[PATCH 2/2] mm_for_maps: take ->cred_guard_mutex to fix the race with exec

[Oleg Nesterov-3]

The problem is minor, but without ->cred_guard_mutex held we can race
with exec() and get the new ->mm but check old creds.

Now we do not need to re-check task->mm after ptrace_may_access(), it
can't be changed to the new mm under us.

Strictly speaking, this also fixes another very minor problem. Unless
security check fails or the task exits mm_for_maps() should never
return NULL, the caller should get either old or new ->mm.

Signed-off-by: Oleg Nesterov <oleg@...>
---

 fs/proc/base.c |   22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

--- WAIT/fs/proc/base.c~2_CRED_MUTEX 2009-07-10 02:05:14.000000000 +0200
+++ WAIT/fs/proc/base.c 2009-07-10 03:23:01.000000000 +0200
@@ -234,19 +234,19 @@ static int check_mem_permission(struct t
 
 struct mm_struct *mm_for_maps(struct task_struct *task)
 {
- struct mm_struct *mm = get_task_mm(task);
+ struct mm_struct *mm;
 
- if (mm && mm != current->mm) {
- /*
- * task->mm can be changed before security check,
- * in that case we must notice the change after.
- */
- if (!ptrace_may_access(task, PTRACE_MODE_READ) ||
-    mm != task->mm) {
- mmput(mm);
- mm = NULL;
- }
+ if (mutex_lock_killable(&task->cred_guard_mutex))
+ return NULL;
+
+ mm = get_task_mm(task);
+ if (mm && mm != current->mm &&
+ !ptrace_may_access(task, PTRACE_MODE_READ)) {
+ mmput(mm);
+ mm = NULL;
  }
+ mutex_unlock(&task->cred_guard_mutex);
+
  return mm;
 }
 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[Serge E. Hallyn]

Quoting Oleg Nesterov (oleg@...):

> The problem is minor, but without ->cred_guard_mutex held we can race
> with exec() and get the new ->mm but check old creds.
>
> Now we do not need to re-check task->mm after ptrace_may_access(), it
> can't be changed to the new mm under us.
>
> Strictly speaking, this also fixes another very minor problem. Unless
> security check fails or the task exits mm_for_maps() should never
> return NULL, the caller should get either old or new ->mm.
>
> Signed-off-by: Oleg Nesterov <oleg@...>

Acked-by: Serge Hallyn <serue@...>

> ---
>
>  fs/proc/base.c |   22 +++++++++++-----------
>  1 file changed, 11 insertions(+), 11 deletions(-)
>
> --- WAIT/fs/proc/base.c~2_CRED_MUTEX 2009-07-10 02:05:14.000000000 +0200
> +++ WAIT/fs/proc/base.c 2009-07-10 03:23:01.000000000 +0200
> @@ -234,19 +234,19 @@ static int check_mem_permission(struct t
>
>  struct mm_struct *mm_for_maps(struct task_struct *task)
>  {
> - struct mm_struct *mm = get_task_mm(task);
> + struct mm_struct *mm;
>
> - if (mm && mm != current->mm) {
> - /*
> - * task->mm can be changed before security check,
> - * in that case we must notice the change after.
> - */
> - if (!ptrace_may_access(task, PTRACE_MODE_READ) ||
> -    mm != task->mm) {
> - mmput(mm);
> - mm = NULL;
> - }
> + if (mutex_lock_killable(&task->cred_guard_mutex))
> + return NULL;
> +
> + mm = get_task_mm(task);
> + if (mm && mm != current->mm &&
> + !ptrace_may_access(task, PTRACE_MODE_READ)) {
> + mmput(mm);
> + mm = NULL;
>   }
> + mutex_unlock(&task->cred_guard_mutex);
> +
>   return mm;
>  }
>

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/