On Thu, Apr 04, 2024 at 01:10:20PM +0200, Gerd Bayer wrote: > > Why can't you use get_free_pages() (or similar) here? (possibly > > rounding up to the relevant page_aligned size). > > Thanks Paolo for your suggestion. However, I wanted to stay as close to > the implementation pre [1] - that used to use __GFP_COMP, too. I'd > rather avoid to change interfaces from "cpu_addr" to "struct page*" at > this point. In the long run, I'd like to drop the requirement for The right interface actually is to simply use folio_alloc, which adds __GFP_COMP and is a fully supported and understood interface. You can just convert the folio to a kernel virtual address using folio_address() right after allocating it. (get_free_pages also retunrs a kernel virtual address, just awkwardly as an unsigned long. In doubt don't use this interface for new code..) > compound pages entirely, since that *appears* to exist primarily for a > simplified handling of the interface to splice_to_pipe() in > net/smc/smc_rx.c. And of course there might be performance > implications... While compounds pages might sound awkward, they are the new normal in form of folios. So just use folios.
