On Tue, Oct 29, 2024 at 08:01:16PM -0700, John Hubbard wrote: > A user-visible consequence has now appeared: user space can no longer > pin more than 2GB of memory anymore on x86_64. That's because, on a 4KB > PAGE_SIZE system, when user space tries to (indirectly, via a device > driver that calls pin_user_pages()) pin 2GB, this requires an allocation > of a folio pointers array of MAX_PAGE_ORDER size, which is the limit for > kmalloc(). Do you have a report whee someone tries to pin that much memor in a single call? What driver is this? Because it seems like a not very smart thing to do.
