[PATCH 0/4] Use __counted_by for ancestor arrays

[PATCH 0/4] Use __counted_by for ancestor arrays

[Michal Koutný]

The trick with utilizing space in cgroup_root for cgroup::ancetors flex
array was an obstacle for kernel reworks for
-Wflex-array-member-not-at-end.

The first patch fixes that, then I wanted to utilize __counted_by for
this flex array which required some more rework how cgroup level is
evaluated.

Similar flex array is also in struct ioc_gq where it was tempting to
simply use __counted_by(level), however, this would be off-by-one as it
has semantics like cgroup's level (0 == root).
Proper adjustment for __counted_by() would either need similar
level/ancestor helpers or abstracted macros for ancestors up/down
iterations.

I only made a simple comment fixup since I'm not sure about benefit of
__counted_by for structs that aren't sized based on direct user input.

Michal Koutný (4):
  cgroup: Eliminate cgrp_ancestor_storage in cgroup_root
  cgroup: Introduce cgroup_level() helper
  cgroup: Use __counted_by for cgroup::ancestors
  blk-iocost: Correct comment ioc_gq::level

 block/bfq-iosched.c           |  2 +-
 block/blk-iocost.c            |  6 ++---
 include/linux/cgroup-defs.h   | 43 +++++++++++++++++++----------------
 include/linux/cgroup.h        | 18 ++++++++++++---
 include/trace/events/cgroup.h |  8 +++----
 kernel/bpf/helpers.c          |  2 +-
 kernel/cgroup/cgroup.c        |  9 ++++----
 net/netfilter/nft_socket.c    |  2 +-
 8 files changed, 53 insertions(+), 37 deletions(-)


base-commit: 8f0b4cce4481fb22653697cced8d0d04027cb1e8
-- 
2.52.0

[PATCH 2/4] cgroup: Introduce cgroup_level() helper

[Michal Koutný]

This is a no functional change to hide physical storage of cgroup's
level and it allows subesequent conversion of the storage.

Signed-off-by: Michal Koutný <mkoutny@suse.com>
---
 block/bfq-iosched.c           |  2 +-
 block/blk-iocost.c            |  4 ++--
 include/linux/cgroup.h        | 18 +++++++++++++++---
 include/trace/events/cgroup.h |  8 ++++----
 kernel/bpf/helpers.c          |  2 +-
 kernel/cgroup/cgroup.c        |  4 ++--
 net/netfilter/nft_socket.c    |  2 +-
 7 files changed, 26 insertions(+), 14 deletions(-)

diff --git a/block/bfq-iosched.c b/block/bfq-iosched.c
index 4a8d3d96bfe49..f293bab068274 100644
--- a/block/bfq-iosched.c
+++ b/block/bfq-iosched.c
@@ -601,7 +601,7 @@ static bool bfqq_request_over_limit(struct bfq_data *bfqd,
 		goto out;
 
 	/* +1 for bfqq entity, root cgroup not included */
-	depth = bfqg_to_blkg(bfqq_group(bfqq))->blkcg->css.cgroup->level + 1;
+	depth = cgroup_level(bfqg_to_blkg(bfqq_group(bfqq))->blkcg->css.cgroup) + 1;
 	if (depth > alloc_depth) {
 		spin_unlock_irq(&bfqd->lock);
 		if (entities != inline_entities)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index a0416927d33dc..b4eebe61dca7f 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -2962,7 +2962,7 @@ static void ioc_cpd_free(struct blkcg_policy_data *cpd)
 static struct blkg_policy_data *ioc_pd_alloc(struct gendisk *disk,
 		struct blkcg *blkcg, gfp_t gfp)
 {
-	int levels = blkcg->css.cgroup->level + 1;
+	int levels = cgroup_level(blkcg->css.cgroup) + 1;
 	struct ioc_gq *iocg;
 
 	iocg = kzalloc_node(struct_size(iocg, ancestors, levels), gfp,
@@ -3003,7 +3003,7 @@ static void ioc_pd_init(struct blkg_policy_data *pd)
 	init_waitqueue_head(&iocg->waitq);
 	hrtimer_setup(&iocg->waitq_timer, iocg_waitq_timer_fn, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
 
-	iocg->level = blkg->blkcg->css.cgroup->level;
+	iocg->level = cgroup_level(blkg->blkcg->css.cgroup)
 
 	for (tblkg = blkg; tblkg; tblkg = tblkg->parent) {
 		struct ioc_gq *tiocg = blkg_to_iocg(tblkg);
diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index bc892e3b37eea..0290878ebad26 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -525,6 +525,18 @@ static inline struct cgroup *cgroup_parent(struct cgroup *cgrp)
 	return NULL;
 }
 
