| Message ID | 20200211213128.73302-6-almasrymina@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v12,1/9] hugetlb_cgroup: Add hugetlb_cgroup reservation counter | expand |
On 2/11/20 1:31 PM, Mina Almasry wrote: > Support MAP_NORESERVE accounting as part of the new counter. > > For each hugepage allocation, at allocation time we check if there is > a reservation for this allocation or not. If there is a reservation for > this allocation, then this allocation was charged at reservation time, > and we don't re-account it. If there is no reserevation for this > allocation, we charge the appropriate hugetlb_cgroup. > > The hugetlb_cgroup to uncharge for this allocation is stored in > page[3].private. We use new APIs added in an earlier patch to set this > pointer. > > Signed-off-by: Mina Almasry <almasrymina@google.com> > > --- > > Changes in v12: > - Minor rebase to new interface for readability. > > Changes in v10: > - Refactored deferred_reserve check. > > --- > mm/hugetlb.c | 27 ++++++++++++++++++++++++++- > 1 file changed, 26 insertions(+), 1 deletion(-) > > diff --git a/mm/hugetlb.c b/mm/hugetlb.c > index a9171c3cbed6b..2d62dd35399db 100644 > --- a/mm/hugetlb.c > +++ b/mm/hugetlb.c > @@ -1342,6 +1342,8 @@ static void __free_huge_page(struct page *page) > clear_page_huge_active(page); > hugetlb_cgroup_uncharge_page(hstate_index(h), > pages_per_huge_page(h), page); > + hugetlb_cgroup_uncharge_page_rsvd(hstate_index(h), > + pages_per_huge_page(h), page); > if (restore_reserve) > h->resv_huge_pages++; > > @@ -2175,6 +2177,7 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, > long gbl_chg; > int ret, idx; > struct hugetlb_cgroup *h_cg; > + bool deferred_reserve; > > idx = hstate_index(h); > /* > @@ -2212,9 +2215,19 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, > gbl_chg = 1; > } > > + /* If this allocation is not consuming a reservation, charge it now. > + */ > + deferred_reserve = map_chg || avoid_reserve || !vma_resv_map(vma); > + if (deferred_reserve) { > + ret = hugetlb_cgroup_charge_cgroup_rsvd( > + idx, pages_per_huge_page(h), &h_cg); > + if (ret) > + goto out_subpool_put; > + } > + > ret = hugetlb_cgroup_charge_cgroup(idx, pages_per_huge_page(h), &h_cg); > if (ret) > - goto out_subpool_put; > + goto out_uncharge_cgroup_reservation; > > spin_lock(&hugetlb_lock); > /* > @@ -2237,6 +2250,14 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, > /* Fall through */ > } > hugetlb_cgroup_commit_charge(idx, pages_per_huge_page(h), h_cg, page); > + /* If allocation is not consuming a reservation, also store the > + * hugetlb_cgroup pointer on the page. > + */ > + if (deferred_reserve) { > + hugetlb_cgroup_commit_charge_rsvd(idx, pages_per_huge_page(h), > + h_cg, page); > + } > + This started before your new code, but those two cgroup_commit_charge calls could/should be done outside the hugetlb_lock. No need to change as it is not a big deal. Those calls only set fields in the page structs. Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
diff --git a/mm/hugetlb.c b/mm/hugetlb.c index a9171c3cbed6b..2d62dd35399db 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1342,6 +1342,8 @@ static void __free_huge_page(struct page *page) clear_page_huge_active(page); hugetlb_cgroup_uncharge_page(hstate_index(h), pages_per_huge_page(h), page); + hugetlb_cgroup_uncharge_page_rsvd(hstate_index(h), + pages_per_huge_page(h), page); if (restore_reserve) h->resv_huge_pages++; @@ -2175,6 +2177,7 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, long gbl_chg; int ret, idx; struct hugetlb_cgroup *h_cg; + bool deferred_reserve; idx = hstate_index(h); /* @@ -2212,9 +2215,19 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, gbl_chg = 1; } + /* If this allocation is not consuming a reservation, charge it now. + */ + deferred_reserve = map_chg || avoid_reserve || !vma_resv_map(vma); + if (deferred_reserve) { + ret = hugetlb_cgroup_charge_cgroup_rsvd( + idx, pages_per_huge_page(h), &h_cg); + if (ret) + goto out_subpool_put; + } + ret = hugetlb_cgroup_charge_cgroup(idx, pages_per_huge_page(h), &h_cg); if (ret) - goto out_subpool_put; + goto out_uncharge_cgroup_reservation; spin_lock(&hugetlb_lock); /* @@ -2237,6 +2250,14 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, /* Fall through */ } hugetlb_cgroup_commit_charge(idx, pages_per_huge_page(h), h_cg, page); + /* If allocation is not consuming a reservation, also store the + * hugetlb_cgroup pointer on the page. + */ + if (deferred_reserve) { + hugetlb_cgroup_commit_charge_rsvd(idx, pages_per_huge_page(h), + h_cg, page); + } + spin_unlock(&hugetlb_lock); set_page_private(page, (unsigned long)spool); @@ -2261,6 +2282,10 @@ struct page *alloc_huge_page(struct vm_area_struct *vma, out_uncharge_cgroup: hugetlb_cgroup_uncharge_cgroup(idx, pages_per_huge_page(h), h_cg); +out_uncharge_cgroup_reservation: + if (deferred_reserve) + hugetlb_cgroup_uncharge_cgroup_rsvd(idx, pages_per_huge_page(h), + h_cg); out_subpool_put: if (map_chg || avoid_reserve) hugepage_subpool_put_pages(spool, 1);
Support MAP_NORESERVE accounting as part of the new counter. For each hugepage allocation, at allocation time we check if there is a reservation for this allocation or not. If there is a reservation for this allocation, then this allocation was charged at reservation time, and we don't re-account it. If there is no reserevation for this allocation, we charge the appropriate hugetlb_cgroup. The hugetlb_cgroup to uncharge for this allocation is stored in page[3].private. We use new APIs added in an earlier patch to set this pointer. Signed-off-by: Mina Almasry <almasrymina@google.com> --- Changes in v12: - Minor rebase to new interface for readability. Changes in v10: - Refactored deferred_reserve check. --- mm/hugetlb.c | 27 ++++++++++++++++++++++++++- 1 file changed, 26 insertions(+), 1 deletion(-) -- 2.25.0.225.g125e21ebc7-goog