[PATCH 0/7 v2] vm, vmscan: enahance vmscan tracepoints

[PATCH 0/7 v2] vm, vmscan: enahance vmscan tracepoints

[Michal Hocko]

Hi,
this is the second version of the patchset [1]. I hope I've addressed all
the review feedback.

While debugging [2] I've realized that there is some room for
improvements in the tracepoints set we offer currently. I had hard times
to make any conclusion from the existing ones. The resulting problem
turned out to be active list aging [3] and we are missing at least two
tracepoints to debug such a problem.

Some existing tracepoints could export more information to see _why_ the
reclaim progress cannot be made not only _how much_ we could reclaim.
The later could be seen quite reasonably from the vmstat counters
already. It can be argued that we are showing too many implementation
details in those tracepoints but I consider them way too lowlevel
already to be usable by any kernel independent userspace. I would be
_really_ surprised if anything but debugging tools have used them.

Any feedback is highly appreciated.

[1] http://lkml.kernel.org/r/20161228153032.10821-1-mhocko@kernel.org
[2] http://lkml.kernel.org/r/20161215225702.GA27944@boerne.fritz.box
[3] http://lkml.kernel.org/r/20161223105157.GB23109@dhcp22.suse.cz

[PATCH 1/7] mm, vmscan: remove unused mm_vmscan_memcg_isolate

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

the trace point is not used since 925b7673cce3 ("mm: make per-memcg LRU
lists exclusive") so it can be removed.

Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/vmscan.h | 31 +------------------------------
 1 file changed, 1 insertion(+), 30 deletions(-)

diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index c88fd0934e7e..39bad8921ca1 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -269,8 +269,7 @@ TRACE_EVENT(mm_shrink_slab_end,
 		__entry->retval)
 );
 
-DECLARE_EVENT_CLASS(mm_vmscan_lru_isolate_template,
-
+TRACE_EVENT(mm_vmscan_lru_isolate,
 	TP_PROTO(int classzone_idx,
 		int order,
 		unsigned long nr_requested,
@@ -311,34 +310,6 @@ DECLARE_EVENT_CLASS(mm_vmscan_lru_isolate_template,
 		__entry->file)
 );
 
-DEFINE_EVENT(mm_vmscan_lru_isolate_template, mm_vmscan_lru_isolate,
-
-	TP_PROTO(int classzone_idx,
-		int order,
-		unsigned long nr_requested,
-		unsigned long nr_scanned,
-		unsigned long nr_taken,
-		isolate_mode_t isolate_mode,
-		int file),
-
-	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_taken, isolate_mode, file)
-
-);
-
-DEFINE_EVENT(mm_vmscan_lru_isolate_template, mm_vmscan_memcg_isolate,
-
-	TP_PROTO(int classzone_idx,
-		int order,
-		unsigned long nr_requested,
-		unsigned long nr_scanned,
-		unsigned long nr_taken,
-		isolate_mode_t isolate_mode,
-		int file),
-
-	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_taken, isolate_mode, file)
-
-);
-
 TRACE_EVENT(mm_vmscan_writepage,
 
 	TP_PROTO(struct page *page),
-- 
2.11.0

[PATCH 2/7] mm, vmscan: add active list aging tracepoint

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

Our reclaim process has several tracepoints to tell us more about how
things are progressing. We are, however, missing a tracepoint to track
active list aging. Introduce mm_vmscan_lru_shrink_active which reports
the number of
	- nr_scanned, nr_taken pages to tell us the LRU isolation
	  effectiveness.
	- nr_referenced pages which tells us that we are hitting referenced
	  pages which are deactivated. If this is a large part of the
	  reported nr_deactivated pages then we might be hitting into
	  the active list too early because they might be still part of
	  the working set. This might help to debug performance issues.
	- nr_activated pages which tells us how many pages are kept on the
	  active list - mostly exec file backed pages. A high number can
	  indicate that we might be trashing on executables.

Changes since v1
- report nr_taken pages as per Minchan
- report nr_activated as per Minchan
- do not report nr_freed pages because that would add a tiny overhead to
  free_hot_cold_page_list which is a hot path
- do not report nr_unevictable because we can report this number via a
  different and more generic tracepoint in putback_lru_page
- fix move_active_pages_to_lru to report proper page count when we hit
  into large pages
- drop nr_scanned because this can be obtained from
  trace_mm_vmscan_lru_isolate as per Minchan

Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/vmscan.h | 36 ++++++++++++++++++++++++++++++++++++
 mm/vmscan.c                   | 18 ++++++++++++++----
 2 files changed, 50 insertions(+), 4 deletions(-)

diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index 39bad8921ca1..087c0b625ba7 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -363,6 +363,42 @@ TRACE_EVENT(mm_vmscan_lru_shrink_inactive,
 		show_reclaim_flags(__entry->reclaim_flags))
 );
 
+TRACE_EVENT(mm_vmscan_lru_shrink_active,
+
+	TP_PROTO(int nid, unsigned long nr_taken,
+		unsigned long nr_activate, unsigned long nr_deactivated,
+		unsigned long nr_referenced, int priority, int file),
+
+	TP_ARGS(nid, nr_taken, nr_activate, nr_deactivated, nr_referenced, priority, file),
+
+	TP_STRUCT__entry(
+		__field(int, nid)
+		__field(unsigned long, nr_taken)
+		__field(unsigned long, nr_activate)
+		__field(unsigned long, nr_deactivated)
+		__field(unsigned long, nr_referenced)
+		__field(int, priority)
+		__field(int, reclaim_flags)
+	),
+
+	TP_fast_assign(
+		__entry->nid = nid;
+		__entry->nr_taken = nr_taken;
+		__entry->nr_activate = nr_activate;
+		__entry->nr_deactivated = nr_deactivated;
+		__entry->nr_referenced = nr_referenced;
+		__entry->priority = priority;
+		__entry->reclaim_flags = trace_shrink_flags(file);
+	),
+
+	TP_printk("nid=%d nr_taken=%ld nr_activated=%ld nr_deactivated=%ld nr_referenced=%ld priority=%d flags=%s",
+		__entry->nid,
+		__entry->nr_taken,
+		__entry->nr_activate, __entry->nr_deactivated, __entry->nr_referenced,
+		__entry->priority,
+		show_reclaim_flags(__entry->reclaim_flags))
+);
+
 #endif /* _TRACE_VMSCAN_H */
 
 /* This part must be outside protection */
diff --git a/mm/vmscan.c b/mm/vmscan.c
index c4abf08861d2..70d1c55463c0 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1846,9 +1846,11 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
  *
  * The downside is that we have to touch page->_refcount against each page.
  * But we had to alter page->flags anyway.
+ *
+ * Returns the number of pages moved to the given lru.
  */
 
-static void move_active_pages_to_lru(struct lruvec *lruvec,
+static unsigned move_active_pages_to_lru(struct lruvec *lruvec,
 				     struct list_head *list,
 				     struct list_head *pages_to_free,
 				     enum lru_list lru)
@@ -1857,6 +1859,7 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
 	unsigned long pgmoved = 0;
 	struct page *page;
 	int nr_pages;
+	int nr_moved = 0;
 
 	while (!list_empty(list)) {
 		page = lru_to_page(list);
@@ -1882,11 +1885,15 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
 				spin_lock_irq(&pgdat->lru_lock);
 			} else
 				list_add(&page->lru, pages_to_free);
+		} else {
+			nr_moved += nr_pages;
 		}
 	}
 
 	if (!is_active_lru(lru))
 		__count_vm_events(PGDEACTIVATE, pgmoved);
+
+	return nr_moved;
 }
 
 static void shrink_active_list(unsigned long nr_to_scan,
@@ -1902,7 +1909,8 @@ static void shrink_active_list(unsigned long nr_to_scan,
 	LIST_HEAD(l_inactive);
 	struct page *page;
 	struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat;
-	unsigned long nr_rotated = 0;
+	unsigned nr_deactivate, nr_activate;
+	unsigned nr_rotated = 0;
 	isolate_mode_t isolate_mode = 0;
 	int file = is_file_lru(lru);
 	struct pglist_data *pgdat = lruvec_pgdat(lruvec);
@@ -1980,13 +1988,15 @@ static void shrink_active_list(unsigned long nr_to_scan,
 	 */
 	reclaim_stat->recent_rotated[file] += nr_rotated;
 
-	move_active_pages_to_lru(lruvec, &l_active, &l_hold, lru);
-	move_active_pages_to_lru(lruvec, &l_inactive, &l_hold, lru - LRU_ACTIVE);
+	nr_activate = move_active_pages_to_lru(lruvec, &l_active, &l_hold, lru);
+	nr_deactivate = move_active_pages_to_lru(lruvec, &l_inactive, &l_hold, lru - LRU_ACTIVE);
 	__mod_node_page_state(pgdat, NR_ISOLATED_ANON + file, -nr_taken);
 	spin_unlock_irq(&pgdat->lru_lock);
 
 	mem_cgroup_uncharge_list(&l_hold);
 	free_hot_cold_page_list(&l_hold, true);
+	trace_mm_vmscan_lru_shrink_active(pgdat->node_id, nr_taken, nr_activate,
+			nr_deactivate, nr_rotated, sc->priority, file);
 }
 
 /*
-- 
2.11.0

[PATCH 3/7] mm, vmscan: show the number of skipped pages in mm_vmscan_lru_isolate

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

mm_vmscan_lru_isolate shows the number of requested, scanned and taken
pages. This is mostly OK but on 32b systems the number of scanned pages
is quite misleading because it includes both the scanned and skipped
pages.  Moreover the skipped part is scaled based on the number of taken
pages. Let's report the exact numbers without any additional logic and
add the number of skipped pages. This should make the reported data much
more easier to interpret.

