README for gptgen v1.2.1

1. Introduction

Gptgen is a tool for converting hard drives partitioned according to the MSDOS-style "MBR" scheme to the GPT (GUID partition table) partitioning scheme, while retaining all data on the hard disk. Extended/logical partitions are supported, though GPT doesn't differentiate between primary and logical partitions, so after conversion, all partitions will be de facto primary.

2. Caveats

• Because GPT's partition types do not map exactly to MBR types, the conversion might not be perfectly accurate, especially when multiple file systems are assigned the same MBR partition type code. This has happened a few times in the past, and is prone to occur again in the future. In case of such a type collision, gptgen assumes the most common use of a given type ID, and transcribes accordingly. So, if you have rare filesystems on your hard disk, then double-check partition type GUIDs after conversion. One specific example of a collision is type 0xAF, which is used both by macOS (to represent HFS and HFS+) and ShagOS (to represent swap space). Gptgen assumes 0xAF to be HFS(+).

• For MBR partition types with no known matching type GUID, gptgen uses a generic GUID, which is as follows: {1575DA16-F2E2-40DE-B715-C6E376663Bxx} where "xx" is the MBR type ID of the partition. Other users of GPT may implement reading this ID, but it is not recommended to rely on this ID for creating new partitions; instead it is better to define a new GUID to represent the desired partition type.

• Gptgen requires administrator privileges to run. On Windows Vista and newer Windows systems with UAC enabled, you may need to run gptgen from an elevated command prompt. On Linux, run with sudo.

• Big-endian support is completely untested, though necessary code is in place.

• The GUID partition table requires that no partition can begin before logical block 34 (number of GPT entries128/block size + 2 - gptgen defaults to 128 entries), and that there must be at least 33 (entry count128/block size + 1) free blocks at the end of the drive before conversion. The first requirement is usually not a problem, since MBR partitions begin in the second "track" of the drive (in the BIOS-supplied fake geometry, not related to the true geometry in modern drives), and a "track" is usually 63 sectors long, making the first partition begin in block 63; however the second limitation might require the last partition of the disk to be shrunk by a cylinder to make way for the backup GPT at the end of the disk. Gptgen will warn and stop if the drive doesn't meet these requirements. (The actual needed sector numbers can vary with the chosen entry count and the drive's block size, "34" and "33" are based on a drive block size of 512 bytes and an entry count of 128.)

• Some implementations of GPT (e.g. Windows) require certain partitions to be present on the disk (e.g. an EFI boot partition). Gptgen doesn't create these partitions, it can only convert the existing partition table's contents. To create such partitions, use a GPT-ready partition manager such as parted.

• Because booting is handled differently with GPT than with MBR, converting a boot drive to GPT may render it unbootable. To make a GPT drive bootable on BIOS, you will need to create a BIOS boot partition and install a GPT-aware boot loader (such as GRUB 2.02 or later) on it.

3. Usage

On Windows (a precompiled Windows binary is included in the package), the syntax of the tool is gptgen [-w] \\.\physicaldriveX, where "X" is the drive number reported by the Disk Management console or the list disk command of the diskpart utility. The -w option makes gptgen write the generated GUID partition tables to the disk, otherwise the primary table will be written to a file named "primary.img", and the secondary table to "secondary.img", in the directory that the program was invoked from. You can then use a raw disk utility such as dd to write the tables to the disk.

On Linux, the syntax is gptgen [-w] /dev/sda, with /dev/sda being the block device representing the physical drive to be converted. The -w option works the same way as on Windows.

You can add the -m parameter to the command line to preserve the existing MBR on the disk, instead of writing a protective MBR. This is not recommended, and may prevent recognition of the drive as GPT on some systems, but it is useful when you want to be extra safe, and verify the newly-written GPT before wiping out the MBR.

Gptgen, by default, builds a GPT consisting of 128 partition entries. You can override this using the -c (--count) parameter, e.g. gptgen -w -c 32 \\.\physicaldriveX. However, this is not recommended, as some implementations have problems recognizing GPTs with other than 128 entries.

The parameter -b <filename> tells gptgen to back up the original MBR of the target drive into the file indicated by <filename>.

4. Compiling and installing

On Linux, you can build gptgen using cmake and make. To install it, run make install. This installs gptgen to /usr/local/sbin by default. To install it to a different location, use cmake -DCMAKE_INSTALL_PREFIX=<prefix> . to generate the Makefiles. make install will then install gptgen to <prefix>/sbin. To relocate the whole installation, use the DESTDIR variable on the make command line (e.g. make DESTDIR=<install_root> install).

Building with the default compiler (GCC or Clang) on Linux:

$ cmake .
$ make
$ sudo make install

On Windows, building with both MinGW and Visual Studio is supported. For building with MinGW, MSYS2 is recommended. For building with Visual Studio, you may use the freely available Visual Studio Community Edition with CMake Project support.

Building with MinGW on Windows (from a MSYS2 Mingw-w64 shell prompt):

$ cmake -DCMAKE_SYSTEM_NAME=Windows .
$ make

Building with Visual Studio Community 2017 from the Command Prompt:

> "C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Auxiliary\Build\vcvarsall.bat" x86_amd64
> set path=%path%;C:\Program Files\CMake\bin
> cmake -G"NMake Makefiles" -DCMAKE_MAKE_PROGRAM=nmake .
> nmake

Building with Visual Studio Community 2017 from the IDE:

1. Open Visual Studio.
2. Go to File->Open->CMake.
3. Open the CMakeLists.txt in the gptgen source directory.
4. Choose the type of build that you would like (e.g. "x64-Release").
5. Run CMake->Build All.
6. Run CMake->Install->gptgen.
7. You can find gptgen.exe in the location show in the CMake build log in the Visual Studio "output" pane.

To build gptgen with debug support, add the following arguments on the CMake command line: -DCMAKE_BUILD_TYPE=Debug -DCMAKE_VERBOSE_MAKEFILE=TRUE