snapshot_set_swap_area() unpins the previously selected swap device and
Subsystems (5)
SJ Park
maintainerv4 cover snapshot_set_swap_area() unpins the previously selected swap device and
Add repin_hibernation_swap_type(), which looks up the new device, clears the old SWP_HIBERNATION flag and sets the new one under a single swap_lock acquisition. The same-device case is short- circuited so userspace can re-select the same swap area without tripping WARN_ON_ONCE and -EBUSY. Switch snapshot_set_swap_area() to the new helper.
A failed lookup now preserves the previous pin instead of dropping it, so a bad SNAPSHOT_SET_SWAP_AREA leaves the prior selection intact. The open and release paths keep using pin_hibernation_swap_type() and unpin_hibernation_swap_type().
The race was identified during AI-assisted review of the SWP_HIBERNATION pinning series.
SJ Park (7):
mm/damon/core: introduce damos_quota_goal->complement mm/damon/core: add complement argument to damos_new_quota_goal() mm/damon/sysfs-schemes: support quota goal complement flag mm/damon/tests/core-kunit: test quota_goal->complement commit Docs/mm/damon/design: document damos quota goal complement flag Docs/admin-guide/mm/damon/usage: update for quota goal complement file Docs/ABI/damon: update for quota goal metric complement sysfs file
.../ABI/testing/sysfs-kernel-mm-damon | 6 ++++ Documentation/admin-guide/mm/damon/usage.rst | 15 +++++---- Documentation/mm/damon/design.rst | 15 ++++++--- include/linux/damon.h | 4 ++- mm/damon/core.c | 33 ++++++++++++++++--- mm/damon/lru_sort.c | 5 +-- mm/damon/reclaim.c | 5 +-- mm/damon/sysfs-schemes.c | 30 +++++++++++++++++ mm/damon/tests/core-kunit.h | 3 ++ samples/damon/mtier.c | 2 +- 10 files changed, 97 insertions(+), 21 deletions(-)
base-commit: 1c2b8d2725f84b43fabe3b3e9628c91db8ca6c65
Signature · 1 line
2.47.3
SJ Park
maintainerRFC PATCH v4 1/7 mm/damon/core: introduce damos_quota_goal->complement
Introduce damos_quota_goal->complement for specifying whether to use a complemented value of the given goal target metric. Add the field to the data structure and implement essential core support. Handle the flag in the quota goal commit and current quota goal metric value retrieval.
--- include/linux/damon.h | 2 ++ mm/damon/core.c | 26 +++++++++++++++++++++++++- 2 files changed, 27 insertions(+), 1 deletion(-)
include/linux/damon.h +2 -0
@@ -175,6 +175,7 @@ enum damos_quota_goal_metric {/** * struct damos_quota_goal - DAMOS scheme quota auto-tuning goal. * @metric: Metric to be used for representing the goal. +* @complement: Use the complement of the metric. * @target_value: Target value of @metric to achieve with the tuning. * @current_value: Current value of @metric. * @nid: Node id.
@@ -197,6 +198,7 @@ enum damos_quota_goal_metric {*/ struct damos_quota_goal {enum damos_quota_goal_metric metric; +bool complement;unsigned long target_value;unsigned long current_value;/* metric-dependent fields */
mm/damon/core.c +25 -1
@@ -1224,6 +1224,7 @@ static int damos_commit_quota_goal(if (!src->target_value)return -EINVAL ; dst->metric = src->metric; + dst->complement = src->complement; dst->target_value = src->target_value;if (dst->metric ==DAMOS_QUOTA_USER_INPUT ) dst->current_value = src->current_value;
@@ -2960,10 +2961,18 @@ static void damos_set_psi_current_val(u64 now_psi_total, struct damos_quota_goal *goal, struct damos *s) { u64 last_psi_total = goal->last_psi_total; +unsigned long val; goal->last_psi_total = now_psi_total;if (last_psi_total !=U64_MAX ) { - goal->current_value = now_psi_total - last_psi_total; + val = now_psi_total - last_psi_total; +if (goal->complement) { +if (val < s->quota.reset_interval *1000 ) + val = s->quota.reset_interval *1000 - val; +else + val =0 ; + } + goal->current_value = val;return ; }/* uninitialized last_psi_total; make no effect this round */
@@ -3255,6 +3264,21 @@ static void damos_set_quota_goal_current_value(struct damon_ctx *c,default :break ; } +if (!goal->complement) +return ; + +/* updte current_value to complemented value */ + +/* for user_input, users set complemented value on their own */ +if (goal->metric ==DAMOS_QUOTA_USER_INPUT ) +return ; +/* damos_set_psi_current_val() handles complement flag itself */ +if (goal->metric ==DAMOS_QUOTA_SOME_MEM_PSI_US ) +return ; +if (goal->current_value <10000 ) + goal->current_value =10000 - goal->current_value; +else + goal->current_value =0 ; }/* Return the highest score since it makes schemes least aggressive */
Signature · 1 line
2.47.3
[email protected]
AI reviewRe: [RFC PATCH v4 1/7] mm/damon/core: introduce damos_quota_goal->complement
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Medium] mm: integer overflow in damos_set_psi_current_val on 32-bit --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] mm: integer overflow in damos_set_psi_current_val on 32-bit
--
commit 4f53a2ca54f02309422119354f04050f15950222
Author: SJ Park <[email protected]>
mm/damon/core: introduce damos_quota_goal->complement
Introduce damos_quota_goal->complement for specifying whether to use a complemented value of the given goal target metric. Add the field to the data structure and implement essential core support. Handle the flag in the quota goal commit and current quota goal metric value retrieval.
