Paragon NTFS, HFS+ Instruction Manual

Paragon Technologie GmbH Leo-Wohleb-Straße 8 • 79098 Freiburg, Germany Tel. +49-761-59018-201 • Fax +49-761-59018-130 Website: www.paragon-software.com E-mail: sales@paragon.software.com

Paragon NTFS&HFS+ for Linux 9.5

User manual

User manual 2

Abstract

This document covers implementation of NTFS and HFS+ ﬁle systems support in Linux operating systems using Paragon NTFS&HFS+ ﬁle system driver. Basic installation procedures are described. Detailed mount options description is given. File system creation (formatting) and checking utilities are described. List of supported NTFS/HFS+ features is given with limitations imposed by Linux. There are also advanced troubleshooting section.

Information

All rights reserved. No parts of this work may be reproduced in any form or by any means - graphic, electronic, or mechanical, including photocopying, recording, taping, or information storage and retrieval systems - without the written permission of the publisher.

Products that are referred to in this document may be either trademarks and/or registered trademarks of the respective owners. The publisher and the author make no claim to these trademarks.

While every precaution has been taken in the preparation of this document, the publisher and the author assume no responsibility for errors or omissions, or for damages resulting from the use of information contained in this document or from the use of programs and source code that may accompany it. In no event shall the publisher and the author be liable for any loss of proﬁt or any other commercial damage caused or alleged to have been caused directly or indirectly by this document.

Printed: September 2017 in Freiburg, Germany.

Special thanks to: All the people who contributed to this document, either by writing text, developing solutions to various issues, performing tests, collecting information or by requested support from our team. To our customers who continue to support us and help us to improve the product by constantly demanding more.

We welcome your feedback

Please send your feedback to your Paragon contact or to sales@paragon-software.com.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

3 Paragon NTFS&HFS+ for Linux 9.5

Contents

1 Introduction 4

1.1 Historical review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.2 Paragon UFSD technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.3 How UFSD works on Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.4 Key features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 System requirements 7

2.1 Hardware requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 Software requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Installation 9

3.1 Shipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.2 Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.3 Installing driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.4 Uninstalling driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4 Using The Driver 12

4.1 Getting started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4.2 Mounting and unmounting partitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

4.3 Dirty ﬂag issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

4.4 GPT issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

4.5 Issues with large HDDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5 Mount options 15

5.1 Mount options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

6 Additional Utilities 19

6.1 ufsd utility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

6.2 chkufsd utility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

6.3 NTFS utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

6.3.1 mkntfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

6.3.2 chkntfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

6.4 HFS+ utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

6.4.1 mkhfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

6.4.2 chkhfs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

7 Troubleshooting 26

7.1 Troubleshooting processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

7.2 Mount troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

7.3 The install.sh script can’t ﬁnd kernel sources . . . . . . . . . . . . . . . . . . . . . . . . 27

7.4 Can’t compile the NTFS/HFS+ for Linux driver . . . . . . . . . . . . . . . . . . . . . . . . 28

7.5 “Can’t load module” message at the end of installation . . . . . . . . . . . . . . . . . . . 28

7.6 ufsd module: kernel-module version mismatch . . . . . . . . . . . . . . . . . . . . . . . 28

7.7 ufsd module: create_module: operation is not permitted . . . . . . . . . . . . . . . . . . 29

7.8 insmod: a module named as ufsd already exists . . . . . . . . . . . . . . . . . . . . . . 29

7.9 insmod: Unknown symbol jnl_op (err0) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

7.10 Can’t mount NTFS/HFS+ volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

7.11 Hardware issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

8 Sysdump utility 31

8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

User manual 4

8.2 Main functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

8.3 Using the sysdump utility to collect both platform system information and metadata . . . 31

8.4 Changing the name of output archive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

8.5 Output ﬁle description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

8.6 Delivering collected data to Paragon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

8.7 Privacy policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

9 UFSD driver compatibility 34

9.1 NTFS features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

9.2 HFS+ features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

10 Frequently Asked Questions 35

10.1 What are ’minor errors’ reported by chkntfs utility? . . . . . . . . . . . . . . . . . . . . . 35

10.2 Warnings on Windows7/Vista when NTFS HDD is reconnected from Linux . . . . . . . . 36

10.3 Recently changed ﬁle has its modiﬁcation time a few hours ahead of or behind the

current system time. Why? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

10.4 Why does mount option A make driver ignore mount option B? . . . . . . . . . . . . . . 39

10.5 Does the driver have an optimization for avoiding data fragmentation? . . . . . . . . . . 39

10.6 Why a lot of memory is used for volume mounting? . . . . . . . . . . . . . . . . . . . . . 40

10.7 Why the disk can’t be dismounted? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

5 Paragon NTFS&HFS+ for Linux 9.5

1 Introduction

1.1 Historical review

Historically, different operating systems supported different ﬁle systems. Sharing ﬁles among different platforms was not an easy task. For instance, documents that were created in Windows and are stored on NTFS partitions may be inaccessible under Linux, because Linux does not include full support for NTFS. For example, open-source NTFS-3G NTFS driver does not support random write access to compressed ﬁles.

Paragon NTFS&HFS+ for Linux 9.5 solves these problems — now everyone can access NTFS and HFS+ partitions from Linux in a usual manner with maximum performance and reliability. The driver allows mounting NTFS and HFS+ partitions, so that programs may work transparently with these mounted partitions — browse contents, open documents, run applications, work with existing ﬁles (delete/copy/modify) and create new ones.

Paragon NTFS&HFS+ for Linux 9.5 is a commercial Linux driver for local access to NTFS and HFS+ volumes. It supports full read/write access. The driver is a Kernel module, which guarantees rapid and transparent access to supported ﬁle systems. Mount volumes manually or insert into fstab, and NTFS/HFS+ partitions will be available like any other directory tree.

Paragon NTFS&HFS+ for Linux 9.5 Professional also includes useful additional utilities that provide the ability to check integrity and create NTFS/HFS+ volumes.

1.2 Paragon UFSD technology

UFSD (Universal File System Driver) is an unique technology developed by Paragon Software to provide full access (read/write, format, etc.) to volumes of the popular ﬁle systems: NTFS, exFAT, HFS+ under various platforms, including Windows, Linux, Mac OS X, etc. in case these ﬁle systems are not otherwise supported.

UFSD technology provides access directly to the physical devices that is why it can process partitions regardless of their support by the current Operating System (OS). With UFSD it is possible to mount NTFS, HFS+ and exFAT partitions under Linux, thus getting access to its contents, just the way it is implemented in the NTFS&HFS+ for Linux driver, and the technology also allows direct access via physical device addressing, the way it is implemented in the driver too.

Paragon UFSDs are designed to be readily integrated into any solution using our UFSD Software Development Kit (UFSD SDK), which includes all of the necessary tools to develop applications with the following main features:

• Access to un-mounted partitions (i.e. drive letter not assigned);

• Access to other ﬁle systems that normally would not be supported by the operating system;

• Platform-independent UFSD API.

This software product contains components, which are partly subject to the license terms of the GNU General Public License or GNU Lesser General Public License ("LGPL"). You can request the modiﬁed source code of this software via a contact request: https://www.paragonsoftware.com/contact.html. The offer is valid for at least 3 years from the date of the publication of the corresponding software product. We deliver the software on CD/DVD or USB stick, whose production costs we claim in return.

User manual 6

Note: NTFS, HFS+ and exFAT drivers for Linux as well as utilities were written using UFSD SDK.

1.3 How UFSD works on Linux

Modern operating systems are based on the concept of Installable File System drivers (IFS). User simply needs to provide an operating system with the proper ﬁle system driver to work with the ﬁle system in usual manner. Paragon NTFS&HFS+ for Linux 9.5 includes NTFS and HFS+ drivers for Linux environment. Once appropriate components of Paragon NTFS&HFS+ for Linux 9.5 are installed, the operating system can mount these ﬁle systems and work with directories/ﬁles stored on the ﬁle systems.

1.4 Key features

Paragon NTFS&HFS+ for Linux 9.5 is released in the Express and Professional Editions. All of the products share the following features:

• Transparent read-write access to NTFS and HFS+ volumes — single Kernel module provides support both NTFS and HFS+ ﬁle systems;

• High performance (in some cases even better than Ext4 FS);

• Easy installation and uninstallation (assistant scripts);

• Support for the latest Linux Kernels and distributions;

• Support for SMP kernels;

• Support for iSCSI and SSD storages;

• File sharing over network via SAMBA;

• Low CPU load during data transfers;

• Unlimited ﬁle and volume size (within NTFS/HFS+ and Kernel limitations).