Acked-by: Minchan Kim <minchan@kernel.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/vmscan.h |  8 ++++++--
 mm/vmscan.c                   | 10 +++++-----
 2 files changed, 11 insertions(+), 7 deletions(-)

diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index 087c0b625ba7..36c999f806bf 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -274,17 +274,19 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		int order,
 		unsigned long nr_requested,
 		unsigned long nr_scanned,
+		unsigned long nr_skipped,
 		unsigned long nr_taken,
 		isolate_mode_t isolate_mode,
 		int file),
 
-	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_taken, isolate_mode, file),
+	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_skipped, nr_taken, isolate_mode, file),
 
 	TP_STRUCT__entry(
 		__field(int, classzone_idx)
 		__field(int, order)
 		__field(unsigned long, nr_requested)
 		__field(unsigned long, nr_scanned)
+		__field(unsigned long, nr_skipped)
 		__field(unsigned long, nr_taken)
 		__field(isolate_mode_t, isolate_mode)
 		__field(int, file)
@@ -295,17 +297,19 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		__entry->order = order;
 		__entry->nr_requested = nr_requested;
 		__entry->nr_scanned = nr_scanned;
+		__entry->nr_skipped = nr_skipped;
 		__entry->nr_taken = nr_taken;
 		__entry->isolate_mode = isolate_mode;
 		__entry->file = file;
 	),
 
-	TP_printk("isolate_mode=%d classzone=%d order=%d nr_requested=%lu nr_scanned=%lu nr_taken=%lu file=%d",
+	TP_printk("isolate_mode=%d classzone=%d order=%d nr_requested=%lu nr_scanned=%lu nr_skipped=%lu nr_taken=%lu file=%d",
 		__entry->isolate_mode,
 		__entry->classzone_idx,
 		__entry->order,
 		__entry->nr_requested,
 		__entry->nr_scanned,
+		__entry->nr_skipped,
 		__entry->nr_taken,
 		__entry->file)
 );
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 70d1c55463c0..31c623d5acb4 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1428,6 +1428,7 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 	unsigned long nr_taken = 0;
 	unsigned long nr_zone_taken[MAX_NR_ZONES] = { 0 };
 	unsigned long nr_skipped[MAX_NR_ZONES] = { 0, };
+	unsigned long skipped = 0, total_skipped = 0;
 	unsigned long scan, nr_pages;
 	LIST_HEAD(pages_skipped);
 
@@ -1479,14 +1480,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 	 */
 	if (!list_empty(&pages_skipped)) {
 		int zid;
-		unsigned long total_skipped = 0;
 
 		for (zid = 0; zid < MAX_NR_ZONES; zid++) {
 			if (!nr_skipped[zid])
 				continue;
 
 			__count_zid_vm_events(PGSCAN_SKIP, zid, nr_skipped[zid]);
-			total_skipped += nr_skipped[zid];
+			skipped += nr_skipped[zid];
 		}
 
 		/*
@@ -1494,13 +1494,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 		 * close to unreclaimable. If the LRU list is empty, account
 		 * skipped pages as a full scan.
 		 */
-		scan += list_empty(src) ? total_skipped : total_skipped >> 2;
+		total_skipped = list_empty(src) ? skipped : skipped >> 2;
 
 		list_splice(&pages_skipped, src);
 	}
-	*nr_scanned = scan;
+	*nr_scanned = scan + total_skipped;
 	trace_mm_vmscan_lru_isolate(sc->reclaim_idx, sc->order, nr_to_scan, scan,
-				    nr_taken, mode, is_file_lru(lru));
+				    skipped, nr_taken, mode, is_file_lru(lru));
 	update_lru_sizes(lruvec, lru, nr_zone_taken, nr_taken);
 	return nr_taken;
 }
-- 
2.11.0

[PATCH 4/7] mm, vmscan: show LRU name in mm_vmscan_lru_isolate tracepoint

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

mm_vmscan_lru_isolate currently prints only whether the LRU we isolate
from is file or anonymous but we do not know which LRU this is.

It is useful to know whether the list is active or inactive, since we
are using the same function to isolate pages from both of them and it's
hard to distinguish otherwise.

Chaneges since v1
- drop LRU_ prefix from names and use lowercase as per Vlastimil
- move and convert show_lru_name to mmflags.h EM magic as per Vlastimil

Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/mmflags.h |  8 ++++++++
 include/trace/events/vmscan.h  | 12 ++++++------
 mm/vmscan.c                    |  2 +-
 3 files changed, 15 insertions(+), 7 deletions(-)

diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h
index aa4caa6914a9..6172afa2fd82 100644
--- a/include/trace/events/mmflags.h
+++ b/include/trace/events/mmflags.h
@@ -240,6 +240,13 @@ IF_HAVE_VM_SOFTDIRTY(VM_SOFTDIRTY,	"softdirty"	)		\
 	IFDEF_ZONE_HIGHMEM(	EM (ZONE_HIGHMEM,"HighMem"))	\
 				EMe(ZONE_MOVABLE,"Movable")
 
+#define LRU_NAMES		\
+		EM (LRU_INACTIVE_ANON, "inactive_anon") \
+		EM (LRU_ACTIVE_ANON, "active_anon") \
+		EM (LRU_INACTIVE_FILE, "inactive_file") \
+		EM (LRU_ACTIVE_FILE, "active_file") \
+		EMe(LRU_UNEVICTABLE, "unevictable")
+
 /*
  * First define the enums in the above macros to be exported to userspace
  * via TRACE_DEFINE_ENUM().
@@ -253,6 +260,7 @@ COMPACTION_STATUS
 COMPACTION_PRIORITY
 COMPACTION_FEEDBACK
 ZONE_TYPE
+LRU_NAMES
 
 /*
  * Now redefine the EM() and EMe() macros to map the enums to the strings
diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index 36c999f806bf..7ec59e0432c4 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -277,9 +277,9 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		unsigned long nr_skipped,
 		unsigned long nr_taken,
 		isolate_mode_t isolate_mode,
-		int file),
+		int lru),
 
-	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_skipped, nr_taken, isolate_mode, file),
+	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_skipped, nr_taken, isolate_mode, lru),
 
 	TP_STRUCT__entry(
 		__field(int, classzone_idx)
@@ -289,7 +289,7 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		__field(unsigned long, nr_skipped)
 		__field(unsigned long, nr_taken)
 		__field(isolate_mode_t, isolate_mode)
-		__field(int, file)
+		__field(int, lru)
 	),
 
 	TP_fast_assign(
@@ -300,10 +300,10 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		__entry->nr_skipped = nr_skipped;
 		__entry->nr_taken = nr_taken;
 		__entry->isolate_mode = isolate_mode;
-		__entry->file = file;
+		__entry->lru = lru;
 	),
 
-	TP_printk("isolate_mode=%d classzone=%d order=%d nr_requested=%lu nr_scanned=%lu nr_skipped=%lu nr_taken=%lu file=%d",
+	TP_printk("isolate_mode=%d classzone=%d order=%d nr_requested=%lu nr_scanned=%lu nr_skipped=%lu nr_taken=%lu lru=%s",
 		__entry->isolate_mode,
 		__entry->classzone_idx,
 		__entry->order,
@@ -311,7 +311,7 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		__entry->nr_scanned,
 		__entry->nr_skipped,
 		__entry->nr_taken,
-		__entry->file)
+		__print_symbolic(__entry->lru, LRU_NAMES))
 );
 
 TRACE_EVENT(mm_vmscan_writepage,
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 31c623d5acb4..13758aaed78b 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1500,7 +1500,7 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 	}
 	*nr_scanned = scan + total_skipped;
 	trace_mm_vmscan_lru_isolate(sc->reclaim_idx, sc->order, nr_to_scan, scan,
-				    skipped, nr_taken, mode, is_file_lru(lru));
+				    skipped, nr_taken, mode, lru);
 	update_lru_sizes(lruvec, lru, nr_zone_taken, nr_taken);
 	return nr_taken;
 }
-- 
2.11.0

[PATCH 5/7] mm, vmscan: extract shrink_page_list reclaim counters into a struct

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

shrink_page_list returns quite some counters back to its caller. Extract
the existing 5 into struct reclaim_stat because this makes the code
easier to follow and also allows further counters to be returned.

While we are at it, make all of them unsigned rather than unsigned long
as we do not really need full 64b for them (we never scan more than
SWAP_CLUSTER_MAX pages at once). This should reduce some stack space.

This patch shouldn't introduce any functional change.

Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 mm/vmscan.c | 61 ++++++++++++++++++++++++++++++-------------------------------
 1 file changed, 30 insertions(+), 31 deletions(-)

diff --git a/mm/vmscan.c b/mm/vmscan.c
index 13758aaed78b..920e47a905c3 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -902,6 +902,14 @@ static void page_check_dirty_writeback(struct page *page,
 		mapping->a_ops->is_dirty_writeback(page, dirty, writeback);
 }
 
+struct reclaim_stat {
+	unsigned nr_dirty;
+	unsigned nr_unqueued_dirty;
+	unsigned nr_congested;
+	unsigned nr_writeback;
+	unsigned nr_immediate;
+};
+
 /*
  * shrink_page_list() returns the number of reclaimed pages
  */
@@ -909,22 +917,18 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 				      struct pglist_data *pgdat,
 				      struct scan_control *sc,
 				      enum ttu_flags ttu_flags,
-				      unsigned long *ret_nr_dirty,
-				      unsigned long *ret_nr_unqueued_dirty,
-				      unsigned long *ret_nr_congested,
-				      unsigned long *ret_nr_writeback,
-				      unsigned long *ret_nr_immediate,
+				      struct reclaim_stat *stat,
 				      bool force_reclaim)
 {
 	LIST_HEAD(ret_pages);
 	LIST_HEAD(free_pages);
 	int pgactivate = 0;
-	unsigned long nr_unqueued_dirty = 0;
-	unsigned long nr_dirty = 0;
-	unsigned long nr_congested = 0;
-	unsigned long nr_reclaimed = 0;
-	unsigned long nr_writeback = 0;
-	unsigned long nr_immediate = 0;
+	unsigned nr_unqueued_dirty = 0;
+	unsigned nr_dirty = 0;
+	unsigned nr_congested = 0;
+	unsigned nr_reclaimed = 0;
+	unsigned nr_writeback = 0;
+	unsigned nr_immediate = 0;
 
 	cond_resched();
 
@@ -1266,11 +1270,13 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 	list_splice(&ret_pages, page_list);
 	count_vm_events(PGACTIVATE, pgactivate);
 
-	*ret_nr_dirty += nr_dirty;
-	*ret_nr_congested += nr_congested;
-	*ret_nr_unqueued_dirty += nr_unqueued_dirty;
-	*ret_nr_writeback += nr_writeback;
-	*ret_nr_immediate += nr_immediate;
+	if (stat) {
+		stat->nr_dirty = nr_dirty;
+		stat->nr_congested = nr_congested;
+		stat->nr_unqueued_dirty = nr_unqueued_dirty;
+		stat->nr_writeback = nr_writeback;
+		stat->nr_immediate = nr_immediate;
+	}
 	return nr_reclaimed;
 }
 
@@ -1282,7 +1288,7 @@ unsigned long reclaim_clean_pages_from_list(struct zone *zone,
 		.priority = DEF_PRIORITY,
 		.may_unmap = 1,
 	};
-	unsigned long ret, dummy1, dummy2, dummy3, dummy4, dummy5;
+	unsigned long ret;
 	struct page *page, *next;
 	LIST_HEAD(clean_pages);
 
@@ -1295,8 +1301,7 @@ unsigned long reclaim_clean_pages_from_list(struct zone *zone,
 	}
 
 	ret = shrink_page_list(&clean_pages, zone->zone_pgdat, &sc,
-			TTU_UNMAP|TTU_IGNORE_ACCESS,
-			&dummy1, &dummy2, &dummy3, &dummy4, &dummy5, true);
+			TTU_UNMAP|TTU_IGNORE_ACCESS, NULL, true);
 	list_splice(&clean_pages, page_list);
 	mod_node_page_state(zone->zone_pgdat, NR_ISOLATED_FILE, -ret);
 	return ret;
@@ -1696,11 +1701,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
 	unsigned long nr_scanned;
 	unsigned long nr_reclaimed = 0;
 	unsigned long nr_taken;
-	unsigned long nr_dirty = 0;
-	unsigned long nr_congested = 0;
-	unsigned long nr_unqueued_dirty = 0;
-	unsigned long nr_writeback = 0;
-	unsigned long nr_immediate = 0;
+	struct reclaim_stat stat = {};
 	isolate_mode_t isolate_mode = 0;
 	int file = is_file_lru(lru);
 	struct pglist_data *pgdat = lruvec_pgdat(lruvec);
@@ -1745,9 +1746,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
 		return 0;
 
 	nr_reclaimed = shrink_page_list(&page_list, pgdat, sc, TTU_UNMAP,
-				&nr_dirty, &nr_unqueued_dirty, &nr_congested,
-				&nr_writeback, &nr_immediate,
-				false);
+				&stat, false);
 
 	spin_lock_irq(&pgdat->lru_lock);
 
