netdev
[Top] [All Lists]

Re: [benc@hawaga.org.uk: Linux "enable EUI-64 token format"]

To: "David S. Miller" <davem@xxxxxxxxxx>
Subject: Re: [benc@hawaga.org.uk: Linux "enable EUI-64 token format"]
From: Keith Owens <kaos@xxxxxxxxxx>
Date: Thu, 03 May 2001 14:26:09 +1000
Cc: netdev@xxxxxxxxxxx
In-reply-to: Your message of "Wed, 02 May 2001 21:20:15 MST." <15088.56447.320844.315012@pizda.ninka.net>
Sender: owner-netdev@xxxxxxxxxxx
On Wed, 2 May 2001 21:20:15 -0700 (PDT), 
"David S. Miller" <davem@xxxxxxxxxx> wrote:
>Keith Owens writes:
> > CONFIG_IPV6_EUI64=y, CONFIG_IPV6_NO_PB=y is the correct configuration.
>
>Then why are you deleting CONFIG_IPV6_NO_PB sections of code in other
>parts of your patch, like this bit in net/ipv4/addrconf.c:
>
>-#ifndef CONFIG_IPV6_NO_PB
>-              if (pinfo->prefix_len == ((sizeof(struct in6_addr) - 
>dev->addr_len)<<3)) {
>-                      memcpy(&addr, &pinfo->prefix, plen);
>-                      memcpy(addr.s6_addr + plen, dev->dev_addr,
>-                             dev->addr_len);
>-                      goto ok;
>-              }
>-#endif

Double negatives are confusing.  CONFIG_IPV6_NO_PB=y in config, 1 in
autoconf.h.  #ifndef CONFIG_IPV6_NO_PB is false when CONFIG_IPV6_NO_PB
is defined so the code is omitted when CONFIG_IPV6_NO_PB=y.  The patch
has the same affect as always setting CONFIG_IPV6_NO_PB, even though
that variable has been removed.


<Prev in Thread] Current Thread [Next in Thread>