Quoted 4 lines
mm/damon/core.c +0
-0
[ ... ]
mm/damon/core.c static void damos_set_psi_current_val(u64 now_psi_total, · line 2971
Quoted 13 lines
mm/damon/core.c +5
-1@@ -2960,10 +2961,18 @@ static void damos_set_psi_current_val(u64 now_psi_total, struct damos_quota_goal *goal, struct damos *s) { u64 last_psi_total = goal->last_psi_total; +unsigned long val; goal->last_psi_total = now_psi_total;if (last_psi_total !=U64_MAX ) { - goal->current_value = now_psi_total - last_psi_total; + val = now_psi_total - last_psi_total; +if (goal->complement) { +if (val < s->quota.reset_interval *1000 ) + val = s->quota.reset_interval *1000 - val;
[Severity: Medium]
Will this calculation overflow on 32-bit platforms?
When reset_interval is configured to a value > 4,294,967 ms (around 1.19 hours), s->quota.reset_interval * 1000 will silently overflow when evaluated as a 32-bit unsigned long multiplication.
If this wraps around, could it corrupt the complemented PSI value and lead to faulty auto-tuning feedback? Does this need a cast or different typing to avoid the overflow?
Quoted 7 lines
mm/damon/core.c +4
-0+else + val =0 ; + } + goal->current_value = val;return ; }/* uninitialized last_psi_total; make no effect this round */
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
SJ Park
maintainerRe: [RFC PATCH v4 1/7] mm/damon/core: introduce damos_quota_goal->complement
Quoted 37 lines · On Sun, 27 Sep 2026 12:17:14 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] mm: integer overflow indamos_set_psi_current_valon 32-bit
--commit
4f53a2ca54f02309422119354f04050f15950222
Author: SJ Park <[email protected]>
mm/damon/core: introducedamos_quota_goal->complementIntroduce
damos_quota_goal->complementfor specifying whether to use a complemented value of the given goal target metric. Add the field to the data structure and implement essential core support. Handle the flag in the quota goal commit and current quota goal metric value retrieval.Quoted 4 lines
mm/damon/core.c +0
-0[ ... ]
mm/damon/core.c static void damos_set_psi_current_val(u64 now_psi_total, · line 2971
Quoted 13 lines
mm/damon/core.c +5
-1@@ -2960,10 +2961,18 @@ static void damos_set_psi_current_val(u64 now_psi_total, struct damos_quota_goal *goal, struct damos *s) { u64 last_psi_total = goal->last_psi_total; +unsigned long val; goal->last_psi_total = now_psi_total;if (last_psi_total !=U64_MAX ) { - goal->current_value = now_psi_total - last_psi_total; + val = now_psi_total - last_psi_total; +if (goal->complement) { +if (val < s->quota.reset_interval *1000 ) + val = s->quota.reset_interval *1000 - val;[Severity: Medium]
Will this calculation overflow on 32-bit platforms?
Quoted 7 lines
When
reset_intervalis configured to a value > 4,294,967 ms (around 1.19 hours),s->quota.reset_interval* 1000 will silently overflow when evaluated as a 32-bit unsigned long multiplication.If this wraps around, could it corrupt the complemented PSI value and lead to faulty auto-tuning feedback? Does this need a cast or different typing to avoid the overflow?
No blocker. Sane users ain't set such large reset interval with complment flag without a test.
Quoted 11 lines
Quoted 8 lines
mm/damon/core.c +4
-0+else + val =0 ; + } + goal->current_value = val;return ; }/* uninitialized last_psi_total; make no effect this round */ Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
Thanks,
SJ
SJ Park
maintainerRFC PATCH v4 2/7 mm/damon/core: add complement argument to damos_new_quota_goal()
In reply to SJ Parkdamos_quota_goal->complement needs to be manually set by each API callers. It is easy to make mistakes. Extend the quota goal constructor, damos_new_quota_goal() to receive and set the complement flag value. Also update all callers to use the new signature.