What’s new in Paragon NTFS&HFS+ for Linux 9.5

• Support for Kernel versions from 2.6.36 up to 4.12.x;

• Added read support of new system compression methods on NTFS from Windows 10/Server 2016;

• Improved xattr and symlink support;

• Overall performance enhancements;

• Stability improvements in unsafe removal scenarios;

• Improved automounting script (Professional version).

NTFS speciﬁc features:

• NTFS versions 1.2, 3.0 and 3.1 (Windows NT 4.0, 2000, XP, 2003, Vista, 7, 8.1, 10, Server 2016);

• Support for compressed ﬁles (random access for reading and writing with no limitations);

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

7 Paragon NTFS&HFS+ for Linux 9.5

• Sparse ﬁles support.

HFS+ speciﬁc features:

• Both case sensitive and case instensitive types of HFS+ ﬁle system are supported;

• During ﬁle copy operation (using cp command) on Linux only ‘data’ fork is copied;

• Support replay native HFS+ journal.

NTFS compatibility information:

File system version Comments

NTFS version 1.2 Originates from Microsoft Windows NT 4.0

NTFS version 3.0 Originates from Microsoft Windows 2000

NTFS version 3.1 Originates from Microsoft Windows XP/2003/ Vista/7/8.1/10 and

Server 2016

Additional features of the Professional Edition:

• Full support of the native HFS+ journal;

• Automatic driver rebuilt for newer supported Kernels (support for the DKMS library);

• Automatic NTFS/HFS+ volume mounting;

• Additional utilities:

– ufsd utility - single binary for all ﬁle system utilities;

– chkufsd utility - single binary for cheking every supported ﬁle system;

– mkntfs - format any partition as NTFS under Linux;

– chkntfs - check NTFS partition integrity and ﬁx errors;

– mkhfs- format any partition as HFS+ under Linux;

– chkhfs - check HFS+ partition for integrity and ﬁx errors;

– Sysdump utility - debug utility for collecting volume metadata image and debug system informa-

tion.

User manual 8

2 System requirements

This topic highlights requirements to hardware and software that may be used to run Paragon NTFS&HFS+ for Linux 9.5.

2.1 Hardware requirements

Minimum hardware requirements:

• Processor: Intel Pentium 300 MHz and higher, or compatible;

• both 32- and 64-bit CPUs are supported;

• 16MB of RAM.

Due to unique technology NTFS&HFS+ for Linux drivers have low system requirements. For example, it is enough for our driver to have 650KB of free RAM to work with NTFS partitions larger than 250 GB. NTFS&HFS+ Kernel modules occupy around 800 Kb of RAM.

Real-life values:

• 200 Kb while executing 5 commands like ’dd if=/dev/zero of=/mnt/sda1/test bs=1M count=1000&’ in background.

• 516 Kb while executing ’rsync -r /home /mnt/sda1’ command.

• 17 Mb while compiling bench test on Desktop Linux system in virtual environment using NTFS ﬁle system: a ﬁle-tree with a size about 220 Mb was created and patched, simulating Linux Kernel installation process.

RAM consumption depends ﬁrst of all on whole amount of memory available in the system. If it is low then the driver wouldn’t keep a lot of descriptors opened to keep the memory usage at minimum.

2.2 Software requirements

Supported Linux Kernel versions

• Linux with kernel versions 2.6.36 and newer;

• Linux with kernel versions up to 4.12.x (NTFS&HFS+ driver was tested with Kernels up to 4.12.5).

Linux distributions the products were tested with:

• Ubuntu 17.10;

• Debian 9.1;

• Fedora 26;

• OpenSuse 42.3;

• CentOS 7.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

9 Paragon NTFS&HFS+ for Linux 9.5

Development Environment

A development environment is required to compile Linux drivers and utilities. Please verify that these tools are all functional. The easiest way is to choose the developer toolkit when installing Linux. What must be installed:

• Kernel source code (recommended) or Kernel header ﬁles (doesn’t always work);

#rpm -qa|grep kernel-source (for RPM based kernel-sources)

• GNU C compiler;

#gcc --version

• GNU C++ compiler (for Professional version only);

#g++ --version

• GNU glibc-static (recommended for Professional version only);

• GNU Make;

#make --version

• GNU ld (binutils);

#ld --version

• Modutils (module-init tools);

#insmod -V

• DKMS library (for Professional version only).

#dkms --version

Limitations

• GNU C compiler (gcc) version 4.9 or higher is required.

• The user should login as root to install the drivers and utilities.

• Correct operation is not guaranteed for customized Linux kernels. Commercial porting service to customized Linux kernels is available from Paragon Software Group — for more information send e-mail to sales@paragon-software.com).

User manual 10

3 Installation

This section describes workﬂows related to installing and using Paragon NTFS&HFS+ for Linux driver.

3.1 Shipment

The setup ﬁles for each product of the family are provided as the downloadable TGZ archives, which can be downloaded from the company site.

3.2 Components

The package includes the following components:

• Source ﬁles for the NTFS&HFS+ for Linux driver;

• Assistant script ﬁles, which are purposed to simplify the installation and uninstallation routines;

• Source ﬁles for additional utilities (for Professional edition only);

• Source ﬁles for DKMS library support (for Professional edition only);

• Source ﬁles for automatic mounting integration (for Professional edition only).

Paragon NTFS&HFS+ Linux driver and utilities must be compiled on the end user’s system for correct conﬁguration. By installing the software you accept the terms of End User License Agreement listed in License ﬁle.

3.3 Installing driver

First, NTFS&HFS+ driver must be built and installed.

Steps to install the NTFS&HFS+ for Linux driver are as follows:

1. Log in as root. This step is obligatory;

2. Build and install the Paragon NTFS&HFS+ for Linux 9.5 using install.sh script. Alternatively,

driver binary module may be built manually using ’configure’ ’make driver’ commands.

3. Install the NTFS & HFS+ driver (this step will make the modules available for use);

4. Activating (loading) the driver. After building and installing, the NTFS & HFS+ driver can be refer-

enced as “universal ﬁle system driver” (ufsd) when mounting NTFS and HFS+ partitions.

The steps 1-3 should be made only once while the step 4 is the standard way of using ﬁle system drivers in Linux environment. NTFS & HFS+ for Linux include a set of assistant script ﬁles for the simpliﬁcation of building, installing and uninstalling procedures. Note that these assistant scripts may fail to work in customized Linux conﬁgurations or unsupported Linux distributions. Use install.sh and uninstall.sh script ﬁles to install and uninstall (correspondingly) NTFS & HFS+ driver and utilities. The sections below describe the installation procedure in details.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

11 Paragon NTFS&HFS+ for Linux 9.5

Unpacking Setup Files

The setup ﬁles of the Linux-based version of the NTFS & HFS+ for Linux are provided in the form of a gzip archive. The archive should be copied to the hard disk and decompressed. For example: For the NTFS & HFS+ for Linux driver and utilities:

• create separate folder:

$ mkdir /usr/tmp/ufsd

• change the current directory to the new one

$ cd /usr/tmp/ufsd

• use tar utility to unpack initial archive

$ tar -xf /path/to/the/initial/archive/ufsd\_*.tar.gz

Next step is to build and install the NTFS & HFS+ for Linux driver.

Using the INSTALL.SH Assistant Script

The assistant script "install.sh" provides the extremely easy and ﬂexible way to make the NTFS & HFS+ for Linux and install driver module in the system. Additionally, the script can reconﬁgure OS so that driver module is automatically rebuilt for another supported Kernel version (Professional edition only) and NTFS/HFS+ volumes are automatically mounted with the UFSD driver (Professional edition only). Please note that development tools and kernel sources are required to present on the system and stay in the default locations to build and install the drivers.

