On Tue, Nov 06, 2018 at 09:37:26AM +0100, Carlos Maiolino wrote: > Not really, fiemap_extent_info is removed at the end of the series, and by doing > it directly, I'd need to write a big patch changing all users of > fiemap_extent_info in a single patch, or break the kernel build between patches. > I usually try to avoid a single patch touching several different filesystems, so > that's why I basically slowly replaced fiemap_extent_info, until it get removed > at the end of the series. Sounds sensible. I also always found fiemap_extent_info rather long and cumbersome. > > The other minor nit about this patch series is that "f_ctx" is not a very > > descriptive variable name. How about "fiemap_ctx"? A bit longer, but it > > is immediately clear to the reader what it is. > > I don't really mind, f_ctx, fmap_ctx, etc. having the pointer the same name as > the struct doesn't look good to me, but particularly I don't mind. Just ctx sounds fine to me.
