Re: [PATCH] cgroup: Kill the parent controller when its last child is killed
From: kernel test robot
Date: Sun Apr 03 2022 - 12:00:43 EST
Hi Bui,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on tj-cgroup/for-next]
[also build test ERROR on v5.17 next-20220401]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/intel-lab-lkp/linux/commits/Bui-Quang-Minh/cgroup-Kill-the-parent-controller-when-its-last-child-is-killed/20220403-215911
base: https://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup.git for-next
config: i386-randconfig-m021 (https://download.01.org/0day-ci/archive/20220403/202204032330.l2wsF3mf-lkp@xxxxxxxxx/config)
compiler: gcc-11 (Debian 11.2.0-19) 11.2.0
reproduce (this is a W=1 build):
# https://github.com/intel-lab-lkp/linux/commit/2bc22feae8a913c7f371bc79ef9967122d8d326c
git remote add linux-review https://github.com/intel-lab-lkp/linux
git fetch --no-tags linux-review Bui-Quang-Minh/cgroup-Kill-the-parent-controller-when-its-last-child-is-killed/20220403-215911
git checkout 2bc22feae8a913c7f371bc79ef9967122d8d326c
# save the config file to linux build tree
mkdir build_dir
make W=1 O=build_dir ARCH=i386 SHELL=/bin/bash kernel/
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@xxxxxxxxx>
All errors (new ones prefixed by >>):
kernel/cgroup/cgroup.c: In function 'css_release_work_fn':
>> kernel/cgroup/cgroup.c:5169:52: error: 'struct cgroup_bpf' has no member named 'refcnt'
5169 | if (!percpu_ref_is_dying(&cgrp->bpf.refcnt))
| ^
vim +5169 kernel/cgroup/cgroup.c
5148
5149 static void css_release_work_fn(struct work_struct *work)
5150 {
5151 struct cgroup_subsys_state *css =
5152 container_of(work, struct cgroup_subsys_state, destroy_work);
5153 struct cgroup_subsys *ss = css->ss;
5154 struct cgroup *cgrp = css->cgroup;
5155 struct cgroup *parent = cgroup_parent(cgrp);
5156
5157 mutex_lock(&cgroup_mutex);
5158
5159 css->flags |= CSS_RELEASED;
5160 list_del_rcu(&css->sibling);
5161
5162 /*
5163 * If parent doesn't have any children, start killing it.
5164 * And don't kill the default root.
5165 */
5166 if (parent && list_empty(&parent->self.children) &&
5167 parent != &cgrp_dfl_root.cgrp &&
5168 !percpu_ref_is_dying(&parent->self.refcnt)) {
> 5169 if (!percpu_ref_is_dying(&cgrp->bpf.refcnt))
5170 cgroup_bpf_offline(parent);
5171 percpu_ref_kill(&parent->self.refcnt);
5172 }
5173
5174 if (ss) {
5175 /* css release path */
5176 if (!list_empty(&css->rstat_css_node)) {
5177 cgroup_rstat_flush(cgrp);
5178 list_del_rcu(&css->rstat_css_node);
5179 }
5180
5181 cgroup_idr_replace(&ss->css_idr, NULL, css->id);
5182 if (ss->css_released)
5183 ss->css_released(css);
5184 } else {
5185 struct cgroup *tcgrp;
5186
5187 /* cgroup release path */
5188 TRACE_CGROUP_PATH(release, cgrp);
5189
5190 cgroup_rstat_flush(cgrp);
5191
5192 spin_lock_irq(&css_set_lock);
5193 for (tcgrp = cgroup_parent(cgrp); tcgrp;
5194 tcgrp = cgroup_parent(tcgrp))
5195 tcgrp->nr_dying_descendants--;
5196 spin_unlock_irq(&css_set_lock);
5197
5198 /*
5199 * There are two control paths which try to determine
5200 * cgroup from dentry without going through kernfs -
5201 * cgroupstats_build() and css_tryget_online_from_dir().
5202 * Those are supported by RCU protecting clearing of
5203 * cgrp->kn->priv backpointer.
5204 */
5205 if (cgrp->kn)
5206 RCU_INIT_POINTER(*(void __rcu __force **)&cgrp->kn->priv,
5207 NULL);
5208 }
5209
5210 mutex_unlock(&cgroup_mutex);
5211
5212 INIT_RCU_WORK(&css->destroy_rwork, css_free_rwork_fn);
5213 queue_rcu_work(cgroup_destroy_wq, &css->destroy_rwork);
5214 }
5215
--
0-DAY CI Kernel Test Service
https://01.org/lkp