svn commit: r236955 - head/sys/net
Randall Stewart
rrs at lakerest.net
Tue Jun 12 13:19:10 UTC 2012
That actually will not effect things. Since the packet is
on the way out. If the csum_data field gets set on the way in
(not out)… so reusing it here I think is safe.
R
On Jun 12, 2012, at 8:53 AM, Bjoern A. Zeeb wrote:
>
> On 12. Jun 2012, at 12:44 , Randall Stewart wrote:
>
>> Author: rrs
>> Date: Tue Jun 12 12:44:17 2012
>> New Revision: 236955
>> URL: http://svn.freebsd.org/changeset/base/236955
>>
>> Log:
>> Note to self. Have morning coffee *before* committing things.
>> There is no mac_addr in the mbuf for BSD.. cheat like
>> we are supposed to and use the csum field since our friend
>> the gif tunnel itself will never use offload.
>
> There are cards that can do checksums for IPIP... with drivers in out tree.
>
>
>>
>> Modified:
>> head/sys/net/if_gif.c
>>
>> Modified: head/sys/net/if_gif.c
>> ==============================================================================
>> --- head/sys/net/if_gif.c Tue Jun 12 12:40:15 2012 (r236954)
>> +++ head/sys/net/if_gif.c Tue Jun 12 12:44:17 2012 (r236955)
>> @@ -388,7 +388,8 @@ keep_going:
>> /* Now pull back the af in packet that
>> * was saved in the address location.
>> */
>> - bcopy(m->m_pkthdr.src_mac_addr, &af, sizeof(af));
>> + af = m->m_pkthdr.csum_data;
>> +
>> if (ifp->if_bridge)
>> af = AF_LINK;
>>
>> @@ -503,10 +504,11 @@ gif_output(ifp, m, dst, ro)
>> dst->sa_family = af;
>> }
>> af = dst->sa_family;
>> - /* Now save the af in the inbound pkt mac
>> - * address location.
>> + /* Now save the af in the inbound pkt csum
>> + * data, this is a cheat since really
>> + * gif tunnels don't do offload.
>> */
>> - bcopy(&af, m->m_pkthdr.src_mac_addr, sizeof(af));
>> + m->m_pkthdr.csum_data = af;
>> if (!(ifp->if_flags & IFF_UP) ||
>> sc->gif_psrc == NULL || sc->gif_pdst == NULL) {
>> m_freem(m);
>
> --
> Bjoern A. Zeeb You have to have visions!
> It does not matter how good you are. It matters what good you do!
>
>
------------------------------
Randall Stewart
803-317-4952 (cell)
More information about the svn-src-head
mailing list