Re: [PATCH 4/6] Docs/mm/damon/design: update for sz_ops_filter_passed_last

From: SJ Park

Date: Mon Nov 16 2026 - 21:26:18 EST


On Mon, 16 Nov 2026 19:10:07 +0000 Mattathias Pinchover via B4 Relay <devnull+mattathias.pinchover.uber.com@xxxxxxxxxx> wrote:

> From: Mattathias Pinchover <mattathias.pinchover@xxxxxxxx>
>
> Update DAMON design document for the newly added DAMOS stat,
> sz_ops_filter_passed_last. Also explain that, unlike other
> stats, it is not accumulated from the beginning of the scheme's
> execution but shows the value of the last apply interval.
>
> Assisted-by: Claude:claude-opus-5-5
> Signed-off-by: Mattathias Pinchover <mattathias.pinchover@xxxxxxxx>
> ---
> Documentation/mm/damon/design.rst | 9 +++++++++
> 1 file changed, 9 insertions(+)
>
> diff --git a/Documentation/mm/damon/design.rst b/Documentation/mm/damon/design.rst
> index b676be697158..e40b8c1a3461 100644
> --- a/Documentation/mm/damon/design.rst
> +++ b/Documentation/mm/damon/design.rst
> @@ -876,6 +876,8 @@ scheme's execution.
> - ``sz_tried``: Total size of regions that the scheme is tried to be applied.
> - ``sz_ops_filter_passed``: Total bytes that passed operations set
> layer-handled DAMOS filters.
> +- ``sz_ops_filter_passed_last``: Bytes that passed operations set
> + layer-handled DAMOS filters in the last ``apply_interval``.
> - ``nr_applied``: Total number of regions that the scheme is applied.
> - ``sz_applied``: Total size of regions that the scheme is applied.
> - ``qt_exceeds``: Total number of times the quota of the scheme has exceeded.
> @@ -903,6 +905,13 @@ to exclude anonymous pages and the region has only anonymous pages, or if the
> action is ``pageout`` while all pages of the region are unreclaimable, applying
> the action to the region will fail.
>
> +Unlike other stats, ``sz_ops_filter_passed_last`` is not
> +accumulated from the beginning of the scheme's execution. It is updated at the
> +end of each :ref:`apply_interval <damon_design_damos>` with the amount of
> +``sz_ops_filter_passed`` increase in the interval. Hence it shows the size of
> +memory that passed the operations set layer-handled filters in the last
> +completed ``apply_interval``.
> +

I think the description is clear enough. Could we drop this? I concern if we
will have more non-accumulated stats but forget updating this in future. We
could add this again if someone really get confused about this in future.

Other than that,

Reviewed-by: SJ Park <sj@xxxxxxxxxx>


Thanks,
SJ

[...]