Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin AT sources DOT redhat DOT com
Message-Id: <200011202245.eAKMjjN27680@plmlir3.mail.eds.com>
From: "Kelley Cook" <Kelley DOT Cook AT home DOT com>
To: "Cygwin mailing list" <Cygwin AT sources DOT redhat DOT com>,
        "GCC Bugs" <gcc-bugs AT gcc DOT gnu DOT org>
Cc: "Zack Weinberg" <zackw AT stanford DOT edu>
Date: Mon, 20 Nov 2000 17:45:26 -0500
Reply-To: "Kelley Cook" <Kelley DOT Cook AT home DOT com>
X-Mailer: PMMail 2000 Standard (2.10.2010) For Windows 2000 (5.0.2195;1)
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Subject: Reason for cygwin GCC 2.97 non-bootstrap found

After much binary searching this weekend, I discovered the reason why
Cygwin hasn't been able to bootstrap since late August.  It currently
fails with a multiple segmentation faults in stage1 -- see:

http://gcc.gnu.org/ml/gcc-bugs/2000-11/msg00392.html
http://gcc.gnu.org/ml/gcc-bugs/2000-09/msg00059.html 
http://gcc.gnu.org/ml/gcc-bugs/2000-08/msg00662.html
http://gcc.gnu.org/ml/gcc-bugs/2000-10/msg00489.html

It was a patch by Zack Weinberg committed on August 24th 2000 -- 
http://gcc.gnu.org/ml/gcc-patches/2000-08/msg01009.html.  

2000-08-24  Zack Weinberg  <zack AT wolery DOT cumb DOT org>

        * ggc-page.c (alloc_page): If HAVE_MMAP_ANYWHERE and we're
        asked for one page, allocate GGC_QUIRE_SIZE of them and put
        the extras on the free list.
        (release_pages): Clean up.
        (ggc_set_mark): Don't adjust G.allocated here...
        (sweep_pages): ... do it here.

Reversing this patch does allow current CVS cygwin to return to
bootstrap land.  Actually all you have to reverse is the first two
hunks (the define and the changes to alloc_page); the cleanup hunks
corresponding to release_pages, gcc_set_mark, and sweep_pages work
fine.

I have tried a few simple workarounds, but it appears that I am not
currently familiar enough with cygwin to attempt a workaround, so
unfortunately I have to leave it up to someone more familiar to fix the
problem.
-- 
Kelley Cook


--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com