| Message ID | 20241204165503.628784-1-koichiro.den@canonical.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | hugetlb: prioritize surplus allocation from current node | expand |
On Thu, Dec 05, 2024 at 01:55:03AM +0900, Koichiro Den wrote: > Previously, surplus allocations triggered by mmap were typically made > from the node where the process was running. On a page fault, the area > was reliably dequeued from the hugepage_freelists for that node. > However, since commit 003af997c8a9 ("hugetlb: force allocating surplus > hugepages on mempolicy allowed nodes"), dequeue_hugetlb_folio_vma() may > fall back to other nodes unnecessarily even if there is no MPOL_BIND > policy, causing folios to be dequeued from nodes other than the current > one. > > Also, allocating from the node where the current process is running is > likely to result in a performance win, as mmap-ing processes often > touch the area not so long after allocation. This change minimizes > surprises for users relying on the previous behavior while maintaining > the benefit introduced by the commit. > > So, prioritize the node the current process is running on when possible. > > Signed-off-by: Koichiro Den <koichiro.den@canonical.com> > --- > mm/hugetlb.c | 20 +++++++++++++++++--- > 1 file changed, 17 insertions(+), 3 deletions(-) > > diff --git a/mm/hugetlb.c b/mm/hugetlb.c > index 5c8de0f5c760..0fa24e105202 100644 > --- a/mm/hugetlb.c > +++ b/mm/hugetlb.c > @@ -2463,7 +2463,13 @@ static int gather_surplus_pages(struct hstate *h, long delta) > long needed, allocated; > bool alloc_ok = true; > int node; > - nodemask_t *mbind_nodemask = policy_mbind_nodemask(htlb_alloc_mask(h)); > + nodemask_t *mbind_nodemask, alloc_nodemask; > + > + mbind_nodemask = policy_mbind_nodemask(htlb_alloc_mask(h)); > + if (mbind_nodemask) > + nodes_and(alloc_nodemask, *mbind_nodemask, cpuset_current_mems_allowed); > + else > + alloc_nodemask = cpuset_current_mems_allowed; > > lockdep_assert_held(&hugetlb_lock); > needed = (h->resv_huge_pages + delta) - h->free_huge_pages; > @@ -2479,8 +2485,16 @@ static int gather_surplus_pages(struct hstate *h, long delta) > spin_unlock_irq(&hugetlb_lock); > for (i = 0; i < needed; i++) { > folio = NULL; > - for_each_node_mask(node, cpuset_current_mems_allowed) { > - if (!mbind_nodemask || node_isset(node, *mbind_nodemask)) { > + > + /* Prioritize current node */ > + if (node_isset(numa_mem_id(), alloc_nodemask)) > + folio = alloc_surplus_hugetlb_folio(h, htlb_alloc_mask(h), > + numa_mem_id(), NULL); > + > + if (!folio) { > + for_each_node_mask(node, alloc_nodemask) { > + if (node == numa_mem_id()) > + continue; > folio = alloc_surplus_hugetlb_folio(h, htlb_alloc_mask(h), > node, NULL); > if (folio) Acked-by: Aristeu Rozanski <aris@ruivo.org>
diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 5c8de0f5c760..0fa24e105202 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -2463,7 +2463,13 @@ static int gather_surplus_pages(struct hstate *h, long delta) long needed, allocated; bool alloc_ok = true; int node; - nodemask_t *mbind_nodemask = policy_mbind_nodemask(htlb_alloc_mask(h)); + nodemask_t *mbind_nodemask, alloc_nodemask; + + mbind_nodemask = policy_mbind_nodemask(htlb_alloc_mask(h)); + if (mbind_nodemask) + nodes_and(alloc_nodemask, *mbind_nodemask, cpuset_current_mems_allowed); + else + alloc_nodemask = cpuset_current_mems_allowed; lockdep_assert_held(&hugetlb_lock); needed = (h->resv_huge_pages + delta) - h->free_huge_pages; @@ -2479,8 +2485,16 @@ static int gather_surplus_pages(struct hstate *h, long delta) spin_unlock_irq(&hugetlb_lock); for (i = 0; i < needed; i++) { folio = NULL; - for_each_node_mask(node, cpuset_current_mems_allowed) { - if (!mbind_nodemask || node_isset(node, *mbind_nodemask)) { + + /* Prioritize current node */ + if (node_isset(numa_mem_id(), alloc_nodemask)) + folio = alloc_surplus_hugetlb_folio(h, htlb_alloc_mask(h), + numa_mem_id(), NULL); + + if (!folio) { + for_each_node_mask(node, alloc_nodemask) { + if (node == numa_mem_id()) + continue; folio = alloc_surplus_hugetlb_folio(h, htlb_alloc_mask(h), node, NULL); if (folio)
Previously, surplus allocations triggered by mmap were typically made from the node where the process was running. On a page fault, the area was reliably dequeued from the hugepage_freelists for that node. However, since commit 003af997c8a9 ("hugetlb: force allocating surplus hugepages on mempolicy allowed nodes"), dequeue_hugetlb_folio_vma() may fall back to other nodes unnecessarily even if there is no MPOL_BIND policy, causing folios to be dequeued from nodes other than the current one. Also, allocating from the node where the current process is running is likely to result in a performance win, as mmap-ing processes often touch the area not so long after allocation. This change minimizes surprises for users relying on the previous behavior while maintaining the benefit introduced by the commit. So, prioritize the node the current process is running on when possible. Signed-off-by: Koichiro Den <koichiro.den@canonical.com> --- mm/hugetlb.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-)