From: Pat LaVarre Subject: free udf.ko newbie days Date: 12 Dec 2003 09:44:50 -0700 Sender: linux-fsdevel-owner@vger.kernel.org Message-ID: <1071247490.7851.31.camel@patibmrh9> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from email-out2.iomega.com ([147.178.1.83]:41671 "EHLO email.iomega.com") by vger.kernel.org with ESMTP id S261217AbTLLQo7 (ORCPT ); Fri, 12 Dec 2003 11:44:59 -0500 Received: from royntex01.iomegacorp.com (unknown [147.178.90.120]) by email.iomega.com (Postfix) with ESMTP id F25F71B93 for ; Fri, 12 Dec 2003 09:44:58 -0700 (MST) To: linux-fsdevel@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org Hello I think I now possess a few days of me to pour into discovering why udf.ko frees blocks twice. Want to tell me how to spend those hours? Already we can reproduce this fault in less than a second of 2.6.0-test11 run time. With tweaked test code I now see consistent dump_stack and printk __FUNCTION__, so I can trigger trace by counting calls since last modprobe -r, thus I see consecutive address_space_operations. prepare_write free overlapping ranges of a block.logicalBlockNum: from bit offset 0 for 3, from bit offset 0 for 4, from bit offset 0 for 5, from bit offset 0 for 6, from bit offset 0 for 7 To help build the patch that fixes this, what should I try next? Pat LaVarre