cgroup: Fix threadgroup_rwsem <-> cpus_read_lock() deadlock

stable inclusion
from stable-v4.19.280
commit e446300968c6bd25d9cd6c33b9600780a39b3975
category: bugfix
bugzilla: https://gitee.com/src-openeuler/kernel/issues/I6TI3Y
CVE: NA

Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=e446300968c6bd25d9cd6c33b9600780a39b3975



--------------------------------

commit 4f7e7236 upstream.

Add #include <linux/cpu.h> to avoid compile error on some architectures.

commit 9a3284fa ("cgroup: Optimize single thread migration") and
commit 671c11f0 ("cgroup: Elide write-locking threadgroup_rwsem
when updating csses on an empty subtree") are not backport. So ignore the
input parameter of cgroup_attach_lock/cgroup_attach_unlock.

original commit message:

Bringing up a CPU may involve creating and destroying tasks which requires
read-locking threadgroup_rwsem, so threadgroup_rwsem nests inside
cpus_read_lock(). However, cpuset's ->attach(), which may be called with
thredagroup_rwsem write-locked, also wants to disable CPU hotplug and
acquires cpus_read_lock(), leading to a deadlock.

Fix it by guaranteeing that ->attach() is always called with CPU hotplug
disabled and removing cpus_read_lock() call from cpuset_attach().

Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-and-tested-by: Imran Khan <imran.f.khan@oracle.com>
Reported-and-tested-by: Xuewen Yan <xuewen.yan@unisoc.com>
Fixes: 05c7b7a9 ("cgroup/cpuset: Fix a race between cpuset_attach() and cpu hotplug")
Cc: stable@vger.kernel.org # v5.17+
Signed-off-by: Cai Xinchen <caixinchen1@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Cai Xinchen <caixinchen1@huawei.com>
Reviewed-by: Wang Weiyang <wangweiyang2@huawei.com>
Signed-off-by: Yongqiang Liu <liuyongqiang13@huawei.com>