On Thu, Apr 14, 2022 at 01:16:26AM +0800, kernel test robot wrote:
> 1664 if (start != end) {
> 1665 if (unlikely(is_vm_hugetlb_page(vma))) {
> 1666 /*
> 1667 * It is undesirable to test vma->vm_file as it
> 1668 * should be non-null for valid hugetlb area.
> 1669 * However, vm_file will be NULL in the error
> 1670 * cleanup path of mmap_region. When
> 1671 * hugetlbfs ->mmap method fails,
> 1672 * mmap_region() nullifies vma->vm_file
> 1673 * before calling this function to clean up.
> 1674 * Since no pte has actually been setup, it is
> 1675 * safe to do nothing in this case.
> 1676 */
> 1677 if (vma->vm_file) {
> > 1678 unsigned long zap_flags = details ?
> 1679 details->zap_flags : 0;
The hugetlb side was overlooked after we re-typed zap_flags into bitmask..
One more patch (attached) needs to be squashed into patch "mm/hugetlb: Only
drop uffd-wp special pte if required" to ease the sparse warning, sorry.
--
Peter Xu
