On Wed, Jul 15, 2020 at 04:18:44PM -0700, Linus Torvalds wrote: > On Wed, Jul 15, 2020 at 4:04 PM Linus Torvalds > <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote: > > > > It *might* be as simple as this incremental thing on top > > No, it needs to be > > + if (*old_addr + *len < old->vm_end) > + return; > > in try_to_align_end(), of course. Okay, this should work, but I'm not convinced that it gives much: number of cases covered by the optimization not going to be high. It can also lead to performance regression: for small mremap() if only one side of the range got aligned and there's no PMD_SIZE range to move, kernel will still iterate over PTEs, but it would need to handle more pte_none()s than without the patch. -- Kirill A. Shutemov
