[PATCH 3/4] proc: Follow p_parent to root for checking startup process

[Damien Zammit]

Since the bootstrap filesystem is no longer a single process,
but can be spread among eg, pci-arbiter and rumpdisk, we cannot
just check the p_parent of the process but need to follow
to the root of the process namespace to check if it belongs to startup.
---
 proc/mgt.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/proc/mgt.c b/proc/mgt.c
index d92bf528..01510f43 100644
--- a/proc/mgt.c
+++ b/proc/mgt.c
@@ -1109,15 +1109,17 @@ S_proc_mark_important (struct proc *p)
   if (!p)
     return EOPNOTSUPP;
 
-  /* Only root may use this interface.  Any children of startup_proc
+  /* Only root may use this interface.  Any bootstrapping processes are
      exempt from this restriction, as startup_proc calls this on their
      behalf.  The kernel process is a notable example of an process
      that needs this exemption.  That is not an problem however, since
-     all children of /hurd/startup are important and we mark them as
-     such anyway.  */
-  if (! check_uid (p, 0) && p->p_parent != startup_proc)
-    return EPERM;
+     all important processes are marked as such anyway.  */
 
+  if (! check_uid (p, 0) && namespace_find_root(p)->p_pid != HURD_PID_STARTUP)
+    {
+      mach_print("proc_mark_important: EPERM (UID!=0 and is not a startup 
process)\n");
+      return EPERM;
+    }
   p->p_important = 1;
   return 0;
 }
-- 
2.31.0