proc: fix the threaded /proc/self

Long ago when the CLONE_THREAD support first went it someone thought it
would be wise to point /proc/self at /proc/<tgid> instead of /proc/<pid>.

Given that /proc/<tgid> can return information about a very different task
(if enough things have been unshared) then our current process /proc/<tgid>
seems blatantly wrong.  So far I have yet to think up an example where the
current behavior would be advantageous, and I can see several places where
it is seriously non-intuitive.

We may be stuck with the current broken behavior for backwards
compatibility reasons but lets try fixing our ancient bug for the 2.6.25
time frame and see if anyone screams.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: "Guillaume Chazarain" <guichaz@yahoo.fr>
Cc: "Pavel Emelyanov" <xemul@openvz.org>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

fs/proc/base.c

@@ -2102,22 +2102,22 @@ static int proc_self_readlink(struct dentry *dentry, char __user *buffer,
 			      int buflen)
 {
 	struct pid_namespace *ns = dentry->d_sb->s_fs_info;
-	pid_t tgid = task_tgid_nr_ns(current, ns);
+	pid_t pid = task_pid_nr_ns(current, ns);
 	char tmp[PROC_NUMBUF];
-	if (!tgid)
+	if (!pid)
 		return -ENOENT;
-	sprintf(tmp, "%d", tgid);
+	sprintf(tmp, "%d", pid);
 	return vfs_readlink(dentry,buffer,buflen,tmp);
 }
 
 static void *proc_self_follow_link(struct dentry *dentry, struct nameidata *nd)
 {
 	struct pid_namespace *ns = dentry->d_sb->s_fs_info;
-	pid_t tgid = task_tgid_nr_ns(current, ns);
+	pid_t pid = task_pid_nr_ns(current, ns);
 	char tmp[PROC_NUMBUF];
-	if (!tgid)
+	if (!pid)
 		return ERR_PTR(-ENOENT);
-	sprintf(tmp, "%d", task_tgid_nr_ns(current, ns));
+	sprintf(tmp, "%d", pid);
 	return ERR_PTR(vfs_follow_link(nd,tmp));
 }