Re: [PATCH] futex:fix robust futex alignment exception

[Peter Zijlstra]

On Fri, Mar 15, 2019 at 03:44:38AM +0000, chenjie6@xxxxxxxxxx wrote:
> From: chen jie <chenjie6@xxxxxxxxxx>

> [11542.215247] [<c017b1d4>] (cmpxchg_futex_value_locked) from [<c017da50>] (handle_futex_death+0x78/0xcc)
> [11542.215259] [<c017da50>] (handle_futex_death) from [<c017dba8>] (exit_robust_list+0x104/0x160)
> [11542.215273] [<c017dba8>] (exit_robust_list) from [<c011b40c>] (mm_release+0x1c/0x108)
> [11542.215287] [<c011b40c>] (mm_release) from [<c011f9a4>] (do_exit+0x218/0x9a4)
> [11542.215299] [<c011f9a4>] (do_exit) from [<c0120250>] (do_group_exit+0xac/0xf4)
> [11542.215311] [<c0120250>] (do_group_exit) from [<c01202a8>] (__wake_up_parent+0x0/0x18)

> Signed-off-by: chen jie <chenjie6@xxxxxxxxxx>

Reviewed-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>

However, should there not also be alignment tests on set_robust_list()?

Also; do_futex() should probably check uaddr and uaddr2.

That is; why aren't there any alignment tests anywhere? Or am I just
gone blind?

> ---
>  kernel/futex.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/kernel/futex.c b/kernel/futex.c
> index a0514e0..70231c4 100644
> --- a/kernel/futex.c
> +++ b/kernel/futex.c
> @@ -3440,6 +3440,9 @@ static int handle_futex_death(u32 __user *uaddr, struct task_struct *curr, int p
>  {
>  	u32 uval, uninitialized_var(nval), mval;
>  
> +	if (((unsigned long)uaddr & 0x3) > 0)
> +		return -1;
> +
>  retry:
>  	if (get_user(uval, uaddr))
>  		return -1;
> -- 
> 1.8.3.4
>