The patch titled
Subject: mm: memcontrol: expose THP events on a per-memcg basis
has been added to the -mm tree. Its filename is
mm-memcontrol-expose-thp-events-on-a-per-memcg-basis.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/mm-memcontrol-expose-thp-events-on-a-per-memcg-basis.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/mm-memcontrol-expose-thp-events-on-a-per-memcg-basis.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Chris Down <chris@xxxxxxxxxxxxxx>
Subject: mm: memcontrol: expose THP events on a per-memcg basis
Currently THP allocation events data is fairly opaque, since you can only
get it system-wide. This patch makes it easier to reason about
transparent hugepage behaviour on a per-memcg basis.
For anonymous THP-backed pages, we already have MEMCG_RSS_HUGE in v1,
which is used for v1's rss_huge [sic]. This is reused here as it's fairly
involved to untangle NR_ANON_THPS right now to make it per-memcg, since
right now some of this is delegated to rmap before we have any memcg
actually assigned to the page. It's a good idea to rework that, but let's
leave untangling THP allocation for a future patch.
Link: http://lkml.kernel.org/r/20190129205852.GA7310@xxxxxxxxxxxxxx
Signed-off-by: Chris Down <chris@xxxxxxxxxxxxxx>
Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>
Cc: Tejun Heo <tj@xxxxxxxxxx>
Cc: Roman Gushchin <guro@xxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
Documentation/admin-guide/cgroup-v2.rst | 14 ++++++++++++++
mm/huge_memory.c | 2 ++
mm/khugepaged.c | 2 ++
mm/memcontrol.c | 13 +++++++++++++
4 files changed, 31 insertions(+)
--- a/Documentation/admin-guide/cgroup-v2.rst~mm-memcontrol-expose-thp-events-on-a-per-memcg-basis
+++ a/Documentation/admin-guide/cgroup-v2.rst
@@ -1195,6 +1195,10 @@ PAGE_SIZE multiple when read back.
Amount of cached filesystem data that was modified and
is currently being written back to disk
+ anon_thp
+ Amount of memory used in anonymous mappings backed by
+ transparent hugepages
+
inactive_anon, active_anon, inactive_file, active_file, unevictable
Amount of memory, swap-backed and filesystem-backed,
on the internal memory management lists used by the
@@ -1254,6 +1258,16 @@ PAGE_SIZE multiple when read back.
Amount of reclaimed lazyfree pages
+ thp_fault_alloc
+
+ Number of transparent hugepages which were allocated to satisfy
+ a page fault, including COW faults
+
+ thp_collapse_alloc
+
+ Number of transparent hugepages which were allocated to
+ allow collapsing an existing range of pages
+
memory.swap.current
A read-only single value file which exists on non-root
cgroups.
--- a/mm/huge_memory.c~mm-memcontrol-expose-thp-events-on-a-per-memcg-basis
+++ a/mm/huge_memory.c
@@ -617,6 +617,7 @@ static vm_fault_t __do_huge_pmd_anonymou
mm_inc_nr_ptes(vma->vm_mm);
spin_unlock(vmf->ptl);
count_vm_event(THP_FAULT_ALLOC);
+ count_memcg_events(memcg, THP_FAULT_ALLOC, 1);
}
return 0;
@@ -1338,6 +1339,7 @@ alloc:
}
count_vm_event(THP_FAULT_ALLOC);
+ count_memcg_events(memcg, THP_FAULT_ALLOC, 1);
if (!page)
clear_huge_page(new_page, vmf->address, HPAGE_PMD_NR);
--- a/mm/khugepaged.c~mm-memcontrol-expose-thp-events-on-a-per-memcg-basis
+++ a/mm/khugepaged.c
@@ -1074,6 +1074,7 @@ static void collapse_huge_page(struct mm
BUG_ON(!pmd_none(*pmd));
page_add_new_anon_rmap(new_page, vma, address, true);
mem_cgroup_commit_charge(new_page, memcg, false, true);
+ count_memcg_events(memcg, THP_COLLAPSE_ALLOC, 1);
lru_cache_add_active_or_unevictable(new_page, vma);
pgtable_trans_huge_deposit(mm, pmd, pgtable);
set_pmd_at(mm, address, pmd, _pmd);
@@ -1502,6 +1503,7 @@ xa_unlocked:
page_ref_add(new_page, HPAGE_PMD_NR - 1);
set_page_dirty(new_page);
mem_cgroup_commit_charge(new_page, memcg, false, true);
+ count_memcg_events(memcg, THP_COLLAPSE_ALLOC, 1);
lru_cache_add_anon(new_page);
/*
--- a/mm/memcontrol.c~mm-memcontrol-expose-thp-events-on-a-per-memcg-basis
+++ a/mm/memcontrol.c
@@ -5586,6 +5586,15 @@ static int memory_stat_show(struct seq_f
seq_printf(m, "file_writeback %llu\n",
(u64)acc.stat[NR_WRITEBACK] * PAGE_SIZE);
+ /*
+ * TODO: We should eventually replace our own MEMCG_RSS_HUGE counter
+ * with the NR_ANON_THP vm counter, but right now it's a pain in the
+ * arse because it requires migrating the work out of rmap to a place
+ * where the page->mem_cgroup is set up and stable.
+ */
+ seq_printf(m, "anon_thp %llu\n",
+ (u64)acc.stat[MEMCG_RSS_HUGE] * PAGE_SIZE);
+
for (i = 0; i < NR_LRU_LISTS; i++)
seq_printf(m, "%s %llu\n", mem_cgroup_lru_names[i],
(u64)acc.lru_pages[i] * PAGE_SIZE);
@@ -5617,6 +5626,10 @@ static int memory_stat_show(struct seq_f
seq_printf(m, "pglazyfree %lu\n", acc.events[PGLAZYFREE]);
seq_printf(m, "pglazyfreed %lu\n", acc.events[PGLAZYFREED]);
+ seq_printf(m, "thp_fault_alloc %lu\n", acc.events[THP_FAULT_ALLOC]);
+ seq_printf(m, "thp_collapse_alloc %lu\n",
+ acc.events[THP_COLLAPSE_ALLOC]);
+
return 0;
}
_
Patches currently in -mm which might be from chris@xxxxxxxxxxxxxx are
mm-create-mem_cgroup_from_seq.patch
mm-extract-memcg-maxable-seq_file-logic-to-seq_show_memcg_tunable.patch
mm-proportional-memorylowmin-reclaim.patch
mm-memcontrol-expose-thp-events-on-a-per-memcg-basis.patch
