tree: https://gitee.com/openeuler/kernel.git openEuler-1.0-LTS head: 26d81348fb134a6bffb49fd1b55d8842cbf6424d commit: 4924308a1ca9cc2f791398836a8744c22078ffbd [1605/1605] cgroup: Fix threadgroup_rwsem <-> cpus_read_lock() deadlock config: x86_64-buildonly-randconfig-2004-20250501 (https://download.01.org/0day-ci/archive/20250703/202507031346.HRMtINKf-lkp@i...) compiler: clang version 20.1.2 (https://github.com/llvm/llvm-project 58df0ef89dd64126512e4ee27b4ac3fd8ddf6247) reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250703/202507031346.HRMtINKf-lkp@i...) If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot <lkp@intel.com> | Closes: https://lore.kernel.org/oe-kbuild-all/202507031346.HRMtINKf-lkp@intel.com/ All warnings (new ones prefixed by >>):

...

kernel/cgroup/cgroup.c:2240: warning: Excess function parameter 'lock_threadgroup' description in 'cgroup_attach_lock' kernel/cgroup/cgroup.c:2250: warning: Excess function parameter 'lock_threadgroup' description in 'cgroup_attach_unlock'

vim +2240 kernel/cgroup/cgroup.c 2215 2216 /** 2217 * cgroup_attach_lock - Lock for ->attach() 2218 * @lock_threadgroup: whether to down_write cgroup_threadgroup_rwsem 2219 * 2220 * cgroup migration sometimes needs to stabilize threadgroups against forks and 2221 * exits by write-locking cgroup_threadgroup_rwsem. However, some ->attach() 2222 * implementations (e.g. cpuset), also need to disable CPU hotplug. 2223 * Unfortunately, letting ->attach() operations acquire cpus_read_lock() can 2224 * lead to deadlocks. 2225 * 2226 * Bringing up a CPU may involve creating and destroying tasks which requires 2227 * read-locking threadgroup_rwsem, so threadgroup_rwsem nests inside 2228 * cpus_read_lock(). If we call an ->attach() which acquires the cpus lock while 2229 * write-locking threadgroup_rwsem, the locking order is reversed and we end up 2230 * waiting for an on-going CPU hotplug operation which in turn is waiting for 2231 * the threadgroup_rwsem to be released to create new tasks. For more details: 2232 * 2233 * http://lkml.kernel.org/r/20220711174629.uehfmqegcwn2lqzu@wubuntu 2234 * 2235 * Resolve the situation by always acquiring cpus_read_lock() before optionally 2236 * write-locking cgroup_threadgroup_rwsem. This allows ->attach() to assume that 2237 * CPU hotplug is disabled on entry. 2238 */ 2239 static void cgroup_attach_lock(void)

2240 { 2241 cpus_read_lock(); 2242 percpu_down_write(&cgroup_threadgroup_rwsem); 2243 } 2244 2245 /** 2246 * cgroup_attach_unlock - Undo cgroup_attach_lock() 2247 * @lock_threadgroup: whether to up_write cgroup_threadgroup_rwsem 2248 */ 2249 static void cgroup_attach_unlock(void) 2250 { 2251 percpu_up_write(&cgroup_threadgroup_rwsem); 2252 cpus_read_unlock(); 2253 } 2254

-- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki