tree: https://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs.git work.mount head: b504486c4efe75e1ac70e0df3f193aa1b33de664 commit: 9f0549cbf6fffe7894735b4fe33691ab5f6d008b [21/24] [experimental] get rid of ->m_count config: openrisc-allnoconfig (https://download.01.org/0day-ci/archive/20250508/202505081831.05K4QvQQ-lkp@xxxxxxxxx/config) compiler: or1k-linux-gcc (GCC) 14.2.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250508/202505081831.05K4QvQQ-lkp@xxxxxxxxx/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@xxxxxxxxx> | Closes: https://lore.kernel.org/oe-kbuild-all/202505081831.05K4QvQQ-lkp@xxxxxxxxx/ All warnings (new ones prefixed by >>): >> fs/namespace.c:2783: warning: Function parameter or struct member 'pinned' not described in 'do_lock_mount' vim +2783 fs/namespace.c b90fa9ae8f51f0 Ram Pai 2005-11-07 2747 6ac392815628f3 Christian Brauner 2023-05-03 2748 /** 6ac392815628f3 Christian Brauner 2023-05-03 2749 * do_lock_mount - lock mount and mountpoint 6ac392815628f3 Christian Brauner 2023-05-03 2750 * @path: target path 6ac392815628f3 Christian Brauner 2023-05-03 2751 * @beneath: whether the intention is to mount beneath @path 6ac392815628f3 Christian Brauner 2023-05-03 2752 * 6ac392815628f3 Christian Brauner 2023-05-03 2753 * Follow the mount stack on @path until the top mount @mnt is found. If 6ac392815628f3 Christian Brauner 2023-05-03 2754 * the initial @path->{mnt,dentry} is a mountpoint lookup the first 6ac392815628f3 Christian Brauner 2023-05-03 2755 * mount stacked on top of it. Then simply follow @{mnt,mnt->mnt_root} 6ac392815628f3 Christian Brauner 2023-05-03 2756 * until nothing is stacked on top of it anymore. 6ac392815628f3 Christian Brauner 2023-05-03 2757 * 6ac392815628f3 Christian Brauner 2023-05-03 2758 * Acquire the inode_lock() on the top mount's ->mnt_root to protect 6ac392815628f3 Christian Brauner 2023-05-03 2759 * against concurrent removal of the new mountpoint from another mount 6ac392815628f3 Christian Brauner 2023-05-03 2760 * namespace. 6ac392815628f3 Christian Brauner 2023-05-03 2761 * 6ac392815628f3 Christian Brauner 2023-05-03 2762 * If @beneath is requested, acquire inode_lock() on @mnt's mountpoint 6ac392815628f3 Christian Brauner 2023-05-03 2763 * @mp on @mnt->mnt_parent must be acquired. This protects against a 6ac392815628f3 Christian Brauner 2023-05-03 2764 * concurrent unlink of @mp->mnt_dentry from another mount namespace 6ac392815628f3 Christian Brauner 2023-05-03 2765 * where @mnt doesn't have a child mount mounted @mp. A concurrent 6ac392815628f3 Christian Brauner 2023-05-03 2766 * removal of @mnt->mnt_root doesn't matter as nothing will be mounted 6ac392815628f3 Christian Brauner 2023-05-03 2767 * on top of it for @beneath. 6ac392815628f3 Christian Brauner 2023-05-03 2768 * 6ac392815628f3 Christian Brauner 2023-05-03 2769 * In addition, @beneath needs to make sure that @mnt hasn't been 6ac392815628f3 Christian Brauner 2023-05-03 2770 * unmounted or moved from its current mountpoint in between dropping 6ac392815628f3 Christian Brauner 2023-05-03 2771 * @mount_lock and acquiring @namespace_sem. For the !@beneath case @mnt 6ac392815628f3 Christian Brauner 2023-05-03 2772 * being unmounted would be detected later by e.g., calling 6ac392815628f3 Christian Brauner 2023-05-03 2773 * check_mnt(mnt) in the function it's called from. For the @beneath 6ac392815628f3 Christian Brauner 2023-05-03 2774 * case however, it's useful to detect it directly in do_lock_mount(). 6ac392815628f3 Christian Brauner 2023-05-03 2775 * If @mnt hasn't been unmounted then @mnt->mnt_mountpoint still points 6ac392815628f3 Christian Brauner 2023-05-03 2776 * to @mnt->mnt_mp->m_dentry. But if @mnt has been unmounted it will 6ac392815628f3 Christian Brauner 2023-05-03 2777 * point to @mnt->mnt_root and @mnt->mnt_mp will be NULL. 6ac392815628f3 Christian Brauner 2023-05-03 2778 * 6ac392815628f3 Christian Brauner 2023-05-03 2779 * Return: Either the target mountpoint on the top mount or the top 6ac392815628f3 Christian Brauner 2023-05-03 2780 * mount's mountpoint. 