Tails

#1 Updated by segfault 2020-01-16 11:08:47

• related to Bug #17384: devel branch FTBFS since linux-image-5.4.0-1-amd64 replaced linux-image-5.3.0-3-amd64 in sid added

#2 Updated by segfault 2020-01-16 11:09:04

• blocks Feature #16209: Core work: Foundations Team added

#3 Updated by segfault 2020-01-16 11:09:48

• Subject changed from Upgrade Linux to 5.4.8+1 to Upgrade Linux to 5.4.8+

#4 Updated by intrigeri 2020-01-27 08:35:44

• related to Bug #17154: Improve entropy gathering added

#5 Updated by intrigeri 2020-01-27 08:39:06

• Assignee set to intrigeri

I’ll give it a shot for the stable branch, if Bug #17384 (for the devel branch) looks good.

#6 Updated by intrigeri 2020-01-28 09:16:13

• Status changed from Confirmed to In Progress

Applied in changeset commit:tails|0afd359c957751a6e1ab63b60c8cc83e560b2a2b.

#7 Updated by intrigeri 2020-01-28 09:17:35

• Feature Branch set to feature/17384-17443-linux-5.4+force-all-tests

(For now, that’s the branch forked off devel.)

#10 Updated by intrigeri 2020-02-03 14:41:06

• Feature Branch changed from feature/17384-17443-linux-5.4+force-all-tests to feature/17384-17443-linux-5.4-for-stable+force-all-tests

intrigeri wrote:
> On a X1 Carbon gen6: boots fine (USB+UEFI), connects to Wi-Fi, YouTube works in Tor Browser, unplugging the boot device triggers a successful emergency shutdown.

Same on HP EliteBook 840G1 and ThinkPad X200.

> Let’s see what Jenkins thinks!

Jenkins test results don’t look any worse than on the stable branch.

All lights are green so I’ll send Bug #17384 to QA (for the devel branch) and will prepare a branch based on stable here.

#11 Updated by intrigeri 2020-02-04 07:59:20

• Status changed from In Progress to Needs Validation
• Assignee deleted (intrigeri)

Jenkins results don’t look worse than on the stable branch.

I had to bump the APT snapshots so I took a look at the diff between .packages files (stable and this branch). Apart of the obvious kernel upgrade, changes are:

• firmware-linux-free 3.4 → 20200122-1: packaging rewrite, new upstream version that only affects the usbdux firmware, no new bug reported
• python3-aiohttp-socks 0.2.2-1 → 0.3.3-1, python3-aiorpcx 0.18.3-1 → 0.18.4-1: we pull them from sid for Electrum, which feels a bit scary; on the branch for Bug #17384 (that has these upgrades too) I could start Electrum, set up a wallet, and it connected to the network, but I don’t know if that’s enough to exercise how Electrum uses these Python modules; let’s cross fingers and hope that all necessary versioned Depends/Breaks/Replaces are in place to support partial upgrades correctly
• virtualbox-guest-* 6.0.12-dfsg-1 → 6.1.2-dfsg-1: necessary to build against Linux 5.4

⇒ good enough IMO, please review & merge into stable.

#12 Updated by intrigeri 2020-02-04 10:26:12

• blocks Bug #17458: Upgrade to Buster 10.3 added

#13 Updated by anonym 2020-02-05 15:18:40

• Assignee set to anonym

#14 Updated by anonym 2020-02-06 16:44:13

• Status changed from Needs Validation to Resolved
• Assignee deleted (anonym)
• % Done changed from 0 to 100

intrigeri wrote:
> Jenkins results don’t look worse than on the stable branch.

I assume you meant the “devel” branch here. :)

> I had to bump the APT snapshots so I took a look at the diff between .packages files (stable and this branch). Apart of the obvious kernel upgrade, changes are:
>
> * firmware-linux-free 3.4 → 20200122-1: packaging rewrite, new upstream version that only affects the usbdux firmware, no new bug reported
> * python3-aiohttp-socks 0.2.2-1 → 0.3.3-1, python3-aiorpcx 0.18.3-1 → 0.18.4-1: we pull them from sid for Electrum, which feels a bit scary; on the branch for Bug #17384 (that has these upgrades too) I could start Electrum, set up a wallet, and it connected to the network, but I don’t know if that’s enough to exercise how Electrum uses these Python modules; let’s cross fingers and hope that all necessary versioned Depends/Breaks/Replaces are in place to support partial upgrades correctly
> * virtualbox-guest-* 6.0.12-dfsg-1 → 6.1.2-dfsg-1: necessary to build against Linux 5.4
>
> ⇒ good enough IMO, please review & merge into stable.

Agreed!

I’ve seen all scenarios pass locally as well, and the code is obviously sound so => merged!