2020-08-19 16:24:56

by Pierre Morel

[permalink] [raw]
Subject: [PATCH v9 1/2] virtio: let arch advertise guest's memory access restrictions

An architecture may restrict host access to guest memory.

Provide a new Kconfig entry the architecture can select,
CONFIG_ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS, when it provides
the arch_has_restricted_virtio_memory_access callback to advertise
VIRTIO common code when the architecture restricts memory access
from the host.

Signed-off-by: Pierre Morel <[email protected]>
---
drivers/virtio/Kconfig | 6 ++++++
drivers/virtio/virtio.c | 15 +++++++++++++++
include/linux/virtio_config.h | 9 +++++++++
3 files changed, 30 insertions(+)

diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
index 5809e5f5b157..509f3b4d8ba1 100644
--- a/drivers/virtio/Kconfig
+++ b/drivers/virtio/Kconfig
@@ -6,6 +6,12 @@ config VIRTIO
bus, such as CONFIG_VIRTIO_PCI, CONFIG_VIRTIO_MMIO, CONFIG_RPMSG
or CONFIG_S390_GUEST.

+config ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS
+ bool
+ help
+ This option is selected by any architecture enforcing
+ VIRTIO_F_IOMMU_PLATFORM
+
menuconfig VIRTIO_MENU
bool "Virtio drivers"
default y
diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c
index a977e32a88f2..57ca41d77b09 100644
--- a/drivers/virtio/virtio.c
+++ b/drivers/virtio/virtio.c
@@ -176,6 +176,21 @@ int virtio_finalize_features(struct virtio_device *dev)
if (ret)
return ret;

+ ret = arch_has_restricted_virtio_memory_access();
+ if (ret) {
+ if (!virtio_has_feature(dev, VIRTIO_F_VERSION_1)) {
+ dev_warn(&dev->dev,
+ "device must provide VIRTIO_F_VERSION_1\n");
+ return -ENODEV;
+ }
+
+ if (!virtio_has_feature(dev, VIRTIO_F_IOMMU_PLATFORM)) {
+ dev_warn(&dev->dev,
+ "device must provide VIRTIO_F_IOMMU_PLATFORM\n");
+ return -ENODEV;
+ }
+ }
+
if (!virtio_has_feature(dev, VIRTIO_F_VERSION_1))
return 0;

diff --git a/include/linux/virtio_config.h b/include/linux/virtio_config.h
index bb4cc4910750..e380664642b4 100644
--- a/include/linux/virtio_config.h
+++ b/include/linux/virtio_config.h
@@ -459,4 +459,13 @@ static inline void virtio_cwrite64(struct virtio_device *vdev,
_r; \
})

+#ifdef CONFIG_ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS
+int arch_has_restricted_virtio_memory_access(void);
+#else
+static inline int arch_has_restricted_virtio_memory_access(void)
+{
+ return 0;
+}
+#endif /* CONFIG_ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS */
+
#endif /* _LINUX_VIRTIO_CONFIG_H */
--
2.25.1


2020-08-21 12:03:01

by Cornelia Huck

[permalink] [raw]
Subject: Re: [PATCH v9 1/2] virtio: let arch advertise guest's memory access restrictions

On Wed, 19 Aug 2020 18:23:17 +0200
Pierre Morel <[email protected]> wrote:

> An architecture may restrict host access to guest memory.

"e.g. IBM s390 Secure Execution or AMD SEV"

Just to make clearer what you are referring to?

>
> Provide a new Kconfig entry the architecture can select,
> CONFIG_ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS, when it provides
> the arch_has_restricted_virtio_memory_access callback to advertise

s/advertise/advertise to/

> VIRTIO common code when the architecture restricts memory access
> from the host.

"The common code can then fail the probe for any device where
VIRTIO_F_IOMMU_PLATFORM is required, but not set."

?

>
> Signed-off-by: Pierre Morel <[email protected]>
> ---
> drivers/virtio/Kconfig | 6 ++++++
> drivers/virtio/virtio.c | 15 +++++++++++++++
> include/linux/virtio_config.h | 9 +++++++++
> 3 files changed, 30 insertions(+)
>
> diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
> index 5809e5f5b157..509f3b4d8ba1 100644
> --- a/drivers/virtio/Kconfig
> +++ b/drivers/virtio/Kconfig
> @@ -6,6 +6,12 @@ config VIRTIO
> bus, such as CONFIG_VIRTIO_PCI, CONFIG_VIRTIO_MMIO, CONFIG_RPMSG
> or CONFIG_S390_GUEST.
>
> +config ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS
> + bool
> + help
> + This option is selected by any architecture enforcing
> + VIRTIO_F_IOMMU_PLATFORM

"This option is selected if the architecture may need to enforce
VIRTIO_F_IOMMU_PLATFORM."

?

> +
> menuconfig VIRTIO_MENU
> bool "Virtio drivers"
> default y

(...)

Reviewed-by: Cornelia Huck <[email protected]>

2020-08-21 13:11:30

by Pierre Morel

[permalink] [raw]
Subject: Re: [PATCH v9 1/2] virtio: let arch advertise guest's memory access restrictions



On 2020-08-21 13:59, Cornelia Huck wrote:
> On Wed, 19 Aug 2020 18:23:17 +0200
> Pierre Morel <[email protected]> wrote:
>
>> An architecture may restrict host access to guest memory.
>
> "e.g. IBM s390 Secure Execution or AMD SEV"
>
> Just to make clearer what you are referring to?

yes, thanks

>
>>
>> Provide a new Kconfig entry the architecture can select,
>> CONFIG_ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS, when it provides
>> the arch_has_restricted_virtio_memory_access callback to advertise
>
> s/advertise/advertise to/

OK

>
>> VIRTIO common code when the architecture restricts memory access
>> from the host.
>
> "The common code can then fail the probe for any device where
> VIRTIO_F_IOMMU_PLATFORM is required, but not set."
>
> ?

Yes, better thanks

>
>>
>> Signed-off-by: Pierre Morel <[email protected]>
>> ---
>> drivers/virtio/Kconfig | 6 ++++++
>> drivers/virtio/virtio.c | 15 +++++++++++++++
>> include/linux/virtio_config.h | 9 +++++++++
>> 3 files changed, 30 insertions(+)
>>
>> diff --git a/drivers/virtio/Kconfig b/drivers/virtio/Kconfig
>> index 5809e5f5b157..509f3b4d8ba1 100644
>> --- a/drivers/virtio/Kconfig
>> +++ b/drivers/virtio/Kconfig
>> @@ -6,6 +6,12 @@ config VIRTIO
>> bus, such as CONFIG_VIRTIO_PCI, CONFIG_VIRTIO_MMIO, CONFIG_RPMSG
>> or CONFIG_S390_GUEST.
>>
>> +config ARCH_HAS_RESTRICTED_VIRTIO_MEMORY_ACCESS
>> + bool
>> + help
>> + This option is selected by any architecture enforcing
>> + VIRTIO_F_IOMMU_PLATFORM
>
> "This option is selected if the architecture may need to enforce
> VIRTIO_F_IOMMU_PLATFORM."
>
> ?

yes, better thanks

>
>> +
>> menuconfig VIRTIO_MENU
>> bool "Virtio drivers"
>> default y
>
> (...)
>
> Reviewed-by: Cornelia Huck <[email protected]>
>

I will make the rewordings.

Thanks,
Pierre

--
Pierre Morel
IBM Lab Boeblingen