Re: [Libevent-users] building libevent from git on *BSD

[Nick Mathewson <nickm@xxxxxxxxxxxxx>]

On Fri, Jan 22, 2010 at 5:10 PM, William Ahern
<william@xxxxxxxxxxxxxxxxxx> wrote:
> On Fri, Jan 22, 2010 at 04:47:43PM -0500, Tom Pusateri wrote:
>> I cloned the git tree today and tried building libevent on FreeBSD 7 and 8.
>>
>> It needs netinet/in6.h but on the *BSD's that include file is in netinet6/in6.h so it can't find the definitions of in6_addr.
>>

Ouch; weird.  I will try to find and fire up a FreeBSD VM some time in
the next week or two to double-check the obvious fix and commit it.
If you want to send in a patch that works for you before then, that
would be excellent.

One thing: I'm a little confused by the line numbers in your compile
output.  In current git master
(the head right now is ff3f6cd42b8), the function evutil_inet_ntop
runs from line 1377 to line 1471.  All of the errors from your output
in that function are between lines 385 and 389.  Is it possible that
you cloned from a wrong source, or that something munged your compiler
output as you were sending it?

The canonical place to clone from, just for reference, is
     git clone git://levent.git.sourceforge.net/gitroot/levent/libevent
You can also do
    git clone git://levent.git.sourceforge.net/gitroot/levent/levent

and get the same respository.



> I thought in6.h is never to be included directly (if it exists), but IPv6
> interfaces should simply be available through <sys/socket.h> and
> <netinet/in.h>.
>
> Are there systems where those two headers don't provide the IPv6 interfaces?

Apparently so; evutil.c (in git master) does indeed include both of
sys/socket.h and netinet/in.h, but still apparently gave Tom those
errors.

-- 
Nick
***********************************************************************
To unsubscribe, send an e-mail to majordomo@xxxxxxxxxxxxx with
unsubscribe libevent-users    in the body.