Re: ARGH pthreads tests failing?

[Christopher Faylor <cgf at redhat dot com>]

On Thu, Jan 23, 2003 at 08:50:03AM +0100, Thomas Pfaff wrote:
>
>
>On Wed, 22 Jan 2003, Christopher Faylor wrote:
>
>> On Wed, Jan 22, 2003 at 10:53:19AM +0100, Thomas Pfaff wrote:
>> >I can only guess here but it might be that the wrong pthread.h is
>> >included in your tests. I remember some mails about -include does not
>> >behave properly with gcc-3, it will always take the headers from
>> >/usr/include first.
>>
>> How about a patch, then?
>
>You might recall a thread in cygwin developer named "Source snapshot fails
>to build" (http://cygwin.com/ml/cygwin-developers/2002-10/msg00012.html)
>where the -isystem problem first appeared and
>"[Bug: gcc-3.2-1] Cygwin DLL (CVS HEAD) build failure - gcc uses wrong
>include path order"
>(http://sources.redhat.com/ml/cygwin/2002-10/msg01532.html).

You may recall a release announcement for gcc-3.2-3 which indicated that
-nostdinc now works.  This was what was causing the problem and IIRC
I got notification from people like Corinna indicating that this release
fixed the problem.  It was one of the motivations for the 3.2-3 release.

>I do have these problems too (gcc is gcc-3.2-3) where w32api headers are
>alway included from /usr/include/w32api instead of w32api in the source
>tree.
>
>Unfortunately i have not the time to fix this, IMHO this is a gcc bug, not
>cygwin.

It's not a gcc bug.  I've just verified that -nostdinc does not load
the w32api headers and that -I works correctly.

cgf