| Age | Commit message (Collapse) | Author |
|---|
| Now flake8 testing is also being running on jammy with python 3.10 there are some new fails that need ignoring of fixing. This is simple modification to achieve a pass. |
| |
| The control interface needed to be specified in the checkbox.conf to execute the gsm connection test. This is a minimal change to now pull that information from mmcli |
| ~bladernr/plainbox-provider-checkbox:1983122-fix-thread_count-stress-ng Fixes a couple bugs with "thread_count" on stressors |
| the list. This fixes a newly introduced bug as well as an old bug that I stumbled on while fixing this. lp: #1983122 |
| ~weichenwu/plainbox-provider-checkbox:CQT-632-bluetooth-test-cases-and-test-plans-modification Add: bluetooth obex job and bluetooth-cert test plan Previously OBEX was under suspend category and it's hard to maintain or rerun Hence put it under bluetooth and add new test plan to in include it for further use 1. add bluetooth_obex_send under bluetooth 2. add obex and beacon cases to bluetooth-cert-automated 3. new test plan after-suspend-bluetooth-cert-automated 4. remove bluetooth test cases from suspend test plan |
| |
| Previously OBEX was under suspend category and it's hard to maintain or rerun Hence put it under bluetooth and add new test plan to in include it for further use 1. add bluetooth_obex_send under bluetooth 2. add obex and beacon cases to bluetooth-cert-automated 3. new test plan after-suspend-bluetooth-cert-automated 4. remove bluetooth test cases from suspend jobs/plans |
| |
| ~pieq/plainbox-provider-checkbox:fix-1980475-fingerprint-lock-screen |
| |
| |
| The light_sensor_test.sh currently only has one type of hid_sensor_als module. It will return a light sensor driver not found even though the function is available on the i2c type. Therefore, I find we can use the value of IIO_SENSOR_PROXY_TYPE to verify whether the system has light sensor hardware component or not. It is a more generic way to have the type with als instead of checking the specific "hid_sensor_als" driver. |
| |
| ~rickwu4444/plainbox-provider-checkbox:revert-suspend-method-for-suspend-stress Fix: Revert suspend method for suspend stress Since this method have different behavior with other job. It will cause for loop keep looping and not wait for rtcwake alert. |
| Since this method have different behavior with other job, and it will cause for loop keep looping and not wait for rtcwake alert |
| ~rickwu4444/plainbox-provider-checkbox:change-suspend-method-if-using-rtcwake Change: Change suspend method if using rtcwake Due to rtcwake is calling kernel power states directly to put system into suspend. That will cause user space services didn't been handle. Therefore, using systemd to handle suspend would be better for handle the user space services. |
| |
| |
| Previous method emulated keystrokes to trigger the GNOME lock screen (Super+L). This does not work with Wayland security model. Instead, use loginctl which allows to lock all the open sessions at once. This works locally as well as through SSH and also with Checkbox remote. LP: #1980475 |
| Release_2022_Week27 |
| |
| |
| |
| |
| ~kissiel/plainbox-provider-checkbox:fix-1980372-instant-usb-insert-pass |
| I think I accidentally removed the code that for non-zapper branch of logic waited for the manipulation to be done. Fixes: LP: #1980372 |
| Due to rtcwake is calling kernel power states directly to put system into suspend. That will cause userspace services didn't been handle. Therefore, using systemd to handle suspend would be better for handle the userspace services. |
| |
| |
| ~pieq/plainbox-provider-checkbox:fix-xorg-jobs-still-being-pulled |
| The resolution change and rotation jobs were still depending on xorg-related job, even though this dependency is not actually required. Removing them. LP: #1968842 |
| Fix: install-method-check.sh to work with MAAS server IDed by IP address |
| Fix some flake8 errors that are triggering failures in merge tests |
| tests LP: #1978355 |
| #1979036) |
| |
| |
| |
| The touchpad_test.py script allows to test all of the four directions at once. After discussing with QA team, there never was a case where only horizontal or only vertical scrolling did not work. It's either all or nothing. touchpad/multitouch-horizontal and touchpad/multitouch-vertical are therefore merged into one touchpad/multitouch job. |
| |
| ~pieq/plainbox-provider-checkbox:1968943-rework-max-resolution-job |
| The graphics_modes_info.py script is used in two jobs that are not called in any of our test plans. Moreover, since it uses xrandr, the output might be incorrect on recent versions of Ubuntu (22.04+) using Wayland as default. |
| On Wayland, xrandr returns the current resolution as the maximum resolution, even when it is not. For instance, if a laptop has a 3840x2160 screen but is currently set to 2560x1440, xrandr would return 2560x1440 as the maximum available resolution. Moreover, the python oneliner to check current resolution breaks on recent versions of Gtk because Gdk.Screen has been deprecated. Introduce a new script that: - retrieves monitor(s) info from sysfs (port, max resolution, status, etc.) - uses up-to-date Gtk 3 methods to grab the current resolution and scale factor for each connected monitors - displays this information for tester to review - compares data from sysfs and Gtk and returns an error if their data does not match (so that the graphics/*_maximum_resolution_* job status is automatically set to Passed or Failed depending of the outcome) Update the graphics/*_maximum_resolution_* job to call this script. LP: #1968943 |
| "scroll-method" was removed back in 2015[1] and replaced with "edge-scrolling-enabled" and "two-finger-scrolling-enabled" keys (both of them boolean values)[2]. Moreover, Wayland and recent versions of X (libinput) dispatch a GDK_SCROLL_SMOOTH event instead of the old GDK_SCROLL_[UP|DOWN|LEFT|RIGHT]. However, using this, the script has to compute the direction using the deltas provided by event.get_scroll_deltas(). This works on default GNOME settings ("Natural Scrolling" enabled). LP: #1969170 [1] https://github.com/GNOME/gsettings-desktop-schemas/commit/656aca6942c8417ddb3216063b97b7a1336076c9 [2] https://github.com/GNOME/gsettings-desktop-schemas/commit/d9dfe56726e49d3a31354466c3fffa2aaedf5a5e |
| "horiz-scroll-enabled" has been removed in GNOME 3.34 and was never an official key to begin with (it had to be enabled manually in dconf). |
| Fix code to avoid deprecation warnings from Gtk in the logs |
| Gtk.STOCK_INDEX shows an icon with a few horizontal lines in recent versions of Ubuntu. Replacing this with Gtk.STOCK_DIALOG_QUESTION, an interrogation mark. |
| |
| ~pieq/plainbox-provider-checkbox:remove-xorg-jobs-from-test-plans |
