On Tue, Jun 06, 2023 at 01:37:48PM +0100, Matthew Wilcox wrote: > > 1. Renamed iomap_page -> iomap_folio & iop -> iof in Patch-1 itself. > > I don't think iomap_folio is the right name. Indeed, I did not believe > that iomap_page was the right name. As I said on #xfs recently ... > > <willy> i'm still not crazy about iomap_page as the name of that > data structure. and calling the variable 'iop' just seems doomed > to be trouble. how do people feel about either iomap_block_state or > folio_block_state ... or even just calling it block_state since it's > local to iomap/buffered-io.c > <willy> we'd then call the variable either ibs or fbs, both of which > have some collisions in the kernel, but none in filesystems > <dchinner> willy - sounds reasonable I'd keep an iomap prefix, but block_state looks fine to me. So iomap_block_state would be my preference.
