On Tue, Dec 08, 2020 at 12:23:16PM +0000, Matthew Wilcox wrote: > On Mon, Dec 07, 2020 at 02:57:03PM -0800, ira.weiny@xxxxxxxxx wrote: > > Placing these functions in 'highmem.h' is suboptimal especially with the > > changes being proposed in the functionality of kmap. From a caller > > perspective including/using 'highmem.h' implies that the functions > > defined in that header are only required when highmem is in use which is > > increasingly not the case with modern processors. Some headers like > > mm.h or string.h seem ok but don't really portray the functionality > > well. 'pagemap.h', on the other hand, makes sense and is already > > included in many of the places we want to convert. > > pagemap.h is for the page cache. It's not for "random page > functionality". Yes, I know it's badly named. No, I don't want to > rename it. These helpers should go in highmem.h along with zero_user(). I could have sworn you suggested pagemap.h. But I can't find the evidence on lore. :-/ hehehe... In the end the code does not care. I have a distaste for highmem.h because it is no longer for 'high memory'. And I think a number of driver writers who are targeting 64bit platforms just don't care any more. So as we head toward memory not being mapped by the kernel for other reasons I think highmem needs to be 'rebranded' if not renamed. Ira