Installation

Just run the install.sh script with root privileges:

# ./install.sh or \$ sudo ./install.sh

The assistant script will automatically perform the following actions:

1. Detect the Linux Kernel version;

2. Find kernel header ﬁles and libraries needed for building the drivers;

3. Add service for rebuilding driver module for supported Kernels via the DKMS library (Professional

edition only);

4. Build driver binary modules (jnl.ko and ufsd.ko);

5. Install the driver;

6. Add automatic mounting settings for NTFS/HFS+ volumes (Professional edition only);

7. Build and install additional utilities (Professional edition only);

INSTALL.SH default mode for the NTFS&HFS+ for Linux driver

The assistant script install.sh always names the NTFS&HFS+ for Linux driver module as ufsd (it is the abbreviation of the project name Universal File System Driver). Now you can mount any NTFS/HFS+ partition:

User manual 12

$ sudo mount -t ufsd <device> <mount_point>.

3.4 Uninstalling driver

To completely remove the drivers and the utilities from the current Kernel, one should dismount all NTFS/HFS+ partitions mounted with the driver, uninstall the drivers and unload binary modules from the Kernel. NTFS&HFS+ for Linux provides tools for the drivers/utilities uninstall automation. The assistant script uninstall.sh completely removes the drivers/utilities from the system.

Using the UNINSTALL.SH Assistant Script

The assistant script uninstall.sh provides the extremely easy and ﬂexible way to deactivate and remove the drivers and utilities from the system. The script performs the correct deactivation, uninstallation and the complete removing of the driver’s and utilities’ ﬁles.

Uninstalling

Unmount all currently mounted NTFS and/or HFS+ partitions and then run the ’uninstall.sh’ script:

$ sudo ./uninstall.sh

The assistant script will automatically perform the following actions:

1. Deactivate the driver modules. If the drivers is still in use, the further script execution is aborted;

2. Uninstall the drivers;

3. Remove all binary and source ﬁles of the driver

4. Uninstall utilities (for Professional version only).

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

13 Paragon NTFS&HFS+ for Linux 9.5

4 Using The Driver

After building and installing Paragon NTFS&HFS+ for Linux 9.5 driver, it can be automatically loaded at the system startup. The driver allows to mount supported partitions and provides access to their whole contents.

4.1 Getting started

The goal of this section is to help system engineers to quickly ﬁnd out how to use the product. It describes general approach to mounting partitions using UFSD ﬁle system driver and helps to avoid common issues. We strongly recommend reading this section before starting using our driver.

To mount volume using UFSD driver, standard mount command is used, with FS type set to ufsd, e.g.:

# mount -t ufsd /dev/sda1 /mnt/sda1

After this command is executed there can be several mount scenarios for a disk (for more information see Mount troubleshooting subsection):

• The disk is “clean” (without any errors), mounted by the driver and ready to use.

• Disk can’t be mounted. In this case can be several scenarios:

1. Disk has “dirty” ﬂag set (for more information see Dirty ﬂag issues subsection):

– Use chkntfs/chkhfs utilities with -a -f options to check the volume for errors and incon-

sistencies and ﬁx them (if any). This is recommended approach (see ?? or ?? subsections);

– Use ‘force’ mount option (see Dirty ﬂag issues subsection).

2. The disk is a GPT-partitioned disk — check GPT issues subsection for more information.

3. Follow other steps on the Mount troubleshooting diagram to ﬁnd the cause of the issue. Update/ﬁx Kernel source code (if necessary) and request new driver for the new Kernel from Paragon Software.

Analyze returned status and check output of (dmesg | tail). In case of failure, follow the ?? to ﬁnd possible causes and and try to mount the partition again using the same or different mount options, if needed (see Mount options subsection).

If there is still a problem mounting the partition ﬁll out Paragon’s online request form from your user account so we could help you with the issue.

4.2 Mounting and unmounting partitions

Mounting NTFS and HFS+ volumes

To mount volume using UFSD driver, use mount command with FS type set to ufsd, e.g.:

# mount -t ufsd /dev/sda1 /mnt/sda1

Sometimes volumes cannot be mounted using ’mount’ command, for example:

# mount -t ufsd /dev/sdd1 /mnt/sdd1 mount: wrong fs type, bad option, bad superblock on /dev/sdd1,

missing codepage or helper program, or other error In some cases useful info is found in syslog - try

User manual 14

dmesg | tail or so

In that case more information can be obtained using the following command:

# dmesg | tail | grep ufsd [ 369.844741] ufsd: volume is dirty and "force" flag is not set

Lines related to UFSD driver start with ’ufsd:’ preﬁx. In case Kernel log is redirected to console, messages may be output to terminal window right after mount command is issued:

# mount -t ufsd /dev/sdd1 /mnt/sdd1 ufsd: volume is dirty and "force" flag is not set mount: wrong fs type, bad option, bad superblock on /dev/sdd1,

missing codepage or helper program, or other error In some cases useful info is found in syslog - try dmesg | tail or so

In this example dmesg output contains information that the volume being mounted is ‘dirty’. See Dirty

ﬂag issues subsection for more information on ‘dirty’ ﬂag and how to resolve the issue.

Dismount volumes

To dismount volumes mounted by UFSD driver issue usual umount command:

# umount /dev/sdd1

See Why the disk can’t be dismounted? subsection for more information, if the volume can’t be dismounted.

4.3 Dirty ﬂag issues

‘Dirty’ ﬂag is special feature implemented in most of the modern ﬁle systems, including NTFS and HFS+. This ﬂag is set after volume is mounted in read/write mode and cleared after volume is correctly unmounted (see notes on ’force’ mount option for more information). Without ‘dirty’ ﬂag it is impossible to tell if given volume was correctly unmounted or not. Detecting incorrectly unmounted volumes helps to detect possible errors as early as possible. Thus, this ﬂag helps to preserve ﬁle system consistency. Please note that even in case ‘dirty’ ﬂag is set on the volume, ﬁle system is not necessarily corrupt.

Paragon NTFS&HFS+ for Linux drivers from version 8 support ‘dirty’ ﬂag on both NTFS and HFS+. By default, driver refuses to mount volumes with ‘dirty’ ﬂag set. Recommended course of action is to check the volume for errors and repair any inconsistencies found using chkntfs/chkhfs utility with

-a -f command line options (see Additional Utilities). Run with -a command line option, the utilities check dirty ﬂag state and in case it is set, they performs all necessary checks. If ‘dirty’ ﬂag is not set, ﬁle system checking utilities exits immediately. If -f command line option is speciﬁed, the utilities repair any errors or inconsistencies that they ﬁnd and ﬁnally clear ‘dirty’ ﬂag. This approach is similar to the way Windows and MacOS handle ‘dirty’ volumes. See corresponding sections on NTFS and HFS+ utilities and ‘File system checking utilities return codes’ section for more information.

To make driver mount dirty volumes without checking for possible errors and correcting them, ’force’ mount option can be used (while it is not recommended). This way, ‘dirty’ ﬂag is not cleared and any possibly existing errors or inconsistencies are not ﬁxed. ‘Dirty’ ﬂag will remain set until volume is checked for errors using Paragon chkntfs/chkhfs with -f command line option or using Windows chkdsk utility with /f switch or MacOS Disk Utility (for NTFS and HFS+ volumes, respectively).

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

15 Paragon NTFS&HFS+ for Linux 9.5

4.4 GPT issues

Some Linux Kernels do not behave correctly when there is EFI partition on GPT-partitioned devices. This is most often the case with HDDs partitioned using MacOS Disk Utility.

This leads to seemingly wrong operation of UFSD driver(s) that refuse to mount partition. In that case fdisk may report that there is only one EFI partition on the device, ignoring some or all of the following NTFS and/or HFS+ partition(s). To work around the issue, attention must be paid to volume type reported by fdisk -l command on GUID-partitioned disks.

Example:

# fdisk -l Disk /dev/sda: 80.0 GB, 80026361856 bytes 255 heads, 63 sectors/track, 9729 cylinders Units = cylinders of 16065*512 = 8225280 bytes

Device Boot Start End Blocks Id System /dev/sda1 1 9730 78150743+ ee EFI GPT

fdisk reports that /dev/sda only contains single EFI partition that spans via entire disk. Neverthe- less, mounting partition /dev/sda2 to /mnt/hda succeedes:

# mount -t ufsd /dev/sda2 /mnt/hda

And after that mount command issued without arguments lists, among others, mounted partition

/dev/sda2 that was not listed by fdisk -l (marked with red below).

