From: Xiu Jianfeng <xiujianfeng@huawei.com>

Currently cgroup1_pidlist_destroy_all() will be called when releasing
cgroup even if the cgroup is on default hierarchy, however it doesn't
make any sense for v2 to destroy pidlist of v1.

Signed-off-by: Xiu Jianfeng <xiujianfeng@huawei.com>

---
v2: move the test to the caller
---
 kernel/cgroup/cgroup.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index a66c088c851c..e32b6972c478 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -5368,7 +5368,8 @@ static void css_free_rwork_fn(struct work_struct *work)
 	} else {
 		/* cgroup free path */
 		atomic_dec(&cgrp->root->nr_cgrps);
-		cgroup1_pidlist_destroy_all(cgrp);
+		if (!cgroup_on_dfl(cgrp))
+			cgroup1_pidlist_destroy_all(cgrp);
 		cancel_work_sync(&cgrp->release_agent_work);
 		bpf_cgrp_storage_free(cgrp);
 
-- 
2.34.1

On Thu, Apr 18, 2024 at 02:19:30AM +0000, Xiu Jianfeng wrote:
> From: Xiu Jianfeng <xiujianfeng@huawei.com>
> 
> Currently cgroup1_pidlist_destroy_all() will be called when releasing
> cgroup even if the cgroup is on default hierarchy, however it doesn't
> make any sense for v2 to destroy pidlist of v1.
> 
> Signed-off-by: Xiu Jianfeng <xiujianfeng@huawei.com>

Applied to cgroup/for-6.10.

Thanks.

-- 
tejun