| Age | Commit message (Collapse) | Author | 
|---|
  |  | Signed-off-by: Maciej Kisielewski <maciej.kisielewski@canonical.com> | 
  |  | Signed-off-by: Maciej Kisielewski <maciej.kisielewski@canonical.com> | 
  |  | ~bladernr/plainbox-provider-checkbox:1723460-enable-underspeed-network-test | 
  |  | the new launcher | 
  |  | plans for embedding. | 
  |  | ~narahuang/plainbox-provider-checkbox/+git/plainbox-provider-checkbox:mac-address-pass-through | 
  |  |  | 
  |  | ~sylvain-pineau/plainbox-provider-checkbox:fix-nm_connection_backup_ids | 
  |  | Using category in identifiers can generate duplicated job ids if more than one wifi device is available on the system: File "/snap/checkbox-foo/z/lib/python3.x/site-packages/plainbox/impl/session/state.py", line 1106, in _add_job_unit raise DependencyDuplicateError(existing_job, new_job) plainbox.impl.depmgr.DependencyDuplicateError: duplicate job id: 'com.canonical.certification::wireless/nm_connection_save_WIRELESS' | 
  |  |  | 
  |  | ~bladernr/plainbox-provider-checkbox:1818898-disk_stats-nvdimms | 
  |  | NVDIMM as NVDIMMs disable iostats by default. LP: #1818898 | 
  |  |  | 
  |  | Sometimes, the time to sleep or resume cannot be computed accurately by FWTS. Instead of returning a numeric value, FWTS then returns "N/A". This commit handles this kind of case better. The job will fail (because it cannot calculate average time), but will provide clearer information instead of a scary stack trace. LP: #1816809 | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  | The network_device_info script was being used by the ethernet/detect and info/network_devices jobs. This clarifies the two roles and removes the unrequired udev/NM matching code. | 
  |  |  | 
  |  | In wifi_nmcli_test on recent nmcli versions a rescan is requested to ensure an up-to-date list of APs is available. However this request can fail if NM has already started a scan, but we immediately attempt the connection. This should let the background scan complete. | 
  |  | ~sylvain-pineau/plainbox-provider-checkbox:fix-oob-inline-comments | 
  |  | When using a snapped validation tool, the simple '# comment' was producing the following error: error: [...] field 'requires', expression did not reference any resources | 
  |  | ~ray.chen/plainbox-provider-checkbox:automated-switch-gpu-job-into-nested-plan modified: units/graphics/test-plan.pxu and units/suspend/suspend-graphics.pxu | 
  |  | ~bladernr/plainbox-provider-checkbox:1815621-add-gpgpu-tests-checkbox Added new GPGPU tests to checkbox | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  | 'suspend_30_cycles_with_reboots' job was originally created because it was found that suspend (S3) may fail after reboots. Adding 'suspend_30_cycles_with_coldboots' to ensure suspend also work properly after cold boots (i.e. poweroff). A new test plan (stress-suspend-30-cycles-with-coldboots-automated) is available. | 
  |  | The 'power-management/suspend_30_cycles_with_reboots' job is used in 'stress-suspend-30-cycles-with-reboots-automated' test plan, yet depends on 'suspend/suspend_advanced' which is manual. Switching to 'suspend/suspend_advanced_auto' instead. | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  | Release_2019_Week4 | 
  |  |  | 
  |  |  | 
  |  | modified: units/graphics/test-plan.pxu Replace {index}_switch_card_{product_slug}_xenial with {index}_auto_switch_card_{product_slug}. Rename test id “_xenial” with “_auto”. modified: units/suspend/suspend-graphics.pxu Rename test id form “_xenial” to “ _auto” accordingly. Modified test cases dependency from “after: graphics/{index}_switch_card_{product_slug}_xenial” to “after: graphics/{index}_auto_switch_card_{product_slug}” | 
  |  | ~bladernr/plainbox-provider-checkbox:1812262-stress-ng-exit-code-catch | 
  |  | return codes. LP: #1812262 | 
  |  |  | 
  |  |  | 
  |  |  | 
  |  | All modern touchpads are multitouch and those two jobs sometimes return false alarms. touchpad/detected-as-mouse is a better test to identify issues. Fixes: lp:1415358 | 
  |  |  | 
  |  | add /etc/media-info as an alternative to buildstamp On some project (*cough* Chelmsford) buildstamp is present in this media-info file. I created Chelmsford-specific buildstamp job to make it work, but Checkbox looks for a hardcoded buildstamp job from com.canonical.certification namespace when creating the report. | 
  |  | On some project (*cough* Chelmsford) buildstamp is present in this media-info file. I created Chelmsford-specific buildstamp job to make it work, but Checkbox looks for a hardcoded buildstamp job from com.canonical.certification namespace when creating the report. Signed-off-by: Maciej Kisielewski <maciej.kisielewski@canonical.com> | 
  |  | ~taihsiangho/plainbox-provider-checkbox:mr-update-xorg-process-testcase |