Bug#1038754: random hangs during boot inside qemu

[Diederik de Haas]

On 22 June 2023 06:16:24 CEST, Ben Hutchings  wrote:
>> Is there a chance to see a stable update with
>> e9523a0d81899361214d118ad60ef76f0e92f71d applied?
>
>That's the *breaking* change, not the fix.
>
>The fix would seem to be
>
>but that hasn't even reached mainline yet.

It now has been merged in Linus' tree:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=dad9774deaf1cf8e8f7483310dfb2690310193d2

Bug#1038754: random hangs during boot inside qemu

[Johannes Schauer Marin Rodrigues]

Control: found -1 6.1.27-1

Quoting Ben Hutchings (2023-06-22 06:16:24)
> Which version are you using - is it 6.3.7-1 from unstable?

I added a uname call to the jenkins job:

https://jenkins.debian.net/job/mmdebstrap-jenkins-worker/316/consoleText

Linux osuosl3-amd64 6.1.0-9-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.27-1 
(2023-05-08) x86_64 GNU/Linux

The problems started happening once Holger updated jenkins to run bookworm. So
if this is indeed a kernel problem (as opposed to a qemu problem), then it's a
stable kernel problem.

> > Is there a chance to see a stable update with
> > e9523a0d81899361214d118ad60ef76f0e92f71d applied?
> 
> That's the *breaking* change, not the fix.

See, this is why I write to you instead of starting to randomly patch things
myself. :)

> The fix would seem to be
> 
> but that hasn't even reached mainline yet.

Thank you for looking into this!

cheers, josch

signature.asc
Description: signature

Bug#1038754: random hangs during boot inside qemu

[Ben Hutchings]

Control: tag -1 upstream

On Tue, 2023-06-20 at 23:36 +0200, Johannes Schauer Marin Rodrigues
wrote:
> Source: linux
> Severity: normal
> Tags: patch
> 

Which version are you using - is it 6.3.7-1 from unstable?

> Hi,
> 
> the jenkins job of my package mmdebstrap spawns ~260 qemu virtual
> machines on amd64 to run the test cases in. For about a week now I'm
> seeing random (about every ~300) hangs during boot. The effect can be
> seen in these logs:
> 
>  - https://jenkins.debian.net/job/mmdebstrap-jenkins-worker/311/consoleText
>  - https://jenkins.debian.net/job/mmdebstrap-jenkins-worker/313/consoleText
>  - https://jenkins.debian.net/job/mmdebstrap-jenkins-worker/314/consoleText
> 
> Ctrl+F the logs for "qemu-system-x86_64: terminating on signal 15 from
> pid XXX (timeout)".
> 
> On #debian-devel, f_g suggested that this is a fixed problem:
> 
> https://lore.kernel.org/stable/c4724b40-89f6-5aa7-720d-c4a4af57c...@amazon.com/#r
> https://lore.kernel.org/all/5a56290d-806e-b9a5-f37c-f21958b5a...@grsecurity.net
> https://lore.kernel.org/stable/20230615091830.rxmv2...@linutronix.de/
> 
> Is there a chance to see a stable update with
> e9523a0d81899361214d118ad60ef76f0e92f71d applied?

That's the *breaking* change, not the fix.

The fix would seem to be

but that hasn't even reached mainline yet.

Ben.

-- 
Ben Hutchings
All the simple programs have been written, and all the good names taken



signature.asc
Description: This is a digitally signed message part

Bug#1038754: random hangs during boot inside qemu

[Johannes Schauer Marin Rodrigues]

Source: linux
Severity: normal
Tags: patch

Hi,

the jenkins job of my package mmdebstrap spawns ~260 qemu virtual
machines on amd64 to run the test cases in. For about a week now I'm
seeing random (about every ~300) hangs during boot. The effect can be
seen in these logs:

 - https://jenkins.debian.net/job/mmdebstrap-jenkins-worker/311/consoleText
 - https://jenkins.debian.net/job/mmdebstrap-jenkins-worker/313/consoleText
 - https://jenkins.debian.net/job/mmdebstrap-jenkins-worker/314/consoleText

Ctrl+F the logs for "qemu-system-x86_64: terminating on signal 15 from
pid XXX (timeout)".

On #debian-devel, f_g suggested that this is a fixed problem:

https://lore.kernel.org/stable/c4724b40-89f6-5aa7-720d-c4a4af57c...@amazon.com/#r
https://lore.kernel.org/all/5a56290d-806e-b9a5-f37c-f21958b5a...@grsecurity.net
https://lore.kernel.org/stable/20230615091830.rxmv2...@linutronix.de/

Is there a chance to see a stable update with
e9523a0d81899361214d118ad60ef76f0e92f71d applied?

Thanks!

cheers, josch