damon_nr_accesses_mvsum() wraps damon_mvsum() with the monitoring
intervals of the context to compute the pseudo moving sum of a region's
access frequency, with a special case for when the whole aggregation
window remains. damon_mvsum() itself is already covered by
damon_test_mvsum(), but the wrapper is not.

Add a table-driven KUnit test that exercises the full-window-remaining
boundary (with both reset and not-yet-reset nr_accesses), partially
elapsed windows, and the no-window-remaining case.

Signed-off-by: Song Hu <husong@kylinos.cn>
---
Changes in v2:
- Drop the confusing 'bp=...:' annotation from each test-case comment.
- Wrap the function documentation comment to the 80-column limit.
---
 mm/damon/tests/core-kunit.h | 49 +++++++++++++++++++++++++++++++++++++
 1 file changed, 49 insertions(+)

diff --git a/mm/damon/tests/core-kunit.h b/mm/damon/tests/core-kunit.h
index 0124f83b39b8..d085c90484d3 100644
--- a/mm/damon/tests/core-kunit.h
+++ b/mm/damon/tests/core-kunit.h
@@ -683,6 +683,54 @@ static void damon_test_mvsum(struct kunit *test)
 	}
 }
 
+/*
+ * Test damon_nr_accesses_mvsum(), which wraps damon_mvsum() with the
+ * monitoring intervals of the context.  With a sample interval of 1 and an
+ * aggregation interval of 10, an aggregation window is 10 sample intervals
+ * long.  Each row below specifies the passed sample intervals, the next
+ * aggregation time in sample intervals, the current and last nr_accesses of a
+ * region, and the expected return value.
+ */
+static void damon_test_nr_accesses_mvsum(struct kunit *test)
+{
+	unsigned long input_expects[] = {
+		/* passed, next_aggr, nr_accesses, last_nr_accesses, expect */
+		0, 10, 5, 3, 3,		/* full window left, unreset */
+		0, 10, 0, 7, 7,		/* full window left, reset */
+		5, 10, 3, 10, 8,	/* half window left */
+		8, 10, 3, 10, 5,	/* 20% window left */
+		10, 10, 42, 49, 42,	/* no window left */
+	};
+	struct damon_ctx *c = damon_new_ctx();
+	struct damon_region *r;
+	int i;
+
+	if (!c)
+		kunit_skip(test, "ctx alloc fail");
+
+	r = damon_new_region(0, 4096);
+	if (!r) {
+		damon_destroy_ctx(c);
+		kunit_skip(test, "region alloc fail");
+	}
+
+	c->attrs.sample_interval = 1;
+	c->attrs.aggr_interval = 10;
+
+	for (i = 0; i < ARRAY_SIZE(input_expects); i += 5) {
+		c->passed_sample_intervals = input_expects[i];
+		c->next_aggregation_sis = input_expects[i + 1];
+		r->nr_accesses = input_expects[i + 2];
+		r->last_nr_accesses = input_expects[i + 3];
+
+		KUNIT_EXPECT_EQ(test, (unsigned int)input_expects[i + 4],
+				damon_nr_accesses_mvsum(r, c));
+	}
+
+	damon_free_region(r);
+	damon_destroy_ctx(c);
+}
+
 static void damos_test_new_filter(struct kunit *test)
 {
 	struct damos_filter *filter;
@@ -1576,6 +1624,7 @@ static struct kunit_case damon_test_cases[] = {
 	KUNIT_CASE(damon_test_update_monitoring_result),
 	KUNIT_CASE(damon_test_set_attrs),
 	KUNIT_CASE(damon_test_mvsum),
+	KUNIT_CASE(damon_test_nr_accesses_mvsum),
 	KUNIT_CASE(damos_test_new_filter),
 	KUNIT_CASE(damos_test_commit_quota_goal),
 	KUNIT_CASE(damos_test_commit_quota_goals),
-- 
2.43.0

On Mon, 13 Jul 2026 17:46:48 +0800 Song Hu <husong@kylinos.cn> wrote:

> damon_nr_accesses_mvsum() wraps damon_mvsum() with the monitoring
> intervals of the context to compute the pseudo moving sum of a region's
> access frequency, with a special case for when the whole aggregation
> window remains. damon_mvsum() itself is already covered by
> damon_test_mvsum(), but the wrapper is not.
> 
> Add a table-driven KUnit test that exercises the full-window-remaining
> boundary (with both reset and not-yet-reset nr_accesses), partially
> elapsed windows, and the no-window-remaining case.

Thank you for this patch!

> 
> Signed-off-by: Song Hu <husong@kylinos.cn>

Reviewed-by: SJ Park <sj@kernel.org>

> ---
> Changes in v2:
> - Drop the confusing 'bp=...:' annotation from each test-case comment.
> - Wrap the function documentation comment to the 80-column limit.

And the v1 is available [1] at mailing list.

Thank you for adding this changelog.  Please add links [2] to the previous
revisions too, from next time.  For other readers, v1 

This patch is applied to damon/next [3] tree.  If this patch is not added to
mm.git in short term (~1 week?), I will ask mm.git maintainer (Andrew Morton)
to pick this.  So, no action from your side is needed for now.  If it seems I
also forgot doing that or you cannot wait for my action, please feel free to
directly ask that to Andrew.

[1] https://lore.kernel.org/20260710041530.557455-1-husong@kylinos.cn
[2] https://docs.kernel.org/process/submitting-patches.html#commentary
[3] https://origin.kernel.org/doc/html/latest/mm/damon/maintainer-profile.html#scm-trees


Thanks,
SJ

[...]