From patchwork Tue Feb 27 18:01:14 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Jan Kara X-Patchwork-Id: 10245935 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id 49BB160362 for ; Tue, 27 Feb 2018 18:02:32 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 38BEA28514 for ; Tue, 27 Feb 2018 18:02:32 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 2D151285E1; Tue, 27 Feb 2018 18:02:32 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 8279528514 for ; Tue, 27 Feb 2018 18:02:31 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751632AbeB0SC2 (ORCPT ); Tue, 27 Feb 2018 13:02:28 -0500 Received: from mx2.suse.de ([195.135.220.15]:51843 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751675AbeB0SBQ (ORCPT ); Tue, 27 Feb 2018 13:01:16 -0500 X-Virus-Scanned: by amavisd-new at test-mx.suse.de Received: from relay2.suse.de (charybdis-ext.suse.de [195.135.220.254]) by mx2.suse.de (Postfix) with ESMTP id CD49AADB9; Tue, 27 Feb 2018 18:01:14 +0000 (UTC) Received: by quack2.suse.cz (Postfix, from userid 1000) id 663D91E04E1; Tue, 27 Feb 2018 19:01:14 +0100 (CET) Date: Tue, 27 Feb 2018 19:01:14 +0100 From: Jan Kara To: Pali =?iso-8859-1?Q?Roh=E1r?= Cc: Jan Kara , Steve Kenton , =?utf-8?Q?Vojt=C4=9Bch?= Vladyka , linux-fsdevel@vger.kernel.org Subject: Re: UDF & open integrity type Message-ID: <20180227180114.yyezafu5bqvylg2o@quack2.suse.cz> References: <20180226193225.k2og3lu6dw46h74w@pali> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <20180226193225.k2og3lu6dw46h74w@pali> User-Agent: NeoMutt/20170421 (1.8.2) Sender: linux-fsdevel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-fsdevel@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP Hi, On Mon 26-02-18 20:32:25, Pali Rohár wrote: > Hi! After doing more experiments with UDF Integrity Type stored in UDF > Logical Volume Integrity Descriptor, it looks like for me that kernel > udf fs driver does not implement handling of open integrity correctly. > > Open Integrity Type (except on VAT disk) mean that filesystem was last > time improperly unmounted, e.g. removable media was disconnected when > some write operation was in progress or when umount on Linux was not > issued. > > Filesystem with Open Integrity type should be checked for error and > ideally also repaired prior next write (or maybe also read) operations. > Probably Open Integrity type can be treated on Linux as Dirty bit in FAT > filesystem. > > But it looks like that kernel udf driver automatically change Open > Integrity Type to Closed when doing unmount, even when Type was Open > prior to UDF mount. For me it does not seem to be correct as kernel > itself does not implement any repair/recovery for UDF filesystem and > automatically lost any information that chkdsk/fsck should be run. > > As there is planned fsck.udf, it is really important to have correct > information of successful umount command stored in filesystem itself. > This can help fsck to know last state of filesystem, like Dirty bit in > FAT or clean state of journal for ext*. Yeah, it makes sense to keep LVID in open state if it was like that when we first saw it. Attached patch should do what you ask for. I'll just probably silence the warning until fsck.udf actually works... Honza From 8aece18ef57588e275be5e2fdcdc00473ee386f3 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Tue, 27 Feb 2018 18:55:31 +0100 Subject: [PATCH] udf: Do not mark possibly inconsistent filesystems as closed If logical volume integrity descriptor contains non-closed integrity type when mounting the volume, there are high chances that the volume is not consistent (device was detached before the filesystem was unmounted). Warn when mounting such volume and don't touch integrity type of the volume so that fsck can recognize it and check such filesystem. Signed-off-by: Jan Kara --- fs/udf/super.c | 11 +++++++++-- fs/udf/udf_sb.h | 1 + 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/fs/udf/super.c b/fs/udf/super.c index 2d4929fa884d..16861ee3b5d5 100644 --- a/fs/udf/super.c +++ b/fs/udf/super.c @@ -1988,7 +1988,13 @@ static void udf_open_lvid(struct super_block *sb) lvidiu->impIdent.identSuffix[1] = UDF_OS_ID_LINUX; ktime_get_real_ts(&ts); udf_time_to_disk_stamp(&lvid->recordingDateAndTime, ts); - lvid->integrityType = cpu_to_le32(LVID_INTEGRITY_TYPE_OPEN); + if (le32_to_cpu(lvid->integrityType) == LVID_INTEGRITY_TYPE_CLOSE) { + lvid->integrityType = cpu_to_le32(LVID_INTEGRITY_TYPE_OPEN); + } else { + UDF_SET_FLAG(sb, UDF_FLAG_NEEDCHECK); + udf_warn(sb, "volume need not be in consistent state. Running " + "fsck is recommended.\n"); + } lvid->descTag.descCRC = cpu_to_le16( crc_itu_t(0, (char *)lvid + sizeof(struct tag), @@ -2028,7 +2034,8 @@ static void udf_close_lvid(struct super_block *sb) lvidiu->minUDFReadRev = cpu_to_le16(sbi->s_udfrev); if (sbi->s_udfrev > le16_to_cpu(lvidiu->minUDFWriteRev)) lvidiu->minUDFWriteRev = cpu_to_le16(sbi->s_udfrev); - lvid->integrityType = cpu_to_le32(LVID_INTEGRITY_TYPE_CLOSE); + if (!UDF_QUERY_FLAG(sb, UDF_FLAG_NEEDCHECK)) + lvid->integrityType = cpu_to_le32(LVID_INTEGRITY_TYPE_CLOSE); lvid->descTag.descCRC = cpu_to_le16( crc_itu_t(0, (char *)lvid + sizeof(struct tag), diff --git a/fs/udf/udf_sb.h b/fs/udf/udf_sb.h index 9dcb475fc74e..02e3aa95d491 100644 --- a/fs/udf/udf_sb.h +++ b/fs/udf/udf_sb.h @@ -29,6 +29,7 @@ #define UDF_FLAG_SESSION_SET 15 #define UDF_FLAG_LASTBLOCK_SET 16 #define UDF_FLAG_BLOCKSIZE_SET 17 +#define UDF_FLAG_NEEDCHECK 18 #define UDF_PART_FLAG_UNALLOC_BITMAP 0x0001 #define UDF_PART_FLAG_UNALLOC_TABLE 0x0002 -- 2.13.6