Now that we have a function substitute in , let me add a test of it
in C++ mode.
#if GNULIB_TEST_GETADDRINFO
SIGNATURE_CHECK (GNULIB_NAMESPACE::getaddrinfo, int,
(const char *, const char *, const struct addrinfo *,
struct addrinfo **));
SIGNATURE_CHECK (GNULIB_N
Bruno Haible writes:
> Hi Simon,
>
> A week ago I proposed:
>> > I propose to generate netdb.h always, like we do
>> > with stdlib.h and unistd.h for 3 years and for many others for
>> > 2 months. Then we can also add GNULIB_POSIXCHECK advice to .
>
>
Hi Simon,
A week ago I proposed:
> > I propose to generate netdb.h always, like we do
> > with stdlib.h and unistd.h for 3 years and for many others for
> > 2 months. Then we can also add GNULIB_POSIXCHECK advice to .
Here's the second part: The patch to add GNULIB_PO
;. This
> is not needed: all systems that have a define 'struct hostent'
> in it.
I have pushed the patch below, I appreciate review if it can be made
simpler and whether I understood the idioms properly.
The patch fixes the problem on MinGW where 'struct hostent' w
Hi Simon,
Good to see more include files covered!
> +#if @HAVE_NETDB_H@
> +
> +/* Declarations for a platform that has . */
> +
> +#else
> +
> +/* Declarations for a platform that has . */
The second 'has' should be a 'lacks', no?
> + AC_CACHE_CHECK([whether is self-contained],
> +[gl_c
I noticed that I have some HAVE_NETDB_H checks to cover mingw32 in
gsasl. The declarations in getaddrinfo.h should really be defined in
netdb.h, as far as I can tell, too.
This is a first patch that merely create a small dummy netdb.h
replacement file. Pushed, but I do appreciate review and
