Re: [External] Re: [PATCH 4/6] mm: hugetlb: add return -EAGAIN for dissolve_free_huge_page

[Muchun Song <songmuchun@bytedance.com>]

On Tue, Jan 5, 2021 at 9:33 AM Mike Kravetz <mike.kravetz@oracle.com> wrote:
>
> On 1/3/21 10:58 PM, Muchun Song wrote:
> > When dissolve_free_huge_page() races with __free_huge_page(), we can
> > do a retry. Because the race window is small.
>
> In general, I agree that the race window is small.  However, worst case
> would be if the freeing of the page is put on a work queue.  Is it acceptable
> to keep retrying in that case?  In addition, the 'Free some vmemmap' series
> may slow the free_huge_page path even more.

I also consider the 'Free some vmemmap' series case. In my next
version series, I will flush the work before dissolve_free_huge_page
returns when encountering this race. So the retry is acceptable.
Right?

Thanks.

>
> In these worst case scenarios, I am not sure we want to just spin retrying.
>
> --
> Mike Kravetz
>
> >
> > Signed-off-by: Muchun Song <songmuchun@bytedance.com>
> > ---
> >  mm/hugetlb.c | 16 +++++++++++-----
> >  1 file changed, 11 insertions(+), 5 deletions(-)
> >
> > diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> > index 72608008f8b4..db00ae375d2a 100644
> > --- a/mm/hugetlb.c
> > +++ b/mm/hugetlb.c
> > @@ -1763,10 +1763,11 @@ static int free_pool_huge_page(struct hstate *h, nodemask_t *nodes_allowed,
> >   * nothing for in-use hugepages and non-hugepages.
> >   * This function returns values like below:
> >   *
> > - *  -EBUSY: failed to dissolved free hugepages or the hugepage is in-use
> > - *          (allocated or reserved.)
> > - *       0: successfully dissolved free hugepages or the page is not a
> > - *          hugepage (considered as already dissolved)
> > + *  -EAGAIN: race with __free_huge_page() and can do a retry
> > + *  -EBUSY:  failed to dissolved free hugepages or the hugepage is in-use
> > + *           (allocated or reserved.)
> > + *       0:  successfully dissolved free hugepages or the page is not a
> > + *           hugepage (considered as already dissolved)
> >   */
> >  int dissolve_free_huge_page(struct page *page)
> >  {
> > @@ -1815,8 +1816,10 @@ int dissolve_free_huge_page(struct page *page)
> >                * We should make sure that the page is already on the free list
> >                * when it is dissolved.
> >                */
> > -             if (unlikely(!PageHugeFreed(head)))
> > +             if (unlikely(!PageHugeFreed(head))) {
> > +                     rc = -EAGAIN;
> >                       goto out;
> > +             }
> >
> >               /*
> >                * Move PageHWPoison flag from head page to the raw error page,
> > @@ -1857,7 +1860,10 @@ int dissolve_free_huge_pages(unsigned long start_pfn, unsigned long end_pfn)
> >
> >       for (pfn = start_pfn; pfn < end_pfn; pfn += 1 << minimum_order) {
> >               page = pfn_to_page(pfn);
> > +retry:
> >               rc = dissolve_free_huge_page(page);
> > +             if (rc == -EAGAIN)
> > +                     goto retry;
> >               if (rc)
> >                       break;
> >       }
> >