[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: lvm module can't see USB disks' volume group

From: Andrei Borzenkov
Subject: Re: lvm module can't see USB disks' volume group
Date: Sat, 11 Feb 2017 10:32:09 +0300
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.7.0

11.02.2017 01:37, Alexandre Oliva пишет:
> On Feb  8, 2017, Alexandre Oliva <address@hidden> wrote:
>> 5TB disk, USB enclosure that exposes disk as 512B logical sectors: GRUB2
>> can see the (GPT) partition table, but it won't expose the PV contents,
>> nor can it access files in a separate 256MB btrfs partition created for
>> testing at the very end of the disk.  It can, however, read the GRUB
>> core embedded in the boot partition in the beginning of the disk.
> I think I figured out what the problem is.  Once I created a filesystem
> on a smallish partition closer to the beginning of the disk and copied
> boot to it, ls (hd*,gpt*)/boot said it couldn't get the C/H/S for the
> disk.

Could you show exact message?

> Now I gotta figure out where it's trying to get the info; it could be
> from the BIOS, in which case it probably won't get it because the disk
> probably won't supply it, or it could be from the boot sector, that I
> suppose doesn't have the information because nothing put it there.
> Presumably to load the embedded image, it just reads from the first few
> blocks, enough to not depend on C/H/S.  Well, I'm guessing a lot here,
> but it kind of makes sense to me, given the error I got today and the
> others I got earlier.  In the end, I might still be out of luck.

It is from BIOS. Assuming the actual error was "hd0 cannot get C/H/S
values", it is output when INT13 AH=08 fails and disk does not support
LBA (i.e. INT13 AH=41 or AH=48 failed previously as well).

As BIOS is capable of loading core.img itself off this disk, it is
obviously capable of working with it. So we need to figure out what BIOS
returns and how to work around it. If you can compile GRUB we can try
add debug output to see what happens.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]