Re: [PATCH -V2] mm, swap: Remove unnecessary smp_rmb() in swap_type_to_swap_info()

[Huang, Ying]

Hi, Will,

Will Deacon <will@xxxxxxxxxx> writes:

> On Thu, May 20, 2021 at 03:33:01PM +0800, Huang Ying wrote:
>> Before commit c10d38cc8d3e ("mm, swap: bounds check swap_info array
>> accesses to avoid NULL derefs"), the typical code to reference the
>> swap_info[] is as follows,
>> 
>>   type = swp_type(swp_entry);
>>   if (type >= nr_swapfiles)
>>           /* handle invalid swp_entry */;
>>   p = swap_info[type];
>>   /* access fields of *p.  OOPS! p may be NULL! */
>> 
>> Because the ordering isn't guaranteed, it's possible that
>> swap_info[type] is read before "nr_swapfiles".  And that may result
>> in NULL pointer dereference.
>> 
>> So after commit c10d38cc8d3e, the code becomes,
>> 
>>   struct swap_info_struct *swap_type_to_swap_info(int type)
>>   {
>> 	  if (type >= READ_ONCE(nr_swapfiles))
>> 		  return NULL;
>> 	  smp_rmb();
>> 	  return READ_ONCE(swap_info[type]);
>>   }
>> 
>>   /* users */
>>   type = swp_type(swp_entry);
>>   p = swap_type_to_swap_info(type);
>>   if (!p)
>> 	  /* handle invalid swp_entry */;
>>   /* dereference p */
>> 
>> Where the value of swap_info[type] (that is, "p") is checked to be
>> non-zero before being dereferenced.  So, the NULL deferencing
>> becomes impossible even if "nr_swapfiles" is read after
>> swap_info[type].  Therefore, the "smp_rmb()" becomes unnecessary.
>> 
>> And, we don't even need to read "nr_swapfiles" here.  Because the
>> non-zero checking for "p" is sufficient.  We just need to make sure we
>> will not access out of the boundary of the array.  With the change,
>> nr_swapfiles will only be accessed with swap_lock held, except in
>> swapcache_free_entries().  Where the absolute correctness of the value
>> isn't needed, as described in the comments.
>> 
>> We still need to guarantee swap_info[type] is read before being
>> dereferenced.  That can be satisfied via the data dependency ordering
>> enforced by READ_ONCE(swap_info[type]).  This needs to be paired with
>> proper write barriers.  So smp_store_release() is used in
>> alloc_swap_info() to guarantee the fields of *swap_info[type] is
>> initialized before swap_info[type] itself being written.  Note that
>> the fields of *swap_info[type] is initialized to be 0 via kvzalloc()
>> firstly.  The assignment and deferencing of swap_info[type] is like
>> rcu_assign_pointer() and rcu_dereference().
>> 
>> Signed-off-by: "Huang, Ying" <ying.huang@xxxxxxxxx>
>> Cc: Daniel Jordan <daniel.m.jordan@xxxxxxxxxx>
>> Cc: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
>> Cc: Andrea Parri <andrea.parri@xxxxxxxxxxxxxxxxxxxx>
>> Cc: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
>> Cc: Andi Kleen <ak@xxxxxxxxxxxxxxx>
>> Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
>> Cc: Omar Sandoval <osandov@xxxxxx>
>> Cc: Paul McKenney <paulmck@xxxxxxxxxx>
>> Cc: Tejun Heo <tj@xxxxxxxxxx>
>> Cc: Will Deacon <will.deacon@xxxxxxx>
>> Cc: Miaohe Lin <linmiaohe@xxxxxxxxxx>
>> 
>> v2:
>> 
>> - Revise the patch description and comments per Peter's comments.
>> 
>> ---
>>  mm/swapfile.c | 15 ++++++---------
>>  1 file changed, 6 insertions(+), 9 deletions(-)
>> 
>> diff --git a/mm/swapfile.c b/mm/swapfile.c
>> index 2aad85751991..65dd979a0f94 100644
>> --- a/mm/swapfile.c
>> +++ b/mm/swapfile.c
>> @@ -100,11 +100,10 @@ atomic_t nr_rotate_swap = ATOMIC_INIT(0);
>>  
>>  static struct swap_info_struct *swap_type_to_swap_info(int type)
>>  {
>> -	if (type >= READ_ONCE(nr_swapfiles))
>> +	if (type >= MAX_SWAPFILES)
>>  		return NULL;
>>  
>> -	smp_rmb();	/* Pairs with smp_wmb in alloc_swap_info. */
>> -	return READ_ONCE(swap_info[type]);
>> +	return READ_ONCE(swap_info[type]); /* rcu_dereference() */
>>  }
>>  
>>  static inline unsigned char swap_count(unsigned char ent)
>> @@ -2884,14 +2883,12 @@ static struct swap_info_struct *alloc_swap_info(void)
>>  	}
>>  	if (type >= nr_swapfiles) {
>>  		p->type = type;
>> -		WRITE_ONCE(swap_info[type], p);
>>  		/*
>> -		 * Write swap_info[type] before nr_swapfiles, in case a
>> -		 * racing procfs swap_start() or swap_next() is reading them.
>> -		 * (We never shrink nr_swapfiles, we never free this entry.)
>> +		 * Publish the swap_info_struct after initializing it.
>> +		 * Note that kvzalloc() above zeroes all its fields.
>>  		 */
>> -		smp_wmb();
>> -		WRITE_ONCE(nr_swapfiles, nr_swapfiles + 1);
>> +		smp_store_release(&swap_info[type], p); /* rcu_assign_pointer() */
>> +		nr_swapfiles++;
>
> Although I like this change, I comment you are removing refers to some
> dodgy-looking code. For example, swap_start() has this loop:
>
> 	for (type = 0; (si = swap_type_to_swap_info(type)); type++) {
> 		if (!(si->flags & SWP_USED) || !si->swap_map)
> 			continue;
>
> so won't this just end up dereferencing NULL if nr_swapfiles < MAX_SWAPFILES?

 	for (type = 0; (si = swap_type_to_swap_info(type)); type++) {
                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Because this is the second sub-statement inside "for ()", I think that "si"
will be checked to be non-NULL before executing the statements inside
"{}" follows "for ()"?


> I think you need to check all callers of swap_type_to_swap_info() are
> either validating the 'type' they pass in or check the returned pointer.

Yes.  I have checked all callers of swap_type_to_swap_info().  One
suspecting caller is swap_cluster_readahead().  But after the following
patch in mmotm tree,

mm/swapfile: use percpu_ref to serialize against concurrent swapoff

get/put_swap_device() will enclose the swap_cluster_readahead() to
check the swap entry beforehand.

Best Regards,
Huang, Ying