ether-wake.c, ether_hostton() and uClibc

Peter S. Mazinger ps.m at gmx.net
Tue Apr 18 19:06:40 UTC 2006


On Tue, 18 Apr 2006, Robert P. J. Day wrote:

> On Tue, 18 Apr 2006, Peter S. Mazinger wrote:
> 
> > On Sat, 15 Apr 2006, Robert P. J. Day wrote:
> >
> > >
> > >   ether-wake.c, line 239:
> > >
> > > #if !defined(__UCLIBC__)
> > >     } else if (ether_hostton(hostid, eaddr) == 0) {
> > >         bb_debug_msg("Station address for hostname %s is %s\n\n", hostid, ether_ntoa(eaddr));
> > > #else
> > > # warning Need to implement ether_hostton() for uClibc
> > >
> > >   doesn't uClibc already support ether_hostton()?  from uClibc-0.9.28,
> > > include/netinet/ether.h:
> > >
> > > /* Map HOSTNAME to 48 bit Ethernet address.  */
> > > extern int ether_hostton (__const char *__hostname, struct ether_addr *__addr)
> > >      __THROW;
> > >
> > >
> > > rday
> >
> > that header was copied w/o adaption from glibc, uClibc does not
> > support ether_hostton()
> 
> yes, mike f. already pointed that out.  so, a trivial question, then

have read it too late, sorry for the duplicate

> -- should header files be cleaned of references to unsupported
> routines?  just asking.
> rday

Normally we use #if 0 /* uClibc: disabled|unsupported */ or similar 
marking within headers, so that we see that it was an uClibc intentional 
change.
Read uClibc's TODO, the first item is exactly this. I have disabled a lot of 
these, but definitely not all of them.
Thanks anyways pointing it out, I have corrected it locally in my tree, 
can't commit it though, because my svn access wasn't reenabled.

Peter 

-- 
Peter S. Mazinger <ps dot m at gmx dot net>           ID: 0xA5F059F2
Key fingerprint = 92A4 31E1 56BC 3D5A 2D08  BB6E C389 975E A5F0 59F2




More information about the busybox mailing list