--- include/linux/damon.h | 2 +- mm/damon/core.c | 7 ++++--- mm/damon/lru_sort.c | 5 +++-- mm/damon/reclaim.c | 5 +++-- mm/damon/sysfs-schemes.c | 2 +- mm/damon/tests/core-kunit.h | 1 + samples/damon/mtier.c | 2 +- 7 files changed, 14 insertions(+), 10 deletions(-)
include/linux/damon.h +1 -1
@@ -1098,7 +1098,7 @@ bool damos_filter_for_ops(enum damos_filter_type type);void damos_destroy_filter (struct damos_filter *f);struct damos_quota_goal *damos_new_quota_goal ( -enum damos_quota_goal_metric metric, +enum damos_quota_goal_metric metric,bool complement,unsigned long target_value);void damos_add_quota_goal (struct damos_quota *q,struct damos_quota_goal *g);void damos_destroy_quota_goal (struct damos_quota_goal *goal);
mm/damon/core.c +4 -3
@@ -691,7 +691,7 @@ void damos_destroy_filter(struct damos_filter *f) }struct damos_quota_goal *damos_new_quota_goal ( -enum damos_quota_goal_metric metric, +enum damos_quota_goal_metric metric,bool complement,unsigned long target_value) {struct damos_quota_goal *goal;
@@ -700,6 +700,7 @@ struct damos_quota_goal *damos_new_quota_goal(if (!goal)return NULL ; goal->metric = metric; + goal->complement = complement; goal->target_value = target_value;if (metric ==DAMOS_QUOTA_SOME_MEM_PSI_US ) goal->last_psi_total =U64_MAX ;
@@ -1262,8 +1263,8 @@ int damos_commit_quota_goals(struct damos_quota *dst, struct damos_quota *src)damos_for_each_quota_goal_safe (src_goal, next, src) {if (j++ < i)continue ; - new_goal =damos_new_quota_goal ( - src_goal->metric, src_goal->target_value); + new_goal =damos_new_quota_goal (src_goal->metric, + src_goal->complement, src_goal->target_value);if (!new_goal)return -ENOMEM ; err =damos_commit_quota_goal (new_goal, src_goal);
mm/damon/lru_sort.c +3 -2
@@ -233,12 +233,13 @@ static int damon_lru_sort_add_quota_goals(struct damos *hot_scheme,if (!active_mem_bp)return 0 ; - goal =damos_new_quota_goal (DAMOS_QUOTA_ACTIVE_MEM_BP , active_mem_bp); + goal =damos_new_quota_goal (DAMOS_QUOTA_ACTIVE_MEM_BP , false, + active_mem_bp);if (!goal)return -ENOMEM ;damos_add_quota_goal (&hot_scheme->quota, goal);/* aim 0.2 % goal conflict, to keep little ping pong */ - goal =damos_new_quota_goal (DAMOS_QUOTA_INACTIVE_MEM_BP , + goal =damos_new_quota_goal (DAMOS_QUOTA_INACTIVE_MEM_BP , false,10000 - active_mem_bp +2 );if (!goal)return -ENOMEM ;
mm/damon/reclaim.c +3 -2
@@ -233,7 +233,7 @@ static int damon_reclaim_apply_parameters(void)damon_set_schemes (param_ctx, &scheme,1 );if (quota_mem_pressure_us) { - goal =damos_new_quota_goal (DAMOS_QUOTA_SOME_MEM_PSI_US , + goal =damos_new_quota_goal (DAMOS_QUOTA_SOME_MEM_PSI_US , false, quota_mem_pressure_us);if (!goal) goto out;
@@ -241,7 +241,8 @@ static int damon_reclaim_apply_parameters(void) }if (quota_autotune_feedback) { - goal =damos_new_quota_goal (DAMOS_QUOTA_USER_INPUT ,10000 ); + goal =damos_new_quota_goal (DAMOS_QUOTA_USER_INPUT , false, +10000 );if (!goal) goto out; goal->current_value = quota_autotune_feedback;
mm/damon/sysfs-schemes.c +1 -1
@@ -2869,7 +2869,7 @@ static int damos_sysfs_add_quota_score(if (!sysfs_goal->target_value)continue ; - goal =damos_new_quota_goal (sysfs_goal->metric, + goal =damos_new_quota_goal (sysfs_goal->metric, false, sysfs_goal->target_value);if (!goal)return -ENOMEM ;
mm/damon/tests/core-kunit.h +1 -0
@@ -930,6 +930,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal =damos_new_quota_goal (dst_goals[i].metric, + dst_goals[i].complementfalse, dst_goals[i].target_value);if (!goal) goto out;
samples/damon/mtier.c +1 -1
@@ -163,7 +163,7 @@ static struct damon_ctx *damon_sample_mtier_build_ctx(bool promote)damon_set_schemes (ctx, &scheme,1 ); quota_goal =damos_new_quota_goal ( promote ?DAMOS_QUOTA_NODE_MEM_USED_BP : -DAMOS_QUOTA_NODE_MEM_FREE_BP , +DAMOS_QUOTA_NODE_MEM_FREE_BP , false, promote ? node0_mem_used_bp : node0_mem_free_bp);if (!quota_goal) goto free_out;
Signature · 1 line
2.47.3
[email protected]
AI reviewRe: [RFC PATCH v4 2/7] mm/damon/core: add complement argument to damos_new_quota_goal()
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Low] damon: build failure in KUnit tests due to 'complementfalse' typo --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] damon: build failure in KUnit tests due to 'complementfalse' typo --
commit c13d1579fc49797d69a40d8e22286757e16cdd20
Author: SJ Park <[email protected]>
mm/damon/core: add complement argument to damos_new_quota_goal()
Extend the quota goal constructor, damos_new_quota_goal(), to receive and set the complement flag value. This commit also updates all callers to use the new signature.
mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 933
Quoted 9 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +930,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,
[Severity: Low]
Does this code cause a build failure when compiling the KUnit tests?
In damos_test_commit_quota_goals_for() within mm/damon/tests/core-kunit.h, it looks like there is a typo (complementfalse) which will break the build when CONFIG_DAMON_KUNIT_TEST=y is set.
Quoted 3 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
SJ Park
maintainerRe: [RFC PATCH v4 2/7] mm/damon/core: add complement argument to damos_new_quota_goal()
Quoted 26 lines · On Sun, 27 Sep 2026 12:12:29 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] damon: build failure in KUnit tests due to 'complementfalse' typo --commit
c13d1579fc49797d69a40d8e22286757e16cdd20
Author: SJ Park <[email protected]>
mm/damon/core: add complement argument todamos_new_quota_goal()Extend the quota goal constructor,
damos_new_quota_goal(), to receive and set the complement flag value. This commit also updates all callers to use the new signature.mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 933
Quoted 9 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +930,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,[Severity: Low]
Does this code cause a build failure when compiling the KUnit tests?
