Debian Bug report logs -
#918480
squashfs-tools: Please move to "squashfskit" fork
Reported by: Chris Lamb <lamby@debian.org>
Date: Sun, 6 Jan 2019 14:30:02 UTC
Severity: wishlist
Tags: moreinfo
Found in version squashfs-tools/1:4.3-7
Reply or subscribe to this bug.
Toggle useless messages
Report forwarded
to debian-bugs-dist@lists.debian.org, lynxis@fe80.eu, intrigeri@boum.org, reproducible-builds@lists.alioth.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Sun, 06 Jan 2019 14:30:04 GMT) (full text, mbox, link).
Acknowledgement sent
to Chris Lamb <lamby@debian.org>:
New Bug report received and forwarded. Copy sent to lynxis@fe80.eu, intrigeri@boum.org, reproducible-builds@lists.alioth.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Sun, 06 Jan 2019 14:30:04 GMT) (full text, mbox, link).
Message #5 received at submit@bugs.debian.org (full text, mbox, reply):
Package: squashfs-tools
Version: 1:4.3-7
Severity: wishlist
X-Debbugs-CC: lynxis@fe80.eu, intrigeri@boum.org, reproducible-builds@lists.alioth.debian.org
Hi László,
First, thanks for maintaining squashfs-tools! As I understand it,
upstream has been replaced with a more-active fork maintained by
Lynxis Couzens:
https://github.com/squashfskit/squashfskit
§
This contains a number of patches that I, Intrigeri & Tails are using
for generating reproducibile output, such as:
From: Chris Lamb <lamby@debian.org>
Date: Sun, 6 Jan 2019 11:03:42 +0100
Subject: If SOURCE_DATE_EPOCH is set, use that timestamp for the mkfs time.
See https://reproducible-builds.org/specs/source-date-epoch/ for more
information about this environment variable.
Based on a patch by Alexander Couzens <lynxis@fe...> posted on
https://sourceforge.net/p/squashfs/mailman/message/34673610/
---
squashfs-tools/mksquashfs.c | 33 +++++++++++++++++++++++++++++++++
1 file changed, 33 insertions(+)
and:
From: intrigeri <intrigeri@boum.org>
Date: Sun, 6 Jan 2019 11:07:08 +0100
Subject: If SOURCE_DATE_EPOCH is set,
also clamp content timestamps with that value.
Based on a patch by Alexander Couzens <lynxis@fe...> posted on
https://sourceforge.net/p/squashfs/mailman/message/34673610/
Updated by Chris Lamb <lamby@debian.org> to use (time_t)0 as a
sentinel value over -1.
---
squashfs-tools/mksquashfs.c | 13 ++++++++++---
1 file changed, 10 insertions(+), 3 deletions(-)
Although I do note this latter one would require a slightly different
rebase due to, at least:
[ Michael Vogt <michael.vogt@ubuntu.com> ]
* debian/patches/0005-add-fstime.patch: initializing the time_t variable
with (time_t)0 instead of (time_t)-1 to avoid creating all filesystems
on "Wed Dec 31 23:59:59 1969"
… that was part of the squashfs-tools 1:4.3-6 upload.
§
Anyway, thanks again for maintaining this package.
Best wishes,
--
,''`.
: :' : Chris Lamb
`. `'` lamby@debian.org / chris-lamb.co.uk
`-
Information forwarded
to debian-bugs-dist@lists.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Sun, 06 Jan 2019 14:39:03 GMT) (full text, mbox, link).
Acknowledgement sent
to Chris Lamb <lamby@debian.org>:
Extra info received and forwarded to list. Copy sent to Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Sun, 06 Jan 2019 14:39:03 GMT) (full text, mbox, link).
Message #10 received at 918480@bugs.debian.org (full text, mbox, reply):
Chris Lamb wrote:
> Updated by Chris Lamb <lamby@debian.org> to use (time_t)0 as a
> sentinel value over -1.
I've at least pushed/poked this upstream so it's somewhere a bit
more obvious:
https://github.com/squashfskit/squashfskit/pull/2
Regards,
--
,''`.
: :' : Chris Lamb
`. `'` lamby@debian.org / chris-lamb.co.uk
`-
Information forwarded
to debian-bugs-dist@lists.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Sun, 06 Jan 2019 15:30:03 GMT) (full text, mbox, link).
Acknowledgement sent
to László Böszörményi (GCS) <gcs@debian.org>:
Extra info received and forwarded to list. Copy sent to Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Sun, 06 Jan 2019 15:30:03 GMT) (full text, mbox, link).
Message #15 received at 918480@bugs.debian.org (full text, mbox, reply):
Hi Chris,
On Sun, Jan 6, 2019 at 3:30 PM Chris Lamb <lamby@debian.org> wrote:
> First, thanks for maintaining squashfs-tools! As I understand it,
> upstream has been replaced with a more-active fork maintained by
> Lynxis Couzens:
>
> https://github.com/squashfskit/squashfskit
Checked the activity of this project, compared to the upstream one[1].
