Bug 449047

Summary:	Fullscreen firefox locks screen refresh
Product:	[Plasma] kwin	Reporter:	Ian Kidd <kde-bugs>
Component:	platform-drm	Assignee:	KWin default assignee <kwin-bugs-null>
Status:	RESOLVED FIXED
Severity:	normal	Keywords:	wayland-only
Priority:	NOR
Version First Reported In:	5.23.90
Target Milestone:	---
Platform:	Arch Linux
OS:	Linux
See Also:	https://bugs.kde.org/show_bug.cgi?id=448818
Latest Commit:	https://invent.kde.org/plasma/kwin/-/commit/db1f7414b746aec69d6ba951acf9568598fef513	Version Fixed/Implemented In:	5.24.0
Sentry Crash Report:
Attachments:	wayland-session.log qdbus.txt

Description Ian Kidd 2022-01-24 06:32:39 UTC

STEPS TO REPRODUCE
1. open firefox 
2. hit F11 for fullscreen
3. 

OBSERVED RESULT
Screen refresh stops and page becomes stuck.  It will take input but does not refresh screen.  Hitting F11 again doesn't refresh screen to regular size.  Manually closing window in task manager does not refresh screen.  Other windows dragged on to that screen are not visible, though they do respond to input if you guess the position correctly.  Requires log out or shell reload to drop last shown view.

EXPECTED RESULT

fullscreen then return to normal

ADDITIONAL INFORMATION

May be because using MOZ_ENABLE_WAYLAND switch when running Firefox so it's not using xwayland?

Comment 1 Ian Kidd 2022-01-24 06:35:19 UTC

Correction: shell reload with `kquitapp5 plasmashell && kstart plasmashell` does not refresh that screen.

Comment 2 Ian Kidd 2022-01-24 06:41:59 UTC

Logged error when hitting F11:  radeon 0000:02:00.0: cannot be used for peer-to-peer DMA as the client and provider (0000:01:00.0) do not share an upstream bridge or whitelisted host bridge 

Can go to task bar and choose Move in that window's menu, move out from under frozen refresh and application is working fine.

Comment 3 Vlad Zahorodnii 2022-01-24 10:04:28 UTC

Can you attach the session logs `~/.local/share/sddm/wayland-session.log` after making firefox fullscreen and the output of the following command `qdbus org.kde.KWin /KWin supportInformation` to this bug report please?

and do you have a laptop with a discrete gpu?

Comment 4 Ian Kidd 2022-01-24 14:54:23 UTC

Created attachment 145857 [details]
wayland-session.log

Attached is wayland-session.log, but it doesn't seem to log anything for the F11 push.  I also notice it locks up the other FF window on another monitor attached to the same GPU, but not the FF sessions displayed by a second GPU.

This isn't a laptop, it's a desktop with a RX560 and a HD7770 (which is the one the problems seem attached to), each running 3 monitors.  I'm using the latest Arch opensource drivers for the cards (mesa 21.3.4-1).

Comment 5 Ian Kidd 2022-01-24 14:55:29 UTC

Created attachment 145858 [details]
qdbus.txt

Comment 6 Ian Kidd 2022-01-24 14:57:39 UTC

Also, if I do an F11 on the RX560 attached FF session, there is no problem with it.

`lspci -v | grep VGA`

01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Baffin [Radeon RX 550 640SP / RX 560/560X] (rev cf) (prog-if 00 [VGA controller])
02:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Cape Verde XT [Radeon HD 7770/8760 / R7 250X] (prog-if 00 [VGA controller])

Comment 7 Vlad Zahorodnii 2022-01-24 15:39:54 UTC

(In reply to Ian Kidd from comment #4)
> Created attachment 145857 [details]
> wayland-session.log
> 
> Attached is wayland-session.log, but it doesn't seem to log anything for the
> F11 push.  I also notice it locks up the other FF window on another monitor
> attached to the same GPU, but not the FF sessions displayed by a second GPU.
> 
> This isn't a laptop, it's a desktop with a RX560 and a HD7770 (which is the
> one the problems seem attached to), each running 3 monitors.  I'm using the
> latest Arch opensource drivers for the cards (mesa 21.3.4-1).

Do you use systemd boot?

Comment 8 Ian Kidd 2022-01-24 20:17:15 UTC

(In reply to Vlad Zahorodnii from comment #7)

> Do you use systemd boot?

Grub on mbr

Comment 9 Vlad Zahorodnii 2022-01-25 10:42:32 UTC

Given "Logged error when hitting F11:  radeon 0000:02:00.0: cannot be used for peer-to-peer DMA as the client and provider (0000:01:00.0) do not share an upstream bridge or whitelisted host bridge " in comment 2, this bug should be already fixed.

Please reopen this bug report if the issue is back.