Quoted 3 lines
In
damos_test_commit_quota_goals_for()withinmm/damon/tests/core-kunit.h, it looks like there is a typo (complementfalse) which will break the build whenCONFIG_DAMON_KUNIT_TEST=yis set.
Good catch. I will fix in the next revision.
Quoted 7 lines
Quoted 4 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
Thanks,
SJ
SJ Park
maintainerRFC PATCH v4 3/7 mm/damon/sysfs-schemes: support quota goal complement flag
In reply to SJ ParkAdd a new sysfs file, complement, under the quota goal directory. It works for setting and getting the quota goal metric complement flag value.
--- mm/damon/sysfs-schemes.c | 32 +++++++++++++++++++++++++++++++- 1 file changed, 31 insertions(+), 1 deletion(-)
mm/damon/sysfs-schemes.c +31 -1
@@ -1220,6 +1220,7 @@ static const struct kobj_type damon_sysfs_watermarks_ktype = {struct damos_sysfs_quota_goal {struct kobject kobj;enum damos_quota_goal_metric metric; +bool complement;unsigned long target_value;unsigned long current_value;int nid;
@@ -1316,6 +1317,30 @@ static ssize_t target_metric_store(struct kobject *kobj,return -EINVAL ; } +static ssize_t complement_show (struct kobject *kobj, +struct kobj_attribute *attr,char *buf) +{ +struct damos_sysfs_quota_goal *goal =container_of (kobj, + struct damos_sysfs_quota_goal, kobj); + +return sysfs_emit (buf,"%c\n" , goal->complement ?'Y' :'N' ); +} + +static ssize_t complement_store (struct kobject *kobj, +struct kobj_attribute *attr,const char *buf,size_t count) +{ +struct damos_sysfs_quota_goal *goal =container_of (kobj, + struct damos_sysfs_quota_goal, kobj); +bool complement; +int err =kstrtobool (buf, &complement); + +if (err) +return err; + + goal->complement = complement; +return count; +} +static ssize_t target_value_show (struct kobject *kobj,struct kobj_attribute *attr,char *buf) {
@@ -1424,6 +1449,9 @@ static void damos_sysfs_quota_goal_release(struct kobject *kobj)static struct kobj_attribute damos_sysfs_quota_goal_target_metric_attr =__ATTR_RW_MODE (target_metric,0600 ); +static struct kobj_attribute damos_sysfs_quota_goal_complement_attr = +__ATTR_RW_MODE (complement,0600 ); +static struct kobj_attribute damos_sysfs_quota_goal_target_value_attr =__ATTR_RW_MODE (target_value,0600 );
@@ -1438,6 +1466,7 @@ static struct kobj_attribute damos_sysfs_quota_goal_path_attr =static struct attribute *damos_sysfs_quota_goal_attrs[] = { &damos_sysfs_quota_goal_target_metric_attr.attr, + &damos_sysfs_quota_goal_complement_attr.attr, &damos_sysfs_quota_goal_target_value_attr.attr, &damos_sysfs_quota_goal_current_value_attr.attr, &damos_sysfs_quota_goal_nid_attr.attr,
@@ -2869,7 +2898,8 @@ static int damos_sysfs_add_quota_score(if (!sysfs_goal->target_value)continue ; - goal =damos_new_quota_goal (sysfs_goal->metric, false, + goal =damos_new_quota_goal (sysfs_goal->metric, + sysfs_goal->complement, sysfs_goal->target_value);if (!goal)return -ENOMEM ;
Signature · 1 line
2.47.3
[email protected]
AI reviewRe: [RFC PATCH v4 3/7] mm/damon/sysfs-schemes: support quota goal complement flag
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Low] damon/tests: build failure due to complementfalse typo --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] damon/tests: build failure due to complementfalse typo
--
commit c13d1579fc49797d69a40d8e22286757e16cdd20
Author: SJ Park <[email protected]>
mm/damon/core: add complement argument to damos_new_quota_goal()
This commit extends the damos_new_quota_goal() constructor to receive and set the complement flag value. It also updates all API callers to use the new signature to prevent manual configuration mistakes.
[ ... ]
mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 933
Quoted 9 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +930,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,
[Severity: Low]
Is this a typo? The struct damos_quota_goal does not have a complementfalse member, which causes a build failure in damos_test_commit_quota_goals_for() when the damon kunit test configuration is enabled.
Quoted 3 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=3
SJ Park
maintainerRe: [RFC PATCH v4 3/7] mm/damon/sysfs-schemes: support quota goal complement flag
Quoted 15 lines · On Sun, 27 Sep 2026 12:19:52 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] damon/tests: build failure due to complementfalse typo
--commit
c13d1579fc49797d69a40d8e22286757e16cdd20
Author: SJ Park <[email protected]>
mm/damon/core: add complement argument todamos_new_quota_goal()This commit extends the
damos_new_quota_goal()constructor to receive and set the complement flag value. It also updates all API callers to use the new signature to prevent manual configuration mistakes.[ ... ]
Quoted 14 lines
mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 933
Quoted 9 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +930,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,[Severity: Low]
Is this a typo? The structdamos_quota_goaldoes not have a complementfalse member, which causes a build failure indamos_test_commit_quota_goals_for()when the damon kunit test configuration is enabled.
