Testcase Confirm Prompt will not update after checkbox resume

[Yung Shen]

During the testcase: graphics/*_switch_card , it will requires user to logout the session which kills the checkbox-gui at the same time.

After loging again and open the canonical-certifcation-client, click on [Rerun last test] button, test case confirm prompt will not change after resume from switching graphic card, it only shows [Graphics/VESA_driver_not_in_use] in the confirm prompt even when following tests.

ProblemType: Bug
DistroRelease: Ubuntu 14.04
Package: linux-image-3.13.0-24-generic 3.13.0-24.46 [modified: boot/vmlinuz-3.13.0-24-generic]
ProcVersionSignature: Ubuntu 3.13.0-24.46-generic 3.13.9
Uname: Linux 3.13.0-24-generic x86_64
ApportVersion: 2.14.1-0ubuntu3
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: ubuntu 6796 F.... pulseaudio
CurrentDesktop: Unity
Date: Fri May 2 06:30:29 2014
HibernationDevice: RESUME=UUID=7d581eac-e128-4c90-9bc8-a8a2d10ee33c
InstallationDate: Installed on 2014-04-30 (2 days ago)
InstallationMedia: Ubuntu 14.04 LTS "Trusty Tahr" - Release amd64 (20140417)
MachineType: Hewlett-Packard HP Envy dv6 Notebook PC
ProcFB: 0 inteldrmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-3.13.0-24-generic root=UUID=c7681f0c-39b6-4438-81e2-d80c9a2f1c7a ro rootdelay=60 quiet splash initcall_debug
RelatedPackageVersions:
 linux-restricted-modules-3.13.0-24-generic N/A
 linux-backports-modules-3.13.0-24-generic N/A
 linux-firmware 1.127
RfKill:
 0: phy0: Wireless LAN
  Soft blocked: no
  Hard blocked: no
SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 11/02/2012
dmi.bios.vendor: Insyde
dmi.bios.version: F.22
dmi.board.asset.tag: Type2 - Board Asset Tag
dmi.board.name: 181B
dmi.board.vendor: Hewlett-Packard
dmi.board.version: 52.24
dmi.chassis.asset.tag: Chassis Asset Tag
dmi.chassis.type: 10
dmi.chassis.vendor: Hewlett-Packard
dmi.chassis.version: Chassis Version
dmi.modalias: dmi:bvnInsyde:bvrF.22:bd11/02/2012:svnHewlett-Packard:pnHPEnvydv6NotebookPC:pvr0896110000005110000620100:rvnHewlett-Packard:rn181B:rvr52.24:cvnHewlett-Packard:ct10:cvrChassisVersion:
dmi.product.name: HP Envy dv6 Notebook PC
dmi.product.version: 0896110000005110000620100
dmi.sys.vendor: Hewlett-Packard

[Zygmunt Krynicki (zyga)]

I'm moving this to checkbox-gui. This is probably another out of countless GUI issues. I'm very sorry you had to experience this.

[Sylvain Pineau (sylvain-pineau)]

I can (finally) reproduce that bug by only selecting the graphics jobs. I had to modify the client cert whitelist as I have only one GPU:

__graphics__
graphics/generator_.*
graphics/2_glxgears_.*
graphics/xorg-version
graphics/2_driver_version_.*
graphics/2_compiz_check_.*
graphics/2_rotation_.*
graphics/2_display_.*
graphics/2_minimum_resolution_.*
graphics/2_maximum_resolution_.*
graphics/xorg-failsafe
graphics/xorg-process
graphics/1_cycle_resolution_.*
graphics/VESA_drivers_not_in_use
graphics/1_switch_card_.*
graphics/1_glxgears_.*
graphics/1_driver_version_.*
graphics/1_compiz_check_.*
graphics/1_rotation_.*
graphics/1_display_.*
graphics/1_minimum_resolution_.*
graphics/1_maximum_resolution_.*
graphics/1_cycle_resolution_.*

Cleaning ~/.cache/plainbox after the whitelist modification helped as I don't see all the tests if I'm trying to resume with a modified whitelist (maybe facing another bug)

[Zygmunt Krynicki (zyga)]

Any ideas with what may be causing this? Incorrect order of events sent
from / to QML when resuming?

