Re: [PATCH 02/05] ipv6: RFC4214 Support (2)

[osprey67 <osprey67@yahoo.com>]

YOSHIFUJI Hideaki / 吉藤英明 wrote:
> In article <47337483.3080309@yahoo.com> (at Thu, 08 Nov 2007 12:41:39 -0800), osprey67 <osprey67@yahoo.com> says:
> 
>> From: Fred L. Templin <osprey67@yahoo.com>
>>
>> This is experimental support for the Intra-Site Automatic
>> Tunnel Addressing Protocol (ISATAP) per RFC4214. It uses
>> the SIT module, and is configured using the unmodified
>> "ip" utility with device names beginning with: "isatap".
>>
>> The following diffs are specific to the Linux 2.6.24-rc2
>> kernel distribution.
>>
>> Signed-off-by: Fred L. Templin <osprey67@yahoo.com>
> 
> Hmm...tabs are still mangled, and it's better to have your
> official address.  Anyway...

I have switched over to Thunderbird and POP mail. I submitted
the patches by cutting and pasting from a text file into the
mail message. Is there a way to stop it from clobbering tabs?

>> --- linux-2.6.24-rc2/include/net/addrconf.h.orig        2007-11-08 12:06:17.000000000 -0800
>> +++ linux-2.6.24-rc2/include/net/addrconf.h     2007-11-08 08:27:24.000000000 -0800
>> @@ -241,6 +241,37 @@ static inline int ipv6_addr_is_ll_all_ro
>>                  addr->s6_addr32[3] == htonl(0x00000002));
>>   }
>>
>> +#if defined(CONFIG_IPV6_ISATAP)
>> +static inline int ipv6_isatap_eui64(u8 *eui, __be32 addr)
>> +{
>> +
>> +       /* RFC3330 Special-Use IPv4 Addresses */
>> +       eui[0] = (((addr & htonl(0xFF000000)) == htonl(0x00000000)) ||
>> +                 ((addr & htonl(0xFF000000)) == htonl(0x0A000000)) ||
>> +                 ((addr & htonl(0xFF000000)) == htonl(0x0D000000)) ||
>> +                 ((addr & htonl(0xFF000000)) == htonl(0x18000000)) ||
>> +                 ((addr & htonl(0xFF000000)) == htonl(0x7F000000)) ||
>> +                 ((addr & htonl(0xFFFF0000)) == htonl(0xA9FE0000)) ||
>> +                 ((addr & htonl(0xFFF00000)) == htonl(0xAC100000)) ||
>> +                 ((addr & htonl(0xFFFFFF00)) == htonl(0xC0000200)) ||
>> +                 ((addr & htonl(0xFFFFFF00)) == htonl(0xC0586300)) ||
>> +                 ((addr & htonl(0xFFFF0000)) == htonl(0xC0A80000)) ||
>> +                 ((addr & htonl(0xFFFE0000)) == htonl(0xC6120000)) ||
>> +                 ((addr & htonl(0xF0000000)) == htonl(0xE0000000)) ||
>> +                 ((addr & htonl(0xF0000000)) == htonl(0xF0000000))) ?
>> +                       0x00 : 0x02;
>> +
>> +       eui[1] = 0; eui[2] = 0x5E; eui[3] = 0xFE;
>> +       memcpy (eui+4, &addr, 4);
>> +       return 0;
>> +}
>> +
> 
> Please put this function in net/ipv6/addrconf.c as addrconf_ifid_isatap().
> 
> Please use MULTICAST, LOCALNET etc. (and probaly introduce new macro
> for others).  IMHO, it's better to add a comment for each entry, e.g.,
> 	MULTICAST(addr) ||		/* 224.0.0.0/4 */
> instead of
> 	((addr & htonl(0xF0000000)) == htonl(0xE0000000)) ||

OK - will do.

>> +static inline int ipv6_addr_is_isatap(const struct in6_addr *addr)
>> +{
>> +       return ((addr->s6_addr32[2] | htonl(0x02000000)) == htonl(0x02005EFE));
>> +}
>> +#endif
>> +
> 
> ipv6_addr_isatap(), maybe (to align with ipv6_addr_any() etc.).

Well, if you look in addrconf.h immediately above this the convention
seems to be to call it as "*_is_*" if it is testing an address and
to omit the "*_is_*" if it is setting an address. Since this is an
address test, maybe more consistent to keep the "*_is_*"?

Thanks - Fred
fred.l.templin@boeing.com

> 
> --yoshfuji
>