Re: Many pthread failures in the test suite, one setgroup failure

[Christopher Faylor <cgf at redhat dot com>]

On Sun, Sep 29, 2002 at 10:49:48PM -0400, Pierre A. Humblet wrote:
>At 07:33 PM 9/29/2002 -0400, Christopher Faylor wrote:
>>>I also noticed that allow_ntsec == 1 on WinME,
>>>which at the very least causes useless calls. 
>>
>>How did allow_ntsec become 1 on WinME?  That much at least should be
>>easy to debug.
>
>Right. This afternoon I had noted it was already 1 when entering
>environ_init. An fgrep just revealed in security.cc
>BOOL allow_ntsec = true;
>I would propose a patch setting it to false at the end of environ_init
>on Win9X even if it is "on" in CYGWIN. Ditto for ntea and smbntsec.
>Is there a better idea out there?

I just activated the existing code in environ_init that turned this on
in cygwin for OS's that support it.  That's what I should have done
before.

If someone wants to set it on windows 95, I don't see why we should stop
them.  It might be useful for debugging.

>>Also, another question is:  Are you sure that you are using the latest
>>CVS sources?  This looks suspiciously like the problem which was just
>>fixed.
>
>I have again updated from cvs and run make. Same result in gdb.

Hopefully, you've run "make clean" and then "make".  Just to be safe.

cgf