@@ -1781,7 +1780,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
 	 * of pages under pages flagged for immediate reclaim and stall if any
 	 * are encountered in the nr_immediate check below.
 	 */
-	if (nr_writeback && nr_writeback == nr_taken)
+	if (stat.nr_writeback && stat.nr_writeback == nr_taken)
 		set_bit(PGDAT_WRITEBACK, &pgdat->flags);
 
 	/*
@@ -1793,7 +1792,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
 		 * Tag a zone as congested if all the dirty pages scanned were
 		 * backed by a congested BDI and wait_iff_congested will stall.
 		 */
-		if (nr_dirty && nr_dirty == nr_congested)
+		if (stat.nr_dirty && stat.nr_dirty == stat.nr_congested)
 			set_bit(PGDAT_CONGESTED, &pgdat->flags);
 
 		/*
@@ -1802,7 +1801,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
 		 * the pgdat PGDAT_DIRTY and kswapd will start writing pages from
 		 * reclaim context.
 		 */
-		if (nr_unqueued_dirty == nr_taken)
+		if (stat.nr_unqueued_dirty == nr_taken)
 			set_bit(PGDAT_DIRTY, &pgdat->flags);
 
 		/*
@@ -1811,7 +1810,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
 		 * that pages are cycling through the LRU faster than
 		 * they are written so also forcibly stall.
 		 */
-		if (nr_immediate && current_may_throttle())
+		if (stat.nr_immediate && current_may_throttle())
 			congestion_wait(BLK_RW_ASYNC, HZ/10);
 	}
 
-- 
2.11.0

[PATCH 6/7] mm, vmscan: enhance mm_vmscan_lru_shrink_inactive tracepoint

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

mm_vmscan_lru_shrink_inactive will currently report the number of
scanned and reclaimed pages. This doesn't give us an idea how the
reclaim went except for the overall effectiveness though. Export
and show other counters which will tell us why we couldn't reclaim
some pages.
	- nr_dirty, nr_writeback, nr_congested and nr_immediate tells
	  us how many pages are blocked due to IO
	- nr_activate tells us how many pages were moved to the active
	  list
	- nr_ref_keep reports how many pages are kept on the LRU due
	  to references (mostly for the file pages which are about to
	  go for another round through the inactive list)
	- nr_unmap_fail - how many pages failed to unmap

All these are rather low level so they might change in future but the
tracepoint is already implementation specific so no tools should be
depending on its stability.

Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/vmscan.h | 29 ++++++++++++++++++++++++++---
 mm/vmscan.c                   | 14 ++++++++++++++
 2 files changed, 40 insertions(+), 3 deletions(-)

diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index 7ec59e0432c4..9037c1734294 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -340,14 +340,27 @@ TRACE_EVENT(mm_vmscan_lru_shrink_inactive,
 
 	TP_PROTO(int nid,
 		unsigned long nr_scanned, unsigned long nr_reclaimed,
+		unsigned long nr_dirty, unsigned long nr_writeback,
+		unsigned long nr_congested, unsigned long nr_immediate,
+		unsigned long nr_activate, unsigned long nr_ref_keep,
+		unsigned long nr_unmap_fail,
 		int priority, int file),
 
-	TP_ARGS(nid, nr_scanned, nr_reclaimed, priority, file),
+	TP_ARGS(nid, nr_scanned, nr_reclaimed, nr_dirty, nr_writeback,
+		nr_congested, nr_immediate, nr_activate, nr_ref_keep,
+		nr_unmap_fail, priority, file),
 
 	TP_STRUCT__entry(
 		__field(int, nid)
 		__field(unsigned long, nr_scanned)
 		__field(unsigned long, nr_reclaimed)
+		__field(unsigned long, nr_dirty)
+		__field(unsigned long, nr_writeback)
+		__field(unsigned long, nr_congested)
+		__field(unsigned long, nr_immediate)
+		__field(unsigned long, nr_activate)
+		__field(unsigned long, nr_ref_keep)
+		__field(unsigned long, nr_unmap_fail)
 		__field(int, priority)
 		__field(int, reclaim_flags)
 	),
@@ -356,14 +369,24 @@ TRACE_EVENT(mm_vmscan_lru_shrink_inactive,
 		__entry->nid = nid;
 		__entry->nr_scanned = nr_scanned;
 		__entry->nr_reclaimed = nr_reclaimed;
+		__entry->nr_dirty = nr_dirty;
+		__entry->nr_writeback = nr_writeback;
+		__entry->nr_congested = nr_congested;
+		__entry->nr_immediate = nr_immediate;
+		__entry->nr_activate = nr_activate;
+		__entry->nr_ref_keep = nr_ref_keep;
+		__entry->nr_unmap_fail = nr_unmap_fail;
 		__entry->priority = priority;
 		__entry->reclaim_flags = trace_shrink_flags(file);
 	),
 
-	TP_printk("nid=%d nr_scanned=%ld nr_reclaimed=%ld priority=%d flags=%s",
+	TP_printk("nid=%d nr_scanned=%ld nr_reclaimed=%ld nr_dirty=%ld nr_writeback=%ld nr_congested=%ld nr_immediate=%ld nr_activate=%ld nr_ref_keep=%ld nr_unmap_fail=%ld priority=%d flags=%s",
 		__entry->nid,
 		__entry->nr_scanned, __entry->nr_reclaimed,
-		__entry->priority,
+		__entry->nr_dirty, __entry->nr_writeback,
+		__entry->nr_congested, __entry->nr_immediate,
+		__entry->nr_activate, __entry->nr_ref_keep,
+		__entry->nr_unmap_fail, __entry->priority,
 		show_reclaim_flags(__entry->reclaim_flags))
 );
 
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 920e47a905c3..d05e42bee511 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -908,6 +908,9 @@ struct reclaim_stat {
 	unsigned nr_congested;
 	unsigned nr_writeback;
 	unsigned nr_immediate;
+	unsigned nr_activate;
+	unsigned nr_ref_keep;
+	unsigned nr_unmap_fail;
 };
 
 /*
@@ -929,6 +932,8 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 	unsigned nr_reclaimed = 0;
 	unsigned nr_writeback = 0;
 	unsigned nr_immediate = 0;
+	unsigned nr_ref_keep = 0;
+	unsigned nr_unmap_fail = 0;
 
 	cond_resched();
 
@@ -1067,6 +1072,7 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 		case PAGEREF_ACTIVATE:
 			goto activate_locked;
 		case PAGEREF_KEEP:
+			nr_ref_keep++;
 			goto keep_locked;
 		case PAGEREF_RECLAIM:
 		case PAGEREF_RECLAIM_CLEAN:
@@ -1104,6 +1110,7 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 				(ttu_flags | TTU_BATCH_FLUSH | TTU_LZFREE) :
 				(ttu_flags | TTU_BATCH_FLUSH))) {
 			case SWAP_FAIL:
+				nr_unmap_fail++;
 				goto activate_locked;
 			case SWAP_AGAIN:
 				goto keep_locked;
@@ -1276,6 +1283,9 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 		stat->nr_unqueued_dirty = nr_unqueued_dirty;
 		stat->nr_writeback = nr_writeback;
 		stat->nr_immediate = nr_immediate;
+		stat->nr_activate = pgactivate;
+		stat->nr_ref_keep = nr_ref_keep;
+		stat->nr_unmap_fail = nr_unmap_fail;
 	}
 	return nr_reclaimed;
 }
@@ -1825,6 +1835,10 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
 
 	trace_mm_vmscan_lru_shrink_inactive(pgdat->node_id,
 			nr_scanned, nr_reclaimed,
+			stat.nr_dirty,  stat.nr_writeback,
+			stat.nr_congested, stat.nr_immediate,
+			stat.nr_activate, stat.nr_ref_keep,
+			stat.nr_unmap_fail,
 			sc->priority, file);
 	return nr_reclaimed;
 }
-- 
2.11.0

[PATCH 7/7] mm, vmscan: add mm_vmscan_inactive_list_is_low tracepoint

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

Currently we have tracepoints for both active and inactive LRU lists
reclaim but we do not have any which would tell us why we we decided to
age the active list. Without that it is quite hard to diagnose
active/inactive lists balancing. Add mm_vmscan_inactive_list_is_low
tracepoint to tell us this information.

Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/vmscan.h | 40 ++++++++++++++++++++++++++++++++++++++++
 mm/vmscan.c                   | 23 ++++++++++++++---------
 2 files changed, 54 insertions(+), 9 deletions(-)

diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index 9037c1734294..625a4b7967d0 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -15,6 +15,7 @@
 #define RECLAIM_WB_MIXED	0x0010u
 #define RECLAIM_WB_SYNC		0x0004u /* Unused, all reclaim async */
 #define RECLAIM_WB_ASYNC	0x0008u
+#define RECLAIM_WB_LRU		(RECLAIM_WB_ANON|RECLAIM_WB_FILE)
 
 #define show_reclaim_flags(flags)				\
 	(flags) ? __print_flags(flags, "|",			\
@@ -426,6 +427,45 @@ TRACE_EVENT(mm_vmscan_lru_shrink_active,
 		show_reclaim_flags(__entry->reclaim_flags))
 );
 
+TRACE_EVENT(mm_vmscan_inactive_list_is_low,
+
+	TP_PROTO(int nid, int reclaim_idx,
+		unsigned long total_inactive, unsigned long inactive,
+		unsigned long total_active, unsigned long active,
+		unsigned long ratio, int file),
+
+	TP_ARGS(nid, reclaim_idx, total_inactive, inactive, total_active, active, ratio, file),
+
+	TP_STRUCT__entry(
+		__field(int, nid)
+		__field(int, reclaim_idx)
+		__field(unsigned long, total_inactive)
+		__field(unsigned long, inactive)
+		__field(unsigned long, total_active)
+		__field(unsigned long, active)
+		__field(unsigned long, ratio)
+		__field(int, reclaim_flags)
+	),
+
+	TP_fast_assign(
+		__entry->nid = nid;
+		__entry->reclaim_idx = reclaim_idx;
+		__entry->total_inactive = total_inactive;
+		__entry->inactive = inactive;
+		__entry->total_active = total_active;
+		__entry->active = active;
+		__entry->ratio = ratio;
+		__entry->reclaim_flags = trace_shrink_flags(file) & RECLAIM_WB_LRU;
+	),
+
+	TP_printk("nid=%d reclaim_idx=%d total_inactive=%ld inactive=%ld total_active=%ld active=%ld ratio=%ld flags=%s",
+		__entry->nid,
+		__entry->reclaim_idx,
+		__entry->total_inactive, __entry->inactive,
+		__entry->total_active, __entry->active,
+		__entry->ratio,
+		show_reclaim_flags(__entry->reclaim_flags))
+);
 #endif /* _TRACE_VMSCAN_H */
 
