New CARP patchset

[Matthew George]

Matthew George wrote:
> Matthew George wrote:
> 
>> Max Laier wrote:
>>
> [...]
> 
>>> Anything else? Please let me know!
>>>
>>

I finally had the time to sit down and look at it.  The ifconfig build 
issue I was having was because I had a previous version of ifcarp.c, so 
the patch appended (prepended?) another copy of the file to it.  I *do* 
believe, however, that the first issue (failed hunk in 
sys/netinet/in_proto.c patch) is real.

>>
>> FIRST ISSUE:
>>
>> --------------------------
>> |diff -urP ../dist/sys/netinet/in_proto.c ./sys/netinet/in_proto.c
>> |--- ../dist/sys/netinet/in_proto.c     Sat Nov  6 21:01:08 2004
>> |+++ ./sys/netinet/in_proto.c   Sat Nov  6 18:13:47 2004
>> --------------------------
>> Patching file ./sys/netinet/in_proto.c using Plan A...
>> Hunk #1 succeeded at 36 (offset 1 line).
>> Hunk #2 succeeded at 92 with fuzz 2 (offset 1 line).
>> Hunk #3 failed at 243.
>> Hunk #4 succeeded at 297 with fuzz 1 (offset -5 lines).
>> 1 out of 4 hunks failed--saving rejects to ./sys/netinet/in_proto.c.rej
>>
>>
>> Here's the reject:
>>
>> ***************
>> *** 238,243 ****
>>     &rip_usrreqs
>>   },
>>   #endif        /* DEV_PFSYNC */
>>   /* Spacer n-times for loadable protocols. */
>>   IPPROTOSPACER,
>>   IPPROTOSPACER,
>> --- 243,256 ----
>>     &rip_usrreqs
>>   },
>>   #endif        /* DEV_PFSYNC */
>> + #ifdef DEV_CARP
>> + { SOCK_RAW,   &inetdomain,    IPPROTO_CARP,   PR_ATOMIC|PR_ADDR,
>> +   carp_input, (pr_output_t*)rip_output,       0,      rip_ctloutput,
>> +   0,
>> +   0,          0,              0,              0,
>> +   &rip_usrreqs
>> + },
>> + #endif /* DEV_CARP */
>>   /* Spacer n-times for loadable protocols. */
>>   IPPROTOSPACER,
>>   IPPROTOSPACER,
>>
>>
>> I also see line offsets in the patch for sys/netinet/if_ether.c but 
>> the patch applies anyway.
>>
>>

-- 
Matthew George
SecureWorks Technical Operations