Re: [dpdk-dev] rte_hash thread safe

["Ananyev, Konstantin" <konstantin.ananyev@intel.com>]

> 
> A gentle reminder,
> 
> I am curious to know if/how rte_hash is thread safe for lookups.It is not obvious to me how following code is thread safe:

I don't think it is really thread-safe: you can't do lookup and add/delete simultaneously from different threads.
Some extra synchronization would be needed.
Konstantin

> 
> _rte_hash_lookup_with_hash(const struct rte_hash *h, const void *key,
> 
>                                         hash_sig_t sig, void **data)
> 
> {
> 
> 
> 
> …
> 
>                         if (rte_hash_cmp_eq(key, k->key, h) == 0) {
> 
>                                 if (data != NULL)
> 
>                                         *data = k->pdata;
> 
> }
> 
> a key could be deleted and another key inserted in its slot while the lookup is happening. For example, in the following sequence of events:
> The slot has Key1,V1
> Lookup Thread T1 compares the input key to Key1 and it matches. The thread gets context switched out Thread T2 deletes Key1.
> Thread T2 inserts Key2 with value V2.
> T1 reads the data from the slot and returns V2. This is incorrect.
> 
> If T1 is a worker/data plane thread, which is supposed to be pinned to the CPU, is it required to consider the case of T1 getting pre-empted?
> 
> However, this case can exist with T1 not getting pre-empted. i.e. Key1 can get deleted by thread T2 while T1 is doing the comparison.
> 
> In this case, 'k' is a memory pointer pointing to an array element. The memory itself is not freed. The code will still return the data
> associated with Key1 - assuming Key2 belongs to a different slot ID. But, it is possible that 'data' may not be valid anymore since it is owned
> by the application.
> 
> 
> Regards,
> Brijesh
> 
> On Wed, Apr 11, 2018 at 9:12 PM, Brijesh Singh <brijesh.s.singh@gmail.com> wrote:
> > Hello,
> >
> > I  want to use DPDK's rte_hash library to keep track of tcp flows. The
> > lookups will be done by multiple threads but inserts will be done only
> > on one thread.
> >
> > As per the documentation rte_hash library has thread safe lookups. Key
> > /data inserts should be done on single thread, since those operations
> > are not thread safe. Is this documentation still correct?
> >
> > The lookup code compares the key and returns the data if the key
> > matches, this doesn't look like thread safe. Am I missing something?
> >
> > _rte_hash_lookup_with_hash(const struct rte_hash *h, const void *key,
> >
> >                                         hash_sig_t sig, void **data)
> >
> > {
> >
> >
> >
> > …
> >
> >                         if (rte_hash_cmp_eq(key, k->key, h) == 0) {
> >
> >                                 if (data != NULL)
> >
> >                                         *data = k->pdata;
> >
> > }
> >
> > Regards,
> > Brijesh