diff mbox series

[v8,4/4] vfio: convey kvm that the vfio-pci device is wc safe

Message ID 20240220072926.6466-5-ankita@nvidia.com (mailing list archive)
State New, archived
Headers show
Series kvm: arm64: allow the VM to select DEVICE_* and NORMAL_NC for IO memory | expand

Commit Message

Ankit Agrawal Feb. 20, 2024, 7:29 a.m. UTC
From: Ankit Agrawal <ankita@nvidia.com>

The VM_ALLOW_ANY_UNCACHED flag is implemented for ARM64,
allowing KVM stage 2 device mapping attributes to use Normal-NC
rather than DEVICE_nGnRE, which allows guest mappings
supporting combining attributes (WC). ARM does not architecturally
guarantee this is safe, and indeed some MMIO regions like the GICv2
VCPU interface can trigger uncontained faults if Normal-NC is used.

To safely use VFIO in KVM the platform must guarantee full safety
in the guest where no action taken against a MMIO mapping can
trigger an uncontained failure. We belive that most VFIO PCI
platforms support this for both mapping types, at least in common
flows, based on some expectations of how PCI IP is integrated. So
make vfio-pci set the VM_ALLOW_ANY_UNCACHED flag.

Suggested-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Jason Gunthorpe <jgg@nvidia.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Ankit Agrawal <ankita@nvidia.com>
---
 drivers/vfio/pci/vfio_pci_core.c | 18 +++++++++++++++++-
 1 file changed, 17 insertions(+), 1 deletion(-)

Comments

Catalin Marinas Feb. 20, 2024, 9:56 a.m. UTC | #1
On Tue, Feb 20, 2024 at 12:59:26PM +0530, ankita@nvidia.com wrote:
> diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
> index 1cbc990d42e0..c93bea18fc4b 100644
> --- a/drivers/vfio/pci/vfio_pci_core.c
> +++ b/drivers/vfio/pci/vfio_pci_core.c
> @@ -1862,8 +1862,24 @@ int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
>  	/*
>  	 * See remap_pfn_range(), called from vfio_pci_fault() but we can't
>  	 * change vm_flags within the fault handler.  Set them now.
> +	 *
> +	 * VM_ALLOW_ANY_UNCACHED: The VMA flag is implemented for ARM64,
> +	 * allowing KVM stage 2 device mapping attributes to use Normal-NC
> +	 * rather than DEVICE_nGnRE, which allows guest mappings
> +	 * supporting combining attributes (WC). ARM does not

Nitpick: "supporting write-combining" (if you plan to respin).
Ankit Agrawal Feb. 20, 2024, 12:11 p.m. UTC | #2
>> diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
>> index 1cbc990d42e0..c93bea18fc4b 100644
>> --- a/drivers/vfio/pci/vfio_pci_core.c
>> +++ b/drivers/vfio/pci/vfio_pci_core.c
>> @@ -1862,8 +1862,24 @@ int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
>>       /*
>>        * See remap_pfn_range(), called from vfio_pci_fault() but we can't
>>        * change vm_flags within the fault handler.  Set them now.
>> +      *
>> +      * VM_ALLOW_ANY_UNCACHED: The VMA flag is implemented for ARM64,
>> +      * allowing KVM stage 2 device mapping attributes to use Normal-NC
>> +      * rather than DEVICE_nGnRE, which allows guest mappings
>> +      * supporting combining attributes (WC). ARM does not
>
> Nitpick: "supporting write-combining" (if you plan to respin).

Ack.
Alex Williamson Feb. 22, 2024, 8:52 p.m. UTC | #3
On Tue, 20 Feb 2024 12:59:26 +0530
<ankita@nvidia.com> wrote:

> From: Ankit Agrawal <ankita@nvidia.com>
> 
> The VM_ALLOW_ANY_UNCACHED flag is implemented for ARM64,
> allowing KVM stage 2 device mapping attributes to use Normal-NC
> rather than DEVICE_nGnRE, which allows guest mappings
> supporting combining attributes (WC). ARM does not architecturally
> guarantee this is safe, and indeed some MMIO regions like the GICv2
> VCPU interface can trigger uncontained faults if Normal-NC is used.
> 
> To safely use VFIO in KVM the platform must guarantee full safety
> in the guest where no action taken against a MMIO mapping can
> trigger an uncontained failure. We belive that most VFIO PCI
> platforms support this for both mapping types, at least in common
> flows, based on some expectations of how PCI IP is integrated. So
> make vfio-pci set the VM_ALLOW_ANY_UNCACHED flag.
> 
> Suggested-by: Catalin Marinas <catalin.marinas@arm.com>
> Acked-by: Jason Gunthorpe <jgg@nvidia.com>
> Acked-by: Catalin Marinas <catalin.marinas@arm.com>
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Ankit Agrawal <ankita@nvidia.com>
> ---
>  drivers/vfio/pci/vfio_pci_core.c | 18 +++++++++++++++++-
>  1 file changed, 17 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
> index 1cbc990d42e0..c93bea18fc4b 100644
> --- a/drivers/vfio/pci/vfio_pci_core.c
> +++ b/drivers/vfio/pci/vfio_pci_core.c
> @@ -1862,8 +1862,24 @@ int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
>  	/*
>  	 * See remap_pfn_range(), called from vfio_pci_fault() but we can't
>  	 * change vm_flags within the fault handler.  Set them now.
> +	 *
> +	 * VM_ALLOW_ANY_UNCACHED: The VMA flag is implemented for ARM64,
> +	 * allowing KVM stage 2 device mapping attributes to use Normal-NC
> +	 * rather than DEVICE_nGnRE, which allows guest mappings
> +	 * supporting combining attributes (WC). ARM does not
> +	 * architecturally guarantee this is safe, and indeed some MMIO
> +	 * regions like the GICv2 VCPU interface can trigger uncontained
> +	 * faults if Normal-NC is used.
> +	 *
> +	 * To safely use VFIO in KVM the platform must guarantee full
> +	 * safety in the guest where no action taken against a MMIO
> +	 * mapping can trigger an uncontained failure. We belive that
> +	 * most VFIO PCI platforms support this for both mapping types,
> +	 * at least in common flows, based on some expectations of how
> +	 * PCI IP is integrated. So set VM_ALLOW_ANY_UNCACHED in VMA flags.
>  	 */
> -	vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
> +	vm_flags_set(vma, VM_ALLOW_ANY_UNCACHED | VM_IO | VM_PFNMAP |
> +			VM_DONTEXPAND | VM_DONTDUMP);
>  	vma->vm_ops = &vfio_pci_mmap_ops;
>  
>  	return 0;

Acked-by: Alex Williamson <alex.williamson@redhat.com>
diff mbox series

Patch

diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
index 1cbc990d42e0..c93bea18fc4b 100644
--- a/drivers/vfio/pci/vfio_pci_core.c
+++ b/drivers/vfio/pci/vfio_pci_core.c
@@ -1862,8 +1862,24 @@  int vfio_pci_core_mmap(struct vfio_device *core_vdev, struct vm_area_struct *vma
 	/*
 	 * See remap_pfn_range(), called from vfio_pci_fault() but we can't
 	 * change vm_flags within the fault handler.  Set them now.
+	 *
+	 * VM_ALLOW_ANY_UNCACHED: The VMA flag is implemented for ARM64,
+	 * allowing KVM stage 2 device mapping attributes to use Normal-NC
+	 * rather than DEVICE_nGnRE, which allows guest mappings
+	 * supporting combining attributes (WC). ARM does not
+	 * architecturally guarantee this is safe, and indeed some MMIO
+	 * regions like the GICv2 VCPU interface can trigger uncontained
+	 * faults if Normal-NC is used.
+	 *
+	 * To safely use VFIO in KVM the platform must guarantee full
+	 * safety in the guest where no action taken against a MMIO
+	 * mapping can trigger an uncontained failure. We belive that
+	 * most VFIO PCI platforms support this for both mapping types,
+	 * at least in common flows, based on some expectations of how
+	 * PCI IP is integrated. So set VM_ALLOW_ANY_UNCACHED in VMA flags.
 	 */
-	vm_flags_set(vma, VM_IO | VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
+	vm_flags_set(vma, VM_ALLOW_ANY_UNCACHED | VM_IO | VM_PFNMAP |
+			VM_DONTEXPAND | VM_DONTDUMP);
 	vma->vm_ops = &vfio_pci_mmap_ops;
 
 	return 0;