Re: [PATCH v11] pidfd: add ioctl to retrieve pid info

[kernel test robot <lkp@intel.com>]

Hi,

kernel test robot noticed the following build errors:

[auto build test ERROR on 8cf0b93919e13d1e8d4466eb4080a4c4d9d66d7b]

url:    https://github.com/intel-lab-lkp/linux/commits/luca-boccassi-gmail-com/pidfd-add-ioctl-to-retrieve-pid-info/20241010-235621
base:   8cf0b93919e13d1e8d4466eb4080a4c4d9d66d7b
patch link:    https://lore.kernel.org/r/20241010155401.2268522-1-luca.boccassi%40gmail.com
patch subject: [PATCH v11] pidfd: add ioctl to retrieve pid info
config: alpha-randconfig-r054-20241013 (https://download.01.org/0day-ci/archive/20241013/202410131455.zyQFJN9p-lkp@intel.com/config)
compiler: alpha-linux-gcc (GCC) 13.3.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20241013/202410131455.zyQFJN9p-lkp@intel.com/reproduce)

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/202410131455.zyQFJN9p-lkp@intel.com/

All errors (new ones prefixed by >>):

   fs/pidfs.c: In function 'pidfd_info':
>> fs/pidfs.c:162:34: error: 'pids_cgrp_id' undeclared (first use in this function); did you mean 'misc_cgrp_id'?
     162 |         cgrp = task_cgroup(task, pids_cgrp_id);
         |                                  ^~~~~~~~~~~~
         |                                  misc_cgrp_id
   fs/pidfs.c:162:34: note: each undeclared identifier is reported only once for each function it appears in


vim +162 fs/pidfs.c

   129	
   130		if (!uinfo)
   131			return -EINVAL;
   132		if (usize < sizeof(struct pidfd_info))
   133			return -EINVAL; /* First version, no smaller struct possible */
   134	
   135		if (copy_from_user(&request_mask, &uinfo->request_mask, sizeof(request_mask)))
   136			return -EFAULT;
   137	
   138		c = get_task_cred(task);
   139		if (!c)
   140			return -ESRCH;
   141	
   142		/* Unconditionally return identifiers and credentials, the rest only on request */
   143	
   144		user_ns = current_user_ns();
   145		kinfo.ruid = from_kuid_munged(user_ns, c->uid);
   146		kinfo.rgid = from_kgid_munged(user_ns, c->gid);
   147		kinfo.euid = from_kuid_munged(user_ns, c->euid);
   148		kinfo.egid = from_kgid_munged(user_ns, c->egid);
   149		kinfo.suid = from_kuid_munged(user_ns, c->suid);
   150		kinfo.sgid = from_kgid_munged(user_ns, c->sgid);
   151		kinfo.fsuid = from_kuid_munged(user_ns, c->fsuid);
   152		kinfo.fsgid = from_kgid_munged(user_ns, c->fsgid);
   153		kinfo.request_mask |= PIDFD_INFO_CREDS;
   154		put_cred(c);
   155	
   156	#ifdef CONFIG_CGROUPS
   157		/*
   158		 * The cgroup id cannot be retrieved anymore after the task has exited
   159		 * (even if it has not been reaped yet), contrary to other fields. Set
   160		 * the flag only if we can still access it. */
   161		rcu_read_lock();
 > 162		cgrp = task_cgroup(task, pids_cgrp_id);
   163		if (cgrp) {
   164			kinfo.cgroupid = cgroup_id(cgrp);
   165			kinfo.request_mask |= PIDFD_INFO_CGROUPID;
   166		}
   167		rcu_read_unlock();
   168	#endif
   169	
   170		/*
   171		 * Copy pid/tgid last, to reduce the chances the information might be
   172		 * stale. Note that it is not possible to ensure it will be valid as the
   173		 * task might return as soon as the copy_to_user finishes, but that's ok
   174		 * and userspace expects that might happen and can act accordingly, so
   175		 * this is just best-effort. What we can do however is checking that all
   176		 * the fields are set correctly, or return ESRCH to avoid providing
   177		 * incomplete information. */
   178	
   179		kinfo.ppid = task_ppid_nr_ns(task, NULL);
   180		kinfo.tgid = task_tgid_vnr(task);
   181		kinfo.pid = task_pid_vnr(task);
   182		kinfo.request_mask |= PIDFD_INFO_PID;
   183	
   184		if (kinfo.pid == 0 || kinfo.tgid == 0 || (kinfo.ppid == 0 && kinfo.pid != 1))
   185			return -ESRCH;
   186	
   187		/*
   188		 * If userspace and the kernel have the same struct size it can just
   189		 * be copied. If userspace provides an older struct, only the bits that
   190		 * userspace knows about will be copied. If userspace provides a new
   191		 * struct, only the bits that the kernel knows about will be copied.
   192		 */
   193		if (copy_to_user(uinfo, &kinfo, min(usize, sizeof(kinfo))))
   194			return -EFAULT;
   195	
   196		return 0;
   197	}
   198	

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