+/**
+ * cgroup_level - cgroup depth
+ * @cgrp: cgroup
+ *
+ * The depth this cgroup is at.  The root is at depth zero and each step down
+ * the hierarchy increments the level.
+ */
+static inline int cgroup_level(struct cgroup *cgrp)
+{
+	return cgrp->level;
+}
+
 /**
  * cgroup_is_descendant - test ancestry
  * @cgrp: the cgroup to be tested
@@ -537,9 +549,9 @@ static inline struct cgroup *cgroup_parent(struct cgroup *cgrp)
 static inline bool cgroup_is_descendant(struct cgroup *cgrp,
 					struct cgroup *ancestor)
 {
-	if (cgrp->root != ancestor->root || cgrp->level < ancestor->level)
+	if (cgrp->root != ancestor->root || cgroup_level(cgrp) < cgroup_level(ancestor))
 		return false;
-	return cgrp->ancestors[ancestor->level] == ancestor;
+	return cgrp->ancestors[cgroup_level(ancestor)] == ancestor;
 }
 
 /**
@@ -556,7 +568,7 @@ static inline bool cgroup_is_descendant(struct cgroup *cgrp,
 static inline struct cgroup *cgroup_ancestor(struct cgroup *cgrp,
 					     int ancestor_level)
 {
-	if (ancestor_level < 0 || ancestor_level > cgrp->level)
+	if (ancestor_level < 0 || ancestor_level > cgroup_level(cgrp))
 		return NULL;
 	return cgrp->ancestors[ancestor_level];
 }
diff --git a/include/trace/events/cgroup.h b/include/trace/events/cgroup.h
index ba9229af9a343..0a1bc91754b5e 100644
--- a/include/trace/events/cgroup.h
+++ b/include/trace/events/cgroup.h
@@ -67,7 +67,7 @@ DECLARE_EVENT_CLASS(cgroup,
 	TP_fast_assign(
 		__entry->root = cgrp->root->hierarchy_id;
 		__entry->id = cgroup_id(cgrp);
-		__entry->level = cgrp->level;
+		__entry->level = cgroup_level(cgrp);
 		__assign_str(path);
 	),
 
@@ -136,7 +136,7 @@ DECLARE_EVENT_CLASS(cgroup_migrate,
 	TP_fast_assign(
 		__entry->dst_root = dst_cgrp->root->hierarchy_id;
 		__entry->dst_id = cgroup_id(dst_cgrp);
-		__entry->dst_level = dst_cgrp->level;
+		__entry->dst_level = cgroup_level(dst_cgrp);
 		__assign_str(dst_path);
 		__entry->pid = task->pid;
 		__assign_str(comm);
@@ -180,7 +180,7 @@ DECLARE_EVENT_CLASS(cgroup_event,
 	TP_fast_assign(
 		__entry->root = cgrp->root->hierarchy_id;
 		__entry->id = cgroup_id(cgrp);
-		__entry->level = cgrp->level;
+		__entry->level = cgroup_level(cgrp);
 		__assign_str(path);
 		__entry->val = val;
 	),
@@ -221,7 +221,7 @@ DECLARE_EVENT_CLASS(cgroup_rstat,
 	TP_fast_assign(
 		__entry->root = cgrp->root->hierarchy_id;
 		__entry->id = cgroup_id(cgrp);
-		__entry->level = cgrp->level;
+		__entry->level = cgroup_level(cgrp);
 		__entry->cpu = cpu;
 		__entry->contended = contended;
 	),
diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
index db72b96f9c8c8..b825f6e0a1c29 100644
--- a/kernel/bpf/helpers.c
+++ b/kernel/bpf/helpers.c
@@ -2577,7 +2577,7 @@ __bpf_kfunc struct cgroup *bpf_cgroup_ancestor(struct cgroup *cgrp, int level)
 {
 	struct cgroup *ancestor;
 
-	if (level > cgrp->level || level < 0)
+	if (level > cgroup_level(cgrp) || level < 0)
 		return NULL;
 
 	/* cgrp's refcnt could be 0 here, but ancestors can still be accessed */
diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index 554a02ee298ba..e011f1dd6d87f 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -5843,7 +5843,7 @@ static struct cgroup *cgroup_create(struct cgroup *parent, const char *name,
 	struct cgroup_root *root = parent->root;
 	struct cgroup *cgrp, *tcgrp;
 	struct kernfs_node *kn;
-	int i, level = parent->level + 1;
+	int i, level = cgroup_level(parent) + 1;
 	int ret;
 
 	/* allocate the cgroup and its ID, 0 is reserved for the root */
@@ -5884,7 +5884,7 @@ static struct cgroup *cgroup_create(struct cgroup *parent, const char *name,
 		goto out_stat_exit;
 
 	for (tcgrp = cgrp; tcgrp; tcgrp = cgroup_parent(tcgrp))
-		cgrp->ancestors[tcgrp->level] = tcgrp;
+		cgrp->ancestors[cgroup_level(tcgrp)] = tcgrp;
 
 	/*
 	 * New cgroup inherits effective freeze counter, and
diff --git a/net/netfilter/nft_socket.c b/net/netfilter/nft_socket.c
index 36affbb697c2f..a5b0340924efb 100644
--- a/net/netfilter/nft_socket.c
+++ b/net/netfilter/nft_socket.c
@@ -64,7 +64,7 @@ static noinline int nft_socket_cgroup_subtree_level(void)
 	if (IS_ERR(cgrp))
 		return PTR_ERR(cgrp);
 
-	level = cgrp->level;
+	level = cgroup_level(cgrp);
 
 	cgroup_put(cgrp);
 
-- 
2.52.0

Re: [PATCH 2/4] cgroup: Introduce cgroup_level() helper

[bot+bpf-ci]

[-- Attachment #1: Type: text/plain, Size: 1062 bytes --]

> diff --git a/block/blk-iocost.c b/block/blk-iocost.c
> index a0416927d33d..b4eebe61dca7 100644
> --- a/block/blk-iocost.c
> +++ b/block/blk-iocost.c

[ ... ]

> @@ -3003,7 +3003,7 @@ static void ioc_pd_init(struct blkg_policy_data *pd)
>  	init_waitqueue_head(&iocg->waitq);
>  	hrtimer_setup(&iocg->waitq_timer, iocg_waitq_timer_fn, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
>
> -	iocg->level = blkg->blkcg->css.cgroup->level;
> +	iocg->level = cgroup_level(blkg->blkcg->css.cgroup)
                                                              ^
Is this missing a semicolon? The line should end with a semicolon to
make this valid C syntax. This will prevent the kernel from building
in the ioc_pd_init() function.

>
>  	for (tblkg = blkg; tblkg; tblkg = tblkg->parent) {
>  		struct ioc_gq *tiocg = blkg_to_iocg(tblkg);


---
AI reviewed your patch. Please fix the bug or email reply why it's not a bug.
See: https://github.com/kernel-patches/vmtest/blob/master/ci/claude/README.md

CI run summary: https://github.com/kernel-patches/bpf/actions/runs/20310204023

Re: [PATCH 2/4] cgroup: Introduce cgroup_level() helper

[kernel test robot]

Hi Michal,

kernel test robot noticed the following build errors:

[auto build test ERROR on 8f0b4cce4481fb22653697cced8d0d04027cb1e8]

url:    https://github.com/intel-lab-lkp/linux/commits/Michal-Koutn/cgroup-Eliminate-cgrp_ancestor_storage-in-cgroup_root/20251218-004346
base:   8f0b4cce4481fb22653697cced8d0d04027cb1e8
patch link:    https://lore.kernel.org/r/20251217162744.352391-3-mkoutny%40suse.com
patch subject: [PATCH 2/4] cgroup: Introduce cgroup_level() helper
config: sparc64-randconfig-r134-20251218 (https://download.01.org/0day-ci/archive/20251220/202512202230.1uoB5chV-lkp@intel.com/config)
compiler: clang version 20.1.8 (https://github.com/llvm/llvm-project 87f0227cb60147a26a1eeb4fb06e3b505e9c7261)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251220/202512202230.1uoB5chV-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202512202230.1uoB5chV-lkp@intel.com/

All errors (new ones prefixed by >>):

>> block/blk-iocost.c:3006:53: error: expected ';' after expression
    3006 |         iocg->level = cgroup_level(blkg->blkcg->css.cgroup)
         |                                                            ^
         |                                                            ;
   1 error generated.


