Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote: > > Note that I've been > > passing the number of pages to read in rac->_nr_pages, and then saving it > > and changing it certain points, e.g. page_cache_readahead_unbounded(). > > I do not like this. You're essentially mutating the meaning of _nr_pages > as the ractl moves down the stack, and that's going to lead to bugs. > I'd keep it as a separate argument. The meaning isn't specified in linux/pagemap.h. Can you adjust the comments on the struct and on readahead_count() to make it more clear what the value represents? Thanks, David
