2.5 'what to expect'
Quoted 2 lines
We had this discussion before, back when I first submitted the large block device patches. The consensus then was to use EFI, or LDM.
GNU parted has had the GPT code for a couple years now, and CONFIG_EFI_PARTITION has been in both 2.4.x and 2.5.x trees for over a year. To the best of my knowledge and experience, it works equally well on x86 as on IA-64, modulo booting with traditional BIOSs of course.
# parted /dev/sdb print # parted /dev/sdb mklabel gpt # parted /dev/sdb mkpartfs p ext2 0.017 some-really-big-number
partx in util-linux can even tell the kernel about it without requiring a reboot.
I'd definitely be interested to know of cases where it doesn't work, so I can get them addressed.
Thanks,
Matt
Signature · 4 lines
Matt Domsch Sr. Software Engineer, Lead Engineer Dell Linux Solutions www.dell.com/linux Linux on Dell mailing lists @ http://lists.us.dell.com
Quoted 5 lines · [email protected] wrote:
GNU parted has had the GPT code for a couple years now, and
CONFIG_EFI_PARTITIONhas been in both 2.4.x and 2.5.x trees for over a year. To the best of my knowledge and experience, it works equally well on x86 as on IA-64, modulo booting with traditional BIOSs of course.
When I rebuilt my server a few months back, I wanted to use GPT partitioning so I could have UUIDs on the partitions, etc. The stumbling block that I ran into is there are no boot loaders that support GPT partitioning (LILO and GRUB being the obvious ones). Even though my BIOS doesn't care and would load the MBR, that MBR has to accomplish something useful.
Since GPT partitioning is normally only used in IA-64 land, and all IA-64 systems come with EFI stuff that handles the tasks that LILO/GRUB would handle on an x86 system, I guess there hasn't been much demand for GPT support for non-EFI systems. So the BIOS is not really the issue, since "traditional BIOSs" (as Matt referred to them)
don't handle bootloading tasks at all except at the most basic level.