Re: [PATCH v7 0/9] Organize the SMMUv3 invalidation flow so iommupt can use it

[Pranjal Shrivastava <praan@google.com>]

On Mon, Sep 21, 2026 at 08:55:08PM -0300, Jason Gunthorpe wrote:
> [ This is part of the patch pile to move SMMUv3 over to the generic page
> table, the precursor patches have been merged now. What's left:
>  1) Organize the SMMUv3 invalidation flow so iommupt can use it
>  2) Use the generic iommu page table for SMMUv3
> 
> The first patch should ideally go to -rc
> 
> The whole branch is here:
>    https://github.com/jgunthorpe/linux/commits/iommu_pt_arm64/
> ]
> 
> iommupt has a design that focuses on building a single iommu_iotlb_gather
> for arbitary batches of map/unmap operations. The gather uses the free
> list and it captures invalidations of tables, leaves and supports mixed
> levels.
> 
> The introduction of PT_FEAT_DETAILED_GATHER provides some additional
> information that is useful for ARM: the damage bitmaps for the table and
> leaf changes.
> 
> Prior to switching SMMUv3 over to use iommupt prepare for this by
> reworking the internal invalidation to work on the same data format that
> iommupt will produce. Bridge the invalidations generated by io-pgtable
> into the new format. The conversion is simple enough, io-pgtable generates
> invalidation operations that have only a single set bit in
> table_levels_bitmap/leaf_levels_bitmap, so we can convert the io-pgtable
> provided size into the proper level leaf or table bit.
>

[...]

FWIW, I was able to run dma_map_benchmark for 4K & 2M 
(-g 1 / -g 512, 1..256 threads) and I observe no change in map/unmap
latency vs. base, no measurable overhead on the leaf unmap hot path.

Tested-by: Pranjal Shrivastava <praan@google.com>

Thanks,
Praan