Broken autoconf mmap test

Corinna Vinschen
Fri Mar 25 20:41:00 GMT 2011

On Mar 25 10:52, Ken Brown wrote:
> On 3/25/2011 5:34 AM, Corinna Vinschen wrote:
> >If you guys think that cheating for autoconf on 64 bit systems is a good
> >idea, I'm willing to implement it.
> It seems to me that the best long-term solution is for the autoconf
> test to be split into two tests as Eric suggested.  That way
> programs that really need full POSIX compliance (if there are any
> such programs) can test for it.

I agree.  A modified mmap test in autoconf makes most senss in the long

> Until that happens and becomes widely known, however, I think
> cheating is better than the status quo.  As it stands, Cygwin
> packages that are built with cygport will cheat anyway, and other
> packages will have different behavior depending on whether or not
> the build was done on a 64-bit machine.  Maintainers who upgrade
> from a 32-bit machine to a 64-bit machine might not even notice that
> their packages have stopped using mmap.

I applied the patch in modified form.  I ordered the tests for speed
and I fixed the incorrect lazy file length test.


Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

More information about the Cygwin-developers mailing list