On Wed, 17 May 2023 08:35:10 +0200 Petr Tesařík <petr@xxxxxxxxxxx> wrote: >[...] > Anyway, my greatest objection to allocating additional swiotlb chunks is > that _all_ of them must be searched to determine that the physical > address does _not_ belong to a swiotlb, incurring performance penalty I thought about this part again, and I overlooked one option. We can track only the _active_ swiotlbs for each device. If a device never needs a swiotlb, there is no active swiotlb, and is_swiotlb_buffer() short-circuits to false. This should avoid all collateral damage to innocent devices. We would also maintain a (global) list of all allocated swiotlbs, used by swiotlb_map() to find free slots and add the respective swiotlb to the per-device active list. One potential advantage is that we could use mapping size and alignment to choose a swiotlb cleverly and minimize internal fragmentation... OK, I'm dreaming. Let's agree on the general approach first. Petr T
