Re: [PATCH] iommu/vt-d: Fix to flush cache of PASID directory table

[Yanfei Xu]

On 6/16/2023 7:43 AM, Zhang, Tina wrote:
> Hi,
>
> > -----Original Message-----
> > From: Yanfei Xu <yanfei.xu@xxxxxxxxx>
> > Sent: Thursday, June 15, 2023 3:16 PM
> > To: dwmw2@xxxxxxxxxxxxx; baolu.lu@xxxxxxxxxxxxxxx; joro@xxxxxxxxxx;
> > will@xxxxxxxxxx; robin.murphy@xxxxxxx
> > Cc: iommu@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; Xu, Yanfei
> > <yanfei.xu@xxxxxxxxx>
> > Subject: [PATCH] iommu/vt-d: Fix to flush cache of PASID directory table
> >
> > Even the PCI devices don't support pasid capability, PASID table is mandatory
> > for a PCI device in scalable mode. However flushing cache of pasid directory
> > table for these devices are not taken after pasid table is allocated as the "size"
> > of table is zero. Fix to assign it with a page size.
> >
> > Fixes: 194b3348bdbb ("iommu/vt-d: Fix PASID directory pointer coherency")
> > Signed-off-by: Yanfei Xu <yanfei.xu@xxxxxxxxx>
> > ---
> >   drivers/iommu/intel/pasid.c | 4 +++-
> >   1 file changed, 3 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/iommu/intel/pasid.c b/drivers/iommu/intel/pasid.c index
> > c5d479770e12..bde7df055865 100644
> > --- a/drivers/iommu/intel/pasid.c
> > +++ b/drivers/iommu/intel/pasid.c
> > @@ -115,7 +115,9 @@ int intel_pasid_alloc_table(struct device *dev)
> >   				  intel_pasid_max_id);
> >
> >   	size = max_pasid >> (PASID_PDE_SHIFT - 3);
> > -	order = size ? get_order(size) : 0;
> > +	if (!size)
> > +		size = PAGE_SIZE;
> How about merging the logic of the above few lines into this one:
> size = info->pasid_supported ? max_pasid >> (PASID_PDE_SHIFT - 3) : PAGE_SIZE;

Yes, it would be more intuitive. But the prerequisite is if we can
make sure that the value of max_pasid shifted is still greater than
0. I roughly went through the PCIE spec and didn't find out where
defines the smallest PASID value for the PCIE device.

Thanks,
Yanfei

> Though the logic is about the same, the suggested one seems more intuitive.
>
> Regards,
> -Tina
>
> > +	order = get_order(size);
> >   	pages = alloc_pages_node(info->iommu->node,
> >   				 GFP_KERNEL | __GFP_ZERO, order);
> >   	if (!pages) {
> > --
> > 2.34.1