The one you promotes has the following latest commits:
master - Sep 23, 2017
staging - May 25, 2017
feature_versionsh - Nov 1, 2017
Upstream latest commit:
master - Jun 12, 2018
It seems more recent. What do I miss with the one you mentioned? That
seems untouched for more than a year.
> This contains a number of patches that I, Intrigeri & Tails are using
> for generating reproducibile output, such as:
[...]
I don't see commits from this year, only your pull request.
> … that was part of the squashfs-tools 1:4.3-6 upload.
Indeed and I think the package is up to date with all the patches
needed for the Buster release. If you may need additional changes,
please tell me so. I don't see how squashfskit might be better - going
to do a diff of the two projects.
Thanks for the heads-up,
Laszlo/GCS
[1] https://github.com/plougher/squashfs-tools
Information forwarded
to debian-bugs-dist@lists.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Sun, 06 Jan 2019 15:45:03 GMT) (full text, mbox, link).
Acknowledgement sent
to Chris Lamb <lamby@debian.org>:
Extra info received and forwarded to list. Copy sent to Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Sun, 06 Jan 2019 15:45:03 GMT) (full text, mbox, link).
Message #20 received at 918480@bugs.debian.org (full text, mbox, reply):
tags 918480 + moreinfo
thanks
[Adding intrigeri@boum.org & lynxis@fe80.eu to CC]
Hi László,
> It seems more recent. What do I miss with the one you mentioned? That
> seems untouched for more than a year.
Hm, I'm not sure now (!). Lynxis/intrigeri, can you comment? Marking this
bug as "moreinfo" for now.
> Indeed and I think the package is up to date with all the patches
> needed for the Buster release. If you may need additional changes,
> please tell me so.
We would, of course, love to drop Tails' local fork of squashfs-
tools for buster so I would be happy with whatever the route to get
there.
I think the changes that are required are:
https://git-tails.immerda.ch/squashfs-tools/plain/debian/patches/0011-If-SOURCE_DATE_EPOCH-is-set-use-that-timestamp-for-t.patch?h=debian
… and:
https://git-tails.immerda.ch/squashfs-tools/plain/debian/patches/0012-If-SOURCE_DATE_EPOCH-is-set-also-clamp-content-times.patch?h=debian
… but I am not 100% sure just yet -- need to do a full
reproducibility test run here to confirm. :)
Best wishes,
--
,''`.
: :' : Chris Lamb
`. `'` lamby@debian.org / chris-lamb.co.uk
`-
Added tag(s) moreinfo.
Request was from Chris Lamb <lamby@debian.org>
to control@bugs.debian.org.
(Sun, 06 Jan 2019 15:45:04 GMT) (full text, mbox, link).
Information forwarded
to debian-bugs-dist@lists.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Sun, 06 Jan 2019 16:54:07 GMT) (full text, mbox, link).
Acknowledgement sent
to intrigeri <intrigeri@debian.org>:
Extra info received and forwarded to list. Copy sent to Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Sun, 06 Jan 2019 16:54:07 GMT) (full text, mbox, link).
Message #27 received at 918480@bugs.debian.org (full text, mbox, reply):
Hi,
Chris Lamb:
>> It seems more recent. What do I miss with the one you mentioned? That
>> seems untouched for more than a year.
> Hm, I'm not sure now (!). Lynxis/intrigeri, can you comment? Marking this
> bug as "moreinfo" for now.
I'm not directly involved in lynxis' (renamed) fork personally
and have little at stake here, except of course it would be nice
if Tails could use Debian's squashfs-tools instead of maintaining
our delta :)
IIRC the main goal of the fork was to integrate patches that various
distros apply, many of which had been submitted to the official
squashfs-tools mailing list, but not reviewed. I think that this fork
only makes sense if distros actually switch to it. That's why I've
suggested László privately two years ago (<85y3zd1vlb.fsf@boum.org>)
to switch to this new upstream.
I suspect we're in a chicken'n'egg situation:
- If squashfskit is not actively maintained, then distros will
rightfully avoid switching to it.
- If distros don't switch to squashfskit, then I would understand if
lynxis is not utterly motivated to maintain it.
On the Debian front, it feels a bit late in the Buster cycle
to switch. What about this:
0. Lynxis updates squashfskit if/as needed, ideally in the next six
months. Possibly after another attempt at submitting our patches to
the official squashfs-tools upstream, to make sure we've expressed
our needs clearly enough?
1. Early after the Buster release, Debian tentatively switches
to squashfskit.
2. 6-12 months later, Debian evaluates whether squashfskit is
good enough for our needs; if not, revert to the official
squashfs-tools upstream.
?
