Re: [PATCH] ipc/util.c: sysvipc_find_ipc() incorrectly updates position index

[Vasily Averin <vvs@virtuozzo.com>]

On 5/12/20 12:21 PM, Jiri Slaby wrote:
> On 08. 05. 20, 12:01, Vasily Averin wrote:
>> On 5/8/20 9:07 AM, Vasily Averin wrote:
>>> On 5/8/20 6:36 AM, Matthew Wilcox wrote:
>>>> On Thu, May 07, 2020 at 05:02:42PM -0700, Andrew Morton wrote:
>>>>> Here's how I resolved things.  Please check?
>>>>>
>>>>> static struct kern_ipc_perm *sysvipc_find_ipc(struct ipc_ids *ids, loff_t pos,
>>>>> 					      loff_t *new_pos)
>>>>> {
>>>>> 	unsigned long index = pos;
>>>>> 	struct kern_ipc_perm *ipc;
>>>>>
>>>>> 	rcu_read_lock();
>>>>> 	ipc = xa_find(&ids->ipcs, &index, ULONG_MAX, XA_PRESENT);
>>>>> 	if (ipc)
>>>>> 		ipc_lock_object(ipc);
>>>>> 	else
>>>>> 		rcu_read_unlock();
>>>>> 	*new_pos = pos + 1;
>>>>> 	return ipc;
>>>>> }
>>>>
>>>> Surely that should be '*new_pos = index + 1'?  Or did I misunderstand
>>>> the reasoning behind the other patch?
>>>
>>> I'm not sure however it looks like xa_find() can return index < pos
>> it seems, I was wrong here.
>> So I'm agree with Matthew, '*new_pos = index + 1' should be used.
> 
> Any progress on this? 5.7-rc*, 5.4.40, and 5.6.12 are still affected.

Andrew included fix to -mm tree and I hope he'll push it to mainline/stable soon.
https://ozlabs.org/~akpm/mmots/broken-out/ipc-utilc-sysvipc_find_ipc-incorrectly-updates-position-index.patch

> Wouldn't it be better to rebase (apply the originally submitted patch)
> before the XA rewrite and push that one to Linus?
I'm expecting thins too. 

Thank you,
	Vasily Averin