/dev/root on / type squashfs (ro) none on /dev type devfs (rw) none on /proc type proc (rw,nodiratime) devpts on /dev/pts type devpts (rw) none on /sys type sysfs (rw) none on /tmp type ramfs (rw) /dev/mtdblock/2 on /usr/local/etc type yaffs (rw,noatime) /dev/rd/0 on /mnt/rd type vfat (rw,nodiratime,fmask=0022,dmask=0022,

,→codepage=cp437,iocharset=iso8859-1)

/dev/sda2 on /mnt/hda type ufsd

(rw,nodiratime,nls=iso8859-1,uid=0,gid=0,fmask=22,dmask=22,nocase)

/dev/scsi/host2/bus0/target0/lun0/part1 on /tmp/usbmounts/sdb1 type ufsd

,→(ro,nodiratime,nls=utf8,uid=0,gid=0,fmask=0,dmask=0)

4.5 Issues with large HDDs

Though our driver supports partitions larger than 2 Tb (tested on 20 Tb partitions on real hardware and on 25 Tb partitions in virtual environment), not all versions of Linux Kernels support block devices larger than 2 Tb on all possible interfaces. E.g. Ubuntu 10.04 does not support 2.5 Tb SATA HDD attached via USB->SATA converter, while the same HDD with the same converter is mounted OK on Windows 7 and the same HDD connected to Ubuntu 10.04 via SATA interface can be mounted and used successfully.

If there is similar issue, please perform the test cases described above to make sure where the root cause of the issue is (in Paragon’s driver or in Linux Kernel).

User manual 16

5 Mount options

This section describes mount options for mounting supported ﬁle system partitions.

5.1 Mount options

SYNOPSYS

mount -t ufsd [-o options] device mount_point

Option NTFS HFS+ Expected behavior and examples

iocharset or nls

+ + -o iocharset=NAME1[,iocharset=NAME2]

-o nls=NAME1[,nls=NAME2]

-o codepage=NAME1[,codepage=NAME2]

The NTFS/HFS+ ﬁle systems store all ﬁle/directory names in Unicode format (UTF-16), which can represent any character from any language. In case none of these options is set, the default codepage will be used (CONFIG_NLS_DEFAULT). If none of the speciﬁed codepages exist on the system, the default codepage will be used again. This option informs the driver how to interpret path strings and translate them to Unicode and back. Up to 8 different code pages can be speciﬁed. The driver tries to use the codepages from speciﬁed list in order until it manages to translate all the characters in the string. If none of the speciﬁed codepages allows to translate all the characters, Kernel’s default codepages is used.

Note:

• Paragon driver uses extended UTF-8 for Unicode number U+10000 characters support when ’=utf8’ is speciﬁed.

• codepage, nls and iocharset mount option must be used in the form:

codepage=... nls=cp... iocharset=cp...

Examples:

• nls=utf8

• iocharset=utf8

nocase + - -o nocase

All ﬁle and directory operations (open, ﬁnd, rename) are case insensitive. Casing is preserved in the names of existing ﬁles and directories.

showmeta + + -o showmeta

Use this parameter to show all meta-ﬁles (System Files) on a mounted NTFS/HFS+ partition. By default, all meta-ﬁles are hidden.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

17 Paragon NTFS&HFS+ for Linux 9.5

Option NTFS HFS+ Expected behavior and examples

noatime + + -o noatime

All ﬁles and directories will not update their last access time attribute if a NTFS/HFS+ partition is mounted with this parameter. This option can speed up ﬁle system operation.

uid + + -o uid=USERID

By default all existed ﬁles on a mounted NTFS/HFS+ volume are owned by root, while created ﬁles are owned by the user. Specifying the uid parameter you can set an owner of ﬁles. The userid can be any name from /etc/passwd, or any number representing a user id.

File’s owner can be changed by the chown command, but this change will be effective until volume unmount.

gid + + -o gid=GROUPID

By default all existed ﬁles on a mounted NTFS/HFS+ volume are owned by group root, while created ﬁles are owned by the user’s group. Specifying the gid parameter you can set a owner group of the ﬁles. The groupid can be any name from /etc/group, or any number representing a group id.

File’s group can be changed by the chown command, but this change will be effective until volume unmount.

umask + + -o umask={VALUE}

The default permissions given to a mounted NTFS/HFS+ volume are rwx------ (for security reasons). The umask option controls these permissions for ﬁles/directories created after the volume is mounted.

mount -t ufsd /dev/hda1 /mnt/ntfs_0 -o umask=0222

fmask dmask

+ + -o fmask=VALUE

-o dmask=VALUE

umask option changes the permissions for new created ﬁles and

directories; fmask is applied to ﬁles; dmask to directories that already exist on a mounted volume. The effect of these options can be combined. To mount Samba, FTP or NFS shares the combination of umask=000, fmask=000,dmask=000 is usually speciﬁed.

ro + + -o ro

To mount an NTFS/HFS+ volume in read-only mode.

User manual 18

Option NTFS HFS+ Expected behavior and examples

bestcompr + - -o bestcompr

Instructs the driver to use highest compression level when writing compressed ﬁles. High CPU-load.

sparse + - -o sparse

Create new ﬁles as “sparse”. This feature allows creating holes inside new created ﬁles (avoids ﬁlling unwritten space with zeroes). This option is not recommended in case NTFS partition is used for BitTorrent downloads.

force + + -o force

Not recommended for use.

Forces the driver to mount partitions even if ‘dirty’ ﬂag (volume dirty) is set. It is recommended to use Paragon or OS-speciﬁc ﬁle system checking utility before mounting ‘dirty’ partitions to reset the ’dirty’ ﬂag.

Note that if ‘dirty’ volume was mounted with ‘force’ mount option, dirty ﬂag will not be cleared when volume is unmounted using umount command.

nohidden + - -o nohidden

Files with the Windows-speciﬁc HIDDEN attribute will not be shown under Linux.

sys_im-

+ - -o sys_immutable

mutable

Files with the Windows-speciﬁc SYSTEM attribute will be marked as system immutable ﬁles. This option is not supported for symlink/ﬁfo/socket ﬁles.

acl + + -o acl

Support POSIX ACLs (Access Control Lists). Effective if supported by Kernel. Not to be confused with NTFS ACLs.

The option speciﬁed as acl enables support for POSIX ACLs. Supported if driver is built with support for ext2-like handling of ﬁle/directory permissions. Recommended for use with the HFS+ volumes.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

19 Paragon NTFS&HFS+ for Linux 9.5

Option NTFS HFS+ Expected behavior and examples

user_xattr + + -o user_xattr

Support for extended attributes. Effective if supported by the Kernel.

Since driver version 9.4.4, this mount option is considered obsolete and it’s functionality is enabled by default in UFSD driver.

sync + + -o sync

All ﬁle system operations will be done synchronously. When used, UFSD driver performance may be decreased.

This option is recommended for use, when disk can be detached without proper unmount, as it helps to minimize the possibility of the ﬁle system errors, when proper unmount for the volume is not performed.

discard + + -o diskcard

discard is the mount option in the UFSD driver, which is recommended for use with the solid-state drives (SSD) to enable support of the TRIM command for improved performance on delete operations..

* Supported in driver version 9.3 and newer.

User manual 20

6 Additional Utilities

Additional utilities for Paragon NTFS&HFS+ for Linux 9.5 provide the ability to check integrity and create NTFS/HFS+ volumes on block devices from your Linux OS. Additional utilities for Paragon NTFS&HFS+ for Linux 9.5 were developed with Paragon UFSD SDK.

6.1 ufsd utility

ufsd utility is the single ﬁle system utility binary, which is included into the package since Paragon

NTFS&HFS+ for Linux version 9.4.3.

Synopsis

ufsd [utility [arguments]...] ufsd [--version]

E.g.: ’ufsd chkntfs -f /dev/sdb1’

Utilities

chkntfs - check and ﬁx NTFS volume

chkhfs - check and ﬁx HFS volume

chkufsd - check and ﬁx NTFS/HFS+ volume (universal utility)

mkntfs - format volume into NTFS

mkhfs - format volume into HFS+

Description

ufsd is the multi-call binary that combines Paragon ﬁle system formatting and checking utilities into a single executable.

All utilities are actually hardlinks to the ufsd utility. In order to make sure that the ﬁles are not taking unnecessary space, it is advised to unpack provided archive on the target platform, or create hardlinks to the ufsd utility manually. You may also use symlinks to achieve the identical functionality or pass utility name as an argument to the ufsd binary, similar to how busybox binary operates.

