argus-clients-3.0.0.rc.5

[Carter Bullard]

The problem is that on some machines its in ethernet.h, on others
its ether.h, and then ......  The const is there for linux, Mac OS X,  
etc....

I'll try to borrow the configure code from the latest tcpdump, and see
if we can't fix it that way.

Carter

On Jun 12, 2006, at 3:52 PM, Peter Van Epp wrote:

> On Mon, Jun 12, 2006 at 03:21:47PM -0400, Carter Bullard wrote:
>> Hey Peter,
>>     So, I've made all the fixes you suggested,  in some form or  
>> manner,
>> but I'm most interested in how to fix the HAVE_ETHER_HOSTTON,
>> and STRTOF define problems.   ./configure is suppose to figure out
>> if these routines exist, and then we're suppose to do the right  
>> thing.
>>
>>    Seems like the ./configure test has a bug?  Do these machines
>> not have definitions for these routines?  Are we just declaring them
>> with the wrong parameters?
>>
>> Inquiring minds want to know!!!
>>
>> Carter
>
> 	It is defined in /usr/include/net/ethernet.h it just doesn't appear
> to match completely:
>
> #ifndef KERNEL
> #include <sys/cdefs.h>
>
> /*
>  * Ethernet address conversion/parsing routines.
>  */
> __BEGIN_DECLS
>
> int     ether_hostton __P((char *, struct ether_addr *));
> int     ether_line __P((char *, struct ether_addr *, char *));
> char    *ether_ntoa __P((const struct ether_addr *));
> int     ether_ntohost __P((char *, struct ether_addr *));
> __END_DECLS
> #endif /* !KERNEL */
>
>
> #if defined(HAVE_ETHER_HOSTTON)
>    extern int ether_hostton(const char *, struct ether_addr *);
> #endif
>
> 	Removing the "const" above appears to fix the problem although it may
> break other OSes.
>
> Peter Van Epp / Operations and Technical Support
> Simon Fraser University, Burnaby, B.C. Canada
>