Skip to contentSkip to conversation

cgroup-v1: add disabled controller check in cgroup1_parse_param()

PATCH In review Luiz Capitulino 2 participants 2 messages Oct 28, 2022 – Oct 31, 2022 Stable branch

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 cgroup-v1: add disabled controller check in cgroup1_parse_param()

From: Chen Zhou <[email protected]>

Commit 61e960b07b637f0295308ad91268501d744c21b5 upstream.

[ This backport uses invalf() instead of invalfc() since the latter is
only available starting with v5.6 ]

When mounting a cgroup hierarchy with disabled controller in cgroup v1, all available controllers will be attached.
For example, boot with cgroup_no_v1=cpu or cgroup_disable=cpu, and then mount with "mount -t cgroup -ocpu cpu /sys/fs/cgroup/cpu", then all enabled controllers will be attached except cpu.

Fix this by adding disabled controller check in cgroup1_parse_param(). If the specified controller is disabled, just return error with information "Disabled controller xx" rather than attaching all the other enabled controllers.

Fixes f5dfb5315d34 ("cgroup: take options parsing into ->parse_monolithic()")
---
 kernel/cgroup/cgroup-v1.c | 3 +++
 1 file changed, 3 insertions(+)

Reviewers,

Only 5.4-stable is affected. The issue was introduced in 5.1 and fixed by Chen in 5.11 and 5.10-stable.

I tested the same reproducer on Amazon Linux 2 as described in the commit message (well, except that I used the cpuset controller).

kernel/cgroup/cgroup-v1.c +3 -0
index 9e847e71cedd..759a931f278b 100644
@@ -944,6 +944,9 @@ int cgroup1_parse_param(struct fs_context *fc, struct fs_parameter *param)
 		for_each_subsys(ss, i) {
 			if (strcmp(param->key, ss->legacy_name))
 				continue;
+			if (!cgroup_ssid_enabled(i) || cgroup1_ssid_disabled(i))
+				return invalf(fc, "Disabled controller '%s'",
+					       param->key);
 			ctx->subsys_mask |= (1 << i);
 			return 0;
 		}
Signature · 1 line
2.24.4.AMZN
maintainer
#1 lore raw
Quoted 28 lines · Luiz Capitulino

From: Chen Zhou <[email protected]>

Commit 61e960b07b637f0295308ad91268501d744c21b5 upstream.

[ This backport uses invalf() instead of invalfc() since the latter is
only available starting with v5.6 ]

When mounting a cgroup hierarchy with disabled controller in cgroup v1, all available controllers will be attached.
For example, boot with cgroup_no_v1=cpu or cgroup_disable=cpu, and then mount with "mount -t cgroup -ocpu cpu /sys/fs/cgroup/cpu", then all enabled controllers will be attached except cpu.

Fix this by adding disabled controller check in cgroup1_parse_param(). If the specified controller is disabled, just return error with information "Disabled controller xx" rather than attaching all the other enabled controllers.

Fixes f5dfb5315d34 ("cgroup: take options parsing into ->parse_monolithic()")
---
 kernel/cgroup/cgroup-v1.c | 3 +++
 1 file changed, 3 insertions(+)
Quoted 7 lines

Reviewers,

Only 5.4-stable is affected. The issue was introduced in 5.1 and fixed by Chen in 5.11 and 5.10-stable.

I tested the same reproducer on Amazon Linux 2 as described in the commit message (well, except that I used the cpuset controller).

Now queued up, thanks.

greg k-h