mm/damon/core: error damos_commit_quota_goal() for zero target_value

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] mm/damon: incomplete zero target_value check in damos_commit_quota_goal() --

commit ea716ea29282ef733822096401c5c8d92eaaef58
Author: SJ Park <[email protected]>

mm/damon/core: error damos_commit_quota_goal() for zero target_value

This commit validates that a DAMOS quota goal's target_value is non-zero in damos_commit_quota_goal(), returning -EINVAL and propagating the error through damos_commit_quota_goals() to prevent a division-by-zero in damos_quota_score().

Link https://lore.kernel.org/[email protected] open here

Link https://lore.kernel.org/[email protected] open here

Link https://lore.kernel.org/[email protected] [1] open here

Quoted 5 lines

mm/damon/core.c +0 -0

index 4e6e3d7d0a5c0..8fa788b0d709c 100644

mm/damon/core.c static void damos_commit_quota_goal_union( · line 814

Quoted 9 lines

mm/damon/core.c +3 -1

@@ -807,15 +807,18 @@ static void damos_commit_quota_goal_union(
 	}
 }
 
-static void damos_commit_quota_goal(
+static int damos_commit_quota_goal(
 		struct damos_quota_goal *dst, struct damos_quota_goal *src)
 {
+	if (!src->target_value)
+		return  -EINVAL;

[Severity: High]
Does checking for a zero target_value only in damos_commit_quota_goal() cover the path where DAMON is started via sysfs in this stable tree?

In mm/damon/sysfs-schemes.c, target_value_store() updates goal->target_value without holding damon_sysfs_lock:

mm/damon/sysfs-schemes.c:target_value_store() {
	struct damos_sysfs_quota_goal *goal = container_of(kobj, struct
			damos_sysfs_quota_goal, kobj);
	int err = kstrtoul(buf, 0, &goal->target_value);

	return err ? err : count;
}

Meanwhile, damos_sysfs_add_quota_score() reads sysfs_goal->target_value twice without synchronization against target_value_store():

mm/damon/sysfs-schemes.c:damos_sysfs_add_quota_score() {
		if (!sysfs_goal->target_value)
			continue;

		goal = damos_new_quota_goal(sysfs_goal->metric,
				sysfs_goal->target_value);
		...
		damos_add_quota_goal(quota, goal);
}

If target_value_store() writes 0 between those two reads, damos_new_quota_goal() allocates a goal with target_value == 0 and attaches it to quota->goals.

When starting DAMON via state_store() -> damon_sysfs_cmd_request_callback() -> damon_sysfs_handle_cmd() -> damon_sysfs_turn_damon_on(), this branch builds the context and starts kdamond directly without calling damon_commit_ctx() or damos_commit_quota_goals():

mm/damon/sysfs.c:damon_sysfs_turn_damon_on() {
	ctx = damon_sysfs_build_ctx(kdamond->contexts->contexts_arr[0]);
	if (IS_ERR(ctx)) {
		kfree(repeat_call_control);
		return PTR_ERR(ctx);
	}
	err = damon_start(&ctx, 1, false);
}

Here damon_sysfs_build_ctx() populates scheme->quota.goals via damon_sysfs_apply_inputs() -> damon_sysfs_add_schemes() -> damon_sysfs_mk_scheme() -> damos_sysfs_add_quota_score(), bypassing damos_commit_quota_goal() entirely.

When kdamond subsequently runs kdamond_fn() -> kdamond_apply_schemes() -> damos_adjust_quota() -> damos_set_effective_quota() -> damos_quota_score():

mm/damon/core.c:damos_quota_score() {
		highest_score = max(highest_score,
				goal->current_value * 10000 /
				goal->target_value);
}

can this still divide by zero if the race occurs during damon_sysfs_turn_damon_on()?

Signature · 2 lines

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1

Patch series "mm/damon: move zero damos quota target_value handling to the core layer".

Having zero DAMOS quota target value can cause division by zero. DAMON API callers are checking the target value parameters to avoid that. It is easy to make mistakes in some of the multiple API callers. Move that to the core layer.

Patch 1 adds the corner case handling into the core layer DAMON parameters validation logic. Patches 2 and 3 remove no more needed DAMON API callers side handling of the corner case in DAMON_LRU_SORT and DMON_SAMPLIE_MTIER, respectively.

This patch (of 3):

If a DAMOS scheme has a damos_quota_goal of zero target_value, damos_quota_goal() could trigger division-by-zero error. Hence each DAMON API callers should do the zero target_value validation. It is easy to make mistakes. Actually such bugs in DAMON_LRU_SORT and DAMON_SAMPLE_MTIER were found and fixed [1].

It is better to handle the corner case only once in the core layer, instead of multiple places in all DAMON API callers. One straightforward option is using an alternative denominator for the corner case in the damos_quota_goal(). However, the zero target_value is meaningless. In this case, the quota goal is always evaluated as achieved or over-achieved. The quota will only keep being reduced.

Simply avoid using zero target_value by adding a check in the core layer DAMOS quota goal parameters validation/commit path, damos_commit_quota_goal(). Update it to return an error in the case. Also update its caller to propagate the error.

Notes for stable kernel porting. The upstream commit doesn't have
Fixes: and therefore not ported to stable kernels. It turned out it fixes commit 8b549a4fd3c5 ("mm/damon/sysfs-schemes: commit damos quota goals user input to DAMOS") which was merged into v6.8. The user impact is a race condition that can result in divide-by-zero. Hence backport. Reproducing the issue requires sysfs write permission, certain DAMON ABI usages that are supposed to be uncommon, and a rare race timing, though. The issue was discovered [2] by Sashiko as a part [3] of its bug report.

[2] https://sashiko.dev/#/bug/linux-7f814c9a-986f-4103-a55a-ef9537dcf62b
[3] https://lore.kernel.org/[email protected]