Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT sources DOT redhat DOT com Delivered-To: mailing list cygwin AT sources DOT redhat DOT com To: Cygwin Mailinglist Subject: Reporting bugs in GCC: a question of procedure From: Jesper Eskilson Organization: Virtutech AB Date: 14 Nov 2001 13:13:48 +0100 Message-ID: Lines: 33 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.1 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii I have a piece of code which causes g++ to loop: g++ -O2 -c gccloopbug.i By specifying -fno-rerun-cse-after-loop: g++ -O2 -fno-rerun-cse-after-loop -c gccloopbug.i the bug disappears and g++ works nicely. Also, if you specify -dt (dump debugging info after the CSE2-pass), the log file gccloopbug.i.cse2 grows to be very large (200Mb before I killed it). I haven't been able to reproduce this on my Linux-machine; mainly because the preprocessed output doesn't compile on my Linux-machine. I was under the impression that preprocessed output should be portable across different hosts, but no. If somebody wants to give it a try, I can send you the code. The bug appears on all of the 2.95.3 series GCC released with Cygwin and MingW. The latest I've tried is the MingW-released 2.95.3-7 version. Should I report this as a GCC bug, even if I haven't been able to reproduce it using the latest GCC version (i.e. 3.0.2, I guess)? Or should I just switch of the rerun-cse-after-loop option and wait until 3.0.2 hits the Cygwin and/or MingW shelves? /Jesper -- Jesper Eskilson Virtutech -- 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/