Re: [PATCH net] sctp: check duplicate node before inserting a new transport

[Xin Long <lucien.xin@gmail.com>]

On Fri, Feb 17, 2017 at 7:19 PM, Neil Horman <nhorman@tuxdriver.com> wrote:
> On Fri, Feb 17, 2017 at 04:35:24PM +0800, Xin Long wrote:
>> sctp has changed to use rhlist for transport rhashtable since commit
>> 7fda702f9315 ("sctp: use new rhlist interface on sctp transport
>> rhashtable").
>>
>> But rhltable_insert_key doesn't check the duplicate node when inserting
>> a node, unlike rhashtable_lookup_insert_key. It may cause duplicate
>> assoc/transport in rhashtable. like:
>>
>>  client (addr A, B)                 server (addr X, Y)
>>     connect to X           INIT (1)
>>                         ------------>
>>     connect to Y           INIT (2)
>>                         ------------>
>>                          INIT_ACK (1)
>>                         <------------
>>                          INIT_ACK (2)
>>                         <------------
>>
>> After sending INIT (2), one transport will be created and hashed into
>> rhashtable. But when receiving INIT_ACK (1) and processing the address
>> params, another transport will be created and hashed into rhashtable
>> with the same addr Y and EP as the last transport. This will confuse
>> the assoc/transport's lookup.
>>
>> This patch is to fix it by returning err if any duplicate node exists
>> before inserting it.
>>
>> Fixes: 7fda702f9315 ("sctp: use new rhlist interface on sctp transport rhashtable")
>> Reported-by: Fabio M. Di Nitto <fdinitto@redhat.com>
>> Signed-off-by: Xin Long <lucien.xin@gmail.com>
>> ---
>>  net/sctp/input.c | 13 +++++++++++++
>>  1 file changed, 13 insertions(+)
>>
>> diff --git a/net/sctp/input.c b/net/sctp/input.c
>> index 458e506..f65245b 100644
>> --- a/net/sctp/input.c
>> +++ b/net/sctp/input.c
>> @@ -872,6 +872,8 @@ void sctp_transport_hashtable_destroy(void)
>>
>>  int sctp_hash_transport(struct sctp_transport *t)
>>  {
>> +     struct sctp_transport *transport;
>> +     struct rhlist_head *tmp, *list;
>>       struct sctp_hash_cmp_arg arg;
>>       int err;
>>
>> @@ -882,8 +884,19 @@ int sctp_hash_transport(struct sctp_transport *t)
>>       arg.paddr = &t->ipaddr;
>>       arg.lport = htons(t->asoc->base.bind_addr.port);
>>
>> +     list = rhltable_lookup(&sctp_transport_hashtable, &arg,
>> +                            sctp_hash_params);
>> +
>> +     rhl_for_each_entry_rcu(transport, tmp, list, node)
>> +             if (transport->asoc->ep == t->asoc->ep) {
>> +                     err = -EEXIST;
>> +                     goto out;
>> +             }
>> +
> Maybe its just a little early, but I'm rather lost as to how this works.  When
> you insert a key to the rhltable, its a sctp_hash_cmp_arg that gets inserted,
> which is a struct that consists of a sctp_addr union, a struct net, and a u16,
> but when you traverse the list for lookup, you treat the returned list as a list
> of sctp_transport structs that you compare.  That seems very wrong.

The return list is "struct rhlist_head" list, "struct rhlist node" is
a member and
nested in struct sctp_transport, in fact this list consist of  sctp_transport.

you can double check the inserting function "rhltable_insert_key()" below.
the param &t->node would be a node linked into this list.

rhltable_insert_key(struct rhlist_head *list[&t->node])
   __rhashtable_insert_fast(struct rhash_head *obj [list->rhead])

...
                list = container_of(obj, struct rhlist_head, rhead);
                           ^---get back to rhash_list, then insert.
                plist = container_of(head, struct rhlist_head, rhead);

                RCU_INIT_POINTER(list->next, plist);
                head = rht_dereference_bucket(head->next, tbl, hash);
                RCU_INIT_POINTER(list->rhead.next, head);
...


the param *arg is NOT really the node used for inserting, but the param
*obj is.

>
> Neil
>
>>       err = rhltable_insert_key(&sctp_transport_hashtable, &arg,
>>                                 &t->node, sctp_hash_params);
>> +
>> +out:
>>       if (err)
>>               pr_err_once("insert transport fail, errno %d\n", err);
>>
>> --
>> 2.1.0
>>
>>