On Sun, Jul 11, 2021 at 05:33:18AM +0300, Vitaliy Makkoveev wrote:
> This member is never set or used. Also I kept 'SENT_IP6' definition for
> prevent the potential break of third party software. Is it ok to
> redefine it to '0x0002'? At least openswan wants this [1].
>
> 1.
> https://github.com/xelerance/Openswan/blob/master/include/openswan/ipsec_encap.h#L20
I wouldn't worry about third party software. Those defines are within
#ifdef _KERNEL so everyone using those is doing it wrong anyway.
ok with SENT_IP6 changed to 0x0002.
>
> Index: sys/netinet/ip_ipsp.h
> ===================================================================
> RCS file: /cvs/src/sys/netinet/ip_ipsp.h,v
> retrieving revision 1.200
> diff -u -p -r1.200 ip_ipsp.h
> --- sys/netinet/ip_ipsp.h 8 Jul 2021 21:07:19 -0000 1.200
> +++ sys/netinet/ip_ipsp.h 11 Jul 2021 02:25:17 -0000
> @@ -113,8 +113,6 @@ struct sockaddr_encap {
> u_int16_t Sport;
> u_int16_t Dport;
> } Sip6;
> -
> - struct ipsec_policy *PolicyHead; /* SENT_IPSP */
> } Sen;
> };
>
> @@ -208,7 +206,6 @@ struct m_tag;
> #define sen_ip6_sport Sen.Sip6.Sport
> #define sen_ip6_dport Sen.Sip6.Dport
> #define sen_ip6_direction Sen.Sip6.Direction
> -#define sen_ipsp Sen.PolicyHead
>
> /*
> * The "type" is really part of the address as far as the routing
> @@ -219,7 +216,6 @@ struct m_tag;
> */
>
> #define SENT_IP4 0x0001 /* data is two struct in_addr */
> -#define SENT_IPSP 0x0002 /* data as in IP4/6 plus SPI */
> #define SENT_IP6 0x0004
>
> #define SENT_LEN sizeof(struct sockaddr_encap)
>