On Tue, May 13, 2014 at 9:52 AM, Sylvain Pineau
<email address hidden>wrote:

> I can (finally) reproduce that bug by only selecting the graphics jobs.
> I had to modify the client cert whitelist as I have only one GPU:
>
> __graphics__
> graphics/generator_.*
> graphics/2_glxgears_.*
> graphics/xorg-version
> graphics/2_driver_version_.*
> graphics/2_compiz_check_.*
> graphics/2_rotation_.*
> graphics/2_display_.*
> graphics/2_minimum_resolution_.*
> graphics/2_maximum_resolution_.*
> graphics/xorg-failsafe
> graphics/xorg-process
> graphics/1_cycle_resolution_.*
> graphics/VESA_drivers_not_in_use
> graphics/1_switch_card_.*
> graphics/1_glxgears_.*
> graphics/1_driver_version_.*
> graphics/1_compiz_check_.*
> graphics/1_rotation_.*
> graphics/1_display_.*
> graphics/1_minimum_resolution_.*
> graphics/1_maximum_resolution_.*
> graphics/1_cycle_resolution_.*
>
> Cleaning ~/.cache/plainbox after the whitelist modification helped as I
> don't see all the tests if I'm trying to resume with a modified
> whitelist (maybe facing another bug)
>
> --
> You received this bug notification because you are a member of Checkbox
> Bug Wranglers, which is subscribed to checkbox in Ubuntu.
> https://bugs.launchpad.net/bugs/1315337
>
> Title:
> Testcase Confirm Prompt will not update after checkbox resume
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/checkbox-gui/+bug/1315337/+subscriptions
>

[Sylvain Pineau (sylvain-pineau)]

I've finally made some progress on this bug. I can confirm now that the manual job "2013.com.canonical.certification::graphics/1_switch_card_3rd_Gen_Core_processor_Graphics_Controller" (on my system) is never present in the PBTreeNode that we get after resume. Gui-engine does not rerun local jobs (and my 1_switch_card* is generated by a the graphics generator job) and should instead use the test definitions that we stored in the session storage.

It does not seem to work as jobs generated from local jobs are not present after resume hence the shift of index and the wrong test definitions displayed in the gui.

The gui received a good index from the run list but as jobs data are pulled from PBTreeNode (a sort of cache), buit here PBTreeNode is not the one we expected.

[Zygmunt Krynicki (zyga)]

Can you confirm if this is a plainbox core resume bug or not? I'm not quite
sure from your description.

On Fri, May 16, 2014 at 3:25 PM, Sylvain Pineau
<email address hidden>wrote:

> I've finally made some progress on this bug. I can confirm now that the
> manual job
>
> "2013.com.canonical.certification::graphics/1_switch_card_3rd_Gen_Core_processor_Graphics_Controller"
> (on my system) is never present in the PBTreeNode that we get after
> resume. Gui-engine does not rerun local jobs (and my 1_switch_card* is
> generated by a the graphics generator job) and should instead use the
> test definitions that we stored in the session storage.
>
> It does not seem to work as jobs generated from local jobs are not
> present after resume hence the shift of index and the wrong test
> definitions displayed in the gui.
>
> The gui received a good index from the run list but as jobs data are
> pulled from PBTreeNode (a sort of cache), buit here PBTreeNode is not
> the one we expected.
>
> --
> You received this bug notification because you are a member of Checkbox
> Bug Wranglers, which is subscribed to checkbox in Ubuntu.
> https://bugs.launchpad.net/bugs/1315337
>
> Title:
> Testcase Confirm Prompt will not update after checkbox resume
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/checkbox-gui/+bug/1315337/+subscriptions
>

[Zygmunt Krynicki (zyga)]

Can you attach 'plainbox session archive' of that session?

[Sylvain Pineau (sylvain-pineau)]

When resuming a session we never refreshed the PBTreenode list of jobs tha tthe get during gui-engine Init.
As on resume gui-engine does not rerun local jobs, jobs generated from such jobs even if available on dbus were not present in the PBTreenode object. Solved by calling RefreshPBObjects() in GuiEngine::GuiResumeSession()