RE: [PATCH 3/4] iommu/sva: Support reservation of global PASIDs

[Tian, Kevin]

> From: Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx>
> Sent: Thursday, March 2, 2023 9:00 AM
> 
> Global PASID allocation is under IOMMU SVA code since it is the primary
> use case.  However, some architecture such as VT-d, global PASIDs are
> necessary for its internal use of DMA API with PASID.

No, global PASID is not a VT-d restriction. It's from ENQCMD/S hence a
device requirement.

> 
> This patch introduces SVA APIs to reserve and release global PASIDs.
> 
> Link: https://lore.kernel.org/all/20230301235646.2692846-4-
> jacob.jun.pan@xxxxxxxxxxxxxxx/
> Signed-off-by: Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx>
> ---
>  drivers/iommu/iommu-sva.c | 25 +++++++++++++++++++++++++
>  include/linux/iommu.h     | 14 ++++++++++++++
>  2 files changed, 39 insertions(+)
> 
> diff --git a/drivers/iommu/iommu-sva.c b/drivers/iommu/iommu-sva.c
> index 8c92a145e15d..cfdeafde88a9 100644
> --- a/drivers/iommu/iommu-sva.c
> +++ b/drivers/iommu/iommu-sva.c
> @@ -149,6 +149,31 @@ u32 iommu_sva_get_pasid(struct iommu_sva
> *handle)
>  }
>  EXPORT_SYMBOL_GPL(iommu_sva_get_pasid);
> 
> +ioasid_t iommu_sva_reserve_pasid(ioasid_t min, ioasid_t max)
> +{
> +	int ret;
> +
> +	if (min == IOMMU_PASID_INVALID || max ==
> IOMMU_PASID_INVALID ||
> +	    min == 0 || max < min)
> +		return IOMMU_PASID_INVALID;
> +
> +	ret = ida_alloc_range(&iommu_global_pasid_ida, min, max,
> GFP_KERNEL);
> +	if (ret < 0)
> +		return IOMMU_PASID_INVALID;
> +
> +	return ret;
> +}
> +EXPORT_SYMBOL_GPL(iommu_sva_reserve_pasid);
> +

I'm not sure it's the right way. It's not related to SVA.

We should move iommu_global_pasid_ida to iomm.c and then have
another interface for allocation.

Above is pretty generic so probably a general one like:

	ioasid_t iommu_allocate_global_pasid(struct device *dev)

internally it can use [1, dev->iommu->max_pasids] as min/max instead
of passed in from the caller.