6.2 chkufsd utility

chkufsd utility is the combined utility for checking and ﬁxing all ﬁle systems, supported by Paragon

NTFS&HFS+ for Linux 9.5.

Synopsis

chkufsd [option] device

E.g.: ’chkufsd -f /dev/sdb1’

Options

-fs:ntfs Treat target volume as NTFS

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

21 Paragon NTFS&HFS+ for Linux 9.5

-fs:hfs Treat target volume as HFS+

-f Fix errors on the disk

-a Perform checks only if ‘dirty’ ﬂag is set

-b:size Override default block (sector) size. Usage of default settings are strongly

recommended. The ’-b’ parameter can use following values: 512, 1024, 2048, 4096.

-m:size Set a limit for memory usage during ﬁle system check. Usage of this parameter may lower ﬁle system check performance.

-h Display this help

-short Minimum ﬁle system check

-safe Errors are not ﬁxed, dirty ﬂag is cleared if no errors found

-showminors Show minor errors

-no-orphans Do not restore real orphan ﬁles

-trace Turn on UFSD trace

-verbose Explain what is being done

-nopercents Do not print percents during checking process

-version Show version and exit

Description

chkufsd is a combined utility that allows to check and ﬁx ﬁle systems supported by the Paragon driver.

Note: Since version 9.4.3 this tool is actually a hardlink to the ufsd utility. In order to make sure that the ﬁles are not taking unnecessary space, it is advised to unpack provided archive on the target platform, or create hardlinks to the ’ufsd’ utility manually.

6.3 NTFS utilities

There are 2 basic utilities for NTFS ﬁle system:

• mkntfs — format any partition as NTFS under Linux

• chkntfs — check NTFS partition for integrity and (optionally) ﬁx errors

6.3.1 mkntfs

mkntfs utility creates NTFS volumes (1.2, 3.0, 3.1 (Windows NT 4.0/2000/XP/2003/Vista/7/8.1/10) ﬁle system) on user speciﬁed (block) device (disk partition) under Linux OS.

User manual 22

Synopsis

mkntfs [options] device

E.g.: ’mkntfs -f /dev/sdb1’

Options

-v:label Specify volume label.

-c Files created on the new volume will be compressed by default.

-a:size Override the default allocation unit size. Default settings are strongly

recommended for general use. NTFS supports 512, 1024, 2048, 4096, 8192, 16K, 32K, 64K. File compression is not supported for allocation unit size above 4096.

-b:size Override the default block (sector) size. Default settings are strongly

recommended for general use. One can use 512, 1024, 2048, 4096.

-m:size Override default MFT record size. Default settings are strongly recom-

mended for general use. One can use 512, 1024, 2048, 4096.

-j:size Set journal size for Paragon journal on NTFS volume. Supported sizes:

from 1MB to 512MB. By default Paragon journal for NTFS is disabled.

-f Force the format without conﬁrmation.

-s:start Specify “hidden” sectors in the boot area.

-g:tracks:sectors Specify disk geometry that should be written in the boot area.

tracks speciﬁes number of tracks per disk side.

sectors speciﬁes number of sectors per track.

The most popular geometries are:

• NORMAL: 63 sectors per track and 15(16) tracks per cylinder.

• LBA: 63 sectors per track and 255 tracks per cylinder.

Generally Windows uses the LBA geometry (-g:255:63). If -g is not speciﬁed, the utility obtains geometry from OS.

-winxp Create NTFS compatible with Windows XP (default)

-winvista Create NTFS compatible with Windows Vista

-win7 Create NTFS compatible with Windows 7

-help Display this help

-nodiscard Do not discard volume before formatting

-trace Turn on UFSD trace

-verbose Explain what is being done

-nopercents Do not print percents during format process

-version Show the version and exit

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

Paragon Technologie GmbH

23 Paragon NTFS&HFS+ for Linux 9.5

Description

mkntfs is a standalone utility that allows to format NTFS partitions under Linux. It is used to create a NTFS 1.2, 3.0, 3.1 (Windows NT 4.0/2000/XP/2003/Vista/7/8.1/10) ﬁle system on adevice (usually a disk partition).

Notes:

1. mkntfs doesn’t change the MBR (Master Boot Record) when formatting a partition. Therefore, most of Linux commands (like fdisk -l) will be unable to determine that partition’s ﬁles system was changed to NTFS.

2. Since version 9.4.3 this tool is actually a hardlink to the ufsd utility . In order to make sure that the ﬁles are not taking unnecessary space, it is advised to unpack provided archive on the target platform, or create hardlinks to the ufsd utility manually.

6.3.2 chkntfs

chkntfs utility performs consistency checking of NTFS volumes and (optionally) ﬁxes errors.

Synopsis

chkntfs [options] device

E.g.: ’chkntfs -f /dev/sdb1’

Options

-f Fix errors on the disk.

-a Perform checks only if ‘dirty’ ﬂag is set.

-b:size Override default block (sector) size. Usage of default settings are

strongly recommended. The ’-b’ parameter can use following values: 512, 1024, 2048, 4096.

-m:size Set a limit for memory usage during ﬁle system check. Usage of this parameter may lower ﬁle system check performance.

-h Display this help.

-m:size Memory limit used by the utility

-short Minimum ﬁle system check

-safe Errors are not ﬁxed, dirty ﬂag is cleared if no errors found

-showminors Show minor errors.

-no-orphans Do not restore real orphan ﬁles

-trace Turn on UFSD trace.

-verbose Explain what is being done.

-nopercents Do not print percents during checking process.

-version Show version and exit.

User manual 24

Description

chkntfs creates and displays a status report about a NTFS ﬁle system. chkntfs also lists and corrects errors on the disk, if any (-f ﬂag must be speciﬁed).

Notes:

1. when --no-orphans option is used, real orphan ﬁles will be deleted. Without this option chkntfs restores real orphan ﬁles to found.XXX folders.

2. Since version 9.4.3 this tool is actually a hardlink to the ufsd utility. In order to make sure that the ﬁles are not taking unnecessary space, it is advised to unpack provided archive on the target platform, or create hardlinks to the ufsd utility manually.

6.4 HFS+ utilities

There are 2 basic utilities for HFS+ ﬁle system:

• mkhfs — format any partition as HFS+ under Linux

• chkhfs — check HFS+ partition for integrity and (optionally) ﬁx errors

6.4.1 mkhfs

mkhfs ustility creates an HFS+ volume on speciﬁed (block) device (disk partition) under Linux OS.

Synopsis

mkhfs [options] device

E.g.: ’mkhfs -j /dev/sdb1’

Options

-v:label Specify the volume label.

-a:size Override the default allocation unit size. Default settings are strongly recom-

mended for general use. HFS+ supports 512, 1024, 2048, 4096, 8192, 16K, 32K and 64K.

-ne:size Specify extents b-tree node size: 512-32K.

-nc:size Specify catalog b-tree node size: 4K-32K.

-na:size Specify attributes b-tree node size: 4K-32K

-f Force the format without conﬁrmation.

-j Make volume journalized with default journal size

-j:size Make volume journalized, manually setting size of the journal

-c Make volume case-sensitive.

-h Display this help

-help Display this help.

-nodiscard Do not discard volume before formatting

-trace Turn on UFSD trace.

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

Paragon Technologie GmbH

25 Paragon NTFS&HFS+ for Linux 9.5

-verbose Explain what is being done.

-nopercents Do not print percents during format process.

-version Show the version and exit.

Description

mkhfs is a standalone utility that allows to format HFS+ partitions under Linux. It is used to create an HFS+ ﬁle system on a device (usually a disk partition).

6.4.2 chkhfs

chkhfs utility provides consistency checking of a HFS+ volume and ﬁx errors.

Synopsis

chkhfs [options] device

E.g.: ’chkhfs -f /dev/sdb1’

Options

-f Fix errors on the disk.

-a Perform checks only if ‘dirty’ ﬂag is set.

-b:size Override default block (sector) size. Usage of default settings are

strongly recommended. The ’-b’ parameter can use following values: 512, 1024, 2048, 4096.

-m:size Set a limit for memory usage during ﬁle system check. Usage of this parameter may lower ﬁle system check performance.

-h Display this help.

-m:size Memory limit used by the utility

-short Minimum ﬁle system check

-safe Errors are not ﬁxed, dirty ﬂag is cleared if no errors found

-showminors Show minor errors.

