dns.idl ==> build breaks on non-ipv6 systems

[Michael Adam]

Hi David,

thanks a lot for your comments!

Actually my patch just moved the functions that Kai wrote
into a new code file. So I am explicitly cc-ing Kai in this
reply.

After so many people have said, I should do the proper fix,
maybe I will just give it a go... :-)

But I think I'd like to move the functions anyhow, whether
before or after fixing them.

Thanks - Michael

David Holder wrote:
> Hi Michael,
> 
> A few thoughts on the patch. May not be relevant as I don't know this 
> bit of the code well...
> 
> 1) I notice your constant for address length is set to 39. IPv6 
> addresses can be 46 characters long.There is a standard constant for 
> this INET6_ADDRSTRLEN.
> 2) The code assumes the format uses only colons. It ignores the legal 
> formats which have dotted-decimal at the end. I don't know if such 
> addresses will ever appear in this part of the code; has this been 
> considered? For example, 0000:0000:0000:0000:0000:FFFF:192.168.100.200.
> 3) I was wondering if the code could be rewritten to use getaddrinfo() 
> and getnameinfo(). Use of inet_pton and inet_ntop is not best practice 
> as these are protocol dependent. getaddrinfo() and getnameinfo() can be 
> used to do protocol independent text to address conversions and vice 
> versa. With the appropriate flags they can do a lot for you, including 
> address format checking.
> 
> Best Regards,
> David
> ------------------------------------------------------------------------
> Dr David Holder CEng FIET MIEEE
> 
> Erion Ltd, Oakleigh, Upper Sutherland Road, Halifax, HX3 8NT
> 
> Reception: +44 (0)1422 207000
> 
> Direct Dial: +44 (0)131 2026317
> 
> Cell: +44 (0) 7768 456831
> Web: http://www.erion.co.uk
> 
> Registered in England and Wales. Registered Number 3521142
> VAT Number: GB 698 3633 78
> 
> 
> 
> On 06/11/2010 02:05, Michael Adam wrote:
> >Hi Kai and list,
> >
> >it seems the introduction of the dns server in s4
> >has broken the build on non-ipv6 systems by using
> >AD_INET6 unconditionally in the pull/push/print_ipv6
> >functions in ndr_basic.c
> >
> >Since these functions are in ndr_basic.c this also
> >breaks the build of s3 on non-ipv6-systems.
> >
> >Attached find a patch that moves these 3 functions
> >to a new module ndr_ipv6.c that is only used in s4.
> >So this fixes the build in s3. Andrew argued that
> >this is no proper fixing, and he is right, but I
> >think that these functions deserve their own module
> >anyways. The s4 case needs fixing, but I would
> >appreciate if you could push (or ACK) this patch first,
> >Kai.
> >
> >Cheers - Michael
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 206 bytes
Desc: not available
URL: <http://lists.samba.org/pipermail/samba-technical/attachments/20101106/de7624c9/attachment.pgp>