On 2016年12月31日 19:24, Jouni Malinen wrote:
> /* FIX: Why is this loop here? This function is called for a specific
> * STA while the loop here adds PMKSA cache entries for all STAs.
> * Shouldn't this addition be done when starting the mesh rather than
> * when going through a STA-specific SAE initialization? */
Thank you !
Yes. This should be moved to SAE initialization code. I moved it to
mesh_rsn_auth_init().
> /* FIX: Wouldn't this lost the last reference to
> * entry->pmksa_cache and by doing so, leak memory?
> */
Yes. I fixed it to free() even if malloc() was failed. In addition, I
moved error message output code into wpa_auth_pmksa_add_entry() to
display peer address.
> /* FIX: free entry->pmksa_cache (need pmksa_cache_auth.c
> * changes to do that, e.g., by allowing
> * _pmksa_cache_free_entry() to be called from here or
> * maybe more likely by adding a new wrapper function
> * for this special external cache case). */
Yes. I must free all unused entries... I fixed it.
Masashi Honma.
_______________________________________________
Hostap mailing list
Hostap@xxxxxxxxxxxxxxxxxxx
http://lists.infradead.org/mailman/listinfo/hostap
