From: David Gibson <david@gibson.dropbear.id.au> To: "Michael S. Tsirkin" <mst@redhat.com> Cc: dgilbert@redhat.com, pair@us.ibm.com, qemu-devel@nongnu.org, brijesh.singh@amd.com, pasic@linux.ibm.com, pragyansri.pathi@intel.com, Greg Kurz <groug@kaod.org>, richard.henderson@linaro.org, berrange@redhat.com, David Hildenbrand <david@redhat.com>, mdroth@linux.vnet.ibm.com, kvm@vger.kernel.org, Marcel Apfelbaum <marcel.apfelbaum@gmail.com>, pbonzini@redhat.com, mtosatti@redhat.com, borntraeger@de.ibm.com, Cornelia Huck <cohuck@redhat.com>, qemu-ppc@nongnu.org, qemu-s390x@nongnu.org, thuth@redhat.com, frankja@linux.ibm.com, jun.nakajima@intel.com, andi.kleen@intel.com, Eduardo Habkost <ehabkost@redhat.com> Subject: Re: [PATCH v8 12/13] confidential guest support: Alter virtio default properties for protected guests Date: Wed, 3 Feb 2021 15:53:04 +1100 [thread overview] Message-ID: <20210203045304.GE2251@yekko.fritz.box> (raw) In-Reply-To: <20210202180328-mutt-send-email-mst@kernel.org> [-- Attachment #1: Type: text/plain, Size: 3187 bytes --] On Tue, Feb 02, 2021 at 06:06:34PM -0500, Michael S. Tsirkin wrote: > On Tue, Feb 02, 2021 at 03:13:14PM +1100, David Gibson wrote: > > The default behaviour for virtio devices is not to use the platforms normal > > DMA paths, but instead to use the fact that it's running in a hypervisor > > to directly access guest memory. That doesn't work if the guest's memory > > is protected from hypervisor access, such as with AMD's SEV or POWER's PEF. > > > > So, if a confidential guest mechanism is enabled, then apply the > > iommu_platform=on option so it will go through normal DMA mechanisms. > > Those will presumably have some way of marking memory as shared with > > the hypervisor or hardware so that DMA will work. > > > > Signed-off-by: David Gibson <david@gibson.dropbear.id.au> > > Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> > > Reviewed-by: Cornelia Huck <cohuck@redhat.com> > > Reviewed-by: Greg Kurz <groug@kaod.org> > > > > --- > > hw/core/machine.c | 13 +++++++++++++ > > 1 file changed, 13 insertions(+) > > > > diff --git a/hw/core/machine.c b/hw/core/machine.c > > index 94194ab82d..497949899b 100644 > > --- a/hw/core/machine.c > > +++ b/hw/core/machine.c > > @@ -33,6 +33,8 @@ > > #include "migration/global_state.h" > > #include "migration/vmstate.h" > > #include "exec/confidential-guest-support.h" > > +#include "hw/virtio/virtio.h" > > +#include "hw/virtio/virtio-pci.h" > > > > GlobalProperty hw_compat_5_2[] = {}; > > const size_t hw_compat_5_2_len = G_N_ELEMENTS(hw_compat_5_2); > > @@ -1196,6 +1198,17 @@ void machine_run_board_init(MachineState *machine) > > * areas. > > */ > > machine_set_mem_merge(OBJECT(machine), false, &error_abort); > > + > > + /* > > + * Virtio devices can't count on directly accessing guest > > + * memory, so they need iommu_platform=on to use normal DMA > > + * mechanisms. That requires also disabling legacy virtio > > + * support for those virtio pci devices which allow it. > > + */ > > + object_register_sugar_prop(TYPE_VIRTIO_PCI, "disable-legacy", > > + "on", true); > > + object_register_sugar_prop(TYPE_VIRTIO_DEVICE, "iommu_platform", > > + "on", false); > > So overriding a boolean property always poses a problem: > if user does set iommu_platform=off we are ignoring this > silently. No, we don't. That's why this is register_sugar_prop() rather than an outright set_prop(). An explicitly given option will take precedence. > Can we change iommu_platform to on/off/auto? Then we can > change how does auto behave. I've never had a satisfactory explanation of what the semantics of "auto" need to be. > > Bonus points for adding "access_platform" and making it > a synonym of platform_iommu. > > > } > > > > machine_class->init(machine); > -- David Gibson | I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 833 bytes --]
WARNING: multiple messages have this Message-ID (diff)
