[RFC PATCH 00/12] mm/damos/stat: introduce nr_snapshots, max_nr_snapshots and tracepoint

[RFC PATCH 00/12] mm/damos/stat: introduce nr_snapshots, max_nr_snapshots and tracepoint

[SeongJae Park]

Introduce three changes for improving DAMOS stat's provided information,
deterministic control, and reading usability.

DAMOS provides stats that are important for understanding its behavior.
It lacks information about how many DAMON-generated monitoring output
snapshots it has worked on.  Add a new stat, nr_snapshots, to show the
information.

Users can control DAMOS schemes in multiple ways.  Using the online
parameters commit feature, they can install and uninstall DAMOS schemes
whenever they want while keeping DAMON runs.  DAMOS quotas and
watermarks can be used for manually or automatically turning on/off or
adjusting the aggressiveness of the scheme.  DAMOS filters can be used
for applying the scheme to specific memory entities based on their types
and locations.  Some users want their DAMOS scheme to be applied to only
specific number of DAMON snapshots, for more deterministic control.  One
example use case is tracepoint based snapshot reading.  Add a new knob,
max_nr_snapshots, to support this.  If the nr_snapshots parameter
becomes same to or greater than the value of this parameter, the scheme
is deactivated.

Users can read DAMOS stats via DAMON's sysfs interface.  For deep level
investigations on environments having advanced tools like perf and
bpftrace, exposing the stats via a tracepoint can be useful.  Implement
a new tracepoint, namely damon:damos_stat_after_apply_interval.

First five patches (patches 1-5) of this series implement the new stat,
nr_snapshots, on the core layer (patch 1), expose on DAMON sysfs user
interface (patch 2), and update documents (patches 3-5).

Following six patches (patches 6-11) are for the new stat based DAMOS
deactivation (max_nr_snapshots).  The first one (patch 6) of this group
updates a kernel-doc comment before making further changes.  Then an
implementation of it on the core layer (patch 7), an introduction of a
new DAMON sysfs interface file for users of the feature (patch 8), and
three updates of the documents (patches 9-11) follow.

The final one (patch 12) introduces the new tracepoint that exposes the
DAMOS stat values for each scheme apply interval.

SeongJae Park (12):
  mm/damon/core: introduce nr_snapshots damos stat
  mm/damon/sysfs-schemes: introduce nr_snapshots damos stat file
  Docs/mm/damon/design: update for nr_snapshots damos stat
  Docs/admin-guide/mm/damon/usage: update for nr_snapshots damos stat
  Docs/ABI/damon: update for nr_snapshots damos stat
  mm/damon: update damos kerneldoc for stat field
  mm/damon/core: implement max_nr_snapshots
  mm/damon/sysfs-schemes: implement max_nr_snapshots file
  Docs/mm/damon/design: update for max_nr_snapshots
  Docs/admin-guide/mm/damon/usage: update for max_nr_snapshots
  Docs/ABI/damon: update for max_nr_snapshots
  mm/damon/core: add trace point for damos stat per apply interval

 .../ABI/testing/sysfs-kernel-mm-damon         | 13 ++++++
 Documentation/admin-guide/mm/damon/usage.rst  | 11 ++---
 Documentation/mm/damon/design.rst             |  7 +++
 include/linux/damon.h                         | 12 ++++--
 include/trace/events/damon.h                  | 41 ++++++++++++++++++
 mm/damon/core.c                               | 37 ++++++++++++++--
 mm/damon/sysfs-schemes.c                      | 43 +++++++++++++++++++
 7 files changed, 153 insertions(+), 11 deletions(-)


base-commit: 522f1a9fc5f1f0a11af39ba5bedae446322ca0a2
-- 
2.47.3

[RFC PATCH 12/12] mm/damon/core: add trace point for damos stat per apply interval

[SeongJae Park]

DAMON users can read DAMOS stats via DAMON sysfs interface.  It enables
efficient, simple and flexible usages of the stats.  Especially for
systems not having advanced tools like perf or bpftrace, that can be
useful.  But if the advanced tools are available, exposing the stats via
tracepoint can reduce unnecessary reimplementation of the wheels.  Add a
new tracepoint for DAMOS stats, namely damos_stat_after_apply_interval.
The tracepoint is triggered for each scheme's apply interval and exposes
the whole stat values.  If the user needs sub-apply interval information
for any chance, damos_before_apply tracepoint could be used.

