From: sashiko-bot@kernel.org
To: "SJ Park" <sj@kernel.org>
Cc: damon@lists.linux.dev
Subject: Re: [RFC PATCH v1.2 2/4] mm/damon/core: handle uninitialized damos_quota_goal->last_psi_total
Date: Fri, 28 Aug 2026 01:55:23 +0000 [thread overview]
Message-ID: <20260828015524.5F0AA1F000E9@smtp.kernel.org> (raw)
In-Reply-To: <20260828014659.121058-3-sj@kernel.org>
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Integer underflow when initializing last_psi_total causes severe quota throttling instead of maintaining the quota.
--
commit 216061cc4d3a0304d1961ab1798b65cbf748aa6c
Author: SJ Park <sj@kernel.org>
mm/damon/core: handle uninitialized damos_quota_goal->last_psi_total
This commit handles an uninitialized last_psi_total for PSI metric DAMOS
quota goals by initializing it to U64_MAX. During the first tuning round,
it detects this state and attempts to set last_psi_total such that the
first iteration has no unexpected effect on the quota.
> diff --git a/mm/damon/core.c b/mm/damon/core.c
> index 644daf5a16560..b0c8e22ab5404 100644
> --- a/mm/damon/core.c
> +++ b/mm/damon/core.c
[ ... ]
> @@ -3018,6 +3022,10 @@ static void damos_set_quota_goal_current_value(struct damon_ctx *c,
> break;
> case DAMOS_QUOTA_SOME_MEM_PSI_US:
> now_psi_total = damos_get_some_mem_psi_total();
> + /* uninitialized last_psi_total; make no effect this round */
> + if (goal->last_psi_total == U64_MAX)
> + goal->last_psi_total = now_psi_total +
> + goal->target_value;
> goal->current_value = now_psi_total - goal->last_psi_total;
[Severity: High]
Does this arithmetic result in an integer underflow on the first tuning
round?
If goal->last_psi_total is set to (now_psi_total + goal->target_value), the
calculation on the very next line becomes:
goal->current_value = now_psi_total - (now_psi_total + goal->target_value)
This evaluates to -goal->target_value. Since current_value is an unsigned
long, this underflows to an astronomically large value. Will the feedback
loop interpret this as a massive over-achievement and drastically throttle
the quota to its minimum, causing the exact opposite of the intended
"no effect this round" behavior?
> goal->last_psi_total = now_psi_total;
> break;
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260828014659.121058-1-sj@kernel.org?part=2
next prev parent reply other threads:[~2026-08-28 1:55 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-08-28 1:46 [RFC PATCH v1.2 0/4] mm/damon: fix misc bugs in kunit, quota goals and sysfs refresh_ms SJ Park
2026-08-28 1:46 ` [RFC PATCH v1.2 1/4] mm/damon/tests/core-kunit: test committing psi goal to psi goal SJ Park
2026-08-28 1:54 ` sashiko-bot
2026-08-28 1:46 ` [RFC PATCH v1.2 2/4] mm/damon/core: handle uninitialized damos_quota_goal->last_psi_total SJ Park
2026-08-28 1:55 ` sashiko-bot [this message]
2026-08-28 2:22 ` SJ Park
2026-08-28 1:46 ` [RFC PATCH v1.2 3/4] mm/damon/core: copy nid for eligible_mem_bp damos quota goal commit SJ Park
2026-08-28 1:56 ` sashiko-bot
2026-08-28 2:23 ` SJ Park
2026-08-28 1:46 ` [RFC PATCH v1.2 4/4] mm/damon/sysfs: set next refresh jiffies per sysfs context SJ Park
2026-08-28 1:57 ` sashiko-bot
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=20260828015524.5F0AA1F000E9@smtp.kernel.org \
--to=sashiko-bot@kernel.org \
--cc=damon@lists.linux.dev \
--cc=sashiko-reviews@lists.linux.dev \
--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