Bug #16073

Upgrade Linux to 4.19

Added by intrigeri 2018-10-25 09:05:49 . Updated 2019-01-30 11:52:02 .

Status:
Resolved
Priority:
Elevated
Assignee:
Category:
Target version:
Tails_3.12
Start date:
2018-10-25
Due date:
% Done:

100%

Feature Branch:
feature/16073-linux-4.19+force-all-tests
Type of work:
Code
Blueprint:

Starter:
Affected tool:
Deliverable for:

Description

Subtasks

Related issues

Related to Tails - Bug #16145: Upgrade Linux to 4.18.20 Resolved 2018-11-22
Blocks Tails - Bug #16072: Enable protected_fifos and protected_regular Resolved 2018-10-25
Blocks Tails - Bug #16224: Black screen after the boot menu with Intel GPU (i915) Resolved 2018-12-13
Blocks Tails - Feature #14991: Remove /usr/share/live/config/xserver-xorg/intel.ids Resolved 2017-11-18
Blocks Tails - Feature #15507: Core work 2019Q1: Foundations Team Resolved 2018-04-08
Blocked by Tails - Bug #16349: Stick to Tor 0.3.4 in Tails 3.12 Resolved 2019-01-12
Blocks Tails - Bug #16352: Fix systemd vulnerabilities: CVE-2018-16864, CVE-2018-16865 and CVE-2018-16866 Resolved 2019-01-13

History

#1 Updated by intrigeri 2018-10-25 09:05:59

  • blocks Bug #16072: Enable protected_fifos and protected_regular added

#2 Updated by intrigeri 2018-10-25 09:06:22

#3 Updated by segfault 2018-11-06 15:40:44

  • Assignee changed from intrigeri to segfault

#4 Updated by segfault 2018-11-06 15:41:24

  • Assignee changed from segfault to intrigeri
  • QA Check set to Info Needed

#5 Updated by intrigeri 2018-11-07 12:48:10

  • Assignee changed from intrigeri to segfault
  • QA Check deleted (Info Needed)

There you go: https://tails.boum.org/contribute/Linux_kernel/

#6 Updated by intrigeri 2018-11-22 09:01:17

  • related to Bug #16145: Upgrade Linux to 4.18.20 added

#7 Updated by intrigeri 2018-12-07 10:23:55

  • Target version changed from Tails_3.11 to Tails_3.12

4.19 was not uploaded to sid yet so given where we’re at in the Tails 3.11 cycle, let’s assume we won’t have to do this and instead focus on Bug #16145.

#8 Updated by intrigeri 2018-12-19 12:56:11

4.19 is now in sid, please proceed early enough in the 3.12 cycle so we have time to identify any regression :)

#9 Updated by intrigeri 2019-01-02 05:10:06

  • blocks Bug #16224: Black screen after the boot menu with Intel GPU (i915) added

#10 Updated by intrigeri 2019-01-02 09:04:29

  • Priority changed from Normal to Elevated

(This blocks Bug #16224.)

Can you please give an ETA here? Maybe you need help?

#11 Updated by intrigeri 2019-01-04 10:16:13

  • blocks Feature #14991: Remove /usr/share/live/config/xserver-xorg/intel.ids added

#12 Updated by intrigeri 2019-01-04 10:28:28

  • Assignee changed from segfault to CyrilBrulebois

kibi has to do the first steps of this work for feature/buster so let’s do it based on devel :)

#13 Updated by CyrilBrulebois 2019-01-04 10:39:09

  • Status changed from Confirmed to In Progress

Applied in changeset commit:tails|bb76938cca399e6307aa9931bf54dd8a06e03425.

#14 Updated by CyrilBrulebois 2019-01-04 10:41:53

I’ve pushed bb76938cca399e6307aa9931bf54dd8a06e03425 with feature/buster since the toolchain needed for 4.18 wasn’t available in buster, and I’ll prepare a separate branch based on devel to prepare the same upgrade for 3.12.

#15 Updated by CyrilBrulebois 2019-01-04 10:51:11

  • Feature Branch set to feature/16073-linux-4.19+force-all-tests

#16 Updated by intrigeri 2019-01-04 15:03:33

#17 Updated by intrigeri 2019-01-04 15:04:03

  • blocked by deleted (Feature #15506: Core work 2018Q4: Foundations Team)

#18 Updated by intrigeri 2019-01-05 09:17:03

FTBFS:

01:40:56 The following packages have unmet dependencies:
01:40:56  linux-headers-4.19.0-1-amd64 : Depends: linux-compiler-gcc-8-x86 but it is not going to be installed
01:40:56 E: Unable to correct problems, you have held broken packages.
01:40:56 E: config/chroot_local-hooks/12-kernel-modules-build-environment failed (exit non-zero). You should check for errors.