Yes, it is a typo. I will fix it in the next spin.
Quoted 8 lines
Quoted 4 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=3
Thanks,
SJ
SJ Park
maintainerRFC PATCH v4 4/7 mm/damon/tests/core-kunit: test quota_goal->complement commit
In reply to SJ ParkExtend existing DAMOS quota goal commit unit test to test the complement flag. Set the source complement flag and confirm the destination is updated to the given input.
--- mm/damon/tests/core-kunit.h | 2 ++ 1 file changed, 2 insertions(+)
mm/damon/tests/core-kunit.h +2 -0
@@ -819,6 +819,7 @@ static void damos_test_commit_quota_goal_for(struct kunit *test,damos_commit_quota_goal (dst, src);KUNIT_EXPECT_EQ (test, dst->metric, src->metric); +KUNIT_EXPECT_EQ (test, dst->complement, src->complement);KUNIT_EXPECT_EQ (test, dst->target_value, src->target_value);if (src->metric ==DAMOS_QUOTA_USER_INPUT )KUNIT_EXPECT_EQ (test, dst->current_value, src->current_value);
@@ -862,6 +863,7 @@ static void damos_test_commit_quota_goal(struct kunit *test)damos_test_commit_quota_goal_for (test, &dst, &(struct damos_quota_goal ){ .metric =DAMOS_QUOTA_USER_INPUT , + .complement = true, .target_value =789 , .current_value =12 }); damos_test_commit_quota_goal_for(test, &dst,
Signature · 1 line
2.47.3
[email protected]
AI reviewRe: [RFC PATCH v4 4/7] mm/damon/tests/core-kunit: test quota_goal->complement commit
Sashiko has reviewed this patch and found no issues. It looks great!
Quoted 5 lines
Extend existing DAMOS quota goal commit unit test to test the complement flag. Set the source complement flag and confirm the destination is updated to the given input.
Signed-off-by SJ Park <[email protected]>
Sashiko has reviewed this patch and found no issues. It looks great!
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=4
SJ Park
maintainerRFC PATCH v4 5/7 Docs/mm/damon/design: document damos quota goal complement flag
In reply to SJ ParkUpdate DAMON design document for the newly added damos quota goal metric value complement flag.
--- Documentation/mm/damon/design.rst | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-)
Documentation/mm/damon/design.rst +11 -4
@@ -702,10 +702,10 @@ There are two such tuning algorithms that users can select as they need. that the zero quota is a valid quota, and therefore ``qt_exceeds`` :ref:`stat <damon_design_damos_stat>` will keep increasing in this case. -The goal can be specified with five parameters, namely ``target_metric``, -``target_value``, ``current_value``, ``nid`` and ``path``. The auto-tuning -mechanism tries to make ``current_value`` of ``target_metric`` be same to -``target_value``. +The goal can be specified with six parameters, namely ``target_metric``, +``complement``, ``target_value``, ``current_value``, ``nid`` and ``path``. The +auto-tuning mechanism tries to make ``current_value`` of ``complement``-ed +``target_metric`` be same to ``target_value``. - ``user_input``: User-provided value. Users could use any metric that they has interest in for the value. Use space main workload's latency or
@@ -733,6 +733,13 @@ mechanism tries to make ``current_value`` of ``target_metric`` be same to - ``hugepage_mem_bp``: Total huge page to total used memory ratio in bp (1/10,000). +``complement`` is a boolean parameter that determines whether to use +complemented value of the target metric. For example, if ``complement`` is set +and target metric is ``active_mem_bp``, it is effectively same to +``inactive_mem_bp``. ``complement`` is no-op when the target metric type is +``user_input``. For the metric type, the user should be able to emit +complemented metric values on their own. + ``nid`` is optionally required for ``node_mem_used_bp``, ``node_mem_free_bp``, ``node_memcg_used_bp``, ``node_memcg_free_bp`` and ``node_eligible_mem_bp`` to point the specific NUMA node.
Signature · 1 line
2.47.3
[email protected]
AI reviewRe: [RFC PATCH v4 5/7] Docs/mm/damon/design: document damos quota goal complement flag
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Low] kselftest/damon: missing sysfs test coverage for quota goal complement --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] kselftest/damon: missing sysfs test coverage for quota goal complement --
commit d3c6455ce7af443b98334d4d8484051c2827e265
Author: SJ Park <[email protected]>
Docs/mm/damon/design: document damos quota goal complement flag
Update DAMON design document for the newly added damos quota goal metric value complement flag.
Quoted 6 lines
Documentation/mm/damon/design.rst +0
-0@@ -702,10 +702,10 @@ There are two such tuning algorithms that users can select as they need. that the zero quota is a valid quota, and therefore ``qt_exceeds`` :ref:`stat
Documentation/mm/damon/design.rst There are two such tuning algorithms that users can select as they need. · line 708
Quoted 10 lines
Documentation/mm/damon/design.rst +4
-4<damon_design_damos_stat>` will keep increasing in this case. -The goal can be specified with five parameters, namely ``target_metric``, -``target_value``, ``current_value``, ``nid`` and ``path``. The auto-tuning -mechanism tries to make ``current_value`` of ``target_metric`` be same to -``target_value``. +The goal can be specified with six parameters, namely ``target_metric``, +``complement``, ``target_value``, ``current_value``, ``nid`` and ``path``. The +auto-tuning mechanism tries to make ``current_value`` of ``complement``-ed +``target_metric`` be same to ``target_value``.