I don't know what's going on for other distros.
Cheers,
--
intrigeri
Information forwarded
to debian-bugs-dist@lists.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Sun, 06 Jan 2019 22:21:03 GMT) (full text, mbox, link).
Acknowledgement sent
to Chris Lamb <lamby@debian.org>:
Extra info received and forwarded to list. Copy sent to Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Sun, 06 Jan 2019 22:21:03 GMT) (full text, mbox, link).
Message #32 received at 918480@bugs.debian.org (full text, mbox, reply):
Chris Lamb wrote:
> https://git-tails.immerda.ch/squashfs-tools/plain/debian/patches/0011-If-SOURCE_DATE_EPOCH-is-set-use-that-timestamp-for-t.patch?h=debian
>
> … and:
>
> https://git-tails.immerda.ch/squashfs-tools/plain/debian/patches/0012-If-SOURCE_DATE_EPOCH-is-set-also-clamp-content-times.patch?h=debian
>
> … but I am not 100% sure just yet -- need to do a full
> reproducibility test run here to confirm.
I can now confirm that these make our images reproducible when a
valid SOURCE_DATE_EPOCH is set in the environment. Could you apply
them and upload?
I can file another bug report that you can close if you wish; just
let me know.
Best wishes,
--
,''`.
: :' : Chris Lamb
`. `'` lamby@debian.org / chris-lamb.co.uk
`-
Information forwarded
to debian-bugs-dist@lists.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Sun, 06 Jan 2019 22:54:03 GMT) (full text, mbox, link).
Acknowledgement sent
to László Böszörményi (GCS) <gcs@debian.org>:
Extra info received and forwarded to list. Copy sent to Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Sun, 06 Jan 2019 22:54:03 GMT) (full text, mbox, link).
Message #37 received at 918480@bugs.debian.org (full text, mbox, reply):
On Sun, Jan 6, 2019 at 11:18 PM Chris Lamb <lamby@debian.org> wrote:
> Chris Lamb wrote:
> > https://git-tails.immerda.ch/squashfs-tools/plain/debian/patches/0011-If-SOURCE_DATE_EPOCH-is-set-use-that-timestamp-for-t.patch?h=debian
> >
> > … and:
> >
> > https://git-tails.immerda.ch/squashfs-tools/plain/debian/patches/0012-If-SOURCE_DATE_EPOCH-is-set-also-clamp-content-times.patch?h=debian
> >
> > … but I am not 100% sure just yet -- need to do a full
> > reproducibility test run here to confirm.
>
> I can now confirm that these make our images reproducible when a
> valid SOURCE_DATE_EPOCH is set in the environment. Could you apply
> them and upload?
Just uploaded, may need a minute or two to be accepted.
> I can file another bug report that you can close if you wish; just
> let me know.
There's no need for such a round-trip, thanks.
Regards,
Laszlo/GCS
Information forwarded
to debian-bugs-dist@lists.debian.org, Laszlo Boszormenyi (GCS) <gcs@debian.org>:
Bug#918480; Package squashfs-tools.
(Wed, 27 Feb 2019 03:03:19 GMT) (full text, mbox, link).
Acknowledgement sent
to Alexander Couzens <lynxis@fe80.eu>:
Extra info received and forwarded to list. Copy sent to Laszlo Boszormenyi (GCS) <gcs@debian.org>.
(Wed, 27 Feb 2019 03:03:19 GMT) (full text, mbox, link).
Message #42 received at 918480@bugs.debian.org (full text, mbox, reply):
[Message part 1 (text/plain, inline)]
Hi Laszlo,
I've released squashfskit v4.14.
v4.14, so it's .10 minor version ahead of squashfs-tools. It should be
more clear if it's a bug in squashfs-tools or squashfskit.
Moving to squashfskit would reduce the debian patches from 17 to 3
patches:
0005-add-fstime.patch
0013-use-macros-not-raw-octal-with-chmod.patch
0014-also-set-stickybit-as-non-root.patch
Best Regards,
lynxis
--
Alexander Couzens
mail: lynxis@fe80.eu
jabber: lynxis@fe80.eu
mobile: +4915123277221
gpg: 390D CF78 8BF9 AA50 4F8F F1E2 C29E 9DA6 A0DF 8604
[Message part 2 (application/pgp-signature, inline)]
Send a report that this bug log contains spam.
Debian bug tracking system administrator <owner@bugs.debian.org>.
Last modified:
Wed May 17 12:04:51 2023;
Machine Name:
buxtehude
Debian Bug tracking system
Debbugs is free software and licensed under the terms of the GNU
Public License version 2. The current version can be obtained
from https://bugs.debian.org/debbugs-source/.
Copyright © 1999 Darren O. Benham,
1997,2003 nCipher Corporation Ltd,
1994-97 Ian Jackson,
2005-2017 Don Armstrong, and many other contributors.