Android has mounted the v1 cpuset controller using filesystem type
"cpuset" (not "cgroup") since 2015 [1], and depends on the resulting
behavior where the controller name is not added as a prefix for cgroupfs
files. [2]

Later, a problem was discovered where cpu hotplug onlining did not
affect the cpuset/cpus files, which Android carried an out-of-tree patch
to address for a while. An attempt was made to upstream this patch, but
the recommendation was to use the "cpuset_v2_mode" mount option
instead. [3]

An effort was made to do so, but this fails with "cgroup: Unknown
parameter 'cpuset_v2_mode'" because commit e1cba4b85daa ("cgroup: Add
mount flag to enable cpuset to use v2 behavior in v1 cgroup") did not
update the special cased cpuset_mount(), and only the cgroup (v1)
filesystem type was updated.

Add cgroup v1 parameter parsing to the cpuset filesystem type so that it
works like the cgroup filesystem type:

$ mkdir /dev/cpuset
$ mount -t cpuset -ocpuset_v2_mode none /dev/cpuset
$ mount|grep cpuset
none on /dev/cpuset type cgroup (rw,relatime,cpuset,noprefix,cpuset_v2_mode,release_agent=/sbin/cpuset_release_agent)

[1] https://cs.android.com/android/_/android/platform/system/core/+/b769c8d24fd7be96f8968aa4c80b669525b930d3
[2] https://cs.android.com/android/platform/superproject/main/+/main:system/core/libprocessgroup/setup/cgroup_map_write.cpp;drc=2dac5d89a0f024a2d0cc46a80ba4ee13472f1681;l=192
[3] https://lore.kernel.org/lkml/f795f8be-a184-408a-0b5a-553d26061385@redhat.com/T/

Fixes: e1cba4b85daa ("cgroup: Add mount flag to enable cpuset to use v2 behavior in v1 cgroup")
Signed-off-by: T.J. Mercier <tjmercier@google.com>
---
 kernel/cgroup/cgroup.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index 3caf2cd86e65..ceedaf47d494 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -2356,6 +2356,7 @@ static struct file_system_type cgroup2_fs_type = {
 static const struct fs_context_operations cpuset_fs_context_ops = {
 	.get_tree	= cgroup1_get_tree,
 	.free		= cgroup_fs_context_free,
+	.parse_param	= cgroup1_parse_param,
 };
 
 /*
@@ -2392,6 +2393,7 @@ static int cpuset_init_fs_context(struct fs_context *fc)
 static struct file_system_type cpuset_fs_type = {
 	.name			= "cpuset",
 	.init_fs_context	= cpuset_init_fs_context,
+	.parameters		= cgroup1_fs_parameters,
 	.fs_flags		= FS_USERNS_MOUNT,
 };
 #endif
-- 
2.49.0.604.gff1f9ca942-goog

Hello.

On Mon, Apr 14, 2025 at 04:28:41PM +0000, "T.J. Mercier" <tjmercier@google.com> wrote:
> Add cgroup v1 parameter parsing to the cpuset filesystem type so that it
> works like the cgroup filesystem type:

Nothing against 'cpuset_v2_mode' for the cpuset_fs_type (when it's
available on cgroup v1) but isn't it too benevolent reusing all of
cgroup1_fs_parameters? AFAICS, this would allow overriding release agent
also for cpuset fs hierarchies among other options from
cgroup1_fs_parameters.

(This would likely end up with a separate .parse_param callback but I
think that's better than adding so many extra features to cpuset fs.)


Thanks,
Michal

On 4/14/25 2:26 PM, Michal Koutný wrote:
> Hello.
>
> On Mon, Apr 14, 2025 at 04:28:41PM +0000, "T.J. Mercier" <tjmercier@google.com> wrote:
>> Add cgroup v1 parameter parsing to the cpuset filesystem type so that it
>> works like the cgroup filesystem type:
> Nothing against 'cpuset_v2_mode' for the cpuset_fs_type (when it's
> available on cgroup v1) but isn't it too benevolent reusing all of
> cgroup1_fs_parameters? AFAICS, this would allow overriding release agent
> also for cpuset fs hierarchies among other options from
> cgroup1_fs_parameters.
>
> (This would likely end up with a separate .parse_param callback but I
> think that's better than adding so many extra features to cpuset fs.)

I concur. It should be a separate cpuset_fs_parameters() to handle it 
instead of reusing cgroup1_fs_parameters() to allow so many other maybe 
irrelevant cgroup1 parameters.

Cheers,
Longman

On Mon, Apr 14, 2025 at 1:11 PM Waiman Long <llong@redhat.com> wrote:
>
>
> On 4/14/25 2:26 PM, Michal Koutný wrote:
> > Hello.
> >
> > On Mon, Apr 14, 2025 at 04:28:41PM +0000, "T.J. Mercier" <tjmercier@google.com> wrote:
> >> Add cgroup v1 parameter parsing to the cpuset filesystem type so that it
> >> works like the cgroup filesystem type:
> > Nothing against 'cpuset_v2_mode' for the cpuset_fs_type (when it's
> > available on cgroup v1) but isn't it too benevolent reusing all of
> > cgroup1_fs_parameters? AFAICS, this would allow overriding release agent
> > also for cpuset fs hierarchies among other options from
> > cgroup1_fs_parameters.
> >
> > (This would likely end up with a separate .parse_param callback but I
> > think that's better than adding so many extra features to cpuset fs.)
>
> I concur. It should be a separate cpuset_fs_parameters() to handle it
> instead of reusing cgroup1_fs_parameters() to allow so many other maybe
> irrelevant cgroup1 parameters.
>
> Cheers,
> Longman

Hi Michal and Longman,

Yes, that's fair. I'll send a v2 tomorrow.

Thanks,
T.J.