[Severity: Low]
Does the kselftest framework need to be updated to cover this new parameter?
Looking at the selftests, test_goal() in tools/testing/selftests/damon/sysfs.sh explicitly tests for target_metric and other parameters, but lacks a check for the newly added complement parameter:
ensure_file "$goal_dir/target_metric" "exist" "600"
Additionally, DamosQuotaGoal.stage() in tools/testing/selftests/damon/_damon_sysfs.py provides no fields or logic to set complement:
err = write_file(os.path.join(self.sysfs_dir(), 'target_metric'),
self.metric)
Could we add these test parameters to ensure the new sysfs feature is covered by the existing framework?
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=5
SJ Park
maintainerRe: [RFC PATCH v4 5/7] Docs/mm/damon/design: document damos quota goal complement flag
Quoted 42 lines · On Sun, 27 Sep 2026 12:13:15 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] kselftest/damon: missing sysfs test coverage for quota goal complement --commit
d3c6455ce7af443b98334d4d8484051c2827e265
Author: SJ Park <[email protected]>Docs/mm/damon/design: document damos quota goal complement flag
Update DAMON design document for the newly added damos quota goal metric value complement flag.
Quoted 6 lines
Documentation/mm/damon/design.rst +0
-0@@ -702,10 +702,10 @@ There are two such tuning algorithms that users can select as they need. that the zero quota is a valid quota, and therefore ``qt_exceeds`` :ref:`statDocumentation/mm/damon/design.rst There are two such tuning algorithms that users can select as they need. · line 708
Quoted 10 lines
Documentation/mm/damon/design.rst +4
-4<damon_design_damos_stat>` will keep increasing in this case. -The goal can be specified with five parameters, namely ``target_metric``, -``target_value``, ``current_value``, ``nid`` and ``path``. The auto-tuning -mechanism tries to make ``current_value`` of ``target_metric`` be same to -``target_value``. +The goal can be specified with six parameters, namely ``target_metric``, +``complement``, ``target_value``, ``current_value``, ``nid`` and ``path``. The +auto-tuning mechanism tries to make ``current_value`` of ``complement``-ed +``target_metric`` be same to ``target_value``.[Severity: Low]
Does the kselftest framework need to be updated to cover this new parameter?Looking at the selftests,
test_goal()intools/testing/selftests/damon/sysfs.shexplicitly tests fortarget_metricand other parameters, but lacks a check for the newly added complement parameter:ensure_file "$goal_dir/target_metric" "exist" "600"Additionally,
DamosQuotaGoal.stage()intools/testing/selftests/damon/_damon_sysfs.pyprovides no fields or logic to set complement:
Quoted 5 lines
err = write_file(os.path.join(self.sysfs_dir(), '
target_metric'),
self.metric)Could we add these test parameters to ensure the new sysfs feature is covered by the existing framework?
good catches. Not blockers of this series, though.
Quoted 4 lines
Signature · 2 lines
-- Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=5
Thanks,
SJ
SJ Park
maintainerRFC PATCH v4 6/7 Docs/admin-guide/mm/damon/usage: update for quota goal complement file
In reply to SJ ParkUpdate DAMON usage document for the newly added quota goal metric complement sysfs file.
--- Documentation/admin-guide/mm/damon/usage.rst | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-)
Documentation/admin-guide/mm/damon/usage.rst +8 -7
@@ -98,7 +98,8 @@ comma (","). │ │ │ │ │ │ │ fail_charge_num,fail_charge_denom │ │ │ │ │ │ │ │ weights/sz_permil,nr_accesses_permil,age_permil │ │ │ │ │ │ │ │ :ref:`goals <sysfs_schemes_quota_goals>`/nr_goals - │ │ │ │ │ │ │ │ │ 0/target_metric,target_value,current_value,nid,path + │ │ │ │ │ │ │ │ │ 0/target_metric,complement,target_value, + │ │ │ │ │ │ │ │ │ current_value,nid,path │ │ │ │ │ │ │ :ref:`watermarks <sysfs_watermarks>`/metric,interval_us,high,mid,low │ │ │ │ │ │ │ :ref:`{core_,ops_,}filters <sysfs_filters>`/nr_filters │ │ │ │ │ │ │ │ 0/type,matching,allow,memcg_path,addr_start,addr_end,damon_target_idx,min,max
@@ -492,12 +493,12 @@ number (``N``) to the file creates the number of child directories named ``0`` to ``N-1``. Each directory represents each goal and current achievement. Among the multiple feedback, the best one is used. -Each goal directory contains five files, namely ``target_metric``, -``target_value``, ``current_value``, ``nid``, and ``path``. Users can set and -get the five parameters for the quota auto-tuning goals that specified on the -:ref:`design doc <damon_design_damos_quotas_auto_tuning>` by writing to and -reading from each of the files. Because the kernel does not update -``current_value``, reading it only makes sense when ``target_metric`` is +Each goal directory contains six files, namely ``target_metric``, +``complement``, ``target_value``, ``current_value``, ``nid``, and ``path``. +Users can set and get the six parameters for the quota auto-tuning goals that +specified on the :ref:`design doc <damon_design_damos_quotas_auto_tuning>` by +writing to and reading from each of the files. Because the kernel does not +update ``current_value``, reading it only makes sense when ``target_metric`` is ``user_input``. Note that users should further write ``commit_schemes_quota_goals`` to the ``state`` file of the :ref:`kdamond directory <sysfs_kdamond>` to pass the feedback to DAMON.
