Debian Bug report logs - #807457
Add command to unpack initramfs

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: Osamu Aoki <osamu@debian.org>

To: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: lsinitramfs(8): multiple-segmented initramfs supported

Date: Sat, 27 Jun 2015 08:53:59 +0900

[Message part 1 (text/plain, inline)]

Package: initramfs-tools
Version: 0.120
Severity: normal
Tags: patch

From initramfs-tools 0.118 (Oct. 2014):
  * [d5f4cd6] lsinitramfs: Parse and list files that include early microcode.
    Thanks to Brett Parker and Olivier Berger. (Closes: #717805)

So the manpage description as the following is incorrect.
| BUGS
|        lsinitramfs  cannot deal with multiple-segmented initramfs images, such
|        as those created when an early  (uncompressed)  initramfs  with  system
|        firmware is prepended to the regular compressed initrams, or when over‐
|        lay data is appended to the initramfs.

So please knock off these lines.

In addition to this, I wonder why not offer utility to extract the
regular compressed initrams by using the same code with trivial changes.
It is pain to manually extract the new initrd image content.

I attach patch here one as example against the git master.

Hmmm...
 https://wiki.debian.org/InitramfsDebug
This needs to be updated.

[0001-multi-segment-initramfs-supported.patch (text/x-diff, attachment)]

[0002-getinitramfs-command.patch (text/x-diff, attachment)]

[0003-Add-getinitramfs-8.patch (text/x-diff, attachment)]

[0004-update-packaging-for-getinitramfs.patch (text/x-diff, attachment)]

Message #10 received at 790095@bugs.debian.org (full text, mbox, reply):

From: Osamu Aoki <osamu@debian.org>

To: 790095@bugs.debian.org

Subject: Re: Bug#790095: getinitramfs code

Date: Sat, 27 Jun 2015 10:03:52 +0900

[Message part 1 (text/plain, inline)]

Hi,

Here is the getinitramfs script itself for testing just in case patch is
inconvenient for testing.

Osamu

[getinitramfs (text/plain, attachment)]

Message #21 received at 807457@bugs.debian.org (full text, mbox, reply):

From: Ben Hutchings <ben@decadent.org.uk>

To: Osamu Aoki <osamu@debian.org>

Cc: 807457@bugs.debian.org

Subject: Re: lsinitramfs(8): multiple-segmented initramfs supported

Date: Wed, 09 Dec 2015 03:23:11 +0000

[Message part 1 (text/plain, inline)]

On Sat, 27 Jun 2015 08:53:59 +0900 Osamu Aoki <osamu@debian.org> wrote:
[...]
> In addition to this, I wonder why not offer utility to extract the
> regular compressed initrams by using the same code with trivial changes.
> It is pain to manually extract the new initrd image content.
> 
> I attach patch here one as example against the git master.
[...]

Rather than duplicating code, please combine this with lsinitramfs
(checking $0 to determine which command was invoked) or move the common
code into a shell library.

Most decompression commands are named after the compressor plus 'un' so
I prefer the name 'unmkinitramfs'.

Ben.

-- 
Ben Hutchings
I'm always amazed by the number of people who take up solipsism because
they heard someone else explain it. - E*Borg on alt.fan.pratchett

[signature.asc (application/pgp-signature, inline)]

Message #28 received at 807457@bugs.debian.org (full text, mbox, reply):

From: Kevin Locke <kevin@kevinlocke.name>

To: 807457@bugs.debian.org

Subject: Re: initramfs-tools: Add command to unpack initramfs

Date: Tue, 9 Feb 2016 20:18:16 -0800

[Message part 1 (text/plain, inline)]

On Wed, 2015-12-09 at 03:23 +0000, Ben Hutchings wrote:
> Rather than duplicating code, please combine this with lsinitramfs
> (checking $0 to determine which command was invoked) or move the common
> code into a shell library.

I've taken a slightly different tact and implemented uninitramfs as a
thin wrapper for cpio.  It is designed as a drop-in replacement for
cpio wherever an initramfs may be present, and passes through all
options unmodified.  With this, I changed lsinitramfs to call
uninitramfs instead of cpio.

See what you think.

Thanks,
Kevin

P.S.  I named it "uninitramfs" instead of "unmkinitramfs" since it
does not depend on mkinitramfs specifically, and it seemed to fit with
"lsinitramfs" and the "initramfs" format.  But if you would prefer
"unmkinitramfs", I can change it.

-- 
Cheers,      |  kevin@kevinlocke.name    | XMPP: kevin@kevinlocke.name
Kevin        |  https://kevinlocke.name  | IRC:   kevinoid on freenode

[0001-Create-uninitramfs-command.patch (text/x-diff, attachment)]

Message #39 received at 807457-done@bugs.debian.org (full text, mbox, reply):

From: Ben Hutchings <ben@decadent.org.uk>

To: 807457-done@bugs.debian.org

Subject: Re: initramfs-tools: Add command to unpack initramfs

Date: Thu, 15 Dec 2016 23:13:03 +0000

[Message part 1 (text/plain, inline)]

Version: 0.126

On Tue, 9 Feb 2016 20:18:16 -0800 Kevin Locke <kevin@kevinlocke.name> wrote:
> On Wed, 2015-12-09 at 03:23 +0000, Ben Hutchings wrote:
> > Rather than duplicating code, please combine this with lsinitramfs
> > (checking $0 to determine which command was invoked) or move the common
> > code into a shell library.
> 
> I've taken a slightly different tact and implemented uninitramfs as a
> thin wrapper for cpio.  It is designed as a drop-in replacement for
> cpio wherever an initramfs may be present, and passes through all
> options unmodified.  With this, I changed lsinitramfs to call
> uninitramfs instead of cpio.
>
> See what you think.

I didn't like the way it passed all options through to cpio:
- It's inconsistent with the other commands
- It makes it difficult to implement any additional options
- cpio's default of unpacking into the current directory is dangerous

So I made the initramfs file and output directory arguments, and added
sensible options to cpio.  The --verbose option can be passed through,
and if there are other important ones we can add those later.

> P.S.  I named it "uninitramfs" instead of "unmkinitramfs" since it
> does not depend on mkinitramfs specifically, and it seemed to fit with
> "lsinitramfs" and the "initramfs" format.  But if you would prefer
> "unmkinitramfs", I can change it.

But it is meant to work with the initramfs images that mkinitramfs
creates and doesn't support arbitrary concatenations of maybe-
compressed cpio archives.  So, unmkinitramfs it is.

Ben.

-- 
Ben Hutchings
It is easier to change the specification to fit the program than vice
versa.

[signature.asc (application/pgp-signature, inline)]

Send a report that this bug log contains spam.