 /* This part must be outside protection */
diff --git a/mm/vmscan.c b/mm/vmscan.c
index d05e42bee511..2a5c6c3fed2d 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2039,11 +2039,11 @@ static void shrink_active_list(unsigned long nr_to_scan,
  *   10TB     320        32GB
  */
 static bool inactive_list_is_low(struct lruvec *lruvec, bool file,
-						struct scan_control *sc)
+						struct scan_control *sc, bool trace)
 {
 	unsigned long inactive_ratio;
-	unsigned long inactive;
-	unsigned long active;
+	unsigned long total_inactive, inactive;
+	unsigned long total_active, active;
 	unsigned long gb;
 	struct pglist_data *pgdat = lruvec_pgdat(lruvec);
 	int zid;
@@ -2055,8 +2055,8 @@ static bool inactive_list_is_low(struct lruvec *lruvec, bool file,
 	if (!file && !total_swap_pages)
 		return false;
 
-	inactive = lruvec_lru_size(lruvec, file * LRU_FILE);
-	active = lruvec_lru_size(lruvec, file * LRU_FILE + LRU_ACTIVE);
+	total_inactive = inactive = lruvec_lru_size(lruvec, file * LRU_FILE);
+	total_active = active = lruvec_lru_size(lruvec, file * LRU_FILE + LRU_ACTIVE);
 
 	/*
 	 * For zone-constrained allocations, it is necessary to check if
@@ -2085,6 +2085,11 @@ static bool inactive_list_is_low(struct lruvec *lruvec, bool file,
 	else
 		inactive_ratio = 1;
 
+	if (trace)
+		trace_mm_vmscan_inactive_list_is_low(pgdat->node_id,
+				sc->reclaim_idx,
+				total_inactive, inactive,
+				total_active, active, inactive_ratio, file);
 	return inactive * inactive_ratio < active;
 }
 
@@ -2092,7 +2097,7 @@ static unsigned long shrink_list(enum lru_list lru, unsigned long nr_to_scan,
 				 struct lruvec *lruvec, struct scan_control *sc)
 {
 	if (is_active_lru(lru)) {
-		if (inactive_list_is_low(lruvec, is_file_lru(lru), sc))
+		if (inactive_list_is_low(lruvec, is_file_lru(lru), sc, true))
 			shrink_active_list(nr_to_scan, lruvec, sc, lru);
 		return 0;
 	}
@@ -2223,7 +2228,7 @@ static void get_scan_count(struct lruvec *lruvec, struct mem_cgroup *memcg,
 	 * lruvec even if it has plenty of old anonymous pages unless the
 	 * system is under heavy pressure.
 	 */
-	if (!inactive_list_is_low(lruvec, true, sc) &&
+	if (!inactive_list_is_low(lruvec, true, sc, false) &&
 	    lruvec_lru_size(lruvec, LRU_INACTIVE_FILE) >> sc->priority) {
 		scan_balance = SCAN_FILE;
 		goto out;
@@ -2448,7 +2453,7 @@ static void shrink_node_memcg(struct pglist_data *pgdat, struct mem_cgroup *memc
 	 * Even if we did not try to evict anon pages at all, we want to
 	 * rebalance the anon lru active/inactive ratio.
 	 */
-	if (inactive_list_is_low(lruvec, false, sc))
+	if (inactive_list_is_low(lruvec, false, sc, true))
 		shrink_active_list(SWAP_CLUSTER_MAX, lruvec,
 				   sc, LRU_ACTIVE_ANON);
 }
@@ -3098,7 +3103,7 @@ static void age_active_anon(struct pglist_data *pgdat,
 	do {
 		struct lruvec *lruvec = mem_cgroup_lruvec(pgdat, memcg);
 
-		if (inactive_list_is_low(lruvec, false, sc))
+		if (inactive_list_is_low(lruvec, false, sc, true))
 			shrink_active_list(SWAP_CLUSTER_MAX, lruvec,
 					   sc, LRU_ACTIVE_ANON);
 
-- 
2.11.0

Re: [PATCH 0/7 v2] vm, vmscan: enahance vmscan tracepoints

[Michal Hocko]

On Wed 04-01-17 11:19:35, Michal Hocko wrote:
> Hi,
> this is the second version of the patchset [1]. I hope I've addressed all
> the review feedback.

I forgot to mention that this is based on the latest mmotm +
http://lkml.kernel.org/r/20161220130135.15719-1-mhocko@kernel.org which
are sitting in the mm tree but haven't been released as a mmmotm yet.

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 2/7] mm, vmscan: add active list aging tracepoint

[Vlastimil Babka]

On 01/04/2017 11:19 AM, Michal Hocko wrote:
> From: Michal Hocko <mhocko@suse.com>
> 
> Our reclaim process has several tracepoints to tell us more about how
> things are progressing. We are, however, missing a tracepoint to track
> active list aging. Introduce mm_vmscan_lru_shrink_active which reports
> the number of
> 	- nr_scanned, nr_taken pages to tell us the LRU isolation
> 	  effectiveness.

Well, this point is no longer true, is it...

> 	- nr_referenced pages which tells us that we are hitting referenced
> 	  pages which are deactivated. If this is a large part of the
> 	  reported nr_deactivated pages then we might be hitting into
> 	  the active list too early because they might be still part of
> 	  the working set. This might help to debug performance issues.
> 	- nr_activated pages which tells us how many pages are kept on the

"nr_activated" is slightly misleading? They remain active, they are not
being activated (that's why the pgactivate vmstat is also not increased
on them, right?). I guess rename to "nr_active" ? Or something like
"nr_remain_active" although that's longer.

[...]

> @@ -1857,6 +1859,7 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
>  	unsigned long pgmoved = 0;
>  	struct page *page;
>  	int nr_pages;
> +	int nr_moved = 0;
>  
>  	while (!list_empty(list)) {
>  		page = lru_to_page(list);
> @@ -1882,11 +1885,15 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
>  				spin_lock_irq(&pgdat->lru_lock);
>  			} else
>  				list_add(&page->lru, pages_to_free);
> +		} else {
> +			nr_moved += nr_pages;
>  		}
>  	}
>  
>  	if (!is_active_lru(lru))
>  		__count_vm_events(PGDEACTIVATE, pgmoved);

So we now have pgmoved and nr_moved. One is used for vmstat, other for
tracepoint, and the only difference is that vmstat includes pages where
we raced with page being unmapped from all pte's (IIUC?) and thus
removed from lru, which should be rather rare? I guess those are being
counted into vmstat only due to how the code evolved from using pagevec.
If we don't consider them in the tracepoint, then I'd suggest we don't
count them into vmstat either, and simplify this.

> +
> +	return nr_moved;
>  }
>  
>  static void shrink_active_list(unsigned long nr_to_scan,
> @@ -1902,7 +1909,8 @@ static void shrink_active_list(unsigned long nr_to_scan,
>  	LIST_HEAD(l_inactive);
>  	struct page *page;
>  	struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat;
> -	unsigned long nr_rotated = 0;
> +	unsigned nr_deactivate, nr_activate;
> +	unsigned nr_rotated = 0;
>  	isolate_mode_t isolate_mode = 0;
>  	int file = is_file_lru(lru);
>  	struct pglist_data *pgdat = lruvec_pgdat(lruvec);
> @@ -1980,13 +1988,15 @@ static void shrink_active_list(unsigned long nr_to_scan,
>  	 */
>  	reclaim_stat->recent_rotated[file] += nr_rotated;
>  
> -	move_active_pages_to_lru(lruvec, &l_active, &l_hold, lru);
> -	move_active_pages_to_lru(lruvec, &l_inactive, &l_hold, lru - LRU_ACTIVE);
> +	nr_activate = move_active_pages_to_lru(lruvec, &l_active, &l_hold, lru);
> +	nr_deactivate = move_active_pages_to_lru(lruvec, &l_inactive, &l_hold, lru - LRU_ACTIVE);
>  	__mod_node_page_state(pgdat, NR_ISOLATED_ANON + file, -nr_taken);
>  	spin_unlock_irq(&pgdat->lru_lock);
>  
>  	mem_cgroup_uncharge_list(&l_hold);
>  	free_hot_cold_page_list(&l_hold, true);
> +	trace_mm_vmscan_lru_shrink_active(pgdat->node_id, nr_taken, nr_activate,
> +			nr_deactivate, nr_rotated, sc->priority, file);
>  }
>  
>  /*
> 

Re: [PATCH 2/7] mm, vmscan: add active list aging tracepoint

[Michal Hocko]

On Wed 04-01-17 13:52:24, Vlastimil Babka wrote:
> On 01/04/2017 11:19 AM, Michal Hocko wrote:
> > From: Michal Hocko <mhocko@suse.com>
> > 
> > Our reclaim process has several tracepoints to tell us more about how
> > things are progressing. We are, however, missing a tracepoint to track
> > active list aging. Introduce mm_vmscan_lru_shrink_active which reports
> > the number of
> > 	- nr_scanned, nr_taken pages to tell us the LRU isolation
> > 	  effectiveness.
> 
> Well, this point is no longer true, is it...

ups, leftover
	- nr_take - the number of isolated pages

> > 	- nr_referenced pages which tells us that we are hitting referenced
> > 	  pages which are deactivated. If this is a large part of the
> > 	  reported nr_deactivated pages then we might be hitting into
> > 	  the active list too early because they might be still part of
> > 	  the working set. This might help to debug performance issues.
> > 	- nr_activated pages which tells us how many pages are kept on the
> 
> "nr_activated" is slightly misleading? They remain active, they are not
> being activated (that's why the pgactivate vmstat is also not increased
> on them, right?). I guess rename to "nr_active" ? Or something like
> "nr_remain_active" although that's longer.

will go with nr_active

> 
> [...]
> 
> > @@ -1857,6 +1859,7 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
> >  	unsigned long pgmoved = 0;
> >  	struct page *page;
> >  	int nr_pages;
> > +	int nr_moved = 0;
> >  
> >  	while (!list_empty(list)) {
> >  		page = lru_to_page(list);
> > @@ -1882,11 +1885,15 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
> >  				spin_lock_irq(&pgdat->lru_lock);
> >  			} else
> >  				list_add(&page->lru, pages_to_free);
> > +		} else {
> > +			nr_moved += nr_pages;
> >  		}
> >  	}
> >  
> >  	if (!is_active_lru(lru))
> >  		__count_vm_events(PGDEACTIVATE, pgmoved);
> 
> So we now have pgmoved and nr_moved. One is used for vmstat, other for
> tracepoint, and the only difference is that vmstat includes pages where
> we raced with page being unmapped from all pte's (IIUC?) and thus
> removed from lru, which should be rather rare? I guess those are being
> counted into vmstat only due to how the code evolved from using pagevec.
> If we don't consider them in the tracepoint, then I'd suggest we don't
> count them into vmstat either, and simplify this.

