locks: Filter /proc/locks output on proc pid ns

On busy container servers reading /proc/locks shows all the locks
created by all clients. This can cause large latency spikes. In my
case I observed lsof taking up to 5-10 seconds while processing around
50k locks. Fix this by limiting the locks shown only to those created
in the same pidns as the one the proc fs was mounted in. When reading
/proc/locks from the init_pid_ns proc instance then perform no
filtering

[ jlayton: reformat comments for 80 columns ]

Signed-off-by: Nikolay Borisov <kernel@kyup.com>
Suggested-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>

diff --git a/fs/locks.c b/fs/locks.c
index ee1b15f6fc135c33e2e0564eb5e10db927080533..7e428b78be074506b4cf6ca03b15c16497b2abbb 100644 (file)
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -2574,9 +2574,20 @@ static void lock_get_status(struct seq_file *f, struct file_lock *fl,
        struct inode *inode = NULL;
        unsigned int fl_pid;
 
-       if (fl->fl_nspid)
-               fl_pid = pid_vnr(fl->fl_nspid);
-       else
+       if (fl->fl_nspid) {
+               struct pid_namespace *proc_pidns = file_inode(f->file)->i_sb->s_fs_info;
+
+               /* Don't let fl_pid change based on who is reading the file */
+               fl_pid = pid_nr_ns(fl->fl_nspid, proc_pidns);
+
+               /*
+                * If there isn't a fl_pid don't display who is waiting on
+                * the lock if we are called from locks_show, or if we are
+                * called from __show_fd_info - skip lock entirely
+                */
+               if (fl_pid == 0)
+                       return;
+       } else
                fl_pid = fl->fl_pid;
 
        if (fl->fl_file != NULL)
@@ -2648,9 +2659,13 @@ static int locks_show(struct seq_file *f, void *v)
 {
        struct locks_iterator *iter = f->private;
        struct file_lock *fl, *bfl;
+       struct pid_namespace *proc_pidns = file_inode(f->file)->i_sb->s_fs_info;
 
        fl = hlist_entry(v, struct file_lock, fl_link);
 
+       if (fl->fl_nspid && !pid_nr_ns(fl->fl_nspid, proc_pidns))
+               return 0;
+
        lock_get_status(f, fl, iter->li_pos, "");
 
        list_for_each_entry(bfl, &fl->fl_block, fl_block)