All of lore.kernel.org
 help / color / mirror / Atom feed
From: Gene Czarcinski <gene@czarc.net>
To: "linux-btrfs@vger.kernel.org list" <linux-btrfs@vger.kernel.org>
Subject: Re: /boot as a btrfs subvolume
Date: Sun, 06 Jan 2013 09:44:22 -0500	[thread overview]
Message-ID: <50E98DC6.3030304@czarc.net> (raw)
In-Reply-To: <EEB5FCDE-687D-4EEF-9A85-68703266704C@colorremedies.com>

On 01/05/2013 04:27 PM, Chris Murphy wrote:
> On Jan 5, 2013, at 2:17 PM, Gene Czarcinski <gene@czarc.net> wrote:
>
>> As of the latest updates to anaconda and grub2 for Fedora 18, it is now possible to install with /boot as a btrfs subvolume.  The way that grub2 is handling this is the "reach down" to the files it needs as if the subvolume was a directory.
>>
>> Is this OK?
> Mostly, but getting GRUB2 to handle subvolid I think would be better, because then subvols can be moved/renamed and things still would work.
>
> Small problem currently is Fedora 18 still depends on grubby to make grub.cfg changes, and grubby is being kinda dumb about updating grub.cfg when /boot is on a boot subvol - ergo there's an error and it doesn't update the grub.cfg correctly.
> https://bugzilla.redhat.com/show_bug.cgi?id=864198
> Work around after any kernel update is to run grub2-mkconfig -o /boot/grub2/grub.cfg manually, and then the grub.cfg is correct.
>
>> At this point I am not worried about snapshots or any other complexities.  If the subvolume name is known. should grub2 be able to reliably "reach down" as if that subvolume was a directory?  It needs some of its configuration files and, of course, the linux kernel.
> It's two parts. GRUB's core.img code works as you describe. But grub-install and grub-mkconfig depend on /etc/fstab. If I use subvolid= in /etc/fstab and update either core.img or grub.cfg (with -install or -mkconfig), then the system is no longer bootable.
So, I conclude from the above that the "safe" thing to do is to keep 
/boot as a separate ext2/3/4 partition.

Gene

  reply	other threads:[~2013-01-06 14:44 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-01-05 21:17 /boot as a btrfs subvolume Gene Czarcinski
2013-01-05 21:27 ` Chris Murphy
2013-01-06 14:44   ` Gene Czarcinski [this message]
2013-01-06 19:11     ` Chris Murphy
2013-01-06 19:17       ` Swâmi Petaramesh
2013-01-06 20:05         ` Gene Czarcinski
2013-01-07  2:00           ` Chris Murphy
2013-01-07 15:42             ` Gene Czarcinski
2013-01-07 18:42               ` Chris Murphy
2013-01-07 21:41                 ` Gene Czarcinski
2013-01-07 22:04                   ` Chris Murphy
2013-01-07 17:48           ` Swâmi Petaramesh
2013-01-07 19:28             ` Chris Murphy
2013-01-05 21:32 ` Chris Murphy
2013-01-06 14:37   ` Gene Czarcinski

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=50E98DC6.3030304@czarc.net \
    --to=gene@czarc.net \
    --cc=linux-btrfs@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.