OK, but I would prefer to have this in a separate patch, OK?
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 2/7] mm, vmscan: add active list aging tracepoint

[Vlastimil Babka]

On 01/04/2017 02:16 PM, Michal Hocko wrote:
> On Wed 04-01-17 13:52:24, Vlastimil Babka wrote:
>> On 01/04/2017 11:19 AM, Michal Hocko wrote:
>>> From: Michal Hocko <mhocko@suse.com>
>>>
>>> Our reclaim process has several tracepoints to tell us more about how
>>> things are progressing. We are, however, missing a tracepoint to track
>>> active list aging. Introduce mm_vmscan_lru_shrink_active which reports
>>> the number of
>>> 	- nr_scanned, nr_taken pages to tell us the LRU isolation
>>> 	  effectiveness.
>>
>> Well, this point is no longer true, is it...
> 
> ups, leftover
> 	- nr_take - the number of isolated pages

nr_taken

> 
>>> 	- nr_referenced pages which tells us that we are hitting referenced
>>> 	  pages which are deactivated. If this is a large part of the
>>> 	  reported nr_deactivated pages then we might be hitting into
>>> 	  the active list too early because they might be still part of
>>> 	  the working set. This might help to debug performance issues.
>>> 	- nr_activated pages which tells us how many pages are kept on the
>>
>> "nr_activated" is slightly misleading? They remain active, they are not
>> being activated (that's why the pgactivate vmstat is also not increased
>> on them, right?). I guess rename to "nr_active" ? Or something like
>> "nr_remain_active" although that's longer.
> 
> will go with nr_active

OK.

> 
>>
>> [...]
>>
>>> @@ -1857,6 +1859,7 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
>>>  	unsigned long pgmoved = 0;
>>>  	struct page *page;
>>>  	int nr_pages;
>>> +	int nr_moved = 0;
>>>  
>>>  	while (!list_empty(list)) {
>>>  		page = lru_to_page(list);
>>> @@ -1882,11 +1885,15 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
>>>  				spin_lock_irq(&pgdat->lru_lock);
>>>  			} else
>>>  				list_add(&page->lru, pages_to_free);
>>> +		} else {
>>> +			nr_moved += nr_pages;
>>>  		}
>>>  	}
>>>  
>>>  	if (!is_active_lru(lru))
>>>  		__count_vm_events(PGDEACTIVATE, pgmoved);
>>
>> So we now have pgmoved and nr_moved. One is used for vmstat, other for
>> tracepoint, and the only difference is that vmstat includes pages where
>> we raced with page being unmapped from all pte's (IIUC?) and thus
>> removed from lru, which should be rather rare? I guess those are being
>> counted into vmstat only due to how the code evolved from using pagevec.
>> If we don't consider them in the tracepoint, then I'd suggest we don't
>> count them into vmstat either, and simplify this.
> 
> OK, but I would prefer to have this in a separate patch, OK?

Sure, thanks!

Re: [PATCH 2/7] mm, vmscan: add active list aging tracepoint

[Michal Hocko]

With fixed triggered by Vlastimil it should be like this.
---
>From b3a1480b54bf10924a9cd09c6d8b274fc81ca4ad Mon Sep 17 00:00:00 2001
From: Michal Hocko <mhocko@suse.com>
Date: Tue, 27 Dec 2016 13:18:20 +0100
Subject: [PATCH] mm, vmscan: add active list aging tracepoint

Our reclaim process has several tracepoints to tell us more about how
things are progressing. We are, however, missing a tracepoint to track
active list aging. Introduce mm_vmscan_lru_shrink_active which reports
the number of
	- nr_taken is number of isolated pages from the active list
	- nr_referenced pages which tells us that we are hitting referenced
	  pages which are deactivated. If this is a large part of the
	  reported nr_deactivated pages then we might be hitting into
	  the active list too early because they might be still part of
	  the working set. This might help to debug performance issues.
	- nr_active pages which tells us how many pages are kept on the
	  active list - mostly exec file backed pages. A high number can
	  indicate that we might be trashing on executables.

Changes since v1
- report nr_taken pages as per Minchan
- report nr_activated as per Minchan
- do not report nr_freed pages because that would add a tiny overhead to
  free_hot_cold_page_list which is a hot path
- do not report nr_unevictable because we can report this number via a
  different and more generic tracepoint in putback_lru_page
- fix move_active_pages_to_lru to report proper page count when we hit
  into large pages
- drop nr_scanned because this can be obtained from
  trace_mm_vmscan_lru_isolate as per Minchan

Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/vmscan.h | 36 ++++++++++++++++++++++++++++++++++++
 mm/vmscan.c                   | 18 ++++++++++++++----
 2 files changed, 50 insertions(+), 4 deletions(-)

diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index 39bad8921ca1..c295d8f1b67a 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -363,6 +363,42 @@ TRACE_EVENT(mm_vmscan_lru_shrink_inactive,
 		show_reclaim_flags(__entry->reclaim_flags))
 );
 
+TRACE_EVENT(mm_vmscan_lru_shrink_active,
+
+	TP_PROTO(int nid, unsigned long nr_taken,
+		unsigned long nr_active, unsigned long nr_deactivated,
+		unsigned long nr_referenced, int priority, int file),
+
+	TP_ARGS(nid, nr_taken, nr_active, nr_deactivated, nr_referenced, priority, file),
+
+	TP_STRUCT__entry(
+		__field(int, nid)
+		__field(unsigned long, nr_taken)
+		__field(unsigned long, nr_active)
+		__field(unsigned long, nr_deactivated)
+		__field(unsigned long, nr_referenced)
+		__field(int, priority)
+		__field(int, reclaim_flags)
+	),
+
+	TP_fast_assign(
+		__entry->nid = nid;
+		__entry->nr_taken = nr_taken;
+		__entry->nr_active = nr_active;
+		__entry->nr_deactivated = nr_deactivated;
+		__entry->nr_referenced = nr_referenced;
+		__entry->priority = priority;
+		__entry->reclaim_flags = trace_shrink_flags(file);
+	),
+
+	TP_printk("nid=%d nr_taken=%ld nr_active=%ld nr_deactivated=%ld nr_referenced=%ld priority=%d flags=%s",
+		__entry->nid,
+		__entry->nr_taken,
+		__entry->nr_active, __entry->nr_deactivated, __entry->nr_referenced,
+		__entry->priority,
+		show_reclaim_flags(__entry->reclaim_flags))
+);
+
 #endif /* _TRACE_VMSCAN_H */
 
 /* This part must be outside protection */
diff --git a/mm/vmscan.c b/mm/vmscan.c
index c4abf08861d2..70d1c55463c0 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1846,9 +1846,11 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec,
  *
  * The downside is that we have to touch page->_refcount against each page.
  * But we had to alter page->flags anyway.
+ *
+ * Returns the number of pages moved to the given lru.
  */
 
-static void move_active_pages_to_lru(struct lruvec *lruvec,
+static unsigned move_active_pages_to_lru(struct lruvec *lruvec,
 				     struct list_head *list,
 				     struct list_head *pages_to_free,
 				     enum lru_list lru)
@@ -1857,6 +1859,7 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
 	unsigned long pgmoved = 0;
 	struct page *page;
 	int nr_pages;
+	int nr_moved = 0;
 
 	while (!list_empty(list)) {
 		page = lru_to_page(list);
@@ -1882,11 +1885,15 @@ static void move_active_pages_to_lru(struct lruvec *lruvec,
 				spin_lock_irq(&pgdat->lru_lock);
 			} else
 				list_add(&page->lru, pages_to_free);
+		} else {
+			nr_moved += nr_pages;
 		}
 	}
 
 	if (!is_active_lru(lru))
 		__count_vm_events(PGDEACTIVATE, pgmoved);
+
+	return nr_moved;
 }
 
 static void shrink_active_list(unsigned long nr_to_scan,
@@ -1902,7 +1909,8 @@ static void shrink_active_list(unsigned long nr_to_scan,
 	LIST_HEAD(l_inactive);
 	struct page *page;
 	struct zone_reclaim_stat *reclaim_stat = &lruvec->reclaim_stat;
-	unsigned long nr_rotated = 0;
+	unsigned nr_deactivate, nr_activate;
+	unsigned nr_rotated = 0;
 	isolate_mode_t isolate_mode = 0;
 	int file = is_file_lru(lru);
 	struct pglist_data *pgdat = lruvec_pgdat(lruvec);
@@ -1980,13 +1988,15 @@ static void shrink_active_list(unsigned long nr_to_scan,
 	 */
 	reclaim_stat->recent_rotated[file] += nr_rotated;
 
-	move_active_pages_to_lru(lruvec, &l_active, &l_hold, lru);
-	move_active_pages_to_lru(lruvec, &l_inactive, &l_hold, lru - LRU_ACTIVE);
+	nr_activate = move_active_pages_to_lru(lruvec, &l_active, &l_hold, lru);
+	nr_deactivate = move_active_pages_to_lru(lruvec, &l_inactive, &l_hold, lru - LRU_ACTIVE);
 	__mod_node_page_state(pgdat, NR_ISOLATED_ANON + file, -nr_taken);
 	spin_unlock_irq(&pgdat->lru_lock);
 
 	mem_cgroup_uncharge_list(&l_hold);
 	free_hot_cold_page_list(&l_hold, true);
+	trace_mm_vmscan_lru_shrink_active(pgdat->node_id, nr_taken, nr_activate,
+			nr_deactivate, nr_rotated, sc->priority, file);
 }
 
 /*
-- 
2.11.0

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 5/7] mm, vmscan: extract shrink_page_list reclaim counters into a struct

[Vlastimil Babka]

On 01/04/2017 11:19 AM, Michal Hocko wrote:
> From: Michal Hocko <mhocko@suse.com>
> 
> shrink_page_list returns quite some counters back to its caller. Extract
> the existing 5 into struct reclaim_stat because this makes the code
> easier to follow and also allows further counters to be returned.
> 
> While we are at it, make all of them unsigned rather than unsigned long
> as we do not really need full 64b for them (we never scan more than
> SWAP_CLUSTER_MAX pages at once). This should reduce some stack space.
> 
> This patch shouldn't introduce any functional change.

[...]

> @@ -1266,11 +1270,13 @@ static unsigned long shrink_page_list(struct list_head *page_list,
>  	list_splice(&ret_pages, page_list);
>  	count_vm_events(PGACTIVATE, pgactivate);
>  
> -	*ret_nr_dirty += nr_dirty;
> -	*ret_nr_congested += nr_congested;
> -	*ret_nr_unqueued_dirty += nr_unqueued_dirty;
> -	*ret_nr_writeback += nr_writeback;
> -	*ret_nr_immediate += nr_immediate;
> +	if (stat) {
> +		stat->nr_dirty = nr_dirty;
> +		stat->nr_congested = nr_congested;
> +		stat->nr_unqueued_dirty = nr_unqueued_dirty;
> +		stat->nr_writeback = nr_writeback;
> +		stat->nr_immediate = nr_immediate;
> +	}

This change of '+=' to '=' raised my eybrows, but it seems both callers
don't care so this is indeed no functional change and potentially faster.

Re: [PATCH 5/7] mm, vmscan: extract shrink_page_list reclaim counters into a struct

[Michal Hocko]

On Wed 04-01-17 15:51:43, Vlastimil Babka wrote:
[...]
> > @@ -1266,11 +1270,13 @@ static unsigned long shrink_page_list(struct list_head *page_list,
> >  	list_splice(&ret_pages, page_list);
> >  	count_vm_events(PGACTIVATE, pgactivate);
> >  
> > -	*ret_nr_dirty += nr_dirty;
> > -	*ret_nr_congested += nr_congested;
> > -	*ret_nr_unqueued_dirty += nr_unqueued_dirty;
> > -	*ret_nr_writeback += nr_writeback;
> > -	*ret_nr_immediate += nr_immediate;
> > +	if (stat) {
> > +		stat->nr_dirty = nr_dirty;
> > +		stat->nr_congested = nr_congested;
> > +		stat->nr_unqueued_dirty = nr_unqueued_dirty;
> > +		stat->nr_writeback = nr_writeback;
> > +		stat->nr_immediate = nr_immediate;
> > +	}
> 
> This change of '+=' to '=' raised my eybrows, but it seems both callers
> don't care so this is indeed no functional change and potentially faster.

