On Tue, Oct 29, 2024 at 04:35:16PM +0000, Pavel Begunkov wrote:

I see, the reply is about your phrase about additional memory abstractions:

"... don't really need to build memory buffer abstraction over memory buffer abstraction."

Yes, over the exsting memory buffer abstraction (dma_buf).

If you mean internals, making up a dmabuf that has never existed in the picture in the first place is not cleaner or easier in any way. If that changes, e.g. there is more code to reuse in the future, we can unify it then.

I'm not sure what "making up" means here, they are all made up :)

...

with pre-registering the memry with the iommu to get good performance in IOMMU-enabled setups.

The page pool already does that just like it handles the normal path without providers.

In which case is basically is a dma-buf. If you'd expose it as such we could actually use to communicate between subsystems in the kernel.