On 12/17/2015 02:16 PM, Eric Blake wrote:
> On 12/17/2015 02:01 PM, Corinna Vinschen wrote:
> 
>>> Here's what happens:
>>>
>>> One of the Gnulib modules includes sys/types.h, which includes sys/select.h
>>> because of the recent changes.  This brings in Gnulib's sys/select.h, which
>>> includes signal.h.  We then get the errors I posted because we haven't yet
>>> finished including sys/types.h.
> 
> Gnulib has been taught to work around early inclusion problems before;
> sounds like this will be another case where gnulib has to make sure the
> system header is complete before its own replacements kick in.
> 
>>>
>>> All the build errors disappear if I remove '#include <sys/select.h>' from
>>> sys/types.h.  You said above that the macros related to select don't really
>>> belong in sys/types.h.  So why does the latter include sys/select.h?

Another data point: POSIX does NOT allow <sys/types.h> to pollute the
namespace with symbols from <sys/select.h>.  True, the use of
__BSD_VISIBLE says that POSIX is not in play, but I'm suspecting that
very few programs are written that use sys/select.h functionality but
were relying on the BSD headers to indirectly include it via
sys/types.h, since such programs would fail to compile on other systems
where the indirect include is not present (more likely, any clients of
sys/select.h are directly including it).

So at this point, I'm leaning towards fixing the cygwin header to not
include sys/select.h from sys/types.h.

-- 
Eric Blake   eblake redhat com    +1-919-301-3266
Libvirt virtualization library http://libvirt.org