[Bug 44831] New: Resizing Wine on a Chromebook prevents Wine from launching again until a reboot

[wine-bugs at winehq.org]

https://bugs.winehq.org/show_bug.cgi?id=44831

            Bug ID: 44831
           Summary: Resizing Wine on a Chromebook prevents Wine from
                    launching again until a reboot
           Product: Wine
           Version: 3.4
          Hardware: x86
                OS: Android
            Status: NEW
          Severity: normal
          Priority: P2
         Component: -unknown
          Assignee: wine-bugs at winehq.org
          Reporter: duboisj at codeweavers.com

Created attachment 60870
  --> https://bugs.winehq.org/attachment.cgi?id=60870
logcat output

Hardware: Google Pixelbook Chrombook
OS: Stable channel Chrome OS (M64.0.3282.190) Android subsystem v. 7.1.1
Wine package: wine 3.4 package from winehq Downloads page

On first launching Wine's Android package on a Chromebook, the Wine desktop is
smaller than a user might want.  Like other Android apps, the Wine desktop has
Chrome OS window decorations on top, which include a resize control.

Using this control to ask that Wine be put in a maximized window results in a
message saying that Wine will be restarted in order to change the window size.

Some percent of the time, this appears to work (whether the desktop can truly
cope with the resize is a separate matter, but wine will at least run after the
resize).  

A high percent of the time, however, Wine dies after the restart and cannot be
restarted again until a reboot.  What happens, in this state, is that you click
on the Wine icon and a window briefly flashes, then disappears.  

If you have a machine in developer mode, you can see that after each click on
the Wine icon, new instances of cmd / wineconsole are being launched and stay
running.  The original wineserver (from before the resize) appears to also stay
running, and never exits.  When wine is running in a sane state, there is also
a process for the java vm / explorer - labelled 'org.winehq.wine' in the output
of a 'ps'.  That process is the one which appears to be killed by the resize
and never manages to restart.

-- 
Do not reply to this email, post in Bugzilla using the
above URL to reply.
You are receiving this mail because:
You are watching all bug changes.