On Mon, Aug 22, 2022 at 9:14 AM John Hubbard <jhubbard@xxxxxxxxxx> wrote:
On 8/21/22 11:35, Shigeru Yoshida wrote:
> __get_user_pages_locked() may return the number of pages less than
> nr_pages. So __gup_longterm_locked() have to return the number of
> pages __get_user_pages_locked() returns if it succeeded, not nr_pages
> passed.
s/passed/requested/
>
> Fixes: 61c63c2076d9 (mm/gup.c: simplify and fix check_and_migrate_movable_pages() return codes)
> CC: Alistair Popple <apopple@xxxxxxxxxx>
> Reported-by: syzbot+616ff0452fec30f4dcfd@xxxxxxxxxxxxxxxxxxxxxxxxx
> Signed-off-by: Shigeru Yoshida <syoshida@xxxxxxxxxx>
> ---
> mm/gup.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
Reviewed-by: John Hubbard <jhubbard@xxxxxxxxxx>
Hi John,
Thank you so much for your review.
I'll send v2 patch.
Shigeru
(with a couple of nits about line length, below)
>
> diff --git a/mm/gup.c b/mm/gup.c
> index 5aa7531a703b..542cf74c59b0 100644
> --- a/mm/gup.c
> +++ b/mm/gup.c
> @@ -2068,22 +2068,22 @@ static long __gup_longterm_locked(struct mm_struct *mm,
> unsigned int gup_flags)
> {
> unsigned int flags;
> - long rc;
> + long rc, nr_pinned_pages;
>
> if (!(gup_flags & FOLL_LONGTERM))
> return __get_user_pages_locked(mm, start, nr_pages, pages, vmas,
> NULL, gup_flags);
> flags = memalloc_pin_save();
> do {
> - rc = __get_user_pages_locked(mm, start, nr_pages, pages, vmas,
> - NULL, gup_flags);
> - if (rc <= 0)
> + nr_pinned_pages = __get_user_pages_locked(mm, start, nr_pages, pages, vmas,
> + NULL, gup_flags);
Can you please wrap at 80 cols, though?
> + if (nr_pinned_pages <= 0)
> break;
> - rc = check_and_migrate_movable_pages(rc, pages, gup_flags);
> + rc = check_and_migrate_movable_pages(nr_pinned_pages, pages, gup_flags);
Also here.
> } while (rc == -EAGAIN);
> memalloc_pin_restore(flags);
>
> - return rc ? rc : nr_pages;
> + return rc ? rc : nr_pinned_pages;
> }
>
> static bool is_valid_gup_flags(unsigned int gup_flags)
thanks,
--
John Hubbard
NVIDIA