[PATCH v8 26/29] vfio-pci: Register an iommu fault handler

[Jean-Philippe Brucker]

On 05/06/2019 23:45, Jacob Pan wrote:
> On Tue, 4 Jun 2019 18:11:08 +0200
> Auger Eric <eric.auger at redhat.com> wrote:
> 
>> Hi Alex,
>>
>> On 6/4/19 12:31 AM, Alex Williamson wrote:
>>> On Sun, 26 May 2019 18:10:01 +0200
>>> Eric Auger <eric.auger at redhat.com> wrote:
>>>   
>>>> This patch registers a fault handler which records faults in
>>>> a circular buffer and then signals an eventfd. This buffer is
>>>> exposed within the fault region.
>>>>
>>>> Signed-off-by: Eric Auger <eric.auger at redhat.com>
>>>>
>>>> ---
>>>>
>>>> v3 -> v4:
>>>> - move iommu_unregister_device_fault_handler to vfio_pci_release
>>>> ---
>>>>  drivers/vfio/pci/vfio_pci.c         | 49
>>>> +++++++++++++++++++++++++++++ drivers/vfio/pci/vfio_pci_private.h
>>>> |  1 + 2 files changed, 50 insertions(+)
>>>>
>>>> diff --git a/drivers/vfio/pci/vfio_pci.c
>>>> b/drivers/vfio/pci/vfio_pci.c index f75f61127277..520999994ba8
>>>> 100644 --- a/drivers/vfio/pci/vfio_pci.c
>>>> +++ b/drivers/vfio/pci/vfio_pci.c
>>>> @@ -30,6 +30,7 @@
>>>>  #include <linux/vfio.h>
>>>>  #include <linux/vgaarb.h>
>>>>  #include <linux/nospec.h>
>>>> +#include <linux/circ_buf.h>
>>>>  
>>>>  #include "vfio_pci_private.h"
>>>>  
>>>> @@ -296,6 +297,46 @@ static const struct vfio_pci_regops
>>>> vfio_pci_fault_prod_regops = { .add_capability =
>>>> vfio_pci_fault_prod_add_capability, };
>>>>  
>>>> +int vfio_pci_iommu_dev_fault_handler(struct iommu_fault_event
>>>> *evt, void *data) +{
>>>> +	struct vfio_pci_device *vdev = (struct vfio_pci_device *)
>>>> data;
>>>> +	struct vfio_region_fault_prod *prod_region =
>>>> +		(struct vfio_region_fault_prod
>>>> *)vdev->fault_pages;
>>>> +	struct vfio_region_fault_cons *cons_region =
>>>> +		(struct vfio_region_fault_cons
>>>> *)(vdev->fault_pages + 2 * PAGE_SIZE);
>>>> +	struct iommu_fault *new =
>>>> +		(struct iommu_fault *)(vdev->fault_pages +
>>>> prod_region->offset +
>>>> +			prod_region->prod *
>>>> prod_region->entry_size);
>>>> +	int prod, cons, size;
>>>> +
>>>> +	mutex_lock(&vdev->fault_queue_lock);
>>>> +
>>>> +	if (!vdev->fault_abi)
>>>> +		goto unlock;
>>>> +
>>>> +	prod = prod_region->prod;
>>>> +	cons = cons_region->cons;
>>>> +	size = prod_region->nb_entries;
>>>> +
>>>> +	if (CIRC_SPACE(prod, cons, size) < 1)
>>>> +		goto unlock;
>>>> +
>>>> +	*new = evt->fault;
>>>> +	prod = (prod + 1) % size;
>>>> +	prod_region->prod = prod;
>>>> +	mutex_unlock(&vdev->fault_queue_lock);
>>>> +
>>>> +	mutex_lock(&vdev->igate);
>>>> +	if (vdev->dma_fault_trigger)
>>>> +		eventfd_signal(vdev->dma_fault_trigger, 1);
>>>> +	mutex_unlock(&vdev->igate);
>>>> +	return 0;
>>>> +
>>>> +unlock:
>>>> +	mutex_unlock(&vdev->fault_queue_lock);
>>>> +	return -EINVAL;
>>>> +}
>>>> +
>>>>  static int vfio_pci_init_fault_region(struct vfio_pci_device
>>>> *vdev) {
>>>>  	struct vfio_region_fault_prod *header;
>>>> @@ -328,6 +369,13 @@ static int vfio_pci_init_fault_region(struct
>>>> vfio_pci_device *vdev) header = (struct vfio_region_fault_prod
>>>> *)vdev->fault_pages; header->version = -1;
>>>>  	header->offset = PAGE_SIZE;
>>>> +
>>>> +	ret =
>>>> iommu_register_device_fault_handler(&vdev->pdev->dev,
>>>> +
>>>> vfio_pci_iommu_dev_fault_handler,
>>>> +					vdev);
>>>> +	if (ret)
>>>> +		goto out;
>>>> +
>>>>  	return 0;
>>>>  out:
>>>>  	kfree(vdev->fault_pages);
>>>> @@ -570,6 +618,7 @@ static void vfio_pci_release(void *device_data)
>>>>  	if (!(--vdev->refcnt)) {
>>>>  		vfio_spapr_pci_eeh_release(vdev->pdev);
>>>>  		vfio_pci_disable(vdev);
>>>> +
>>>> iommu_unregister_device_fault_handler(&vdev->pdev->dev);  
>>>
>>>
>>> But this can fail if there are pending faults which leaves a device
>>> reference and then the system is broken :(  
>> This series only features unrecoverable errors and for those the
>> unregistration cannot fail. Now unrecoverable errors were added I
>> admit this is confusing. We need to sort this out or clean the
>> dependencies.
> As Alex pointed out in 4/29, we can make
> iommu_unregister_device_fault_handler() never fail and clean up all the
> pending faults in the host IOMMU belong to that device. But the problem
> is that if a fault, such as PRQ, has already been injected into the
> guest, the page response may come back after handler is unregistered
> and registered again.

I'm trying to figure out if that would be harmful in any way. I guess it
can be a bit nasty if we handle the page response right after having
injected a new page request that uses the same PRGI. In any other case we
discard the page response, but here we forward it to the endpoint and:

* If the response status is success, endpoint retries the translation. The
guest probably hasn't had time to handle the new page request and
translation will fail, which may lead the endpoint to give up (two
unsuccessful translation requests). Or send a new request

* otherwise the endpoint won't retry the access, and could also disable
PRI if the status is failure.

> We need a way to reject such page response belong
> to the previous life of the handler. Perhaps a sync call to the guest
> with your fault queue eventfd? I am not sure.

We could simply expect the device driver not to send any page response
after unregistering the fault handler. Is there any reason VFIO would need
to unregister and re-register the fault handler on a live guest?

Thanks,
Jean