On Sat, Apr 11, 2020 at 03:02:40PM -0700, Linus Torvalds wrote: > On Sat, Apr 11, 2020 at 2:48 PM Matthew Wilcox <willy@xxxxxxxxxxxxx> wrote: > > > > I wasn't entirely forthcoming ... I actually want to introduce a new > > > > #define page_offset(page, x) ((unsigned long)(x) & (page_size(page) - 1)) > > No, no, no. > > THAT would be confusing. Re-using a name (even if you renamed it) for > something new and completely different is just bad taste. It would > also be a horrible problem - again - for any stable backport etc. > > Just call that "offset_in_page()" and be done with it. But we _have_ an offset_in_page() and it doesn't take a struct page argument. Reusing the name wouldn't be too bad because it would take two arguments, so nothing would inadvertently apply cleanly. Anyway, if you give me a decision on pgindex_t vs pgidx_t, I'll work that up before -rc1 closes.
