Project

General

Profile

Actions
Copy link

action #169207

open

[qe-core] test fails in ulp_openposix

Added by cvidot 4 months ago. Updated 13 days ago.

Status:
Feedback
Priority:
Normal
Assignee:
dzedro
Category:
Bugs in existing tests
Target version:
QA (public) - QE-Core: Ready
Start date:
2024-11-01
Due date:
% Done:

0%

Estimated time:
Difficulty:
Tags:
bugbusters

Description

Observation

openQA test in scenario sle-15-SP4-Server-DVD-Incidents-Livepatches-x86_64-mau-userspace-livepatch@64bit fails in
ulp_openposix

Test suite description

Testsuite maintained at https://gitlab.suse.de/qa-maintenance/qam-openqa-yml.
Basic regression tests for libpulp0 and glibc-livepatches package updates.
Maintainer: mdoucha@suse.cz

Reproducible

Fails since (at least) Build :33680:libpulp

Expected result

Last good: :28608:libpulp (or more recent)

Further details

Always latest result in this scenario: latest

Actions
Copy link
 #2

Updated by dzedro 4 months ago

  • Status changed from New to In Progress
Actions
Copy link
 #3

Updated by dzedro about 2 months ago

  • Status changed from In Progress to Blocked

Update is now rejected, bug still open.

Actions
Copy link
 #4

Updated by mgrifalconi 27 days ago

  • Status changed from Blocked to Feedback

Hey, bug is open but they asked for info, new requests were also accepted. Maybe we can close bug and poo?

Actions
Copy link
 #5

Updated by dzedro 27 days ago

Accepted request do mean nothing, there is no RR now, so nothing to test, needinfo is for nobody.

Actions
Copy link
 #6

Updated by dzedro 15 days ago

Libpulp update passed https://dashboard.qam.suse.de/incident/37452
But it required multiple restarts
The test is failing on all versions e.g.
https://openqa.suse.de/tests/16740251#step/ulp_openposix/4815
https://openqa.suse.de/tests/16751651
I tried to increase RAM or change ULP_THREAD_COUNT & ULP_THREAD_SLEEP, but that didn't help. https://openqa.suse.de/tests/overview?build=jpupava_lp&distri=sle
@MDoucha could you take a look on this test ?

Actions
Copy link
 #7

Updated by MDoucha 15 days ago · Edited

dzedro wrote in #note-6:

Libpulp update passed https://dashboard.qam.suse.de/incident/37452
But it required multiple restarts
The test is failing on all versions e.g.
https://openqa.suse.de/tests/16740251#step/ulp_openposix/4815
https://openqa.suse.de/tests/16751651
I tried to increase RAM or change ULP_THREAD_COUNT & ULP_THREAD_SLEEP, but that didn't help. https://openqa.suse.de/tests/overview?build=jpupava_lp&distri=sle
@MDoucha could you take a look on this test ?

In the first job, it looks like something broke in the kernel. But otherwise, is guess the real problem is that libpulp prints too much debug output. The job didn't even reach the test module where ULP_THREAD_COUNT or ULP_THREAD_SLEEP would matter.

[  198.870498][    C0] irq 25: nobody cared (try booting with the "irqpoll" option)
[  198.871660][    C0] CPU: 0 PID: 13202 Comm: ulp Kdump: loaded Tainted: G           O  K     5.14.21-150500.55.91-default #1 SLE15-SP5 960328bbe7dff1fc5036db3ad3d2a5ca66174f38
[  198.874013][    C0] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.2-3-gd478f380-rebuilt.opensuse.org 04/01/2014
[  198.875832][    C0] Call Trace:
[  198.876319][    C0]  <IRQ>
[  198.876741][    C0]  dump_stack_lvl+0x58/0x7b
[  198.877413][    C0]  __report_bad_irq+0x30/0xc0
[  198.878116][    C0]  note_interrupt+0x247/0x2a0
[  198.878810][    C0]  handle_irq_event_percpu+0x4e/0x70
[  198.879595][    C0]  handle_irq_event+0x34/0x60
[  198.880288][    C0]  handle_edge_irq+0x7e/0x1a0
[  198.880983][    C0]  __common_interrupt+0x3e/0xb0
[  198.881703][    C0]  common_interrupt+0x58/0xa0
[  198.882412][    C0]  </IRQ>
[  198.882846][    C0]  <TASK>
[  198.883280][    C0]  asm_common_interrupt+0x59/0x80
[  198.884026][    C0] RIP: 0010:finish_task_switch+0x91/0x2d0
Actions
Copy link
 #8

Updated by dzedro 13 days ago · Edited

@MDoucha I created https://github.com/os-autoinst/os-autoinst-distri-opensuse/pull/21241 also I would create PR to increase RAM to 4096 if this change is acceptable.
Test can still fail sometimes, but is more stable than before.

Actions
Copy link

Also available in: Atom PDF