mouse_hide seems to not work anymore
seems like we broke
mouse_hide with recent changes.
https://openqa.suse.de/tests/564660#step/consoletest_finish/8 should not show a mouse cursor but from the logs we can see that it has been hidden just in before:
18:45:30.8155 Debug: /var/lib/openqa/share/tests/sle/tests/console/consoletest_finish.pm:50 called testapi::mouse_hide 18:45:30.8157 30735 <<< testapi::mouse_hide(border_offset=1) 18:45:30.8162 30737 mouse_move 1023, 767 18:45:30.8163 30737 send_pointer_event 0, 1023, 767, 1 18:45:30.8167 Debug: /var/lib/openqa/share/tests/sle/tests/console/consoletest_finish.pm:51 called testapi::assert_screen 18:45:30.8168 30735 <<< testapi::assert_screen(mustmatch='generic-desktop', timeout=30)
probably introduced by commit 400a02e in os-autoinst. Could maybe also be commit 6772373 but that one states that it has been tested locally.
#1 Updated by AdamWill almost 6 years ago
Well, what's wrong there exactly? The mouse gets moved to 1023x767...should it have been moved somewhere else? Where was it getting moved before with this code? Do you have an older log of the same job? I can't access openqa.suse.de , so I'm flying blind unless you have a log from a public instance...
#2 Updated by coolo almost 6 years ago
You're a bit quick with your analysises. https://openqa.suse.de/tests/564660#step/first_boot/2 "shows" a hidden mouse that was still there in /1 before the mouse_hide. So mouse_hide definitely works in general.
#3 Updated by coolo almost 6 years ago
- Status changed from New to Rejected
and that worker has this code running since at least sep 8th. If mouse_hide was conceptually broken, you would have noticed.
and https://openqa.suse.de/tests/527643#step/consoletest_finish/7 predates all changes you mention and also shows the cursor.
But if you had looked closer, you would have noticed that https://openqa.suse.de/tests/564660#step/consoletest_finish/6 fails - and as such the login is not successful. Why the cursor is not following VNC is a good question, but I don't think mouse_hide is the problem.
#4 Updated by okurz almost 6 years ago
But if you had looked closer, you would have noticed that https://openqa.suse.de/tests/564660#step/consoletest_finish/6 fails - and as such the login is not successful.
I certainly saw this. That's how I came to the current observation. I did not say it failed for this reason. This is also why I tagged the job with bsc#999224. I don't know how this is related, though.
Why the cursor is not following VNC is a good question, but I don't think mouse_hide is the problem.
I don't understand what you mean. Can we reopen the ticket and reword the subject then?