Hi Ying, Ok, I will remove that "ifdef ... else", I don't like it either, just worried about a compiler warning but I have tested & it is ok. I think we can pass the node pointer to the "__dnode" anyway (don't need a NULL), so can remove the other "ifdef ... else" at every function calls (total = 8), do you agree?
BR/Tuong -----Original Message----- From: Ying Xue <[email protected]> Sent: Wednesday, November 6, 2019 3:42 PM To: Tuong Lien <[email protected]>; [email protected]; [email protected]; [email protected] Subject: Re: [PATCH v2 4/5] tipc: introduce TIPC encryption & authentication On 11/5/19 6:50 PM, Tuong Lien wrote: > void tipc_bearer_xmit(struct net *net, u32 bearer_id, > struct sk_buff_head *xmitq, > - struct tipc_media_addr *dst) > +#ifdef CONFIG_TIPC_CRYPTO > + struct tipc_media_addr *dst, > + struct tipc_node *__dnode > +#else > + struct tipc_media_addr *dst > +#endif > + ) To make code simpler, I think we can directly add one new parameter (ie, __dnode) no matter whether CONFIG_TIPC_CRYPTO is enabled or not. When CONFIG_TIPC_CRYPTO is disabled, we just pass NULL to "__dnode" parameter. As a consequence, we don't need to use CONFIG_TIPC_CRYPTO config to separate the two different situations of whether CONFIG_TIPC_CRYPTO is enabled or not whenever tipc_bearer_xmit() is called. _______________________________________________ tipc-discussion mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/tipc-discussion
