On Tue, Mar 03, 2020 at 02:54:55PM -0800, Alexander Duyck wrote:
> It looks like you are only doing a partial revert though. The original
> changes in the function were:
> @@ -1681,8 +1683,6 @@ static unsigned int
> khugepaged_scan_mm_slot(unsigned int pages,
> if (khugepaged_scan.address < hstart)
> khugepaged_scan.address = hstart;
> VM_BUG_ON(khugepaged_scan.address & ~HPAGE_PMD_MASK);
> - if (shmem_file(vma->vm_file) && !shmem_huge_enabled(vma))
> - goto skip;
>
> while (khugepaged_scan.address < hend) {
> int ret;
> @@ -1694,9 +1694,12 @@ static unsigned int
> khugepaged_scan_mm_slot(unsigned int pages,
> khugepaged_scan.address + HPAGE_PMD_SIZE >
> hend);
> if (shmem_file(vma->vm_file)) {
> - struct file *file = get_file(vma->vm_file);
> + struct file *file;
> pgoff_t pgoff = linear_page_index(vma,
> khugepaged_scan.address);
> + if (!shmem_huge_enabled(vma))
> + goto skip;
> + file = get_file(vma->vm_file);
> up_read(&mm->mmap_sem);
> ret = 1;
> khugepaged_scan_shmem(mm, file->f_mapping,
>
> You reverted the second piece, but I didn't notice you reverting the
> first. WIth the first piece being reverted it would make more sense as
> you would just be skipping the block that much sooner.
Ah; that was an oversight on my part. Thank you for catching it! I'll
resubmit with that fixed.
