Mailing-List: contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-developers-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin-developers/>
List-Post: <mailto:cygwin-developers AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-developers-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-developers-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin-developers AT sources DOT redhat DOT com
Message-ID: <3B60362A.4030704@ece.gatech.edu>
Date: Thu, 26 Jul 2001 11:24:26 -0400
From: Charles Wilson <cwilson AT ece DOT gatech DOT edu>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.2) Gecko/20010713
X-Accept-Language: en-us
MIME-Version: 1.0
CC: cygwin-developers AT sources DOT redhat DOT com
Subject: Re: Problems with autoconf-2.52 testsuite using current CVS Cygwin
References: <996158432 DOT 5232 DOT ezmlm AT sources DOT redhat DOT com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Oops.  Forgot to change the title on the last message.  Hopefully this 
one will appear within the correct thread.

> The autoconf testsuite creates a lot of temporary directories and
> files in them below /tmp. The directories are named using two
> characters (e.g. "ac") and the PID of the current process.So a
> typical name is `ac1322' on NT/W2K. Let's work out that example.
> 
> After performing the related test, the testsuite removes the
> testdirectory ("/tmp/ac1322") and moves forward to the next
> test.
> 
> Now the problem suddenly shows up:
> 
> The testsuite fails to remove the temporary directory when running
> under 1.3.3.


Well, I just noticed that after running the autoconf test suite under 
1.3.2 -- which passes -- I have a bunch of ac* directories left in my 
/tmp dir, as well.  (I didn't *specifically* notice before, because I 
thought those dirs were leftovers from failed automake or libtool tests, 
not autoconf).

Perhaps 1.3.2 is just "lucky" -- did the mechanism for assining PIDs (as 
seen by cygwin processes) change between 1.3.2 and 1.3.3; that is, are 
PIDs more likely to collide now?

Or, did the return val of mkdir() change when dir-already-exists?

Of course, even if I'm right about the PID issue (and I doubt it) the 
inability to remove dirs under 1.3.3 would still be a problem...


--Chuck