Re: getaddrinfo module conflict

[Simon Josefsson]

> Yoann Vandoorselaere wrote:
>> My suggestion is that unless we can get a full featured replacement of
>> the getaddrinfo function within GnuLib (and thus replace any non
>> conforming system implementation), we should not attempt to redefine any
>> of the flags if the function is available on the system.
>
> Though it would be nice to see a full-featured replacement, for now
> leaving the definitions of AI_* flags which aren't implemented by the
> replacement does sound simplest.  How's the attached patch look?

I think this is a step back, but given that the right choice seem to be to
1) implement support for the specific AI_ flag that was needed but missing
here, and 2) check for availability in the system for that feature, and
that we don't have a patch that does this now, I won't object.  (And your
white-space fixing patch is fine too, of course.)

I'm currently on vacation, but when I get back, I'll try to implement the
missing flags and test the system implementation for them.  It would be
useful to document exactly which flag(s) were needed now, and the
workaround for it.  Yoann?

/Simon