[reSIProcate] compat.hxx broken for IPPROTO_IPV6

[Scott Godin]

V6 should always supported if building for Windows XP SP1 and above - so I
think Jeremy's workaround is safe (with a slight mode).  If I use exactly
what Jeremy is proposing then I get compiler errors for multiple definitions
of IPPROTO_IPV6.  Jeremy - would the following work for you as well?
#ifdef USE_IPV6
#ifndef IPPROTO_IPV6
#if(_WIN32_WINNT >= 0x0501)   // Some versions of the windows SDK define
IPPROTO_IPV6 differently - always enable IP v6 if USE_IPV6 and _WIN32_WINNT
>= 0x0501
#define IPPROTO_IPV6 ::IPPROTO_IPV6
#else
#ifdef _MSC_VER
#define __STR2__(x) #x
#define __STR1__(x) __STR2__(x)
#define __LOC__ __FILE__ "("__STR1__(__LINE__)"): "
#pragma message (__LOC__ " IPv6 support requested, but IPPROTO_V6 undefined;
this platform does not appear to support IPv6 ")
#else
#warning IPv6 support requested, but IPPROTO_IPV6 undefined; this platform
does not appear to support IPv6
#endif
#undef USE_IPV6
#endif
#endif
#endif

Scott

On Thu, Apr 23, 2009 at 12:44 PM, Jeremy Geras <jgeras at counterpath.com>wrote:

>  They have #if(_WIN32_WINNT >= 0x0501) just before the entry for IPPROTO_IPV6
> in the version of ws2def.h that I've got, so nope, it isn't always in the
> enum.
>
> I've changed compat.hxx to the following in my local copy to make things
> work -- note that this is slightly different from what I proposed
> originally, since there actually are places that check the value of
> IPPROTO_IPV6 (not just an #ifdef test), and it picks up the definition from
> the #define instead of the enum:
>
>
> #ifdef
> USE_IPV6
>
> #if
> (_WIN32_WINNT >= 0x0501)
>
> #define
> IPPROTO_IPV6 ::IPPROTO_IPV6
>
> #else
>
> #ifdef _MSC_VER
>
> #define __STR2__(x) #x
>
> #define __STR1__(x) __STR2__(x)
>
> #define __LOC__ __FILE__ "("__STR1__(__LINE__)"): "
>
> #pragma message (__LOC__ " IPv6 support requested, but IPPROTO_V6
> undefined; this platform does not appear to support IPv6 ")
>
> #else
>
> #warning IPv6 support requested, but IPPROTO_IPV6 undefined; this platform
> does not appear to support IPv6
>
> #endif
>
> #undef USE_IPV6
>
> #endif
>
> #endif
>
>
>
> Jeremy
>
>
>  ------------------------------
> *From:* Byron Campen [bcampen at estacado.net]
> *Sent:* April 23, 2009 9:02 AM
> *To:* Jeremy Geras
> *Cc:* Scott Godin; resiprocate-devel at list.resiprocate.org
> *Subject:* Re: [reSIProcate] compat.hxx broken for IPPROTO_IPV6
>
>   Ick. What an irritating thing to do. Presumably, this is in the enum
> whether or not V6 is actually supported? Or is V6 always supported in this
> SDK?
>
>  Best regards,
> Byron Campen
>
>   So it looks like this is only a problem for folks who have updated their
> Platform SDK to some version beyond that which comes with VS 2005 ...  I see
> the #define for IPPROTO_IPV6 in the old winsock2.h that came with VS .NET
> 2003, but it definitely is not there in the version that got installed with
> the latest Platform SDK.
>
>  *From:* slgodin at gmail.com [mailto:slgodin at gmail.com <slgodin at gmail.com>]
> *On Behalf Of *Scott Godin
> *Sent:* March-25-09 5:33 AM
> *To:* Jeremy Geras
> *Cc:* resiprocate-devel at list.resiprocate.org
> *Subject:* Re: [reSIProcate] compat.hxx broken for IPPROTO_IPV6
>
>  I have IPPPROTO_IPV6 defined in WinSock2.h at c:\Program Files\Microsoft
> Visual Studio 8\VC\PlatformSDK\Include
>  Scott
>
>  2009/3/24 Jeremy Geras <jgeras at counterpath.com>
>   Hi,
>
>   Unless I'm missing some #include that everyone else on Windows has, I
> don't think that IPPROTO_IPV6 can be checked with an #ifndef in
> rutil\compat.hxx the way that it is right now on the trunk...  IPPROTO_IPV6
> is part of an enum, it isn't #define'd.
>
>   So
>   #ifndef IPPROTO_IPV6
>
>   should actually be something more along the lines of
>
>   #if(_WIN32_WINNT >= 0x0501)
> #define IPPROTO_IPV6
> #else
>   ...
>
>   Or maybe we should just #undef USE_IPV6 ?
>
>   In case it's relevant, I'm on Windows XP SP3 with the latest Platform
> SDK (6.1) compiling in VS 2008.
>
>   Jeremy
>
>
> _______________________________________________
> resiprocate-devel mailing list
> resiprocate-devel at resiprocate.org
> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>
>  _______________________________________________
> resiprocate-devel mailing list
> resiprocate-devel at resiprocate.org
> https://list.resiprocate.org/mailman/listinfo/resiprocate-devel
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://list.resiprocate.org/pipermail/resiprocate-devel/attachments/20090520/d5ad42b9/attachment.htm>