Hi Kirill,
On Tue, Jul 14, 2020 at 10:33:06AM +0300, Kirill A. Shutemov wrote:
> On Sun, Jul 12, 2020 at 03:58:06PM -0700, Linus Torvalds wrote:
> > Anybody else have any opinions?
>
> Maybe we just shouldn't allow move_normal_pmd() if ranges overlap?
>
> Other option: pass 'overlaps' down to move_normal_pmd() and only WARN() if
> see establised PMD without overlaps being true.
I was thinking we should not call move_page_tables() with overlapping ranges
at all, just to keep things simple. I am concerned about other issues such as
if you move a range forward, you will end up overwriting part of the source
range.
Allow me some time to develop a proper patch, I have it on my list. I will
try to get to it this week.
I think we can also add a patch to detect the overlap as you did and warn in
such situation.
Thoughts?
thanks,
- Joel
> Untested patch:
>
> diff --git a/mm/mremap.c b/mm/mremap.c
> index 5dd572d57ca9..e33fcee541fe 100644
> --- a/mm/mremap.c
> +++ b/mm/mremap.c
> @@ -245,6 +245,18 @@ unsigned long move_page_tables(struct vm_area_struct *vma,
> unsigned long extent, next, old_end;
> struct mmu_notifier_range range;
> pmd_t *old_pmd, *new_pmd;
> + bool overlaps;
> +
> + /*
> + * shift_arg_pages() can call move_page_tables() on overlapping ranges.
> + * In this case we cannot use move_normal_pmd() because destination pmd
> + * might be established page table: move_ptes() doesn't free page
> + * table.
> + */
> + if (old_addr > new_addr)
> + overlaps = old_addr - new_addr < len;
> + else
> + overlaps = new_addr - old_addr < len;
>
> old_end = old_addr + len;
> flush_cache_range(vma, old_addr, old_end);
> @@ -282,7 +294,7 @@ unsigned long move_page_tables(struct vm_area_struct *vma,
> split_huge_pmd(vma, old_pmd, old_addr);
> if (pmd_trans_unstable(old_pmd))
> continue;
> - } else if (extent == PMD_SIZE) {
> + } else if (!overlaps && extent == PMD_SIZE) {
> #ifdef CONFIG_HAVE_MOVE_PMD
> /*
> * If the extent is PMD-sized, try to speed the move by
> --
> Kirill A. Shutemov