-no-orphans Do not restore real orphan ﬁles

-trace Turn on UFSD trace.

-verbose Explain what is being done.

-nopercents Do not print percents during checking process.

-version Show version and exit.

User manual 26

Description

chkhfs creates and displays a status report about a HFS+ ﬁle system. chkhfs also lists and corrects errors on the disk, if any (-f ﬂag must be speciﬁed).

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

27 Paragon NTFS&HFS+ for Linux 9.5

7 Troubleshooting

This section highlights troubleshooting processes.

7.1 Troubleshooting processes

Step 1. Consult Documentation

Please consult documentation to make sure that encountered behavior is not by design, with special attention given to the part related to binary module installation, testing and troubleshooting as well as to section on Hardware requirements and Software requirements. Please also review Mount

troubleshooting and Using The Driver subsection.

Step 2. Make sure the issue is not related to Linux itself

Now, make sure that root cause of the issue is not related to Linux itself. For example, if an issue is discovered while performing certain ﬁle system-related operation on a volume mounted with Paragon NTFS&HFS+ for Linux 9.5 driver, make sure the same issue is not observed when the same operation is performed on ‘native’ ﬁle system like Ext2fs, Ext3fs or FAT (except, of course, for operations speciﬁc to NTFS/HFS+ ﬁle systems or to Paragon’s driver itself, e.g. IOCTLs, additional utilities and so on).

Step 3. Prepare to report the issue

After performing previous steps and making sure that the issue is related to Paragon NTFS&HFS+ for Linux 9.5 driver, prepare to report the issue to Paragon.

CO LLE CT ALL IN FOR MATION O N T HE ISS UE The most important point in issue resolution process is quickly obtaining all the information related to the issue. Quick collection of required information is the key to resolving an issue faster.

Step 4. Assist Paragon engineers to resolve the issue quickly

Try to provide as detailed information on the issue, as possible.

User manual 28

7.2 Mount troubleshooting

Use our mount troubleshooting diagram for faster mount issue resolution.

7.3 The install.sh script can’t ﬁnd kernel sources

1. Read Software requirements section, make sure all tools are functional. For more information, please read kernel documentation.

2. Linux Kernel must be conﬁgured correctly.

3. Make sure that you have kernel sources, for example, in the /usr/src/linux-x.x.xx directory, where x.x.xx is your kernel version (for example, 4.4.9). Type uname -r in the command line to know your current kernel version.

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

Paragon Technologie GmbH

29 Paragon NTFS&HFS+ for Linux 9.5

4. Create a symbolic link from the /usr/src/linux-x.x.xx directory to /usr/src/linux. To create the link type

ls -s /usr/src/linux-$(uname-) /usr/src/linux

5. Make sure that you have the config-x.x.xx ﬁle, for the booted Linux kernel, in the /boot directory. If you haven’t the config-x.x.xx ﬁle then type

ls -s /usr/src/linux-$(uname-r)/.config /boot/config-$(uname-r)

to create a symbolic link to the conﬁg ﬁle.

Note: There are cases when the kernel sources may be located in other directories. In these cases you should create a symbolic link to /usr/src/linux, for example,

ls -s /lib/modules/$(uname-r)/build /usr/src/linux

If you still have the same problem i.e. install.sh script can’t ﬁnd the kernel sources it is better to rebuild your kernel or download and build a stable kernel from the www.kernel.org site.

7.4 Can’t compile the NTFS/HFS+ for Linux driver

1. Read Software requirements section, make sure all tools are functional. For more information, please read kernel documents.

2. Linux kernel must be conﬁgured correctly.

3. The /boot directory must contain the conﬁg-(kernel version) ﬁle. If the ﬁle is missing you should execute the following command:

ls -s /usr/src/linux-$(uname-r)/.config /boot/config-$(uname-r)

7.5 “Can’t load module” message at the end of installation

1. Make sure that you use the same version of GCC compiler that was used for kernel compilation.

2. Make sure that Makefile of the kernel (you can ﬁnd Makefile in the directory where the kernel sources are located) have the correct kernel version at the beginning of the ﬁle. For example: if your loaded kernel version is 4.4.9-300.fc23.x86_64 then the following lines must be found at the beginning of the Makeﬁle:

VERSION = 4 PATCHLEVEL = 4 SUBLEVEL = 9 EXTRAVERSION = -300.fc23.x86_64

7.6 ufsd module: kernel-module version mismatch

That means kernel version mismatch.

1. Check kernel source version in /usr/include/linux/version.h

2. Check the currently running kernel version: uname -r

3. Both versions must match.

User manual 30

4. If they don’t match, please restore Kernel conﬁguration or recompile kernel anew (advanced).

7.7 ufsd module: create_module: operation is not permitted

That means you must have root privilege to load driver.

7.8 insmod: a module named as ufsd already exists

1. Make sure that you use the same version of GCC compiler that was used for kernel compilation.

VERSION = 4 PATCHLEVEL = 4 SUBLEVEL = 9 EXTRAVERSION = -300.fc23.x86_64

7.9 insmod: Unknown symbol jnl_op (err0)

That means one is trying to load ufsd.ko module into the Linux Kernel before jnl.ko module has been loaded:

# insmod ufsd.ko insmod: ERROR: could not insert module ufsd.ko: Unknown symbol in module # dmesg | tail [84199.673358] ufsd: Unknown symbol jnl_op (err 0)

Load jnl.ko module ﬁrst and ufsd.ko afterwards:

# insmod jnl.ko # insmod ufsd.ko # lsmod | grep ufsd ufsd 614400 0 jnl 36864 1 ufsd

7.10 Can’t mount NTFS/HFS+ volume

1. Make sure that the driver is activated (loaded into the Kernel) via following command:

lsmod | grep ufsd

2. Make sure that the driver supports ﬁle system mounted partition is formatted with checking its version:

cat /proc/fs/ufsd/version

3. The volume is dirty. Use chkntfs/chkhfs utility with -a -f command line options to reset ’dirty’ ﬂag. Alternatively, use ’force’ mount options to make the driver ignore ’dirty’ ﬂag.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

31 Paragon NTFS&HFS+ for Linux 9.5

7.11 Hardware issues

Sometimes storage issues are caused by a faulty hardware. Hardware issues can be classiﬁed by their place of origin:

1. storage device (HDD, SSD, USB ﬂash drive, SD card, etc.)

2. target platform (hardware board which utilizes the UFSD driver)

3. data medium (e.g. USB cable)

Often this kind of issues can be identiﬁed based on the system log output (e.g. dmesg). The basic principle is to look for an error messages from the underlying subsystem on top of which the UFSD driver operates (e.g. USB controller). Some examples:

» usb*-*: USB disconnect, device number

(in case the USB device was not actually disconnected)

» end_request: I/O error, dev sdX, sector » Buffer I/O error on device sdX, logical block » mmcblk*: error -110 transferring data, sector » sd*:*:*:*: [sdX] Unhandled error code

» sd*:*:*:*: [sdX] Spinning up disk........not responding...

» sd*:*:*:*: [sdX] Device not ready » xhci_hcd*:*:*: WARN: Stalled endpoint

***

*** ***

Identiﬁcation of the faulty hardware cannot be done solely on the UFSD driver messages because the driver operates on a higher level than HW drivers. Still UFSD messages can be helpful for obtaining a general picture of the issue. Here are some typical messages in case of hardware issues:

» ufsd: failed to read block 0xNNNN » ufsd: failed to map block 0xNNNN » ufsd: failed to write block 0xNNNN » ufsd: bio read I/O error

A hardware problem cannot be solved at the ﬁle system driver level. Thereby that kind of issues is out of the Paragon’s scope of responsibility. The best solution is to contact the faulty hardware’s manufacturer technical support. One should contact us only if the hardware fault is followed by a system hang or kernel panic caused by the UFSD driver.

User manual 32

8 Sysdump utility

8.1 Introduction

In case Paragon team requires additional information on the test storage device and/or test platform itself for the issue troubleshooting with speciﬁc NTFS/HFS+ volume(s), the sysdump utility can be used to capture very compact images of volumes with ﬁle system inconsistencies. Additionally it could be used to collect test platform system information: storage device summary, Kernel version, loaded UFSD driver version and so on.

As volume images only include ﬁle system metadata, the risk of leaking sensitive data when using sysdump is minimized.

