Re: on CONFIG_MM_OWNER=y, kernel panic is possible.

[Balbir Singh <balbir@linux.vnet.ibm.com>]

KOSAKI Motohiro wrote:
>>> That is not possible. If you look at where mm_update_next_owner() is called
>>> from, we call it from
>>>
>>> exit_mm() and exec_mmap()
>>>
>>> In both cases, we ensure that the task's mm has changed (to NULL and the new mm
>>> respectively), before we call mm_update_next_owner(), hence c->mm can never be
>>> equal to p->mm.
>> if so, following patch is needed instead.
> 
> and, one more.
> 
> comment of owner member of mm_struct is bogus.
> that is not guranteed point to thread-group-leader.
> 
> 
> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
> 
> ---
>  include/linux/mm_types.h |    3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> Index: b/include/linux/mm_types.h
> ===================================================================
> --- a/include/linux/mm_types.h  2008-05-04 22:56:52.000000000 +0900
> +++ b/include/linux/mm_types.h  2008-05-06 15:53:04.000000000 +0900
> @@ -231,8 +231,7 @@ struct mm_struct {
>         rwlock_t                ioctx_list_lock;        /* aio lock */
>         struct kioctx           *ioctx_list;
>  #ifdef CONFIG_MM_OWNER
> -       struct task_struct *owner;      /* The thread group leader that */
> -                                       /* owns the mm_struct.          */
> +       struct task_struct *owner;      /* point to one of task that owns the mm_struct. */
>  #endif
> 
>  #ifdef CONFIG_PROC_FS
> 
> 
> 

How about just, the task that owns the mm_struct? One of, implies multiple owners.


-- 
	Warm Regards,
	Balbir Singh
	Linux Technology Center
	IBM, ISTL