Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote: > AFS tracks one dirty range per folio, but it first brings the folio uptodate > by reading it from the server before overwriting it (I suppose that's a > fourth option). I'm intending on moving afs towards the nfs way of doing things when writing to as-yet unread folios - unless a cache is in operation, then we read it anyway and store the folio(s) into the cache unless the entire cache granule is going to be overwritten unless we're supporting disconnected mode. I know that's exceptions-to-exceptions. David
