I came across this program in e2fsprogs and it doesn't seem to have a
man page. What is its purpose? Looking at the code, it appears to
extract the fs type listed in the /etc/fstab of the root fs before it is
mounted, but why?
On 4/11/12 10:46 AM, Phillip Susi wrote:
> I came across this program in e2fsprogs and it doesn't seem to have a
> man page. What is its purpose? Looking at the code, it appears to
> extract the fs type listed in the /etc/fstab of the root fs before it
> is mounted, but why?
+2004-09-18 Theodore Ts'o <[email protected]>
+
+ * e2initrd_helper.c: New program which allows initrd scripts to
+ check on the contents of /etc/fstab without (a) needing to
+ mount the root filesystem, and (b) needing /bin/awk in the
+ initrd environment. (Addresses Debian Bug #247775)
+
Other than that, I dunno, but Ted might. :)
-Eric
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Apr 11, 2012 at 10:55:17AM -0500, Eric Sandeen wrote:
> On 4/11/12 10:46 AM, Phillip Susi wrote:
> > I came across this program in e2fsprogs and it doesn't seem to have a
> > man page. What is its purpose? Looking at the code, it appears to
> > extract the fs type listed in the /etc/fstab of the root fs before it
> > is mounted, but why?
>
> +2004-09-18 Theodore Ts'o <[email protected]>
> +
> + * e2initrd_helper.c: New program which allows initrd scripts to
> + check on the contents of /etc/fstab without (a) needing to
> + mount the root filesystem, and (b) needing /bin/awk in the
> + initrd environment. (Addresses Debian Bug #247775)
> +
Debian used to have a system where we could transparently update the
root file system to use ext4 file system features as part of the
initrd.
At one point I had hopes that we could fix up the boot scripts so that
if an initrd was present, we could run e2fsck out of the initrd, so we
wouldn't have to play games with mounting the file system read-only.
So at one point the plan was this program would also extract all of
/etc/fstab for the initrd, and other things that might be useful
before the root file system was mounted.
I've since decided that this would involve getting fsck all-too-bound
up in distribution-specific init systems, and might actually aid and
abet the further dominance of GNOME-OS (or plymouth/upstart; I dispise
all of the modern init/initscripts systems equally), so I've since
stopped trying to implement things in that particular design
direction.
It's why I'll be abandoning Ubuntu when the new Ivy Bridge laptops
become available, and switching not to Fedora, but to Debian Testing
--- but I'm sure Mark Shuttleworth really doesn't care what distro
people like me choose to use.... :-)
- Ted
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 04/11/2012 02:27 PM, Ted Ts'o wrote:
> Debian used to have a system where we could transparently update the
> root file system to use ext4 file system features as part of the
> initrd.
>
> At one point I had hopes that we could fix up the boot scripts so that
> if an initrd was present, we could run e2fsck out of the initrd, so we
> wouldn't have to play games with mounting the file system read-only.
> So at one point the plan was this program would also extract all of
> /etc/fstab for the initrd, and other things that might be useful
> before the root file system was mounted.
>
> I've since decided that this would involve getting fsck all-too-bound
> up in distribution-specific init systems, and might actually aid and
> abet the further dominance of GNOME-OS (or plymouth/upstart; I dispise
> all of the modern init/initscripts systems equally), so I've since
> stopped trying to implement things in that particular design
> direction.
So does this mean it is now cruft and should be removed at some point?
> It's why I'll be abandoning Ubuntu when the new Ivy Bridge laptops
> become available, and switching not to Fedora, but to Debian Testing
> --- but I'm sure Mark Shuttleworth really doesn't care what distro
> people like me choose to use.... :-)
AFAIK, Ubuntu uses Debian's initramfs system... are you saying this is not the case? If so, what's the difference? When I attended the last UDS there was actually a push coming from the arm folks to do away with the initrd on systems that do not require it ( no LVM ) to accelerate boot time by not wasting time loading another file, which apparently is rather slow on the arm boot loader.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iQEcBAEBAgAGBQJPhgNZAAoJEJrBOlT6nu75bGkIALYdwv2542miponYdnt6bX3x
D/rc0oersb7VNe12DV6++Xp2/PWFQRZCdDWDMWUw6gVyE9PYElK6a39MMoKQtBod
RkbWFpI20hG0lkIxNGt55w16+xPMUSD9KRY7MT2R3qXoLd1sKkbummVOdrNJnrT/
WHeK1BpVrXbE6qqlD8DdJtHbBAkCDTz08IKjkEXBMsovmAeIq8PVvG2L9foUCX7o
HDzttaLXdBcQ42/ytWm3xRNJET0Bh/Cmg7TBGBHwkxpylIRgEg5ZgvLNPaWTA5eJ
l6eQsaL18r9Yn8lvDTq5fpxukbAh229NhhzzrcPl82E3P+cBC4AuSUX/W2hQqPc=
=P8AI
-----END PGP SIGNATURE-----
On Wed, Apr 11, 2012 at 06:19:05PM -0400, Phillip Susi wrote:
> AFAIK, Ubuntu uses Debian's initramfs system... are you saying this is not the case? If so, what's the difference? When I attended the last UDS there was actually a push coming from the arm folks to do away with the initrd on systems that do not require it ( no LVM ) to accelerate boot time by not wasting time loading another file, which apparently is rather slow on the arm boot loader.
Both share initramfs-tools, but they don't use the same version and
really do things subtily different. I've just spent far too much time
making something work in the Debian version just to notice it needs
major updates for Ubuntu.