Re: Need for a unique Linux GPT GUID type code (PATCH included)

[KESHAV P.R.]

Hi,
     I have been following this discussion closely. Rod Smith, thanks
for bringing this up. I would also suggesta similar "linuxdata" flag
to set a partitoon as Linux Data type irrespective of the unserlyting
FS. I alos recommend changing "boot" flag in parted for GPT disk to
something more meaningful like efisys. Many devs (of distro
installers) are confused as to how to set a partition as EFI SYSTEM
type in GPT disks and they do not know that "boot" in GPT disks sets a
partition as EFISYS unlike the same flag in MBR disks.

Also libparted has few other issues with GPT disks.

1. libparted (before 3.0) segfaulted when gpt disk having >128
partition entries in GPT table is accessed. I tried this with a USB
pendrive with 140 paritition entries using gdisk, with about 7
partitions created in the disk, and parted segfaulted with print
option. I have not tested this with libparted 3.0 .

2. libparted seems to erase a partition's GPT attributes when that
partition is accessed in any way. This may cause problem for a person
who uses syslinux as the boot loader, as it uses "Legacy BIOS
bootable" gpt attribute to identify a bootable partition. This problem
does not exist with grub2 as it uses a defined partitio type
(bios_grub) instead of partition attribute like syslinux does.

This also lead me to think that there is no way to distinguish between
setting a GPT partition type and a GPT partition attribute when a flag
is set on a partition. While internally libparted does this correstly,
externally (to the user) parted does not seem to distinguish between a
gpt partition type and a gpt partition attribute.

//offtopic: With all FS related operations removed in parted 3.0, is
there any FAT resizing tool (and same for HFS+ too) that does not rely
on libparted. THis is partly due to
https://bugzilla.gnome.org/show_bug.cgi?id=651559 .
http://sourceforge.net/projects/fatresize/ seems to rely on libparted
and is therefore useless with libparted 3.0 . Thanks in advance.

Regards.

Keshav