public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Rob Landley <rob@landley.net>
To: Felix Oxley <lkml@oxley.org>
Cc: linux-kernel@vger.kernel.org
Subject: Re: [PATCH 1/1] Kconfig help text for RAM Disk & initrd
Date: Mon, 17 Oct 2005 02:33:46 -0500	[thread overview]
Message-ID: <200510170233.46811.rob@landley.net> (raw)
In-Reply-To: <200510170812.11590.lkml@oxley.org>

On Monday 17 October 2005 02:12, Felix Oxley wrote:
> On Monday 17 October 2005 06:37, Rob Landley wrote:
> > Actually if this is a patch against 2.6, between ramfs (including
> > initramfs) and the ability to loopback mount files, I personally consider
> > ramdisks semi-obsolete.  (This might be _why_ it says most normal users
> > won't need them.)
>
> Well, you may well know better than I, however on my Suse 10 box,
> initrd is used and I see only CONFIG_INITRAMFS_SOURCE="" in my .config.

There's no config symbol for it because you can't disable it.  It's always 
hardwired in.  (Most of the code for ramfs is just the normal page cache 
stuff anyway.  It's a filesystem that uses the disk cache mechanism to store 
its data.  Quiet clever, really.)  Initramfs is called "rootfs" in the mount 
list, and is always there.  (Check proc/mounts on any 2.6 kernel.)  Getting 
rid of rootfs would cause an immediate deadlock becuase it's used as a search 
terminator in the doubly linked list of mounts inside the kernel.

And yes, moving off of initramfs is non-obvious because of this.  I talked 
about that a bit here:

http://www.busybox.net/lists/busybox/2005-October/016597.html

By default, the build creates an empty archive to initialize initramfs (this 
archive gets linked into the kernel image, it's not a separate file), and the 
boot process extracts it (a NOP) and then tries to exec /init in rootfs 
(which is empty, so this fails), and then falls through to the normal root 
partition finding behavior.

> I made this patch because omiting intrd caused my system to be unable to
> boot.

Then your system is setup to boot via initrd.  Probably your vendor has 
an /sbin/installkernel script that makes one.  (See arch/i386/boot/install.sh 
for details.  When you run "make install" that script looks for an 
"installkernel" script on the current system and runs it if it finds it, so 
that any vendor weirdness can apply to your new kernel automatically...)

> regards,
> Felix

Rob

  reply	other threads:[~2005-10-17  7:34 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2005-10-17  0:02 [PATCH 1/1] Kconfig help text for RAM Disk & initrd Felix Oxley
2005-10-17  5:37 ` Rob Landley
2005-10-17  7:12   ` Felix Oxley
2005-10-17  7:33     ` Rob Landley [this message]
2005-10-17 12:44       ` Felix Oxley
2005-10-17 21:32         ` Rob Landley

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=200510170233.46811.rob@landley.net \
    --to=rob@landley.net \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lkml@oxley.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox