Re: [PATCH 1/1] iommu/vt-d: Drain PRQs when domain removed from RID

[Baolu Lu]

On 10/31/24 22:56, Joel Granados wrote:
> On Thu, Oct 31, 2024 at 05:51:39PM +0800, Lu Baolu wrote:
> > As this iommu driver now supports page faults for requests without
> > PASID, page requests should be drained when a domain is removed from
> > the RID2PASID entry.
> >
> > This results in the intel_iommu_drain_pasid_prq() call being moved to
> > intel_pasid_tear_down_entry(). This indicates that when a translation
> > is removed from any PASID entry and the PRI has been enabled on the
> > device, page requests are flushed in the domain detachment path.
> >
> > The intel_iommu_drain_pasid_prq() helper has been modified to support
> > sending device TLB invalidation requests for both PASID and non-PASID
> > cases.
> >
> > Signed-off-by: Lu Baolu<baolu.lu@xxxxxxxxxxxxxxx>
> > ---
> >   drivers/iommu/intel/iommu.c |  1 -
> >   drivers/iommu/intel/pasid.c |  1 +
> >   drivers/iommu/intel/prq.c   | 22 +++++++---------------
> >   3 files changed, 8 insertions(+), 16 deletions(-)
> >
> > diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c
> > index 87a3563dfe54..3878f35be09d 100644
> > --- a/drivers/iommu/intel/iommu.c
> > +++ b/drivers/iommu/intel/iommu.c
> > @@ -4069,7 +4069,6 @@ static void intel_iommu_remove_dev_pasid(struct device *dev, ioasid_t pasid,
> >   	intel_iommu_debugfs_remove_dev_pasid(dev_pasid);
> >   	kfree(dev_pasid);
> >   	intel_pasid_tear_down_entry(iommu, dev, pasid, false);
> > -	intel_iommu_drain_pasid_prq(dev, pasid);
> >   }
> >   
> >   static int intel_iommu_set_dev_pasid(struct iommu_domain *domain,
> > diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c
> > index 7e76062a7ad2..31665fb62e1c 100644
> > --- a/drivers/iommu/intel/pasid.c
> > +++ b/drivers/iommu/intel/pasid.c
> > @@ -265,6 +265,7 @@ void intel_pasid_tear_down_entry(struct intel_iommu *iommu, struct device *dev,
> >   		iommu->flush.flush_iotlb(iommu, did, 0, 0, DMA_TLB_DSI_FLUSH);
> >   
> >   	devtlb_invalidation_with_pasid(iommu, dev, pasid);
> > +	intel_iommu_drain_pasid_prq(dev, pasid);
> >   }
> >   
> >   /*
> > diff --git a/drivers/iommu/intel/prq.c b/drivers/iommu/intel/prq.c
> > index 3c50c848893f..ae7f6f34462f 100644
> > --- a/drivers/iommu/intel/prq.c
> > +++ b/drivers/iommu/intel/prq.c
> > @@ -66,12 +66,8 @@ void intel_iommu_drain_pasid_prq(struct device *dev, u32 pasid)
> >   	struct pci_dev *pdev;
> >   	int head, tail;
> >   	u16 sid, did;
> > -	int qdep;
> >   
> >   	info = dev_iommu_priv_get(dev);
> > -	if (WARN_ON(!info || !dev_is_pci(dev)))
> You do this on purpose because
> 1. It is not possible to go through this path without an iommu set in
>     the device
> 2. PRI within the intel driver is only relevant for PCI
> 3. There will always be a struct device_domain_info related to the
>     device iommu when doing PRI.
> right?

Yes.

--
baolu