Autor: Christopher M Smith Data: A: exim-users Assumpte: Re: [Exim] Building exim-4.10 for OSX 10.2 / Darwin 6.0
On Wednesday, August 28, 2002, at 04:34 AM, Philip Hazel wrote:
> On Tue, 27 Aug 2002, Christopher M. Smith wrote:
>
>> easily fixed by adding #define NO_IP_VAR_H to the OS/os.h-Darwin
>> header file.
>
> Would this also work for other versions of Darwin? Or am I going to
> have
> to set up different OS files for different releases? Are there any
> experts out there?
A few of the folks at work have 10.1 still installed (Darwin 5.0), and
it
has the netinet/ip_var.h header file. So this would have to be
different.
>> I also modified the CC variable set in OS/Makefile-Darwin to use gcc
>> instead
>> of cc.
>
> That is a change you can make in Local/Makefile. No need to modify the
> OS file.
True, I was just trying to be consistent with what I saw in the
codebase.
I mean, it may not be necessary to change as that it was pointed out
earlier that cc is simply linked to gcc for backwards compatibility
reasons.
>> smtp_in.c: In function `smtp_start_session':
>> smtp_in.c:1059: storage size of `ipoptblock' isn't known
>> smtp_in.c:1105: dereferencing pointer to incomplete type
>> smtp_in.c:1137: dereferencing pointer to incomplete type
>
> Surprise, surprise, the structure for ipoptblock is usually defined in
> /usr/include/netinet/ip_var.h.
>
> You can cut out all the IP options processing by setting
>
> #define NO_IP_OPTIONS
>
> in OS/os.h-Darwin. This loses you the check for IP option settings
> (which should not occur), but some OS do this automatically for you. I
> don't know if that is true of Darwin.
I still don't know the answer to this yet, but as that I am not running
a full
blown mail relay and using Exim solely for outbound mail on my laptop,
I am not too worried about this. By defining both the
#define NO_IP_VAR_H
#define NO_IP_OPTIONS
in the header file, the software built fine and is working great.
>
> Always welcome. Though a straight patch may not make sense unless it
> also works on other versions of Darwin.
Once I have the answer to your IP options question above, I will
post my findings to the list and you can let me know what you'd like
to do and I'll post whatever is appropriate.