Re: 2.4+ptrace exploit fix breaks root's ability to strace

[Russell King]

On Sat, Mar 22, 2003 at 02:58:51PM +0000, Alan Cox wrote:
> On Sat, 2003-03-22 at 10:31, Russell King wrote:
> > Are the authors of the ptrace patch aware that, in addition to closing the
> > hole, the "fix" also prevents a ptrace-capable task (eg, strace started by
> > root) from ptracing user threads?
> > 
> > For example, you can't strace vsftpd processes started from xinetd.
> > 
> > Is this intended behaviour?
> 
> Its an unintended side effect, nobody has sent a patch to fix it yet.

How about this fix?  PT_PTRACE_CAP is set when we attach to a process
and the process doing the attaching has the ptrace capability.

Note that this patch is against the 2.4.19 version of ptrace.c with the
2.4.20 ptrace patch applied.

--- orig/kernel/ptrace.c	Wed Mar 19 15:54:45 2003
+++ linux/kernel/ptrace.c	Sat Mar 22 10:14:01 2003
@@ -22,7 +22,7 @@
 int ptrace_check_attach(struct task_struct *child, int kill)
 {
 	mb();
-	if (!is_dumpable(child))
+	if (!is_dumpable(child) && !(child->ptrace & PT_PTRACE_CAP))
 		return -EPERM;
 
 	if (!(child->ptrace & PT_PTRACED))
@@ -140,7 +140,7 @@
 	/* Worry about races with exit() */
 	task_lock(tsk);
 	mm = tsk->mm;
-	if (!is_dumpable(tsk) || (&init_mm == mm))
+	if ((!is_dumpable(tsk) || (&init_mm == mm)) && !(tsk->ptrace & PT_PTRACE_CAP))
 		mm = NULL;
 	if (mm)
 		atomic_inc(&mm->mm_users);

-- 
Russell King ([e-mail])                The developer of ARM Linux
             http://www.arm.linux.org.uk/personal/aboutme.html