#19 Updated by CyrilBrulebois 2019-01-06 18:19:20

Right, this is due to https://salsa.debian.org/kernel-team/linux/commit/01900fbc7319c7a427886141b0f051154534b56a →

commit 01900fbc7319c7a427886141b0f051154534b56a
Author: Ben Hutchings <ben@decadent.org.uk>
Date:   Tue Aug 28 22:31:59 2018 +0100

    Compile with gcc-8 on all architectures

    Update the compiler dependencies for x86 to the first version of gcc-8
    with retpoline support.

Will bump the “fake our gcc package” infrastructure from version 7 to 8.

#20 Updated by CyrilBrulebois 2019-01-06 19:25:07

Confirmed to build fine, now waiting for tests results.

#21 Updated by intrigeri 2019-01-09 17:25:40

I took a quick look at test suite jobs 1-4 on Jenkins. Apart of the memory erasure tests that fail on devel too, and a few fragile tests that failed in some runs but not others (so can be mostly safely ignored), “Using Totem ǂ Watching a WebM video over HTTPS” failed every time. Before we call this ready for QA, let’s test this manually on an ISO built from this branch, to ensure it’s not a regression brought by Linux 4.19.

#22 Updated by intrigeri 2019-01-12 11:15:33

  • Assignee changed from CyrilBrulebois to intrigeri

#23 Updated by intrigeri 2019-01-12 15:00:37

intrigeri wrote:
> “Using Totem ǂ Watching a WebM video over HTTPS” failed every time. Before we call this ready for QA, let’s test this manually on an ISO built from this branch, to ensure it’s not a regression brought by Linux 4.19.

“Run” this Totem test case manually on this branch, works fine. And while I was at it I tested the same with Totem started from the Applications menu, which failed due to a buggy build-time hook (the torification tweak was not applied). Testing a fix.

Apart of that, pidgin.feature started failing on this branch on Jenkins. Tracked this down to the upgrade to tor 0.3.5 => let’s postpone that upgrade to Tails 3.13 (Feature #16348, Bug #16349).

#24 Updated by intrigeri 2019-01-12 15:41:20

  • Assignee changed from intrigeri to CyrilBrulebois
  • % Done changed from 0 to 50
  • QA Check set to Ready for QA

Tested on bare metal the branch for Bug #16072, which is based on this one. It boots, Wi-Fi connects, Tor Browser starts, unplugging the USB stick triggers successful shutdown: HP EliteBook 840 G1, ThinkPad X200.

Regarding the Totem issue: with my fix (commit:07e0cbed3053ab0407a982654007641fa5865702) applied, testing manually, Totem can read this WebM video over HTTPs both when started from a terminal and when started from the Applications menu. I don’t know why the test fails in our test suite, but I guess there’s a reason why this test case is tagged @fragile.

So I’m calling this ready for QA. kibi, can you take it?

I suggest batching this review with the one for Bug #16072, whose branch is based on this one. It should be ready for QA in ~3h.

#25 Updated by intrigeri 2019-01-12 15:47:14

  • Assignee deleted (CyrilBrulebois)

(Emailing the whole team to bootstrap some review coordination.)

#26 Updated by intrigeri 2019-01-12 19:54:02

  • blocked by Bug #16349: Stick to Tor 0.3.4 in Tails 3.12 added

#27 Updated by intrigeri 2019-01-13 12:08:17

  • blocks Bug #16352: Fix systemd vulnerabilities: CVE-2018-16864, CVE-2018-16865 and CVE-2018-16866 added

#28 Updated by hefee 2019-01-14 10:01:26

  • Assignee set to hefee

#29 Updated by hefee 2019-01-14 10:56:37

  • Assignee deleted (hefee)

looks fine code-wise - needs bar metal checks.

#30 Updated by intrigeri 2019-01-14 11:14:53

  • Assignee set to kurono

#31 Updated by intrigeri 2019-01-14 17:12:43

Oops, I’ve just merged this into devel by mistake (by merging another branch that was based on this one). Fingers crossed.

#32 Updated by kurono 2019-01-14 18:24:31

  • Assignee changed from kurono to intrigeri

This branch with Linux to 4.19 works fine in my laptop with the Intel Corporation HD Graphics 520 hardware.

#33 Updated by intrigeri 2019-01-14 18:32:12

  • Status changed from In Progress to Fix committed
  • Assignee deleted (intrigeri)
  • % Done changed from 50 to 100
  • QA Check changed from Ready for QA to Pass

Thanks! All right then. anonym, if you can test with NVIDIA, this is still welcome.

#34 Updated by anonym 2019-01-15 13:50:24

I tested on my T430 and everything worked nicely when I (via BIOS) disabled the Intel card and enabled the NVidia card. In Optimus mode Intel was preferred, FWIW.

#35 Updated by intrigeri 2019-01-15 21:07:57

> I tested on my T430 and everything worked nicely when I (via BIOS) disabled the Intel card and enabled the NVidia card. In Optimus mode Intel was preferred, FWIW.

Thanks!

#36 Updated by anonym 2019-01-30 11:52:02

  • Status changed from Fix committed to Resolved