Description of problem:
When certain apps are run, like nautilus or gnome-terminal, the global menu (available in the top bar) can't be used and there's a spinner spinning beside it for a long time, even though the app has already started. This takes 15 seconds for nautilus, 30 seconds for gnome-terminal. After that timeout, the global menu can be finally used.
In some cases, there are other consequences as well. For example, "change background" or "display settings" action on the desktop can't be run again until the timeout is over. So if you close "change background" dialog and try to run it again, nothing happens. Only after 30 seconds you can start it again.
For some reason, I see this in all my virtual machines, but I can't reproduce that on my desktop. That is the same case for my colleague (desktop OK, VMs affected).
This only happens on wayland, on X11 everything is fine.
Version-Release number of selected component (if applicable):
100% in VMs
Steps to Reproduce:
1. run nautilus or terminal or "change background"
2. see the spinner not going away, see the global menu not accessible
Created attachment 1201248 [details]
bug demonstration video
Created attachment 1201249 [details]
rpm -qa output
I'll highlight once again: for some reason I only see this in VMs, not on bare metal.
This happens every time I launch gnome-terminal under Wayland (Fedora 25) on bare metal. System is a Thinkpad X220, Intel HD 3000 graphics.
It happens with other GNOME apps as well, like Nautilus. The global menu is unusable and the app doesn't appear in the Alt-Tab menu.
(In reply to Kamil Páral from comment #3)
> I'll highlight once again: for some reason I only see this in VMs, not on
> bare metal.
Seeing this on bare metal too. Also on Intel HD 3000 graphics and on Intel HD (without a number) graphics. This issue has been present for a while, at least during the whole 3.20 release cycle, probably even before.
I saw this behavior in case of Bijiben too.
It happens with all apps launched with dbus activation (the ones with a desktop file starting with /usr/share/applications/org.gnome.*).
Is this bad enough to potentially go on the must-be-fixed-by-release list? Or not?
This happens to me all the time with Gedit and Terminix in addition to the aforementioned software on bare metal (not tested under VMs due to GNOME Boxes bug). I have noticed since F24 and it hadn't changed yet.
My CPU is Intel Iris Pro 5200 and CPU is Intel core i7 (4th generation).
(In reply to Evandro Giovanini from comment #7)
> It happens with all apps launched with dbus activation (the ones with a
> desktop file starting with /usr/share/applications/org.gnome.*).
According to my testing, it happens with most of them, but not all. So it's probably related, but not more complex.
I can reproduce this with new installations in VMs and bare metal, and also just by running LiveCD. It doesn't happen on my laptop for some reason, maybe because it's an upgraded system over several releases. However, I tried upgrading a clean F24 installation to F25 and the problem occurs there as well.
I'm proposing this as a blocker, under this criterion:
"All elements of the default panel (or equivalent) configuration in all release-blocking desktops must function correctly in typical use. "
The global menu doesn't work for a lot of core default apps during the first 15-30 seconds. During the same time, Alt+Tab doesn't work for them either. Certain actions like "Change background" and "Display settings" can't be started (from the desktop context menu) multiple times during the timeout.
Discussed during the 2016-10-17 blocker review meeting: 
The decision to classify this bug as an "AcceptedBlocker" was made as it violates the following criteria:
"All elements of the default panel (or equivalent) configuration in all release-blocking desktops must function correctly in typical use".
This also seems to be the case in F24. ThinkPad T450s, bare metal, x86_64, Wayland session. Example: nautilus.
rui is going to look into startup notification problems
Why is this bug marked as POST?
(In reply to Christian Stadelmann from comment #14)
> Why is this bug marked as POST?
See the upstream Bug on GNOME Bugzilla, there's a fix on its way.
Go/no go is Thursday: we really need this fixed today or tomorrow for f25. It seems from comments upstream that the proposed patches are basically OK, at least for a short term fix: could we do an f25 build with them backported? Thanks.
mutter-3.22.1-8.fc25 gtk3-3.22.2-2.fc25 has been submitted as an update to Fedora 25. https://bodhi.fedoraproject.org/updates/FEDORA-2016-f3f32649f5
Tested with a current Workstation nightly, confirmed both the bug without the update applied and the fix with the update applied.
gtk3-3.22.2-2.fc25, mutter-3.22.1-8.fc25 has been pushed to the Fedora 25 testing repository. If problems still persist, please make note of it in this bug report.
See https://fedoraproject.org/wiki/QA:Updates_Testing for
instructions on how to install test updates.
You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2016-f3f32649f5
back to verified. damnit, bodhi, stop this.
gtk3-3.22.2-2.fc25, mutter-3.22.1-8.fc25 has been pushed to the Fedora 25 stable repository. If problems still persist, please make note of it in this bug report.
*** Bug 1393950 has been marked as a duplicate of this bug. ***