Project

General

Profile

action #15496

qemu -vga qlx support

Added by dimstar almost 6 years ago. Updated about 5 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
-
Category:
Feature requests
Target version:
-
Start date:
2016-12-14
Due date:
% Done:

0%

Estimated time:
Difficulty:

Description

In order to be able to test GNOME on Wayland (we really should start doing this) - we need a more recent graphic engine than cirrus

As far as I understand, cirrus is not really the recommended mode anymore by qemu devs, and is also not the default. qxl would be replacing it.

I configured a machine 64bit_qcl - which sets QEMUVGA=qxl - and configured this as a machine on GNOME (not wayland yet, but X should also work)
This test fails to launch entirely though: https://openqa.opensuse.org/tests/321195

The error from the log file: 17:22:37.5882 7806 QEMU: ((null):7810): Spice-ERROR **: reds.c:3333:do_spice_init: statistics shm_open failed, Permission denied


Related issues

Related to openQA Tests - action #21786: [functional]proper wayland supportResolved2017-08-04

Related to openQA Tests - action #27062: [sle][functional][sle15][desktop] Add new QEMUVGA types for Wayland testingResolved2017-10-27

Related to openQA Tests - action #53339: [opensuse] test fails in swing due to incorrect rendering on 16bpp framebuffersResolved2019-06-19

History

#1 Updated by mlin7442 almost 6 years ago

uno missing this apparmor file update https://github.com/os-autoinst/openQA/pull/1078

and seems need to add

/dev/shm/* w,

to worker's profile for qxl.

#2 Updated by mlin7442 almost 6 years ago

I updated worker profile on uno then.

#3 Updated by okurz@suse.de almost 6 years ago

is that because of apparmor?
EDIT: Sorry, missed the comment, ignore

#4 Updated by mlin7442 almost 6 years ago

the second run is on ow4 which should not have apparmor problem https://openqa.opensuse.org/tests/321628

the clue from log,

06:39:41.3929 20665 QEMU: ((null):20669): SpiceWorker-Warning *: red_worker.c:11333:dev_create_primary_surface: wrong primary surface creation request
06:39:41.3939 20665 QEMU: ((null):20669): Spice-Warning *
: red_parse_qxl.c:424:bitmap_consistent: image stride too small for width: 2048 < ((1024 * 32 + 7) / 8) (SPICE_BITMAP_FMT_32BIT=8)
06:39:41.4025 20665 QEMU: ((null):20669): SpiceWorker-Warning *: red_worker.c:1309:validate_surface: canvas address is 0x7f048c000bc8 for 0 (and is NULL)
06:39:41.4026 20665 QEMU:
06:39:41.4026 20665 QEMU: ((null):20669): SpiceWorker-Warning *
: red_worker.c:1310:validate_surface: failed on 0
06:39:41.4026 20665 QEMU: ((null):20669): SpiceWorker-Warning **: red_worker.c:163:rendering_incorrect: rendering incorrect from now on: handle_dev_update_async

#5 Updated by coolo almost 6 years ago

this sounds like a problem with the product actually. Is the qxl driver even part of installation-images?

#6 Updated by okurz almost 6 years ago

  • Category set to Feature requests

#7 Updated by okurz about 5 years ago

  • Related to action #21786: [functional]proper wayland support added

#8 Updated by qkzhu about 5 years ago

  • Related to action #27062: [sle][functional][sle15][desktop] Add new QEMUVGA types for Wayland testing added

#9 Updated by qkzhu about 5 years ago

fyi: The kernel bug of QXL has been fixed: https://bugzilla.opensuse.org/show_bug.cgi?id=1057241

I am not sure if the qxl support is implemented by openQA, and o.o.o also uses QEMUVGA=virtio for kde_walyand testing.
So I will use virtio for SLED15 testing in action#27062

#10 Updated by coolo about 5 years ago

  • Status changed from New to Resolved

So I see nothing left to be done for openqa

#11 Updated by okurz over 3 years ago

  • Related to action #53339: [opensuse] test fails in swing due to incorrect rendering on 16bpp framebuffers added

Also available in: Atom PDF