6ac392815628f3 Christian Brauner 2023-05-03 2781 */ 9f0549cbf6fffe Al Viro 2025-04-25 2782 static int do_lock_mount(struct path *path, struct pinned_mountpoint *pinned, bool beneath) b12cea9198fa99 Al Viro 2011-03-18 @2783 { 6ac392815628f3 Christian Brauner 2023-05-03 2784 struct vfsmount *mnt = path->mnt; 64f44b27ae9184 Christian Brauner 2023-05-03 2785 struct dentry *dentry; 0d039eac6e5950 Al Viro 2025-04-23 2786 struct path under = {}; 9f0549cbf6fffe Al Viro 2025-04-25 2787 int err = -ENOENT; 64f44b27ae9184 Christian Brauner 2023-05-03 2788 64f44b27ae9184 Christian Brauner 2023-05-03 2789 for (;;) { 0d039eac6e5950 Al Viro 2025-04-23 2790 struct mount *m = real_mount(mnt); 6ac392815628f3 Christian Brauner 2023-05-03 2791 6ac392815628f3 Christian Brauner 2023-05-03 2792 if (beneath) { 0d039eac6e5950 Al Viro 2025-04-23 2793 path_put(&under); 6ac392815628f3 Christian Brauner 2023-05-03 2794 read_seqlock_excl(&mount_lock); 0d039eac6e5950 Al Viro 2025-04-23 2795 under.mnt = mntget(&m->mnt_parent->mnt); 0d039eac6e5950 Al Viro 2025-04-23 2796 under.dentry = dget(m->mnt_mountpoint); 6ac392815628f3 Christian Brauner 2023-05-03 2797 read_sequnlock_excl(&mount_lock); 0d039eac6e5950 Al Viro 2025-04-23 2798 dentry = under.dentry; 6ac392815628f3 Christian Brauner 2023-05-03 2799 } else { 64f44b27ae9184 Christian Brauner 2023-05-03 2800 dentry = path->dentry; 6ac392815628f3 Christian Brauner 2023-05-03 2801 } 6ac392815628f3 Christian Brauner 2023-05-03 2802 5955102c9984fa Al Viro 2016-01-22 2803 inode_lock(dentry->d_inode); 97216be09efd41 Al Viro 2013-03-16 2804 namespace_lock(); 64f44b27ae9184 Christian Brauner 2023-05-03 2805 0d039eac6e5950 Al Viro 2025-04-23 2806 if (unlikely(cant_mount(dentry) || !is_mounted(mnt))) 0d039eac6e5950 Al Viro 2025-04-23 2807 break; // not to be mounted on 0d039eac6e5950 Al Viro 2025-04-23 2808 0d039eac6e5950 Al Viro 2025-04-23 2809 if (beneath && unlikely(m->mnt_mountpoint != dentry || 0d039eac6e5950 Al Viro 2025-04-23 2810 &m->mnt_parent->mnt != under.mnt)) { 6ac392815628f3 Christian Brauner 2023-05-03 2811 namespace_unlock(); 6ac392815628f3 Christian Brauner 2023-05-03 2812 inode_unlock(dentry->d_inode); 0d039eac6e5950 Al Viro 2025-04-23 2813 continue; // got moved 6ac392815628f3 Christian Brauner 2023-05-03 2814 } 6ac392815628f3 Christian Brauner 2023-05-03 2815 b12cea9198fa99 Al Viro 2011-03-18 2816 mnt = lookup_mnt(path); 0d039eac6e5950 Al Viro 2025-04-23 2817 if (unlikely(mnt)) { 64f44b27ae9184 Christian Brauner 2023-05-03 2818 namespace_unlock(); 64f44b27ae9184 Christian Brauner 2023-05-03 2819 inode_unlock(dentry->d_inode); 64f44b27ae9184 Christian Brauner 2023-05-03 2820 path_put(path); 64f44b27ae9184 Christian Brauner 2023-05-03 2821 path->mnt = mnt; 64f44b27ae9184 Christian Brauner 2023-05-03 2822 path->dentry = dget(mnt->mnt_root); 0d039eac6e5950 Al Viro 2025-04-23 2823 continue; // got overmounted 64f44b27ae9184 Christian Brauner 2023-05-03 2824 } 9f0549cbf6fffe Al Viro 2025-04-25 2825 err = get_mountpoint(dentry, pinned); 9f0549cbf6fffe Al Viro 2025-04-25 2826 if (err) 0d039eac6e5950 Al Viro 2025-04-23 2827 break; 0d039eac6e5950 Al Viro 2025-04-23 2828 if (beneath) { 0d039eac6e5950 Al Viro 2025-04-23 2829 /* 0d039eac6e5950 Al Viro 2025-04-23 2830 * @under duplicates the references that will stay 0d039eac6e5950 Al Viro 2025-04-23 2831 * at least until namespace_unlock(), so the path_put() 0d039eac6e5950 Al Viro 2025-04-23 2832 * below is safe (and OK to do under namespace_lock - 0d039eac6e5950 Al Viro 2025-04-23 2833 * we are not dropping the final references here). 0d039eac6e5950 Al Viro 2025-04-23 2834 */ 0d039eac6e5950 Al Viro 2025-04-23 2835 path_put(&under); 0d039eac6e5950 Al Viro 2025-04-23 2836 } 9f0549cbf6fffe Al Viro 2025-04-25 2837 return 0; 0d039eac6e5950 Al Viro 2025-04-23 2838 } 97216be09efd41 Al Viro 2013-03-16 2839 namespace_unlock(); 5955102c9984fa Al Viro 2016-01-22 2840 inode_unlock(dentry->d_inode); 6ac392815628f3 Christian Brauner 2023-05-03 2841 if (beneath) 0d039eac6e5950 Al Viro 2025-04-23 2842 path_put(&under); 9f0549cbf6fffe Al Viro 2025-04-25 2843 return err; 84d17192d2afd5 Al Viro 2013-03-15 2844 } b12cea9198fa99 Al Viro 2011-03-18 2845 :::::: The code at line 2783 was first introduced by commit :::::: b12cea9198fa99ffd3de1776c323bc7464d26b44 change the locking order for namespace_sem :::::: TO: Al Viro <viro@xxxxxxxxxxxxxxxxxx> :::::: CC: Al Viro <viro@xxxxxxxxxxxxxxxxxx> -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki
