X-Spam-Check-By: sourceware.org From: "Dan Stratila" To: Subject: Re: malloc/new allocate twice as much? Date: Fri, 30 Dec 2005 21:18:02 -0500 Message-ID: <000001c60db0$6e173e20$43055f12@somerville> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Spam-Score: -1.215 Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com Delivered-To: mailing list cygwin AT cygwin DOT com From: Krzysztof Duleba To: cygwin at cygwin dot com Date: Sat, 31 Dec 2005 02:23:20 +0100 Subject: Re: malloc/new allocate twice as much? References: <001e01c60d96$82fc34e0$43055f12 AT somerville> >>This is an issue I've discovered and reproduced while trying to compile >>polymake (see the thread "gcc crash (memory?)"). It seems that both malloc >>and new allocate take up twice the amount of memory needed. >>I wrote a short program to allocate and use memory in chunks (mem-test.cpp), >>and compiled it with g++ 3.4.4. I monitored it using the Windows Task >>Manager. When asked to allocate 100MB in increments of 10MB, the process >>size goes to 100MB and then drops to 0, as expected. However, the total >>commit charge for the machine goes up by 200MB, and then drops back! >Try the snapshot. On my laptop with XP SP2, 20051222 snapshot, 1GB RAM >and 1GB swap file I had no problem allocating 1.8 GB. Yes! This solved the malloc/new problem. Thanks. I sure hope the next release of the cygwin dll will not reintroduce this bug... ;) Now, I am also able to compile polymake, which required g++ to consume up to 600MB of RAM and caused it to crash when I was using the old dll with the buggy malloc/new. Dan -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/