Yes, I was quite surprised as well, maybe we had a code which relied on
the aggregated numbers in the past but I didn't bother to go over git
logs to check. There is no such user anymore...
 

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 2/7] mm, vmscan: add active list aging tracepoint

[Minchan Kim]

On Wed, Jan 04, 2017 at 02:52:47PM +0100, Michal Hocko wrote:
> With fixed triggered by Vlastimil it should be like this.
> ---
> From b3a1480b54bf10924a9cd09c6d8b274fc81ca4ad Mon Sep 17 00:00:00 2001
> From: Michal Hocko <mhocko@suse.com>
> Date: Tue, 27 Dec 2016 13:18:20 +0100
> Subject: [PATCH] mm, vmscan: add active list aging tracepoint
> 
> Our reclaim process has several tracepoints to tell us more about how
> things are progressing. We are, however, missing a tracepoint to track
> active list aging. Introduce mm_vmscan_lru_shrink_active which reports
> the number of
> 	- nr_taken is number of isolated pages from the active list
> 	- nr_referenced pages which tells us that we are hitting referenced
> 	  pages which are deactivated. If this is a large part of the
> 	  reported nr_deactivated pages then we might be hitting into
> 	  the active list too early because they might be still part of
> 	  the working set. This might help to debug performance issues.
> 	- nr_active pages which tells us how many pages are kept on the
> 	  active list - mostly exec file backed pages. A high number can
> 	  indicate that we might be trashing on executables.
> 
> Changes since v1
> - report nr_taken pages as per Minchan
> - report nr_activated as per Minchan
> - do not report nr_freed pages because that would add a tiny overhead to
>   free_hot_cold_page_list which is a hot path
> - do not report nr_unevictable because we can report this number via a
>   different and more generic tracepoint in putback_lru_page
> - fix move_active_pages_to_lru to report proper page count when we hit
>   into large pages
> - drop nr_scanned because this can be obtained from
>   trace_mm_vmscan_lru_isolate as per Minchan
> 
> Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
> Acked-by: Mel Gorman <mgorman@suse.de>
> Signed-off-by: Michal Hocko <mhocko@suse.com>

Acked-by: Minchan Kim <minchan@kernel.org>

Thanks.

Re: [PATCH 4/7] mm, vmscan: show LRU name in mm_vmscan_lru_isolate tracepoint

[Minchan Kim]

On Wed, Jan 04, 2017 at 11:19:39AM +0100, Michal Hocko wrote:
> From: Michal Hocko <mhocko@suse.com>
> 
> mm_vmscan_lru_isolate currently prints only whether the LRU we isolate
> from is file or anonymous but we do not know which LRU this is.
> 
> It is useful to know whether the list is active or inactive, since we
> are using the same function to isolate pages from both of them and it's
> hard to distinguish otherwise.
> 
> Chaneges since v1
> - drop LRU_ prefix from names and use lowercase as per Vlastimil
> - move and convert show_lru_name to mmflags.h EM magic as per Vlastimil
> 
> Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
> Acked-by: Mel Gorman <mgorman@suse.de>
> Signed-off-by: Michal Hocko <mhocko@suse.com>

> ---
>  include/trace/events/mmflags.h |  8 ++++++++
>  include/trace/events/vmscan.h  | 12 ++++++------
>  mm/vmscan.c                    |  2 +-
>  3 files changed, 15 insertions(+), 7 deletions(-)
> 
> diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h
> index aa4caa6914a9..6172afa2fd82 100644
> --- a/include/trace/events/mmflags.h
> +++ b/include/trace/events/mmflags.h
> @@ -240,6 +240,13 @@ IF_HAVE_VM_SOFTDIRTY(VM_SOFTDIRTY,	"softdirty"	)		\
>  	IFDEF_ZONE_HIGHMEM(	EM (ZONE_HIGHMEM,"HighMem"))	\
>  				EMe(ZONE_MOVABLE,"Movable")
>  
> +#define LRU_NAMES		\
> +		EM (LRU_INACTIVE_ANON, "inactive_anon") \
> +		EM (LRU_ACTIVE_ANON, "active_anon") \
> +		EM (LRU_INACTIVE_FILE, "inactive_file") \
> +		EM (LRU_ACTIVE_FILE, "active_file") \
> +		EMe(LRU_UNEVICTABLE, "unevictable")
> +
>  /*
>   * First define the enums in the above macros to be exported to userspace
>   * via TRACE_DEFINE_ENUM().
> @@ -253,6 +260,7 @@ COMPACTION_STATUS
>  COMPACTION_PRIORITY
>  COMPACTION_FEEDBACK
>  ZONE_TYPE
> +LRU_NAMES
>  
>  /*
>   * Now redefine the EM() and EMe() macros to map the enums to the strings
> diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
> index 36c999f806bf..7ec59e0432c4 100644
> --- a/include/trace/events/vmscan.h
> +++ b/include/trace/events/vmscan.h
> @@ -277,9 +277,9 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
>  		unsigned long nr_skipped,
>  		unsigned long nr_taken,
>  		isolate_mode_t isolate_mode,
> -		int file),
> +		int lru),

It may break trace-vmscan-postprocess.pl. Other than that,

Acked-by: Minchan Kim <minchan@kernel.org>

Re: [PATCH 0/7 v2] vm, vmscan: enahance vmscan tracepoints

[Vlastimil Babka]

On 01/04/2017 11:19 AM, Michal Hocko wrote:
> Hi,
> this is the second version of the patchset [1]. I hope I've addressed all
> the review feedback.
> 
> While debugging [2] I've realized that there is some room for
> improvements in the tracepoints set we offer currently. I had hard times
> to make any conclusion from the existing ones. The resulting problem
> turned out to be active list aging [3] and we are missing at least two
> tracepoints to debug such a problem.
> 
> Some existing tracepoints could export more information to see _why_ the
> reclaim progress cannot be made not only _how much_ we could reclaim.
> The later could be seen quite reasonably from the vmstat counters
> already. It can be argued that we are showing too many implementation
> details in those tracepoints but I consider them way too lowlevel
> already to be usable by any kernel independent userspace. I would be
> _really_ surprised if anything but debugging tools have used them.
> 
> Any feedback is highly appreciated.

When patch-specific feedback is addressed, then for the whole series:

Acked-by: Vlastimil Babka <vbabka@suse.cz>

> [1] http://lkml.kernel.org/r/20161228153032.10821-1-mhocko@kernel.org
> [2] http://lkml.kernel.org/r/20161215225702.GA27944@boerne.fritz.box
> [3] http://lkml.kernel.org/r/20161223105157.GB23109@dhcp22.suse.cz
> 
> 

Re: [PATCH 4/7] mm, vmscan: show LRU name in mm_vmscan_lru_isolate tracepoint

[Michal Hocko]

On Thu 05-01-17 15:04:58, Minchan Kim wrote:
> On Wed, Jan 04, 2017 at 11:19:39AM +0100, Michal Hocko wrote:
> > From: Michal Hocko <mhocko@suse.com>
> > 
> > mm_vmscan_lru_isolate currently prints only whether the LRU we isolate
> > from is file or anonymous but we do not know which LRU this is.
> > 
> > It is useful to know whether the list is active or inactive, since we
> > are using the same function to isolate pages from both of them and it's
> > hard to distinguish otherwise.
> > 
> > Chaneges since v1
> > - drop LRU_ prefix from names and use lowercase as per Vlastimil
> > - move and convert show_lru_name to mmflags.h EM magic as per Vlastimil
> > 
> > Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com>
> > Acked-by: Mel Gorman <mgorman@suse.de>
> > Signed-off-by: Michal Hocko <mhocko@suse.com>
> 
> > ---
> >  include/trace/events/mmflags.h |  8 ++++++++
> >  include/trace/events/vmscan.h  | 12 ++++++------
> >  mm/vmscan.c                    |  2 +-
> >  3 files changed, 15 insertions(+), 7 deletions(-)
> > 
> > diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h
> > index aa4caa6914a9..6172afa2fd82 100644
> > --- a/include/trace/events/mmflags.h
> > +++ b/include/trace/events/mmflags.h
> > @@ -240,6 +240,13 @@ IF_HAVE_VM_SOFTDIRTY(VM_SOFTDIRTY,	"softdirty"	)		\
> >  	IFDEF_ZONE_HIGHMEM(	EM (ZONE_HIGHMEM,"HighMem"))	\
> >  				EMe(ZONE_MOVABLE,"Movable")
> >  
> > +#define LRU_NAMES		\
> > +		EM (LRU_INACTIVE_ANON, "inactive_anon") \
> > +		EM (LRU_ACTIVE_ANON, "active_anon") \
> > +		EM (LRU_INACTIVE_FILE, "inactive_file") \
> > +		EM (LRU_ACTIVE_FILE, "active_file") \
> > +		EMe(LRU_UNEVICTABLE, "unevictable")
> > +
> >  /*
> >   * First define the enums in the above macros to be exported to userspace
> >   * via TRACE_DEFINE_ENUM().
> > @@ -253,6 +260,7 @@ COMPACTION_STATUS
> >  COMPACTION_PRIORITY
> >  COMPACTION_FEEDBACK
> >  ZONE_TYPE
> > +LRU_NAMES
> >  
> >  /*
> >   * Now redefine the EM() and EMe() macros to map the enums to the strings
> > diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
> > index 36c999f806bf..7ec59e0432c4 100644
> > --- a/include/trace/events/vmscan.h
> > +++ b/include/trace/events/vmscan.h
> > @@ -277,9 +277,9 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
> >  		unsigned long nr_skipped,
> >  		unsigned long nr_taken,
> >  		isolate_mode_t isolate_mode,
> > -		int file),
> > +		int lru),
> 
> It may break trace-vmscan-postprocess.pl. Other than that,

I wasn't aware of the script. And you are right it will break it. The
following should fix it. Btw. shrink_inactive_list tracepoint changes
will to be synced as well. I do not speak perl much but the following
should just work (untested yet).
---
diff --git a/Documentation/trace/postprocess/trace-vmscan-postprocess.pl b/Documentation/trace/postprocess/trace-vmscan-postprocess.pl
index 8f961ef2b457..ba976805853a 100644
--- a/Documentation/trace/postprocess/trace-vmscan-postprocess.pl
+++ b/Documentation/trace/postprocess/trace-vmscan-postprocess.pl
@@ -112,8 +112,8 @@ my $regex_direct_end_default = 'nr_reclaimed=([0-9]*)';
 my $regex_kswapd_wake_default = 'nid=([0-9]*) order=([0-9]*)';
 my $regex_kswapd_sleep_default = 'nid=([0-9]*)';
 my $regex_wakeup_kswapd_default = 'nid=([0-9]*) zid=([0-9]*) order=([0-9]*)';
-my $regex_lru_isolate_default = 'isolate_mode=([0-9]*) order=([0-9]*) nr_requested=([0-9]*) nr_scanned=([0-9]*) nr_taken=([0-9]*) file=([0-9]*)';
-my $regex_lru_shrink_inactive_default = 'nid=([0-9]*) zid=([0-9]*) nr_scanned=([0-9]*) nr_reclaimed=([0-9]*) priority=([0-9]*) flags=([A-Z_|]*)';
+my $regex_lru_isolate_default = 'isolate_mode=([0-9]*) classzone_idx=([0-9]*) order=([0-9]*) nr_requested=([0-9]*) nr_scanned=([0-9]*) nr_skipped=([0-9]*) nr_taken=([0-9]*) lru=([a-z_]*)';
+my $regex_lru_shrink_inactive_default = 'nid=([0-9]*) nr_scanned=([0-9]*) nr_reclaimed=([0-9]*) nr_dirty=([0-9]*) nr_writeback=([0-9]*) nr_congested=([0-9]*) nr_immediate=([0-9]*) nr_activate=([0-9]*) nr_ref_keep=([0-9]*) nr_unmap_fail=([0-9]*) priority=([0-9]*) flags=([A-Z_|]*)';
 my $regex_lru_shrink_active_default = 'lru=([A-Z_]*) nr_scanned=([0-9]*) nr_rotated=([0-9]*) priority=([0-9]*)';
 my $regex_writepage_default = 'page=([0-9a-f]*) pfn=([0-9]*) flags=([A-Z_|]*)';
 
@@ -205,15 +205,15 @@ $regex_wakeup_kswapd = generate_traceevent_regex(
 $regex_lru_isolate = generate_traceevent_regex(
 			"vmscan/mm_vmscan_lru_isolate",
 			$regex_lru_isolate_default,
-			"isolate_mode", "order",
-			"nr_requested", "nr_scanned", "nr_taken",
-			"file");
+			"isolate_mode", "classzone_idx", "order",
+			"nr_requested", "nr_scanned", "nr_skipped", "nr_taken",
+			"lru");
 $regex_lru_shrink_inactive = generate_traceevent_regex(
 			"vmscan/mm_vmscan_lru_shrink_inactive",
 			$regex_lru_shrink_inactive_default,
-			"nid", "zid",
-			"nr_scanned", "nr_reclaimed", "priority",
-			"flags");
+			"nid", "nr_scanned", "nr_reclaimed", "nr_dirty", "nr_writeback",
+			"nr_congested", "nr_immediate", "nr_activate", "nr_ref_keep",
+			"nr_unmap_fail", "priority", "flags");
 $regex_lru_shrink_active = generate_traceevent_regex(
 			"vmscan/mm_vmscan_lru_shrink_active",
 			$regex_lru_shrink_active_default,
@@ -381,8 +381,8 @@ sub process_events {
 				next;
 			}
 			my $isolate_mode = $1;
-			my $nr_scanned = $4;
-			my $file = $6;
+			my $nr_scanned = $5;
+			my $file = $8;
 
 			# To closer match vmstat scanning statistics, only count isolate_both
 			# and isolate_inactive as scanning. isolate_active is rotation
@@ -391,7 +391,7 @@ sub process_events {
 			# isolate_both     == 3
 			if ($isolate_mode != 2) {
 				$perprocesspid{$process_pid}->{HIGH_NR_SCANNED} += $nr_scanned;
-				if ($file == 1) {
+				if ($file =~ /_file/) {
 					$perprocesspid{$process_pid}->{HIGH_NR_FILE_SCANNED} += $nr_scanned;
 				} else {
 					$perprocesspid{$process_pid}->{HIGH_NR_ANON_SCANNED} += $nr_scanned;
@@ -406,8 +406,8 @@ sub process_events {
 				next;
 			}
 
-			my $nr_reclaimed = $4;
-			my $flags = $6;
+			my $nr_reclaimed = $3;
+			my $flags = $12;
 			my $file = 0;
 			if ($flags =~ /RECLAIM_WB_FILE/) {
 				$file = 1;
 
> Acked-by: Minchan Kim <minchan@kernel.org>

Thanks
 

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 0/7 v2] vm, vmscan: enahance vmscan tracepoints

[Michal Hocko]

Andrew,
it seems that all the patches have been acked. One of the patches has
been refreshed and send as a reply-to original one. One script in the
Documentation directory needs to be updated but I guess this is low
priority.

Should I resubmit what I have with or you are going to pick it up from
here?
-- 
Michal Hocko
SUSE Labs

Re: [PATCH 4/7] mm, vmscan: show LRU name in mm_vmscan_lru_isolate tracepoint

[Mel Gorman]

On Thu, Jan 05, 2017 at 11:16:13AM +0100, Michal Hocko wrote:
> > > diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
> > > index 36c999f806bf..7ec59e0432c4 100644
> > > --- a/include/trace/events/vmscan.h
> > > +++ b/include/trace/events/vmscan.h
> > > @@ -277,9 +277,9 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
> > >  		unsigned long nr_skipped,
> > >  		unsigned long nr_taken,
> > >  		isolate_mode_t isolate_mode,
> > > -		int file),
> > > +		int lru),
> > 
> > It may break trace-vmscan-postprocess.pl. Other than that,
> 
> I wasn't aware of the script. And you are right it will break it. The
> following should fix it. Btw. shrink_inactive_list tracepoint changes
> will to be synced as well. I do not speak perl much but the following
> should just work (untested yet).

It's also optional to remove them. When those were first merged, it was
done to illustrate how multiple tracepoints can be used to aggregate
tracepoint information. They are better ways of gathering the same class
of information. They are of historical interest but not as fully supported
scripts that can never break.

-- 
Mel Gorman
SUSE Labs

Re: [PATCH 4/7] mm, vmscan: show LRU name in mm_vmscan_lru_isolate tracepoint

[Michal Hocko]

On Thu 05-01-17 14:56:23, Mel Gorman wrote:
> On Thu, Jan 05, 2017 at 11:16:13AM +0100, Michal Hocko wrote:
> > > > diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
> > > > index 36c999f806bf..7ec59e0432c4 100644
> > > > --- a/include/trace/events/vmscan.h
> > > > +++ b/include/trace/events/vmscan.h
> > > > @@ -277,9 +277,9 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
> > > >  		unsigned long nr_skipped,
> > > >  		unsigned long nr_taken,
> > > >  		isolate_mode_t isolate_mode,
> > > > -		int file),
> > > > +		int lru),
> > > 
> > > It may break trace-vmscan-postprocess.pl. Other than that,
> > 
> > I wasn't aware of the script. And you are right it will break it. The
> > following should fix it. Btw. shrink_inactive_list tracepoint changes
> > will to be synced as well. I do not speak perl much but the following
> > should just work (untested yet).
> 
> It's also optional to remove them. When those were first merged, it was
> done to illustrate how multiple tracepoints can be used to aggregate
> tracepoint information. They are better ways of gathering the same class
> of information. They are of historical interest but not as fully supported
> scripts that can never break.

Yeah, that was my understanding and why I didn't consider it a priority.
But it seemed like an easy thing to fix even with my anti-perl mindset.
Here is the full patch (untested)
---
>From 1b843a180a1436873aab1fe3819dfc7dbf393870 Mon Sep 17 00:00:00 2001
From: Michal Hocko <mhocko@suse.com>
Date: Thu, 5 Jan 2017 11:34:03 +0100
Subject: [PATCH] trace-vmscan-postprocess: sync with tracepoints updates

Both mm_vmscan_lru_shrink_active and mm_vmscan_lru_isolate have changed
so the script needs to be update to reflect those changes

Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 .../trace/postprocess/trace-vmscan-postprocess.pl  | 26 +++++++++++-----------
 1 file changed, 13 insertions(+), 13 deletions(-)

diff --git a/Documentation/trace/postprocess/trace-vmscan-postprocess.pl b/Documentation/trace/postprocess/trace-vmscan-postprocess.pl
index 8f961ef2b457..ba976805853a 100644
--- a/Documentation/trace/postprocess/trace-vmscan-postprocess.pl
+++ b/Documentation/trace/postprocess/trace-vmscan-postprocess.pl
@@ -112,8 +112,8 @@ my $regex_direct_end_default = 'nr_reclaimed=([0-9]*)';
 my $regex_kswapd_wake_default = 'nid=([0-9]*) order=([0-9]*)';
 my $regex_kswapd_sleep_default = 'nid=([0-9]*)';
 my $regex_wakeup_kswapd_default = 'nid=([0-9]*) zid=([0-9]*) order=([0-9]*)';
-my $regex_lru_isolate_default = 'isolate_mode=([0-9]*) order=([0-9]*) nr_requested=([0-9]*) nr_scanned=([0-9]*) nr_taken=([0-9]*) file=([0-9]*)';
-my $regex_lru_shrink_inactive_default = 'nid=([0-9]*) zid=([0-9]*) nr_scanned=([0-9]*) nr_reclaimed=([0-9]*) priority=([0-9]*) flags=([A-Z_|]*)';
+my $regex_lru_isolate_default = 'isolate_mode=([0-9]*) classzone_idx=([0-9]*) order=([0-9]*) nr_requested=([0-9]*) nr_scanned=([0-9]*) nr_skipped=([0-9]*) nr_taken=([0-9]*) lru=([a-z_]*)';
+my $regex_lru_shrink_inactive_default = 'nid=([0-9]*) nr_scanned=([0-9]*) nr_reclaimed=([0-9]*) nr_dirty=([0-9]*) nr_writeback=([0-9]*) nr_congested=([0-9]*) nr_immediate=([0-9]*) nr_activate=([0-9]*) nr_ref_keep=([0-9]*) nr_unmap_fail=([0-9]*) priority=([0-9]*) flags=([A-Z_|]*)';
 my $regex_lru_shrink_active_default = 'lru=([A-Z_]*) nr_scanned=([0-9]*) nr_rotated=([0-9]*) priority=([0-9]*)';
 my $regex_writepage_default = 'page=([0-9a-f]*) pfn=([0-9]*) flags=([A-Z_|]*)';
 
@@ -205,15 +205,15 @@ $regex_wakeup_kswapd = generate_traceevent_regex(
 $regex_lru_isolate = generate_traceevent_regex(
 			"vmscan/mm_vmscan_lru_isolate",
 			$regex_lru_isolate_default,
-			"isolate_mode", "order",
-			"nr_requested", "nr_scanned", "nr_taken",
-			"file");
+			"isolate_mode", "classzone_idx", "order",
+			"nr_requested", "nr_scanned", "nr_skipped", "nr_taken",
+			"lru");
 $regex_lru_shrink_inactive = generate_traceevent_regex(
 			"vmscan/mm_vmscan_lru_shrink_inactive",
 			$regex_lru_shrink_inactive_default,
-			"nid", "zid",
-			"nr_scanned", "nr_reclaimed", "priority",
-			"flags");
+			"nid", "nr_scanned", "nr_reclaimed", "nr_dirty", "nr_writeback",
+			"nr_congested", "nr_immediate", "nr_activate", "nr_ref_keep",
+			"nr_unmap_fail", "priority", "flags");
 $regex_lru_shrink_active = generate_traceevent_regex(
 			"vmscan/mm_vmscan_lru_shrink_active",
 			$regex_lru_shrink_active_default,
@@ -381,8 +381,8 @@ sub process_events {
 				next;
 			}
 			my $isolate_mode = $1;
-			my $nr_scanned = $4;
-			my $file = $6;
+			my $nr_scanned = $5;
+			my $file = $8;
 