8.2 Main functions

Main functions of the sysdump utility are:

• Capturing metadata image of the test volume

• Collecting HW sample system information.

Note that sysdump utility requires root privileges as they are needed for working with storage devices on Linux platform.

When utility is run without additional options basic help is displayed

# ./sysdump Dump NTFS/HFS/Ext volumes (metafiles only) and collect system information

,→.

Paragon’s Sysdump utility doesn’t collect any personal or user-sensitive

,→information.

Please refer to the ’Privacy policy’ described in the Sysdump utility

,→guide for more information.

Usage: sysdump [-s] [device] [-o <filename>]

-s collect system information

-o new file name (if name without suffix ".tar", then suffix ".tar"

,→will be added automatically)

E.g. sysdump /dev/hdb1

8.3 Using the sysdump utility to collect both platform system information and metadata

1. Extract sysdump utility to a folder with read/write access.

2. In terminal change to the folder with the extracted sysdump utility. Use the following command to capture the metadata image and to collect platform system information:

# ./sysdump -s /path/to/partition

3. This will create an archive named, based on platform parameters <Architecture>.<Kernel version>.<date>.tar with storage device metadata image, its md5 sum and sample system

information:

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

33 Paragon NTFS&HFS+ for Linux 9.5

# ./sysdump -s /dev/sdc1 Scanning NTFS... Added 13 files/dirs from $Extend . Recognized as NTFS. Dumping "/dev/sdc1" (465.76 Gb) ... Dump finished. File size 1248334848 bytes (including tail) Collecting system information... System information was saved to /home/UFSD_utilites/i686

,→.3.13.0.20141105.tar

8.4 Changing the name of output archive

Name of the output archive can be changed by using the additional ’-o’ parameter and adding new ﬁle name. For example:

# ./sysdump -s /dev/sdc1 -o test.tar Scanning NTFS... Added 13 files/dirs from $Extend . Recognized as NTFS. Dumping "/dev/sdc1" (465.76 Gb) ... Dump finished. File size 1248334848 bytes (including tail) Collecting system information... System information was saved to /home/UFSD_utilites/test.tar

8.5 Output ﬁle description

The Sysdump utility creates a single tar archive in the working folder with its’ work results for easier data transfer to Paragon team. This archive can be easily uncompressed with the tar utility, e.g.:

# tar -xf ./x86_64.3.11.9.20131211.tar

There are several ﬁles inside archive:

• dmesg - contents of the ’dmesg’ command output

• dumpbin.gz - compressed metadata image of the test volume

• dumpbin.md5 -md5 checksum for the collected image itself, not the archive ﬁle

• modules - list of all Kernel modules (*.ko ﬁles) found on the platform

• sysinfo - ﬁle with platform system information: list of storage devices, Linux Kernel version, version

of the loaded UFSD driver, CPU, memory information, etc.

8.6 Delivering collected data to Paragon

Please send the tar ﬁle, generated by the ’Sysdump’ utility to your support contact at Paragon Software via e-mail, via ’My account’ page at Paragon Software Support Portal.

8.7 Privacy policy

Paragon’s Sysdump utility doesn’t collect any personal or user-sensitive information. Platform data is obtained via usage of the native Linux utilities and standard system ﬁles (e.g. fdisk, parted,

User manual 34

/proc/cpuinfo, /proc/meminfo and others). The resulted volume metadata image includes only ﬁle system metadata and risk of leaking sensitive data when using sysdump utility is minimized. Nevertheless, collected information is also covered by the mutual Non-disclosure agreement and couldn’t be forwarded to any third party.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

35 Paragon NTFS&HFS+ for Linux 9.5

9 UFSD driver compatibility

This section describes ﬁle system features supported by Paragon NTFS&HFS+ for Linux 9.5 driver, respectively.

9.1 NTFS features

Compressed ﬁles

Reading and writing compressed ﬁles is fully supported in both sequential and random orders.

Encrypted ﬁles

Encrypted ﬁles are read encrypted. During copy operation, ﬁle data streams will be copied encrypted with loss of decryption capability.

Alternate data streams

When copying from NTFS to Linux FS: all additional streams will not be copied, along with compression ﬂag and security attributes.

Hardlinks and symlinks

Any link will be copied as a full ﬁle with its body, losing link information. Maximum ﬁlename length NTFS stores ﬁlenames in UTF-16 encoding. This may cause trouble when very long ﬁlenames containing non-latin characters are used and UTF-8 is selected as default Kernel codepage.

9.2 HFS+ features

Case sensitivity

Both case sensitive and case instensitive types of HFS+ ﬁle system are supported.

Alternate data streams (forks)

During ﬁle copy operation (using cp command) on Linux only ’data’ fork is copied.

User manual 36

10 Frequently Asked Questions

10.1 What are ’minor errors’ reported by chkntfs utility?

Most of information about ﬁles (times, sizes, attributes) in NTFS is duplicated and triplicated. Minor error means that copies does not match original. E.g. "latime" means last access time. The native chkdsk from Microsoft does not show these mismatches and ﬁxes it silently (if /f is speciﬁed) — see http://technet.microsoft.com/en-us/library/cc959914.aspx. Paragon chkntfs utility can also ﬁnd the following minor errors:

mdtime - modiﬁcation time

chtime - last change time

asize - data allocated size

dsize - data size

attrib - attributes

To see more verbose output on minor errors, use -showminors command line option when running chkntfs. For an example output, please see the log below:

# chkntfs --showminors /dev/sda2 WARNING! f parameter not specified. Running chkntfs in read-only mode. Checking Volume /dev/sda2... Verifying 1680 records ... $UpCase file is formatted for use in Windows NT/2K/XP Verifying 161

,→folders ...

minor error " latime" in index 0x5 "." => "admin" minor error " latime" in index 0xb "$Extend" => "$Reparse" minor error " latime" in index 0x1f "public" => "EZ TALK.doc" minor error " latime" in index 0x1f "public" => "Fedora-13-i686-LiveKDE" minor error " latime" in index 0x1f "public" => "Fedora-13-x86_64-Live" minor error " latime" in index 0x1f "public" => "FEDORA~1" minor error " latime" in index 0x1f "public" => "FEDORA~2" minor error " latime" in index 0x1f "public" => "FTP_login _information.

,→doc"

minor error " latime" in index 0x1f "public" => "Reports" minor error " latime" in index 0x1f "public" => "... 2 K.M..b." minor error " latime" in index 0x1f "public" => "...~1" minor error " mdtime chtime latime" in index 0x5bd "_restore{FB5EFA8E-F7E1-4999-B498-21EEC0CF7124}" => "RP83" minor error " latime" in index 0x676 "mungchacha_com .+LC Kung Fu Dunk"

,→=> "DISC2.DAT.bc!"

minor error " latime" in index 0x676 "mungchacha_com .+LC Kung Fu Dunk"

,→=> "DISC2D~1.BC!"

Verifying files security...

4.83 Gb in 1492 files 464 Kb in 163 directories

0 Kb in bad blocks in 0 fragments 90424 Kb in use by the system 65536 Kb occupied by the log file

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

37 Paragon NTFS&HFS+ for Linux 9.5

4096 bytes in each allocation unit 182879943 total allocation units on volume 181590430 allocation units available on volume

The volume /dev/sda2 contains minor error(s).

10.2 Warnings on Windows7/Vista when NTFS HDD is reconnected from Linux

After NTFS volume previously operated by Paragon NTFS&HFS+ for Linux 9.5 driver is attached to Windows Vista/Windows 7 machine, warnings are displayed on the screen. Why?

This is the case when volume was not unmounted correctly before it was detached from Linux system.

This section illustrates ‘dirty’ volumes handling as implemented in Windows 7. For more information on dirty ﬂag and its support in Paragon ﬁle system drivers products see Dirty ﬂag issues subsection.

An USB HDD enclosure with 320 Gb SATA HDD with one NTFS partition was detached from system while ﬁle copy operation was in progress. After the enclosure was attached to Windows 7 PC again, the following dialog was displayed:

After user clicks ‘Scan and ﬁx (recommended)’, scan process begins:

User manual 38

After checking is completed, the following summary window is displayed:

After user clicks ’Details’, the window is expanded and more detailed information is displayed to the user:

In case there are no errors, the following information is displayed:

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

Paragon Technologie GmbH

39 Paragon NTFS&HFS+ for Linux 9.5

10.3 Recently changed ﬁle has its modiﬁcation time a few hours ahead of or behind the current system time. Why?

This offset occurs due to the fact that NTFS stores ﬁle times as UTC time (in contrast to FAT that stores local time) and the system might not have time zone setting that can be read by C library and then used to convert ﬁle times reported by Kernel to local time.

