > Subject: Re: [PATCH 5/5] intel_iommu: do not allow EIM without KVM suppor= t >=20 > 2016-09-23 11:27+0200, Paolo Bonzini: > > Since the whole IOMMU feature is new and somewhat experimental, I think > > it's okay to just make EIM the default for >=3D2.8 machine types if KVM= is > > on (using DEFINE_PROP_ON_OFF_AUTO; auto means true if KVM is on and >=20 > Sounds good. BTW this also means KVM+vIOMMU requires 4.8. Let's remember to document it in the release notes. > > false otherwise, and pc-2.7 would set eim=3Doff). >=20 > What about eim=3Don in pc-2.7, to avoid breaking migration? Yup, just a thinko. Paolo