From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 14417 invoked by alias); 22 Oct 2007 09:26:47 -0000 Received: (qmail 14405 invoked by uid 22791); 22 Oct 2007 09:26:45 -0000 X-Spam-Check-By: sourceware.org Received: from mx2.redhat.com (HELO mx2.redhat.com) (66.187.237.31) by sourceware.org (qpsmtpd/0.31) with ESMTP; Mon, 22 Oct 2007 09:26:43 +0000 Received: from int-mx2.corp.redhat.com (int-mx2.corp.redhat.com [172.16.27.26]) by mx2.redhat.com (8.13.1/8.13.1) with ESMTP id l9M9QfCe008394 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO) for ; Mon, 22 Oct 2007 05:26:41 -0400 Received: from zebedee.littlepinkcloud.COM (vpn-14-3.rdu.redhat.com [10.11.14.3]) by int-mx2.corp.redhat.com (8.13.1/8.13.1) with ESMTP id l9M9QdcW027836; Mon, 22 Oct 2007 05:26:39 -0400 Received: from littlepinkcloud.COM (localhost.localdomain [127.0.0.1]) by zebedee.littlepinkcloud.COM (8.13.8/8.13.5) with ESMTP id l9M9Qa9o028633; Mon, 22 Oct 2007 10:26:37 +0100 Received: (from aph@localhost) by littlepinkcloud.COM (8.13.8/8.13.5/Submit) id l9M9Qap5028630; Mon, 22 Oct 2007 10:26:36 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-ID: <18204.27852.59210.963470@zebedee.pink> Date: Mon, 22 Oct 2007 09:26:00 -0000 From: Andrew Haley To: David Daney Cc: Java Patch List Subject: Re: Patch: Add new _Jv_SetGCFreeSpaceDivisor method to libgcj. In-Reply-To: <471AEA80.4040301@avtrex.com> References: <471AEA80.4040301@avtrex.com> X-Mailer: VM 7.19 under Emacs 22.0.93.1 Mailing-List: contact java-patches-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: java-patches-owner@gcc.gnu.org X-SW-Source: 2007-q4/txt/msg00023.txt.bz2 David Daney writes: > As I mentioned previously, GC_free_space_divisor is no longer exported > from libgcj.so. > > Access to this GC tuning parameter is needed on systems where the > maximum heap size is set. If GC_free_space_divisor is not set to a > value larger than its default value of three, the heap can easily get > fragmented in such a manner that OutOfMemoryErrors are thrown even > though there is plenty of free memory. > > Since the loss of the ability to adjust GC_free_space_divisor is a > regression, this should be committed even though we are in stage three. > > Tested on x86_64-pc-linux-gnu in libjava with no regressions. > > OK to commit? > > 2007-10-20 David Daney > > * include/jvm.h (_Jv_SetGCFreeSpaceDivisor): Declare new function. > * boehm.cc (_Jv_SetGCFreeSpaceDivisor): Define it. Sure. Incidentally, my ceasing to export GC_* wasn't a particularly well thought-out decision: I just looked at the symbols we were exporting and decided on a subset I thought was useful. Maybe there are some GC_ symbols that should be exported. Thoughts? Andrew.