 			# To closer match vmstat scanning statistics, only count isolate_both
 			# and isolate_inactive as scanning. isolate_active is rotation
@@ -391,7 +391,7 @@ sub process_events {
 			# isolate_both     == 3
 			if ($isolate_mode != 2) {
 				$perprocesspid{$process_pid}->{HIGH_NR_SCANNED} += $nr_scanned;
-				if ($file == 1) {
+				if ($file =~ /_file/) {
 					$perprocesspid{$process_pid}->{HIGH_NR_FILE_SCANNED} += $nr_scanned;
 				} else {
 					$perprocesspid{$process_pid}->{HIGH_NR_ANON_SCANNED} += $nr_scanned;
@@ -406,8 +406,8 @@ sub process_events {
 				next;
 			}
 
-			my $nr_reclaimed = $4;
-			my $flags = $6;
+			my $nr_reclaimed = $3;
+			my $flags = $12;
 			my $file = 0;
 			if ($flags =~ /RECLAIM_WB_FILE/) {
 				$file = 1;
-- 
2.11.0

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 3/7] mm, vmscan: show the number of skipped pages in mm_vmscan_lru_isolate

[Michal Hocko]

On Tue 03-01-17 18:21:48, Vlastimil Babka wrote:
> On 12/28/2016 04:30 PM, Michal Hocko wrote:
> > --- a/mm/vmscan.c
> > +++ b/mm/vmscan.c
> > @@ -1428,6 +1428,7 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
> >  	unsigned long nr_taken = 0;
> >  	unsigned long nr_zone_taken[MAX_NR_ZONES] = { 0 };
> >  	unsigned long nr_skipped[MAX_NR_ZONES] = { 0, };
> > +	unsigned long skipped = 0, total_skipped = 0;
> >  	unsigned long scan, nr_pages;
> >  	LIST_HEAD(pages_skipped);
> > 
> > @@ -1479,14 +1480,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
> >  	 */
> >  	if (!list_empty(&pages_skipped)) {
> >  		int zid;
> > -		unsigned long total_skipped = 0;
> > 
> >  		for (zid = 0; zid < MAX_NR_ZONES; zid++) {
> >  			if (!nr_skipped[zid])
> >  				continue;
> > 
> >  			__count_zid_vm_events(PGSCAN_SKIP, zid, nr_skipped[zid]);
> > -			total_skipped += nr_skipped[zid];
> > +			skipped += nr_skipped[zid];
> >  		}
> > 
> >  		/*
> > @@ -1494,13 +1494,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
> >  		 * close to unreclaimable. If the LRU list is empty, account
> >  		 * skipped pages as a full scan.
> >  		 */
> > -		scan += list_empty(src) ? total_skipped : total_skipped >> 2;
> > +		total_skipped = list_empty(src) ? skipped : skipped >> 2;
> 
> Should the tracepoint output reflect this halving heuristic or rather report
> the raw data? Or is each variant inferrable from the other?

I would rather see the raw data because you can always go and check the
_current_ implementation and calculate the heuristic.

-- 
Michal Hocko
SUSE Labs

Re: [PATCH 3/7] mm, vmscan: show the number of skipped pages in mm_vmscan_lru_isolate

[Vlastimil Babka]

On 12/28/2016 04:30 PM, Michal Hocko wrote:
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -1428,6 +1428,7 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
>  	unsigned long nr_taken = 0;
>  	unsigned long nr_zone_taken[MAX_NR_ZONES] = { 0 };
>  	unsigned long nr_skipped[MAX_NR_ZONES] = { 0, };
> +	unsigned long skipped = 0, total_skipped = 0;
>  	unsigned long scan, nr_pages;
>  	LIST_HEAD(pages_skipped);
>
> @@ -1479,14 +1480,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
>  	 */
>  	if (!list_empty(&pages_skipped)) {
>  		int zid;
> -		unsigned long total_skipped = 0;
>
>  		for (zid = 0; zid < MAX_NR_ZONES; zid++) {
>  			if (!nr_skipped[zid])
>  				continue;
>
>  			__count_zid_vm_events(PGSCAN_SKIP, zid, nr_skipped[zid]);
> -			total_skipped += nr_skipped[zid];
> +			skipped += nr_skipped[zid];
>  		}
>
>  		/*
> @@ -1494,13 +1494,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
>  		 * close to unreclaimable. If the LRU list is empty, account
>  		 * skipped pages as a full scan.
>  		 */
> -		scan += list_empty(src) ? total_skipped : total_skipped >> 2;
> +		total_skipped = list_empty(src) ? skipped : skipped >> 2;

Should the tracepoint output reflect this halving heuristic or rather report the 
raw data? Or is each variant inferrable from the other?

>
>  		list_splice(&pages_skipped, src);
>  	}
> -	*nr_scanned = scan;
> +	*nr_scanned = scan + total_skipped;
>  	trace_mm_vmscan_lru_isolate(sc->reclaim_idx, sc->order, nr_to_scan, scan,
> -				    nr_taken, mode, is_file_lru(lru));
> +				    skipped, nr_taken, mode, is_file_lru(lru));
>  	update_lru_sizes(lruvec, lru, nr_zone_taken, nr_taken);
>  	return nr_taken;
>  }
>

Re: [PATCH 3/7] mm, vmscan: show the number of skipped pages in mm_vmscan_lru_isolate

[Hillf Danton]

On Wednesday, December 28, 2016 11:30 PM Michal Hocko wrote:
> From: Michal Hocko <mhocko@suse.com>
> 
> mm_vmscan_lru_isolate shows the number of requested, scanned and taken
> pages. This is mostly OK but on 32b systems the number of scanned pages
> is quite misleading because it includes both the scanned and skipped
> pages.  Moreover the skipped part is scaled based on the number of taken
> pages. Let's report the exact numbers without any additional logic and
> add the number of skipped pages. This should make the reported data much
> more easier to interpret.
> 
> Signed-off-by: Michal Hocko <mhocko@suse.com>
> ---
Acked-by: Hillf Danton <hillf.zj@alibaba-inc.com> 

Re: [PATCH 3/7] mm, vmscan: show the number of skipped pages in mm_vmscan_lru_isolate

[Minchan Kim]

On Wed, Dec 28, 2016 at 04:30:28PM +0100, Michal Hocko wrote:
> From: Michal Hocko <mhocko@suse.com>
> 
> mm_vmscan_lru_isolate shows the number of requested, scanned and taken
> pages. This is mostly OK but on 32b systems the number of scanned pages
> is quite misleading because it includes both the scanned and skipped
> pages.  Moreover the skipped part is scaled based on the number of taken
> pages. Let's report the exact numbers without any additional logic and
> add the number of skipped pages. This should make the reported data much
> more easier to interpret.
> 
> Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Minchan Kim <minchan@kernel.org>

[PATCH 3/7] mm, vmscan: show the number of skipped pages in mm_vmscan_lru_isolate

[Michal Hocko]

From: Michal Hocko <mhocko@suse.com>

mm_vmscan_lru_isolate shows the number of requested, scanned and taken
pages. This is mostly OK but on 32b systems the number of scanned pages
is quite misleading because it includes both the scanned and skipped
pages.  Moreover the skipped part is scaled based on the number of taken
pages. Let's report the exact numbers without any additional logic and
add the number of skipped pages. This should make the reported data much
more easier to interpret.

Signed-off-by: Michal Hocko <mhocko@suse.com>
---
 include/trace/events/vmscan.h |  8 ++++++--
 mm/vmscan.c                   | 10 +++++-----
 2 files changed, 11 insertions(+), 7 deletions(-)

diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h
index d34cc0ced2be..6af4dae46db2 100644
--- a/include/trace/events/vmscan.h
+++ b/include/trace/events/vmscan.h
@@ -274,17 +274,19 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		int order,
 		unsigned long nr_requested,
 		unsigned long nr_scanned,
+		unsigned long nr_skipped,
 		unsigned long nr_taken,
 		isolate_mode_t isolate_mode,
 		int file),
 
-	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_taken, isolate_mode, file),
+	TP_ARGS(classzone_idx, order, nr_requested, nr_scanned, nr_skipped, nr_taken, isolate_mode, file),
 
 	TP_STRUCT__entry(
 		__field(int, classzone_idx)
 		__field(int, order)
 		__field(unsigned long, nr_requested)
 		__field(unsigned long, nr_scanned)
+		__field(unsigned long, nr_skipped)
 		__field(unsigned long, nr_taken)
 		__field(isolate_mode_t, isolate_mode)
 		__field(int, file)
@@ -295,17 +297,19 @@ TRACE_EVENT(mm_vmscan_lru_isolate,
 		__entry->order = order;
 		__entry->nr_requested = nr_requested;
 		__entry->nr_scanned = nr_scanned;
+		__entry->nr_skipped = nr_skipped;
 		__entry->nr_taken = nr_taken;
 		__entry->isolate_mode = isolate_mode;
 		__entry->file = file;
 	),
 
-	TP_printk("isolate_mode=%d classzone=%d order=%d nr_requested=%lu nr_scanned=%lu nr_taken=%lu file=%d",
+	TP_printk("isolate_mode=%d classzone=%d order=%d nr_requested=%lu nr_scanned=%lu nr_skipped=%lu nr_taken=%lu file=%d",
 		__entry->isolate_mode,
 		__entry->classzone_idx,
 		__entry->order,
 		__entry->nr_requested,
 		__entry->nr_scanned,
+		__entry->nr_skipped,
 		__entry->nr_taken,
 		__entry->file)
 );
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 2302a1a58c6e..4f7c0d66d629 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1428,6 +1428,7 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 	unsigned long nr_taken = 0;
 	unsigned long nr_zone_taken[MAX_NR_ZONES] = { 0 };
 	unsigned long nr_skipped[MAX_NR_ZONES] = { 0, };
+	unsigned long skipped = 0, total_skipped = 0;
 	unsigned long scan, nr_pages;
 	LIST_HEAD(pages_skipped);
 
@@ -1479,14 +1480,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 	 */
 	if (!list_empty(&pages_skipped)) {
 		int zid;
-		unsigned long total_skipped = 0;
 
 		for (zid = 0; zid < MAX_NR_ZONES; zid++) {
 			if (!nr_skipped[zid])
 				continue;
 
 			__count_zid_vm_events(PGSCAN_SKIP, zid, nr_skipped[zid]);
-			total_skipped += nr_skipped[zid];
+			skipped += nr_skipped[zid];
 		}
 
 		/*
@@ -1494,13 +1494,13 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
 		 * close to unreclaimable. If the LRU list is empty, account
 		 * skipped pages as a full scan.
 		 */
-		scan += list_empty(src) ? total_skipped : total_skipped >> 2;
+		total_skipped = list_empty(src) ? skipped : skipped >> 2;
 
 		list_splice(&pages_skipped, src);
 	}
-	*nr_scanned = scan;
+	*nr_scanned = scan + total_skipped;
 	trace_mm_vmscan_lru_isolate(sc->reclaim_idx, sc->order, nr_to_scan, scan,
-				    nr_taken, mode, is_file_lru(lru));
+				    skipped, nr_taken, mode, is_file_lru(lru));
 	update_lru_sizes(lruvec, lru, nr_zone_taken, nr_taken);
 	return nr_taken;
 }
-- 
2.10.2