Bug 1611706
| Summary: | virt-viewer 6.0 frequently stops responding to mouse scroll-wheel input (usually when full-screen) | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Robert <redhat-bz> | ||||
| Component: | virt-viewer | Assignee: | Daniel Berrangé <berrange> | ||||
| Status: | CLOSED NOTABUG | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||
| Severity: | medium | Docs Contact: | |||||
| Priority: | unspecified | ||||||
| Version: | 28 | CC: | berrange, cfergeau, fidencio, marcandre.lureau, redhat-bz, victortoso | ||||
| Target Milestone: | --- | ||||||
| Target Release: | --- | ||||||
| Hardware: | x86_64 | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2018-08-24 13:16:00 UTC | Type: | Bug | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Attachments: |
|
||||||
|
Description
Robert
2018-08-02 16:38:36 UTC
I thought I would add that I also have two monitors, so (for me) "full screen" does not imply that the mouse cursor might leave & return... could be an important detail. - The remote-viewer client is running on KDE or GNOME? Under X11 or Wayland? - Does it take too long to reproduce? - Could you please provide the log from your client using --spice-debug which includes the situation that the scroll stopped working? I tried to reproduce for about 5 minutes on X11/GNOME F28 client with X11/GNOME F28 guest, nothing so far. Cheers, Created attachment 1473028 [details]
spice-debug log of brief virt-viewer session effected by loss of scroll wheel input
* I'm running under kde/x11
* I don't think it is time dependent. Sometimes the scroll wheel works for a long time, and other times it abruptly stops working shortly after entering a virt-viewer window. Now I get the impression it is more like a co-incidence of inputs, or something...
* I've attached such a log where (1) the interaction time is short, (2) the scroll wheel stopped working, and (3) upon noticing it stop working, I 'leave' the window to cut the log file.
In case it saves you a bit of time, I *think* these are the messages you are looking for (which seem to replace the 'scroll_event' messages?):
'''
decode-glz.c:352 decode_header: 758x48, id 451, ref 421
decode-glz.c:352 decode_header: 758x48, id 452, ref 421
decode-glz.c:352 decode_header: 758x48, id 453, ref 421
decode-glz.c:352 decode_header: 758x48, id 454, ref 421
decode-glz.c:352 decode_header: 758x48, id 455, ref 422
decode-glz.c:352 decode_header: 758x48, id 456, ref 422
decode-glz.c:352 decode_header: 758x48, id 457, ref 422
decode-glz.c:352 decode_header: 758x48, id 458, ref 422
'''
NB: I have recently found that... *often*... all that is needed to "reset" virt-viewer from this state (i.e. work around the issue) is cursor out-then-back-into the viewer window...
I am increasingly convinced that this is actually a kwin problem, as I have noticed other window-ordering issues, and have even experienced another application fail to accept scroll-wheel input... my guess is that virt-viewer is just more susceptible to whatever the kwin issue is due to my workflow. |