From: David Gibson <david@gibson.dropbear.id.au> To: "Michael S. Tsirkin" <mst@redhat.com> Cc: pair@us.ibm.com, mtosatti@redhat.com, brijesh.singh@amd.com, kvm@vger.kernel.org, David Hildenbrand <david@redhat.com>, qemu-devel@nongnu.org, frankja@linux.ibm.com, borntraeger@de.ibm.com, mdroth@linux.vnet.ibm.com, pasic@linux.ibm.com, pragyansri.pathi@intel.com, andi.kleen@intel.com, thuth@redhat.com, Eduardo Habkost <ehabkost@redhat.com>, richard.henderson@linaro.org, dgilbert@redhat.com, Greg Kurz <groug@kaod.org>, qemu-s390x@nongnu.org, jun.nakajima@intel.com, berrange@redhat.com, Cornelia Huck <cohuck@redhat.com>, qemu-ppc@nongnu.org, pbonzini@redhat.com Subject: Re: [PATCH v8 12/13] confidential guest support: Alter virtio default properties for protected guests Date: Wed, 3 Feb 2021 15:53:04 +1100 [thread overview] Message-ID: <20210203045304.GE2251@yekko.fritz.box> (raw) In-Reply-To: <20210202180328-mutt-send-email-mst@kernel.org> [-- Attachment #1: Type: text/plain, Size: 3187 bytes --] On Tue, Feb 02, 2021 at 06:06:34PM -0500, Michael S. Tsirkin wrote: > On Tue, Feb 02, 2021 at 03:13:14PM +1100, David Gibson wrote: > > The default behaviour for virtio devices is not to use the platforms normal > > DMA paths, but instead to use the fact that it's running in a hypervisor > > to directly access guest memory. That doesn't work if the guest's memory > > is protected from hypervisor access, such as with AMD's SEV or POWER's PEF. > > > > So, if a confidential guest mechanism is enabled, then apply the > > iommu_platform=on option so it will go through normal DMA mechanisms. > > Those will presumably have some way of marking memory as shared with > > the hypervisor or hardware so that DMA will work. > > > > Signed-off-by: David Gibson <david@gibson.dropbear.id.au> > > Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> > > Reviewed-by: Cornelia Huck <cohuck@redhat.com> > > Reviewed-by: Greg Kurz <groug@kaod.org> > > > > --- > > hw/core/machine.c | 13 +++++++++++++ > > 1 file changed, 13 insertions(+) > > > > diff --git a/hw/core/machine.c b/hw/core/machine.c > > index 94194ab82d..497949899b 100644 > > --- a/hw/core/machine.c > > +++ b/hw/core/machine.c > > @@ -33,6 +33,8 @@ > > #include "migration/global_state.h" > > #include "migration/vmstate.h" > > #include "exec/confidential-guest-support.h" > > +#include "hw/virtio/virtio.h" > > +#include "hw/virtio/virtio-pci.h" > > > > GlobalProperty hw_compat_5_2[] = {}; > > const size_t hw_compat_5_2_len = G_N_ELEMENTS(hw_compat_5_2); > > @@ -1196,6 +1198,17 @@ void machine_run_board_init(MachineState *machine) > > * areas. > > */ > > machine_set_mem_merge(OBJECT(machine), false, &error_abort); > > + > > + /* > > + * Virtio devices can't count on directly accessing guest > > + * memory, so they need iommu_platform=on to use normal DMA > > + * mechanisms. That requires also disabling legacy virtio > > + * support for those virtio pci devices which allow it. > > + */ > > + object_register_sugar_prop(TYPE_VIRTIO_PCI, "disable-legacy", > > + "on", true); > > + object_register_sugar_prop(TYPE_VIRTIO_DEVICE, "iommu_platform", > > + "on", false); > > So overriding a boolean property always poses a problem: > if user does set iommu_platform=off we are ignoring this > silently. No, we don't. That's why this is register_sugar_prop() rather than an outright set_prop(). An explicitly given option will take precedence. > Can we change iommu_platform to on/off/auto? Then we can > change how does auto behave. I've never had a satisfactory explanation of what the semantics of "auto" need to be. > > Bonus points for adding "access_platform" and making it > a synonym of platform_iommu. > > > } > > > > machine_class->init(machine); > -- David Gibson | I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 833 bytes --]
