On Mon, Jan 09, 2023 at 08:59:11AM +1100, Dave Chinner wrote: > Indeed, we also have this same "iomap valid check" functionality in the > writeback code as cached iomaps can become stale due to racing > writeback, truncated, etc. But you wouldn't know it by looking at the iomap > writeback code - this is currently hidden by XFS by embedding > the checks into the iomap writeback ->map_blocks function.
And that's in many ways a good thing, as it avoids various callouts that are expensive and confusing. Just like how this patch gets it right by not having a mess of badly interacting callbacks, but one that ensures that the page is ready. > Hence I think removing ->iomap_valid is a big step backwards for the > iomap core code - the iomap core needs to be able to formally verify > the iomap is valid at any point in time, not just at the point in > time a folio in the page cache has been locked... For using it anywhere else but the buffered write path it is in the wrong place to start with, and nonwithstanding my above concern I can't relaly think of a good place and prototype for such a valid callback to actually cover all use cases.
