All of lore.kernel.org
 help / color / mirror / Atom feed
From: matt black <cache22@webkorner.com>
To: linux-raid@vger.kernel.org
Subject: boot root no /dev
Date: Thu, 21 Dec 2006 11:35:54 -0500	[thread overview]
Message-ID: <458AB7EA.7010903@webkorner.com> (raw)

Greetings helpful people, welcome to me your kindly list  :)

I set up raid-1 on new drive /dev/sdb, and copied files from /dev/sda. 
I test boot to it with following grub.conf:

title  Gentoo 2.6.18-r4-md
root   (hd1,0)
kernel /2.6.18-gentoo-r4-md root=/dev/md1 dxs_support=5 
video=mtrr,vesafb:1024x768

/dev/md1 corresponds to /dev/sd*2 which is root part. (/dev/sd*1 is /boot)

Kernel autodetects and reiserfs looks happy with it, then I get 
"Warning: unable to open an initial console."  This an error that occurs 
in init/init.c after setting up the initial rootfs and freeing initial 
memory.  After the warning init.c tries to run /sbin/init, /etc/init, 
/bin/init, or /bin/sh.

I believe it succeeds because then udev errors scroll by endlessly and 
killing things with magic-sysrq give me the impression that bash is 
running.  So I get the impression that /dev/md1 got mounted 
automagically.  I did create an empty /dev dir on md1 for /dev to mount 
to, but I didn't copy anything from original /dev (shouldn't have to 
right?)  It just occurred to me to cd-boot and see if the original 
system is cheating with static /dev/ entries--it was originally a devfs 
system.

The kernel docs imply (it wouldn't be appropriate to be crystal clear) 
that there is always a default initramfs built in if not overridden, 
correct?  In my source it looks like that would be:

---

# This is a very simple, default initramfs

dir /dev 0755 0 0
nod /dev/console 0600 0 0 c 5 1
dir /root 0700 0 0
# file /kinit usr/kinit/kinit 0755 0 0
# slink /init kinit 0755 0 0

---

in the course of my wanderings I tried making my own initramfs list with 
the addition of "nod /dev/md1 0600 0 0 b 9 1" which makes no difference.

in my kernel config:

CONFIG_INITRAMFS_SOURCE="/root/my_initramfs_list"
CONFIG_INITRAMFS_ROOT_UID=0
CONFIG_INITRAMFS_ROOT_GID=0

# CONFIG_BLK_DEV_RAM is not set
CONFIG_BLK_DEV_INITRD=y

Note that sometime in 2.6.17 the CONFIG_BLK_DEV_RAM option was given to 
break out ramdisk from INITRD.  CONFIG_BLK_DEV_INITRD now enables only 
initramfs by itself but it has to be enabled along with 
CONFIG_BLK_DEV_RAM if you want a ramdisk.  Maybe I'll build a 2.6.16 md 
kernel and see what happens.

Is it required that this initramfs be built-in so that I have /dev 
nodes?  How can I verify it is doing so?  In init/initramfs.c it seems I 
should get a message from  printk(KERN_INFO "Unpacking initramfs..."); 
which I never see even when successfully booting this kernel from 
/dev/sda.

Thanks for any assistance...



                 reply	other threads:[~2006-12-21 16:35 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=458AB7EA.7010903@webkorner.com \
    --to=cache22@webkorner.com \
    --cc=linux-raid@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.