Signed-off-by: SeongJae Park <sj@kernel.org>
---
 include/trace/events/damon.h | 41 ++++++++++++++++++++++++++++++++++++
 mm/damon/core.c              | 15 +++++++++++++
 2 files changed, 56 insertions(+)

diff --git a/include/trace/events/damon.h b/include/trace/events/damon.h
index 852d725afea2..24fc402ab3c8 100644
--- a/include/trace/events/damon.h
+++ b/include/trace/events/damon.h
@@ -9,6 +9,47 @@
 #include <linux/types.h>
 #include <linux/tracepoint.h>
 
+TRACE_EVENT(damos_stat_after_apply_interval,
+
+	TP_PROTO(unsigned int context_idx, unsigned int scheme_idx,
+		struct damos_stat *stat),
+
+	TP_ARGS(context_idx, scheme_idx, stat),
+
+	TP_STRUCT__entry(
+		__field(unsigned int, context_idx)
+		__field(unsigned int, scheme_idx)
+		__field(unsigned long, nr_tried)
+		__field(unsigned long, sz_tried)
+		__field(unsigned long, nr_applied)
+		__field(unsigned long, sz_applied)
+		__field(unsigned long, sz_ops_filter_passed)
+		__field(unsigned long, qt_exceeds)
+		__field(unsigned long, nr_snapshots)
+	),
+
+	TP_fast_assign(
+		__entry->context_idx = context_idx;
+		__entry->scheme_idx = scheme_idx;
+		__entry->nr_tried = stat->nr_tried;
+		__entry->sz_tried = stat->sz_tried;
+		__entry->nr_applied = stat->nr_applied;
+		__entry->sz_applied = stat->sz_applied;
+		__entry->sz_ops_filter_passed = stat->sz_ops_filter_passed;
+		__entry->qt_exceeds = stat->qt_exceeds;
+		__entry->nr_snapshots = stat->nr_snapshots;
+	),
+
+	TP_printk("ctx_idx=%u scheme_idx=%u nr_tried=%lu sz_tried=%lu "
+			"nr_applied=%lu sz_tried=%lu sz_ops_filter_passed=%lu "
+			"qt_exceeds=%lu nr_snapshots=%lu",
+			__entry->context_idx, __entry->scheme_idx,
+			__entry->nr_tried, __entry->sz_tried,
+			__entry->nr_applied, __entry->sz_applied,
+			__entry->sz_ops_filter_passed, __entry->qt_exceeds,
+			__entry->nr_snapshots)
+);
+
 TRACE_EVENT(damos_esz,
 
 	TP_PROTO(unsigned int context_idx, unsigned int scheme_idx,
diff --git a/mm/damon/core.c b/mm/damon/core.c
index 36313cd1ff1c..2f212a18e4a0 100644
--- a/mm/damon/core.c
+++ b/mm/damon/core.c
@@ -2256,6 +2256,19 @@ static void damos_adjust_quota(struct damon_ctx *c, struct damos *s)
 	quota->min_score = score;
 }
 
+static void damos_trace_stat(struct damon_ctx *c, struct damos *s)
+{
+	unsigned int cidx = 0, sidx = 0;
+	struct damos *siter;
+
+	damon_for_each_scheme(siter, c) {
+		if (siter == s)
+			break;
+		sidx++;
+	}
+	trace_damos_stat_after_apply_interval(cidx, sidx, &s->stat);
+}
+
 static void kdamond_apply_schemes(struct damon_ctx *c)
 {
 	struct damon_target *t;
@@ -2294,6 +2307,8 @@ static void kdamond_apply_schemes(struct damon_ctx *c)
 			(s->apply_interval_us ? s->apply_interval_us :
 			 c->attrs.aggr_interval) / sample_interval;
 		s->last_applied = NULL;
+		if (trace_damos_stat_after_apply_interval_enabled())
+			damos_trace_stat(c, s);
 	}
 	mutex_unlock(&c->walk_control_lock);
 }
-- 
2.47.3

Re: [RFC PATCH 00/12] mm/damos/stat: introduce nr_snapshots, max_nr_snapshots and tracepoint

[SeongJae Park]

On Sun, 23 Nov 2025 10:43:14 -0800 SeongJae Park <sj@kernel.org> wrote:

> Introduce three changes for improving DAMOS stat's provided information,
> deterministic control, and reading usability.