Signature · 1 line
2.47.3
[email protected]
AI reviewRe: [RFC PATCH v4 6/7] Docs/admin-guide/mm/damon/usage: update for quota goal complement file
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Low] damon: build error in core-kunit.h due to typo complementfalse --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] damon: build error in core-kunit.h due to typo complementfalse
--
commit eec8b1a2e73ee2822f8574a7a1b78a43cc0fed67
Author: SJ Park <[email protected]>
Docs/admin-guide/mm/damon/usage: update for quota goal complement file
Update DAMON usage document for the newly added quota goal metric complement sysfs file.
Quoted 4 lines
mm/damon/tests/core-kunit.h +0
-0
[ ... ]
mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 935
Quoted 5 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +932,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,
[Severity: Low]
Does this regression cause a compilation failure when building with damon kunit tests enabled? It looks like struct damos_quota_goal has a member named complement, but not complementfalse in damos_test_commit_quota_goals_for().
Quoted 3 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=6
SJ Park
maintainerRe: [RFC PATCH v4 6/7] Docs/admin-guide/mm/damon/usage: update for quota goal complement file
Quoted 19 lines · On Sun, 27 Sep 2026 12:13:14 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] damon: build error incore-kunit.hdue to typo complementfalse
--commit
eec8b1a2e73ee2822f8574a7a1b78a43cc0fed67
Author: SJ Park <[email protected]>Docs/admin-guide/mm/damon/usage: update for quota goal complement file
Update DAMON usage document for the newly added quota goal metric complement sysfs file.
Quoted 4 lines
mm/damon/tests/core-kunit.h +0
-0[ ... ]
Quoted 11 lines
mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 935
Quoted 5 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +932,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,[Severity: Low]
Does this regression cause a compilation failure when building with damon kunit tests enabled? It looks like structdamos_quota_goalhas a member named complement, but not complementfalse indamos_test_commit_quota_goals_for().
I will fix it in the next spin.
Quoted 8 lines
Quoted 4 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=6
Thanks,
SJ
SJ Park
maintainerRFC PATCH v4 7/7 Docs/ABI/damon: update for quota goal metric complement sysfs file
In reply to SJ ParkUpdate DAMON ABI document for the newly added damos quota goal metric complement sysfs file.
--- Documentation/ABI/testing/sysfs-kernel-mm-damon | 6 ++++++ 1 file changed, 6 insertions(+)
Documentation/ABI/testing/sysfs-kernel-mm-damon +6 -0
@@ -375,6 +375,12 @@ Contact: SJ Park <[email protected]> Description: Writing to and reading from this file sets and gets the quota auto-tuning goal metric. +What: /sys/kernel/mm/damon/admin/kdamonds/<K>/contexts/<C>/schemes/<S>/quotas/goals/<G>/complement +Date: Sep 2026 +Contact: SJ Park <[email protected]> +Description: Writing to and reading from this file sets and gets whether to + use the complement value of the quota auto-tuning goal metric. + What: /sys/kernel/mm/damon/admin/kdamonds/<K>/contexts/<C>/schemes/<S>/quotas/goals/<G>/target_value Date: Nov 2023 Contact: SJ Park <[email protected]>
Signature · 1 line
2.47.3
[email protected]
AI reviewRe: [RFC PATCH v4 7/7] Docs/ABI/damon: update for quota goal metric complement sysfs file
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Low] mm/damon: build failure due to typo 'complementfalse' in tests --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] mm/damon: build failure due to typo 'complementfalse' in tests
--
commit c13d1579fc49797d69a40d8e22286757e16cdd20
Author: SJ Park <[email protected]>
mm/damon/core: add complement argument to damos_new_quota_goal()
This commit extends the damos_new_quota_goal() constructor to receive and set a complement flag argument. It updates all API callers to use the new signature to prevent manual assignment mistakes.
[ ... ]
mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 933
Quoted 9 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +930,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
[Severity: Low]
Does this code cause a build failure when CONFIG_DAMON_KUNIT_TEST=y is enabled?
The struct damos_quota_goal does not have a member named complementfalse, so accessing dst_goals[i].complementfalse in damos_test_commit_quota_goals_for() fails to compile. Should this be dst_goals[i].complement instead?
Quoted 3 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=7
SJ Park
maintainerRe: [RFC PATCH v4 7/7] Docs/ABI/damon: update for quota goal metric complement sysfs file
Quoted 26 lines · On Sun, 27 Sep 2026 12:14:52 +0000 [email protected] wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Low] mm/damon: build failure due to typo 'complementfalse' in tests
--commit
c13d1579fc49797d69a40d8e22286757e16cdd20
Author: SJ Park <[email protected]>
mm/damon/core: add complement argument todamos_new_quota_goal()This commit extends the
damos_new_quota_goal()constructor to receive and set a complement flag argument. It updates all API callers to use the new signature to prevent manual assignment mistakes.[ ... ]
mm/damon/tests/core-kunit.h static void damos_test_commit_quota_goals_for(struct kunit *test, · line 933
Quoted 9 lines
mm/damon/tests/core-kunit.h +1
-0@@ -930,6 +930,7 @@ static void damos_test_commit_quota_goals_for(struct kunit *test,* Make it kfree()-able. */ goal = damos_new_quota_goal(dst_goals[i].metric, + dst_goals[i].complementfalse,^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Quoted 8 lines
[Severity: Low]
Does this code cause a build failure whenCONFIG_DAMON_KUNIT_TEST=yis enabled?The struct
damos_quota_goaldoes not have a member named complementfalse, so accessing dst_goals[i].complementfalse indamos_test_commit_quota_goals_for()fails to compile. Should this be dst_goals[i].complement instead?
