From: Liew Rui Yan <aethernet65535@gmail.com>
To: SJ Park <sj@kernel.org>
Cc: Liew Rui Yan <aethernet65535@gmail.com>,
damon@lists.linux.dev, linux-mm@kvack.org
Subject: Re: [RFC] mm/damon/damo: saving damon sysfs configuration
Date: Wed, 30 Sep 2026 17:58:54 +0800 [thread overview]
Message-ID: <20260930100031.7338-1-aethernet65535@gmail.com> (raw)
In-Reply-To: <20260929074300.40012-1-sj@kernel.org>
On Tue, 29 Sep 2026 00:43:00 -0700 SJ Park <sj@kernel.org> wrote:
> Hi Liew,
>
> On Tue, 29 Sep 2026 03:13:06 +0800 Liew Rui Yan <aethernet65535@gmail.com> wrote:
>
> > Hi SJ and Community,
> >
> > I've noticed that damo [1] doesn't seem to have a feature for saving
> > DAMON SYSFS configuration. I tried 'damo args damon', but it doesn't
> > save the DAMOS schemes information.
>
> If so, it is a bug. Could you share more details including how the issue can
> be reproduced? Directly sending a PR should also be fine.
Yes, this can be reproduced easily.
# make sure DAMON is started
$ sudo ./damo args damon > damo_args_damon.json
$ sudo ./damo report damon --json > damo_report_damon.json
$ sudo ./damo stop
$ sudo ./damo start --kdamonds damo_args_damon.json
$ sudo ./damo report damon --json > damo_report_damon_after.json
$ diff damo_report_damon.json damo_report_damon_after.json
4c4
< "pid": "4959",
---
> "pid": "5582",
16,17c16,17
< "start": "4,294,967,296",
< "end": "15,807,283,199"
---
> "start": "4,096",
> "end": "15,807,283,200"
25c25
< "ops_update_us": "1 m",
---
> "ops_update_us": "1 s",
46,159c46
< "schemes": [
< {
< "action": "pageout",
< "dests": [],
< "access_pattern": {
< "sz_bytes": {
< "min": "4.000 KiB",
< "max": "max"
< },
< "nr_accesses": {
< "min": "0 samples",
< "max": "0 samples"
< },
< "age": {
< "min": "60,000 aggr_intervals",
< "max": "18,446,744,073,709,551,615 aggr_intervals"
< }
< },
< "apply_interval_us": "0 ns",
< "quotas": {
< "time_ms": "0 ns",
< "sz_bytes": "16.000 MiB",
< "reset_interval_ms": "1 s",
< "fail_charge_num": "0",
< "fail_charge_denom": "0",
< "goals": [
< {
< "metric": "user_input",
< "nid": "0",
< "memcg_path": "",
< "target_value": "3,000",
< "current_value": "7,855"
< }
< ],
< "goal_tuner": "temporal\n",
< "effective_sz_bytes": "0 B",
< "weights": {
< "sz_permil": "0 %",
< "nr_accesses_permil": "0 %",
< "age_permil": "0 %"
< }
< },
< "watermarks": {
< "metric": "none",
< "interval_us": "0 ns",
< "high_permil": "0 %",
< "mid_permil": "0 %",
< "low_permil": "0 %"
< },
< "filters": [],
< "stats": {
< "nr_tried": "0",
< "sz_tried": "0 B",
< "nr_applied": "0",
< "sz_applied": "0 B",
< "sz_ops_filter_passed": "0 B",
< "qt_exceeds": "19",
< "nr_snapshots": "0",
< "max_nr_snapshots": "0"
< }
< },
< {
< "action": "stat",
< "dests": [],
< "access_pattern": {
< "sz_bytes": {
< "min": "0 B",
< "max": "max"
< },
< "nr_accesses": {
< "min": "0 samples",
< "max": "18,446,744,073,709,551,615 samples"
< },
< "age": {
< "min": "0 aggr_intervals",
< "max": "18,446,744,073,709,551,615 aggr_intervals"
< }
< },
< "apply_interval_us": "0 ns",
< "quotas": {
< "time_ms": "0 ns",
< "sz_bytes": "0 B",
< "reset_interval_ms": "0 ns",
< "fail_charge_num": "0",
< "fail_charge_denom": "0",