FYI, I'm not aiming to make this be merged in 6.19-rc1.


Thanks,
SJ

[...]

Re: [RFC PATCH 12/12] mm/damon/core: add trace point for damos stat per apply interval

[Steven Rostedt]

On Sun, 23 Nov 2025 10:43:26 -0800
SeongJae Park <sj@kernel.org> wrote:

> --- a/mm/damon/core.c
> +++ b/mm/damon/core.c
> @@ -2256,6 +2256,19 @@ static void damos_adjust_quota(struct damon_ctx *c, struct damos *s)
>  	quota->min_score = score;
>  }
>  
> +static void damos_trace_stat(struct damon_ctx *c, struct damos *s)
> +{
> +	unsigned int cidx = 0, sidx = 0;
> +	struct damos *siter;
> +
> +	damon_for_each_scheme(siter, c) {
> +		if (siter == s)
> +			break;
> +		sidx++;
> +	}
> +	trace_damos_stat_after_apply_interval(cidx, sidx, &s->stat);
> +}
> +
>  static void kdamond_apply_schemes(struct damon_ctx *c)
>  {
>  	struct damon_target *t;
> @@ -2294,6 +2307,8 @@ static void kdamond_apply_schemes(struct damon_ctx *c)
>  			(s->apply_interval_us ? s->apply_interval_us :
>  			 c->attrs.aggr_interval) / sample_interval;
>  		s->last_applied = NULL;
> +		if (trace_damos_stat_after_apply_interval_enabled())
> +			damos_trace_stat(c, s);
>  	}
>  	mutex_unlock(&c->walk_control_lock);
>  }


I wonder if the above would look better (and still produce good assembly)
if it was:

static inline void damos_trace_stat(struct damon_ctx *c, struct damos *s)
{
	unsigned int cidx = 0, sidx = 0;
	struct damos *siter;

	if (!trace_damos_stat_after_apply_interval_enabled())
		return;

	damon_for_each_scheme(siter, c) {
		if (siter == s)
			break;
		sidx++;
	}
	trace_damos_stat_after_apply_interval(cidx, sidx, &s->stat);
}

static void kdamond_apply_schemes(struct damon_ctx *c)
{
	struct damon_target *t;
> @@ -2294,6 +2307,8 @@ static void kdamond_apply_schemes(struct damon_ctx *c)
			(s->apply_interval_us ? s->apply_interval_us :
			 c->attrs.aggr_interval) / sample_interval;
 		s->last_applied = NULL;
		damos_trace_stat(c, s);
	}
	mutex_unlock(&c->walk_control_lock);
}


I have no real preference. I just think keeping the "if ()" statement out
of the main code as a more aesthetic look. But the above should be
equivalent in actual functionality.

-- Steve

Re: [RFC PATCH 12/12] mm/damon/core: add trace point for damos stat per apply interval

[SeongJae Park]

On Mon, 24 Nov 2025 10:15:45 -0500 Steven Rostedt <rostedt@goodmis.org> wrote:

> On Sun, 23 Nov 2025 10:43:26 -0800
> SeongJae Park <sj@kernel.org> wrote:
> 
> > --- a/mm/damon/core.c
> > +++ b/mm/damon/core.c
[...]
> I wonder if the above would look better (and still produce good assembly)
> if it was:
[...]
> I have no real preference. I just think keeping the "if ()" statement out
> of the main code as a more aesthetic look. But the above should be
> equivalent in actual functionality.

Thank you for this suggestion, Steve!  I'll update this patch as you suggested
in the next version.


Thanks,
SJ

[...]