I will fix it in the next spin.
Quoted 8 lines
Quoted 4 lines
mm/damon/tests/core-kunit.h +0
-0dst_goals[i].target_value);if (!goal) goto out;Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=7
Thanks,
SJ
SJ Park
maintainerv4 cover mm/damon: introduce damos quota goal target metric complement flag (was: "Re: [RFC PATCH v4 0/7] snapshot_set_swap_area() unpins the previously selected swap device and")
In reply to SJ ParkQuoted 5 lines · On Sun, 27 Sep 2026 05:05:24 -0700 SJ Park <[email protected]> wrote:
Add
repin_hibernation_swap_type(), which looks up the new device, clears the oldSWP_HIBERNATIONflag and sets the new one under a singleswap_lockacquisition. The same-device case is short- circuited so userspace can re-select the same swap area without trippingWARN_ON_ONCEand-EBUSY. Switchsnapshot_set_swap_area()to the new helper.
[...]
Oops, this cover letter is completely wrong... Please ignore. My tool was using the commit message of unrelated commit due to my mistake. I'm so sorry for the noise.
Attaching the correct cover letter below.
Thanks,
SJ
=== >8 ===
mm/damon: introduce damos quota goal target metric complement flag
Aim-oriented DAMOS quota auto-tuning assumes the aggressiveness of the scheme (quota) and the quota goal metric are directly proportional. Depending on the scheme setup and usage, keeping the relationship can be challenging.
For example, let's suppose a memory tiering approach for higher upper tier utilization. One common idea for that (TPP) is utilizing two schemes, one for promotion and the other one for demotion. The promotion scheme migrates hot pages from lower tier to upper tier, aiming for high utilization of the upper tier. The demotion scheme migrates cold pages from upper tier to lower tier, aiming for head room free memory of the upper tier. Both schemes and their goals are in direct proportion. However, for this kind of use case, we need to implement two different goal metrics (per-node memory utilization and free memory ratio) while essentially the free memory ratio is just a complemented value of the utilization.
To avoid adding too many new metrics, we are adding metrics that turn out to be really needed for each found use case. For example, some_mem_psi_us, node_eligible_mem_bp and hugepage_mem_bp don't have their complemented value version. But it is not really difficult to expect use cases that their complemented version can be useful. For example, hugepage_mem_bp use case may need a way to reduce the hugepage ratio. That would require a complemented version of hugepage_mem_bp. Adding a new metric for each of such use cases could make the number of metrics unnecessarily high, and discourage flexible usages of DAMOS.
Add a new flag, quota goal complement, to allow flexible tuning goal setup without unnecessarily increasing the number of metrics. The flag specifies whether to use the complemented value of the given goal metric for the tuning. For example, if the complement flag is set, the upper tier memory utilization ratio metric works the same as the free memory ratio metric for the tier.
Note: Below is additional information that is not intended to be added to the final commit message.
Test
====
I tested this feature using the latest version of DAMON user-space tool, damo [1], as below.
On an idle system having 10 GiB single node memory (node id 0), fill up 80% of memory with page cache.
$ dd if=/dev/zero of=foo bs=1M count=$((1024*8))
Run DAMOS with pageout action for 80% of free memory, but using node_mem_used_bp metric like below.
$ sudo ./damo start --damos_action pageout \
--damos_quota_space 100M --damos_quota_interval 1s \
--damos_quota_fail_charge_ratio 1 4096 \
--damos_quota_goal_tuner temporal \
--damos_quota_goal node_mem_used_bp 20% 0Because the system has more than 20% node 0 memory utilization, DAMOS does nothing, as expected.
Run DAMOS with pageout action for 80% of free memory, using node_mem_used_bp metrics with complement flag like below.
$ sudo ./damo start --damos_action pageout \
--damos_quota_space 100M --damos_quota_interval 1s \
--damos_quota_fail_charge_ratio 1 4096 \
--damos_quota_goal_tuner temporal \
--damos_quota_goal node_mem_used_bp 80% complement 0Because the complement flag is given, it is effectively the same as node_mem_freepbp metrics with a target of 80%. DAMOS reclaims the page cache ~100 MiB per second, until the free memory ratio (complemented memory utilization ratio) reaches the goal, 80%.
Changes from RFC v4
- RFC v4: https://lore.kernel.org/[email protected]
- Fix a typo in test code (s/complementfalse/complement/) Changes from RFC v3
- RFC v3: https://lore.kernel.org/[email protected]
- Rebase to the latest mm-new.
- Add tests done to the cover letter.
Changes from RFC v2
- RFC v2: https://lore.kernel.org/[email protected]
- Avoid quota goal value underflow for some_mem_psi.
- Handle complement flag for psi goal with unset last_psi_total.
- Pass complement argument to damos_new_quota_goal() from core and core-kunit.
- Explicitly document no-op complement behavior for user_input. Changes from RFC
- RFC: https://lore.kernel.org/[email protected]
- Avoid quota goal value underflow.
- Fix usage doc for number of files in each goal directory.
- Rebase to the latest mm-new.