vim +3006 block/blk-iocost.c

  2981	
  2982	static void ioc_pd_init(struct blkg_policy_data *pd)
  2983	{
  2984		struct ioc_gq *iocg = pd_to_iocg(pd);
  2985		struct blkcg_gq *blkg = pd_to_blkg(&iocg->pd);
  2986		struct ioc *ioc = q_to_ioc(blkg->q);
  2987		struct ioc_now now;
  2988		struct blkcg_gq *tblkg;
  2989		unsigned long flags;
  2990	
  2991		ioc_now(ioc, &now);
  2992	
  2993		iocg->ioc = ioc;
  2994		atomic64_set(&iocg->vtime, now.vnow);
  2995		atomic64_set(&iocg->done_vtime, now.vnow);
  2996		atomic64_set(&iocg->active_period, atomic64_read(&ioc->cur_period));
  2997		INIT_LIST_HEAD(&iocg->active_list);
  2998		INIT_LIST_HEAD(&iocg->walk_list);
  2999		INIT_LIST_HEAD(&iocg->surplus_list);
  3000		iocg->hweight_active = WEIGHT_ONE;
  3001		iocg->hweight_inuse = WEIGHT_ONE;
  3002	
  3003		init_waitqueue_head(&iocg->waitq);
  3004		hrtimer_setup(&iocg->waitq_timer, iocg_waitq_timer_fn, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
  3005	
> 3006		iocg->level = cgroup_level(blkg->blkcg->css.cgroup)
  3007	
  3008		for (tblkg = blkg; tblkg; tblkg = tblkg->parent) {
  3009			struct ioc_gq *tiocg = blkg_to_iocg(tblkg);
  3010			iocg->ancestors[tiocg->level] = tiocg;
  3011		}
  3012	
  3013		spin_lock_irqsave(&ioc->lock, flags);
  3014		weight_updated(iocg, &now);
  3015		spin_unlock_irqrestore(&ioc->lock, flags);
  3016	}
  3017	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

Re: [PATCH 2/4] cgroup: Introduce cgroup_level() helper

[kernel test robot]

Hi Michal,

kernel test robot noticed the following build errors:

[auto build test ERROR on 8f0b4cce4481fb22653697cced8d0d04027cb1e8]

url:    https://github.com/intel-lab-lkp/linux/commits/Michal-Koutn/cgroup-Eliminate-cgrp_ancestor_storage-in-cgroup_root/20251218-004346
base:   8f0b4cce4481fb22653697cced8d0d04027cb1e8
patch link:    https://lore.kernel.org/r/20251217162744.352391-3-mkoutny%40suse.com
patch subject: [PATCH 2/4] cgroup: Introduce cgroup_level() helper
config: nios2-allmodconfig (https://download.01.org/0day-ci/archive/20251221/202512210532.ziNaxDJf-lkp@intel.com/config)
compiler: nios2-linux-gcc (GCC) 11.5.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20251221/202512210532.ziNaxDJf-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202512210532.ziNaxDJf-lkp@intel.com/

All error/warnings (new ones prefixed by >>):

   block/blk-iocost.c: In function 'ioc_pd_init':
>> block/blk-iocost.c:3006:60: error: expected ';' before 'for'
    3006 |         iocg->level = cgroup_level(blkg->blkcg->css.cgroup)
         |                                                            ^
         |                                                            ;
    3007 | 
    3008 |         for (tblkg = blkg; tblkg; tblkg = tblkg->parent) {
         |         ~~~                                                 
>> block/blk-iocost.c:2988:26: warning: unused variable 'tblkg' [-Wunused-variable]
    2988 |         struct blkcg_gq *tblkg;
         |                          ^~~~~


vim +3006 block/blk-iocost.c

  2981	
  2982	static void ioc_pd_init(struct blkg_policy_data *pd)
  2983	{
  2984		struct ioc_gq *iocg = pd_to_iocg(pd);
  2985		struct blkcg_gq *blkg = pd_to_blkg(&iocg->pd);
  2986		struct ioc *ioc = q_to_ioc(blkg->q);
  2987		struct ioc_now now;
> 2988		struct blkcg_gq *tblkg;
  2989		unsigned long flags;
  2990	
  2991		ioc_now(ioc, &now);
  2992	
  2993		iocg->ioc = ioc;
  2994		atomic64_set(&iocg->vtime, now.vnow);
  2995		atomic64_set(&iocg->done_vtime, now.vnow);
  2996		atomic64_set(&iocg->active_period, atomic64_read(&ioc->cur_period));
  2997		INIT_LIST_HEAD(&iocg->active_list);
  2998		INIT_LIST_HEAD(&iocg->walk_list);
  2999		INIT_LIST_HEAD(&iocg->surplus_list);
  3000		iocg->hweight_active = WEIGHT_ONE;
  3001		iocg->hweight_inuse = WEIGHT_ONE;
  3002	
  3003		init_waitqueue_head(&iocg->waitq);
  3004		hrtimer_setup(&iocg->waitq_timer, iocg_waitq_timer_fn, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
  3005	
> 3006		iocg->level = cgroup_level(blkg->blkcg->css.cgroup)
  3007	
  3008		for (tblkg = blkg; tblkg; tblkg = tblkg->parent) {
  3009			struct ioc_gq *tiocg = blkg_to_iocg(tblkg);
  3010			iocg->ancestors[tiocg->level] = tiocg;
  3011		}
  3012	
  3013		spin_lock_irqsave(&ioc->lock, flags);
  3014		weight_updated(iocg, &now);
  3015		spin_unlock_irqrestore(&ioc->lock, flags);
  3016	}
  3017	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki