RE: [PATCH 3/3] vfio: Report PASID capability via VFIO_DEVICE_FEATURE ioctl

[Duan, Zhenzhong]

>-----Original Message-----
>From: Liu, Yi L <yi.l.liu@xxxxxxxxx>
>Sent: Tuesday, November 28, 2023 11:12 AM
>Subject: Re: [PATCH 3/3] vfio: Report PASID capability via VFIO_DEVICE_FEATURE
>ioctl
>
>On 2023/11/27 15:28, Duan, Zhenzhong wrote:
>>
>>
>>> -----Original Message-----
>>> From: Liu, Yi L <yi.l.liu@xxxxxxxxx>
>>> Sent: Monday, November 27, 2023 2:39 PM
>>> Subject: [PATCH 3/3] vfio: Report PASID capability via VFIO_DEVICE_FEATURE
>>> ioctl
>>>
>>> This reports the PASID capability data to userspace via VFIO_DEVICE_FEATURE,
>>> hence userspace could probe PASID capability by it. This is a bit different
>>> with other capabilities which are reported to userspace when the user reads
>>> the device's PCI configuration space. There are two reasons for this.
>>>
>>> - First, Qemu by default exposes all available PCI capabilities in vfio-pci
>>>    config space to the guest as read-only, so adding PASID capability in the
>>>    vfio-pci config space will make it exposed to the guest automatically while
>>>    an old Qemu doesn't really support it.
>>>
>>> - Second, PASID capability does not exit on VFs (instead shares the cap of
>>>    the PF). Creating a virtual PASID capability in vfio-pci config space needs
>>>    to find a hole to place it, but doing so may require device specific
>>>    knowledge to avoid potential conflict with device specific registers like
>>>    hiden bits in VF config space. It's simpler by moving this burden to the
>>>    VMM instead of maintaining a quirk system in the kernel.
>>>
>>> Suggested-by: Alex Williamson <alex.williamson@xxxxxxxxxx>
>>> Signed-off-by: Yi Liu <yi.l.liu@xxxxxxxxx>
>>> ---
>>> drivers/vfio/pci/vfio_pci_core.c | 47 ++++++++++++++++++++++++++++++++
>>> include/uapi/linux/vfio.h        | 13 +++++++++
>>> 2 files changed, 60 insertions(+)
>>>
>>> diff --git a/drivers/vfio/pci/vfio_pci_core.c b/drivers/vfio/pci/vfio_pci_core.c
>>> index 1929103ee59a..8038aa45500e 100644
>>> --- a/drivers/vfio/pci/vfio_pci_core.c
>>> +++ b/drivers/vfio/pci/vfio_pci_core.c
>>> @@ -1495,6 +1495,51 @@ static int vfio_pci_core_feature_token(struct
>>> vfio_device *device, u32 flags,
>>> 	return 0;
>>> }
>>>
>>> +static int vfio_pci_core_feature_pasid(struct vfio_device *device, u32 flags,
>>> +				       struct vfio_device_feature_pasid __user
>>> *arg,
>>> +				       size_t argsz)
>>> +{
>>> +	struct vfio_pci_core_device *vdev =
>>> +		container_of(device, struct vfio_pci_core_device, vdev);
>>> +	struct vfio_device_feature_pasid pasid = { 0 };
>>> +	struct pci_dev *pdev = vdev->pdev;
>>> +	u32 capabilities = 0;
>>> +	int ret;
>>> +
>>> +	/* We do not support SET of the PASID capability */
>>> +	ret = vfio_check_feature(flags, argsz, VFIO_DEVICE_FEATURE_GET,
>>> +				 sizeof(pasid));
>>> +	if (ret != 1)
>>> +		return ret;
>>> +
>>> +	/*
>>> +	 * Needs go to PF if the device is VF as VF shares its PF's
>>> +	 * PASID Capability.
>>> +	 */
>>> +	if (pdev->is_virtfn)
>>> +		pdev = pci_physfn(pdev);
>>> +
>>> +	if (!pdev->pasid_enabled)
>>> +		goto out;
>>
>> Does a PF bound to VFIO have pasid enabled by default?
>
>Today, host iommu driver (at least intel iommu driver) enables it in the
>time of device probe and seems not changed afterward. So yes, VFIO should
>see it if pasid is enabled.
>
>> Isn't the guest kernel's responsibility to enable pasid cap of an assigned PF?
>
>guest kernel should not have the capability to change host's pasid
>configuration. It can only write to its own vconfig emulated by
>hypervisor.

Understood, thanks Yi.

BRs.
Zhenzhong