next prev parent reply other threads:[~2021-02-03 4:58 UTC|newest] Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top 2021-02-02 4:13 [PATCH v8 00/13] Generalize memory encryption models David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 01/13] qom: Allow optional sugar props David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 02/13] confidential guest support: Introduce new confidential guest support class David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 03/13] sev: Remove false abstraction of flash encryption David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 04/13] confidential guest support: Move side effect out of machine_set_memory_encryption() David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 05/13] confidential guest support: Rework the "memory-encryption" property David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 06/13] sev: Add Error ** to sev_kvm_init() David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 07/13] confidential guest support: Introduce cgs "ready" flag David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-03 10:42 ` Dr. David Alan Gilbert 2021-02-03 10:42 ` Dr. David Alan Gilbert 2021-02-03 16:15 ` Greg Kurz 2021-02-03 16:15 ` Greg Kurz 2021-02-04 2:45 ` David Gibson 2021-02-04 2:45 ` David Gibson 2021-02-10 16:25 ` Venu Busireddy 2021-02-10 16:25 ` Venu Busireddy 2021-02-11 23:48 ` David Gibson 2021-02-11 23:48 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 08/13] confidential guest support: Move SEV initialization into arch specific code David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-03 16:19 ` Greg Kurz 2021-02-03 16:19 ` Greg Kurz 2021-02-02 4:13 ` [PATCH v8 09/13] confidential guest support: Update documentation David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 10/13] spapr: Add PEF based confidential guest support David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-03 17:50 ` Greg Kurz 2021-02-03 17:50 ` Greg Kurz 2021-02-04 2:47 ` David Gibson 2021-02-04 2:47 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 11/13] spapr: PEF: prevent migration David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 12/13] confidential guest support: Alter virtio default properties for protected guests David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-02 23:06 ` Michael S. Tsirkin 2021-02-02 23:06 ` Michael S. Tsirkin 2021-02-03 4:53 ` David Gibson [this message] 2021-02-03 4:53 ` David Gibson 2021-02-02 4:13 ` [PATCH v8 13/13] s390: Recognize confidential-guest-support option David Gibson 2021-02-02 4:13 ` David Gibson 2021-02-03 9:05 ` Christian Borntraeger 2021-02-03 9:05 ` Christian Borntraeger
Reply instructions: You may reply publicly to this message via plain-text email using any one of the following methods: * Save the following mbox file, import it into your mail client, and reply-to-all from there: mbox Avoid top-posting and favor interleaved quoting: https://en.wikipedia.org/wiki/Posting_style#Interleaved_style * Reply using the --to, --cc, and --in-reply-to switches of git-send-email(1): git send-email \ --in-reply-to=20210203045304.GE2251@yekko.fritz.box \ --to=david@gibson.dropbear.id.au \ --cc=andi.kleen@intel.com \ --cc=berrange@redhat.com \ --cc=borntraeger@de.ibm.com \ --cc=brijesh.singh@amd.com \ --cc=cohuck@redhat.com \ --cc=david@redhat.com \ --cc=dgilbert@redhat.com \ --cc=ehabkost@redhat.com \ --cc=frankja@linux.ibm.com \ --cc=groug@kaod.org \ --cc=jun.nakajima@intel.com \ --cc=kvm@vger.kernel.org \ --cc=marcel.apfelbaum@gmail.com \ --cc=mdroth@linux.vnet.ibm.com \ --cc=mst@redhat.com \ --cc=mtosatti@redhat.com \ --cc=pair@us.ibm.com \ --cc=pasic@linux.ibm.com \ --cc=pbonzini@redhat.com \ --cc=pragyansri.pathi@intel.com \ --cc=qemu-devel@nongnu.org \ --cc=qemu-ppc@nongnu.org \ --cc=qemu-s390x@nongnu.org \ --cc=richard.henderson@linaro.org \ --cc=thuth@redhat.com \ /path/to/YOUR_REPLY https://kernel.org/pub/software/scm/git/docs/git-send-email.html * If your mail client supports setting the In-Reply-To header via mailto: links, try the mailto: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.