Mailing-List: contact cygwin-help@sourceware.cygnus.com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner@sources.redhat.com Delivered-To: mailing list cygwin@sources.redhat.com From: tprinceusa@mindspring.com Date: Thu, 12 Jul 2001 16:40:34 -0400 To: Cheuk Cheng Cc: cygwin@cygwin.com Subject: Re: Running Dejagnu under Cygwin Message-ID: X-Originating-IP: 134.134.248.28 Cheuk Cheng wrote: > I had installed the gnupro99r1p1 toolkit on my Win2000 machine. I ran configure on the parallel source directory of gcc (..\win32\gcc) and then installed the newly built binaries after make finished. If I ran all tests to the end, the test summary would say that thousands passed while there would tens of unexpected failures and passes. At this point, I am not sure whether the unexpected ones are normal or not. >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Yes, if you look at the cygwin results which have been appearing on gcc-testsuites, you will see that there are from a dozen to upwards of 80 unexpected g++ testsuite failures, and the unix script is the only one ever used. A few of the failures respond to configure --enable-sjlj-exceptions. I don't know all the secrets which David Billinghurst has found to get his recent excellent results, but I push in other directions, such as configuring at least 8-byte alignments. Yours is the first report I've seen of success with gnu-pro. -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/