< "goals": [],
< "goal_tuner": "consist\n",
< "effective_sz_bytes": "0 B",
< "weights": {
< "sz_permil": "0 %",
< "nr_accesses_permil": "0 %",
< "age_permil": "0 %"
< }
< },
< "watermarks": {
< "metric": "none",
< "interval_us": "0 ns",
< "high_permil": "0 %",
< "mid_permil": "0 %",
< "low_permil": "0 %"
< },
< "filters": [],
< "stats": {
< "nr_tried": "2,000",
< "sz_tried": "1.801 TiB",
< "nr_applied": "0",
< "sz_applied": "0 B",
< "sz_ops_filter_passed": "0 B",
< "qt_exceeds": "0",
< "nr_snapshots": "172",
< "max_nr_snapshots": "0"
< }
< }
< ]
---
> "schemes": []
>
> > 'damo report damon --json' provides
> > more detailed information, but its output cannot be read by 'damo start
> > --kdamonds'.
>
> Maybe this is also an issue that we could fix.
In my opinion, this is not an issue, since there's no documentation
recommending we use the 'damo report damon --json' to generate
configuration for 'damo start --kdamonds'.
>
> >
> > I'd like to know if there is currently any way in damo to save the DAMON
> > SYSFS configuration that can be correctly read by 'damo start
> > --kdamonds'. If not, would it be worth adding one?
>
> I believe it would be useful. Please feel free to send a PR.
>
> Nonetheless, we added the json/yaml kdamonds config support mainly because
> there was no good way to flexibly configure kdamonds with the command line
> interface. Now it is much more improved [1].
Thank you for clarifying this!
>
> I'd recommend using the command line interface. That is, you could copy your
> command line setup in a bash script and modify/reuse it.
>
> Is there a use case that it cannot work or suboptimum?
No, CLI (Command Line Interface) do very convenience. I can write my
configuration in CLI and it is very readable. Thank you for letting me
know this feature!
#!/bin/bash
# update-damon-config.sh
./damo args damon \
--kdamond --damon_ctx --ops paddr \
--monitoring_intervals 5000 100000 1000000 \
--damon_target --regions 4294967296-15807283199 \
--damos_scheme --damos_action pageout \
--damos_sz_region 4096 18446744073709551615 \
--damos_access_rate 0 0 \
--damos_age 600000000 18446744073709551615 \
--damos_quota_goal user_input 3000 10000 \
--damos_quota_goal_tuner temporal \
--damos_quota_interval 1000 \
--damos_quota_time 0 \
--damos_quota_space 16777216 \
--damos_quota_weights 0 0 0 \
> config.json
>
> [1] https://github.com/damonitor/damo/issues/3#issuecomment-5655965736
[...]
Best regards,
Rui Yan
next prev parent reply other threads:[~2026-09-30 10:00 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-28 19:13 [RFC] mm/damon/damo: saving damon sysfs configuration Liew Rui Yan
2026-09-29 7:43 ` SJ Park
2026-09-30 9:58 ` Liew Rui Yan [this message]
2026-10-08 12:09 ` Liew Rui Yan
2026-10-08 12:32 ` SJ Park
2026-10-08 13:44 ` Liew Rui Yan
2026-10-08 15:00 ` SJ Park
2026-10-08 16:14 ` Liew Rui Yan
2026-10-09 8:01 ` SJ Park
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260930100031.7338-1-aethernet65535@gmail.com \
--to=aethernet65535@gmail.com \
--cc=damon@lists.linux.dev \
--cc=linux-mm@kvack.org \
--cc=sj@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox