[PATCH 09/26] mm: Provide a means of invalidation without using launder_folio

[David Howells]

Jeff Layton <jlayton at kernel.org> wrote:

> I'd have liked to have seen the first caller of this function too.

Looking again at my code, I also need to make netfs_unbuffered_write_iter()
use this function too.  I missed it before because that wasn't calling
invalidate_inode_pages2_range() directly, but rather going through
kiocb_invalidate_pages().

So I'll add those changes into this patch to give you a first caller.  I also
then need to make filemap_invalidate_inode() take a byte range.

David