Consequently, if a ﬁle is written to an NTFS volume on Windows with time zone set to, say, UTC+8, and then the volume is connected to the Linux system, C library reports values provided by Kernel ‘as is’ without converting them to local time. However, if a ﬁle is modiﬁed on the Linux system, its modiﬁcation time is written to the ﬁle system as system’s current time and then it is reported correctly. In the latter case, after the ﬁle modiﬁed on the Linux system is brought back to the Windows machine

User manual 40

(with its local time zone set to UTC+8), the ﬁle’s modiﬁcation time will be reported 8 hours ahead of current time (assuming that current time is the same on the Linux system and Windows PC).

There is ’bias’ mount option (see Mount options subsection) that allows to work around the issue on systems that do not have time zone setting readable by C standard library (ﬁrst introduced in version 8.1.023). However, we recommend that time zone setting that can be used by C standard library to convert time values, is added to the Linux system.

10.4 Why does mount option A make driver ignore mount option B?

When you mount disk with several mount options driver may ignore some of them.

This issue can happen when mount command is used with several options coded like:

# mount -t ufsd -o option_A,option_B -o option_C device mount point

In this case driver may ignore options A and B when mounting disk with option C.

To prevent this possibility it is recommended to write your commands with several options like:

# mount -t ufsd -o option_A,option_B,option_C device mount point

10.5 Does the driver have an optimization for avoiding data fragmentation?

We use the driver for recording video data. This writes a huge amount of small data packs to the hard drive. Can you tell me if the driver has an optimization for avoiding data fragmentation? Or will this cause a strong fragmentation to the hard disk?

As of UFSD version 8.4, level of fragmentation mostly depends on number of ﬁles that are written to a volume simultaneously. In case only one ﬁle is written with small chunks of data, no fragmentation will occur beyond fragmentation caused by fragmentation of free space already existing on the volume. In case several ﬁles are written simultaneously, fragmentation will occur if no counter measures are taken.

One of possible countermeasures is to allocate large continuous chunks of disk space for ﬁles, by using:

fallocate (fd, FALLOC_FL_KEEP_SIZE, ...)

This routine allocates disk space for ﬁle, but does not change ﬁle size. Before closing the ﬁle, one could call ftruncate() to reclaim disk space that was allocated but was not written to, but that is up to developer to decide whether it is needed to reclaim the unwritten space (maybe the allocated space will be used in future write sessions, e.g. as in P2P network client software when downloading large ﬁles during several sessions). This approach also reduces CPU load thus improving performance, as there’s no need for FS driver to invoke disk space allocation routines each time small block should be appended to a ﬁle.

UFSD version 8.5 has transparent feature called delayed allocation (available on 2.6.X Kernel versions with writeback_inodes_sb_if_idle routine, see delalloc mount option) to prevent fragmentation even in case writes are performed in small chunks. For this feature to work, the system must have enough memory for disk/ﬁle cache and write operations must be performed in buffered mode.

Paragon Technologie GmbH

Leo-Wohleb-Straße 8, 79098 Freiburg, Germany

41 Paragon NTFS&HFS+ for Linux 9.5

10.6 Why a lot of memory is used for volume mounting?

Let us describe what’s going on when UFSD driver mounts volumes.

First of all, the driver must read ﬁle system boot record, and after verifying it, it must also read some metadata from the mounted volume, namely, parts of $Mft and the entire $Bitmap metaﬁle. For example, if the volume has $Bitmap of 74 MB, the driver has to read not less than 74MB to mount it. When our driver reads data from disk, Kernel keeps the data cached in memory. The amount of memory that Kernel allocates for I/O buffers is printed in line #3 of /proc/meminfo ﬁle (Buffers: XXXX kB). It’s up to Kernel to decide how much memory to allocate for I/O buffers (this can be tuned via Kernel metaﬁles – please see the link below for more information). The memory (allocated for ‘Buffers’) is normally reclaimed by kernel when Kernel or an application needs to allocate some memory for ‘private’ use.

The real-time report on memory allocated by our driver for operations on speciﬁc partition that is currently mounted, is available in line #2 of ﬁle /proc/fs/ufsd/<block_device_name>/volinfo. Peak amount of memory allocated by UFSD driver when mounting NTFS volume is around 250 KB. The amount then reduced to 40 KB after mount operation is completed.

10.7 Why the disk can’t be dismounted?

When you try to dismount the disk with the ’umount’ command the volume is reported ’busy’ and can’t be unmounted. Why?

This issue can happen when there are working processes, that are still using the volume. Therefore, there are several options to remove the conditions that prevent the storage medium from being unmounted:

1. Check if it is possible to safely unmount the storage using the system’s web interface.

2. Check if support for external storage can be disabled from the system’s web interface. Disable the services and retry unmounting the storage medium.

3. Check if the various ﬁle/media sharing services (like multimedia, SAMBA (SMB), AFP, etc) can be disabled from the system’s web interface. Disable the services and retry unmounting the storage medium.

Note: Windows system keeps SAMBA connection to the storage for several minutes. Disable the connection and retry to unmount the storage medium. For example in Windows XP it can be done from

’Disconnect network drive’ menu (e.g. ’My computer’ -> ’Tools’ -> ’Disconnect Network Drive’ menu).

If the volume couldn’t be unmount with these steps, use ’sync’ command to ﬂush buffers to the storage medium before detaching it manually from the device:

# sync

It is recommended also to add ’sync’ command before ’umount’ to the unmounting web-interface script, as it helps to avoid ﬁle system corruption on the storage medium, if unmount script was unsuccessful and the drive was detached manually.

Paragon NTFS, HFS+ Instruction Manual

Specifications and Main Features

Frequently Asked Questions

User Manual

1 Introduction

1.1 Historical review

1.2 Paragon UFSD technology

1.3 How UFSD works on Linux

1.4 Key features

2 System requirements

2.1 Hardware requirements

2.2 Software requirements

3 Installation

3.1 Shipment

3.2 Components

3.3 Installing driver

3.4 Uninstalling driver

4 Using The Driver

4.1 Getting started

4.2 Mounting and unmounting partitions

4.4 GPT issues

4.5 Issues with large HDDs

5 Mount options

5.1 Mount options

6 Additional Utilities

6.1 ufsd utility

6.2 chkufsd utility

6.3 NTFS utilities

mkntfs utility creates NTFS volumes (1.2, 3.0, 3.1 (Windows NT 4.0/2000/XP/2003/Vista/7/8.1/10) ﬁle system) on user speciﬁed (block) device (disk partition) under Linux OS.

chkntfs utility performs consistency checking of NTFS volumes and (optionally) ﬁxes errors.

6.4 HFS+ utilities

mkhfs ustility creates an HFS+ volume on speciﬁed (block) device (disk partition) under Linux OS.

chkhfs utility provides consistency checking of a HFS+ volume and ﬁx errors.

7 Troubleshooting

7.1 Troubleshooting processes

7.2 Mount troubleshooting

7.4 Can’t compile the NTFS/HFS+ for Linux driver

7.5 “Can’t load module” message at the end of installation

7.6 ufsd module: kernel-module version mismatch

7.7 ufsd module: create_module: operation is not permitted

7.8 insmod: a module named as ufsd already exists

7.9 insmod: Unknown symbol jnl_op (err0)

7.10 Can’t mount NTFS/HFS+ volume

7.11 Hardware issues

8 Sysdump utility

8.1 Introduction

8.2 Main functions

8.3 Using the sysdump utility to collect both platform system information and metadata

8.4 Changing the name of output archive

8.6 Delivering collected data to Paragon

8.7 Privacy policy

9 UFSD driver compatibility

9.1 NTFS features

9.2 HFS+ features

10 Frequently Asked Questions

10.2 Warnings on Windows7/Vista when NTFS HDD is reconnected from Linux

10.4 Why does mount option A make driver ignore mount option B?

10.5 Does the driver have an optimization for avoiding data fragmentation?

10.6 Why a lot of memory is used for volume mounting?