On 09/16/2010 04:52 AM, Bruno Haible wrote:
This message talks about the unportability of a field 's6_addr32' in
'struct in6_addr'. Sure this is unportable: POSIX [1] specifies only the
presence of 's6_addr'.
I don't see why gnulib should support this: An application can perfectly
access the 16
On 09/16/2010 06:32 AM, Eric Blake wrote:
I agree with Bruno - the portability problem stems from using a field
member not specified by POSIX in the first place. It should be
sufficiently easy to rewrite the offending libvirt code to use only the
portable interfaces, without burdening gnulib wit
On 09/15/2010 02:29 PM, Justin Clift wrote:
On 09/16/2010 04:52 AM, Bruno Haible wrote:
This message talks about the unportability of a field 's6_addr32' in
'struct in6_addr'. Sure this is unportable: POSIX [1] specifies only the
presence of 's6_addr'.
I don't see why gnulib should support thi
Justin Clift wrote:
> If you can figure a way to address this OSX specific thing in gnulib at
> the same time, that would be nifty:
>
>http://www.redhat.com/archives/libvir-list/2010-September/msg00248.html
This message talks about the unportability of a field 's6_addr32' in
'struct in6_addr'