Fedora 29 onboard video memory problem

[luigicantoni]

My onboard video memory appears to map over itself.
The problem is that I login onto a gnome gui session (CTRL-ALT F2) session once logged in.
After using that session for a while (like say bring up Firefox) then my CTRL-ALT-F3 or F4 etc sessions have a piece of graphic data in the top left corner. Its about 35chr wide and 16 lines long.
It is clearly a bitmap layout of the gui screen. I don't mean a tiny image but if the gui changes so too does that image.
It is basically bits displayed as coloured dots.

What I have worked out is that if I put in a video card in (thus disabling the onboard) I do not see the issue.
This does not mean it is not there its just I don't see it as the card should be using its own memory for the video images.
My big concern is that in fact the issue is there and will (does) cause some problems.

What I have worked out.
I have 4 systems.
1) ASUS mother board with a Pentium CPU (3-5 years old)
2) Gigabit Mother board with I5 CPU (1-2 years old)
3) MSI Mother board with I5 CPU (several months old)
4) MSI Mother board with I3 CPU (weeks old)

I have several of the ASUS and never noticed an issue.
The gigabit board one I think had the problem with earlier kernels as well as F29 but because I wanted more video I had a card in it and so I never noticed the problem after the machine was put together.
The first MSI machine I noticed it but the machine was idle for a bit and I thought I must have a problem but did not investigate.
When I saw the same problem on the second machine and saw it was the same I then started testing.

My conclusion using exactly the same DISK (and various other versions like F28 etc also) the ASUS is not problem at all.
All the other three only have the problem when no video card installed and all don't show it when installed.

Has anyone else seen this problem and have any ideas? I have not found anything anywhere of a problem like this.
I am happy to try things if I can.
At the moment my solution is put in video cards but I really would much rather that the problem is sorted because I am worried about future strange problems.
Maybe it is just some parameter that needs to be set so that the video memory does not overlap.
I have had some other strange issues with all three machines that have problems but they are very intermittent the machines are used by others so I don't exactly know if they are related to this problem or something else. I just don't know at the moment so I am not prepared to say this causes them.

[bobx001]

What video cards are you using on all of these systems ?

[luigicantoni]

What video cards are you using on all of these systems ?

They are the onboard videos.
The failing mother boards are:
MSI H310M PRO-VD LGA1151
MSI B360M Gaming Plus DDR4 Turbo M2 LGA1151
Gigabyte Z370M-D3H LGA1151-CL

I have added an image of what it looks like. The session still appears to run OK. The bit pattern does not clear once it is there and it appears to be over whatever would normally be there.

[antikythera]

What video cards are you using on all of these systems ?

They are the onboard videos.
The failing mother boards are:
MSI H310M PRO-VD LGA1151
MSI B360M Gaming Plus DDR4 Turbo M2 LGA1151
Gigabyte Z370M-D3H LGA1151-CL

I have added an image of what it looks like. The session still appears to run OK. The bit pattern does not clear once it is there and it appears to be over whatever would normally be there.

so Intel i915 graphics integrated into the CPU rather than the board. anyhow, the behaviour seems to match this bug and it's not hardware failure. try editing the grub entry linux command line and adding the arguments in the comments and see if they help at all

https://bugs.freedesktop.org/show_bug.cgi?id=108085

[luigicantoni]

Thanks for the hint. I agree looks like the same thing.
I inserted
i915.enable_fbc=0
just before my rhgb point in the grub bootup line using the inline editor at bootup (config not changed).
Problem still there.
I tried i915.enable_dc=0 and i915.disable_power_well=0 just in case but problem still there.

I agree looks like a related problem.

I am happy to try any other suggestions or debugs. I did not read all the pointers that 108085 points to just the main thread so if I missed something let me know and I'll try that too.

[antikythera]

not sure, I can't recreate the issue with my Intel HD 4000 GPU (an earlier generation of hardware that uses i915 as well) so have no idea what else you could try. I found that thread from searching for the behaviour on the internet

[luigicantoni]

The other group 108085 report:
Hello everyone, just to share that I have the same issue using libreelec with a NUC7PJYH.
It seems everybody using a GLK have the issue:
https://forum.libreelec.tv/thread/12...uks/?pageNo=10

That problem does look similar. Seems like more is happening on 108085 so I'll follow that one until there is a resolution and report that here.

[antikythera]

okay thanks for the update

[luigicantoni]

I think the cause is X11 related.

The problem varies depending on what order you do things and also a bit on what is on the graphical screen as to if you see it or not. The good bit is I appear now to have a method to make it fail etc (and sort of work).

If I have ctrl-alt-F1 with the graphical login screen
ctrl-alt-F2 is a logged in session
ctrl-alt-F3 is the character based login. This is the screen display with the problem visible on it.

Swap from F1<->F3 100% looks OK
Swap F1<->F2 100% OK (always was and never had an issue)
Swap F2<->F3 and my problem is there.
Do F2->F3 (problem there) do F3->F1->F3 problem not there
Or do F2->F1->F3 problem not there.
It fixes up the display if I go through the F1 (graphical login) screen.

I just re-tested my machines with my X11 conf file there and not there and the behavour is as described above. The file makes no difference.

I had not noticed that behavour before as I normally go from F2-F3 and back again without going through F1.

Also something I have noticed is that when I swap out of F2 for a moment the screen goes silly in the top left with the graphical data and F3 does not clear it but F1 seams to clear it out.

My guess is the swap screen code is writing the data out (maybe only as temporary buffer) and F1 (as an initialization code) clear that temporary area out) and that is what is happening and why F1 fixes it.

Might have been helpful to know that.

[antikythera]

so it's switching tty that triggers it for you? also if using a standard Fedora 29 Workstation installation you are using Wayland (the default session type) rather than X11 unless you specifically log into Xorg instead from the login screen.

you could try using Xorg instead of Wayland and see if the problem still occurs but my hunch is that it will.

[luigicantoni]

sorry to ask the stupidest of questions but what is my simplest way to swap from Wayland to Xorg (yes standard workstation install)
I am about to go home and these systems are at work (mine is laptop with no issues).
By the way the failures are sometimes happening and sometimes not. I cannot find a pattern yet.

[antikythera]

on the login screen there should be a cog next to your username. click on that and a drop down menu will appear, Xorg or X11 should be listed.

[luigicantoni]

OK i also found out how to do that so I have changed to Xorg in gdm/custom.conf (which it turns out I had already made changes).
I did a couple of quick tests with
[daemon]
WaylandEnable=true
DefaultSession=gnome-xorg.desktop
I'll check them out more on Monday but my quick tests gave no improvement.
I'll check out your method although I don't remember seeing that before in the cog menu.

Thanks for your help so far I'll report back Monday then.

[bobx001]

I have another small mini-pc which gives me problems with the i915 , however they are different issues (cannot reboot, always goes into CPU Lockup ....).
Looks like the i915 driver needs urgent attention.

I would try using "lxdm" as a display manager (forcing you to go Xorg), and say xfce, or lxde as a desktop, and see if that improves your video memory handling.
All these "new" interfaces (Wayland) are moving into "composition" (fire and forgetting commands to the video cards instead of handling it on the driver), rather than "rasterization" (they're trying to reinvent the wheel LOL), and I guess this maybe some part of the issue.

[luigicantoni]

Hi All,
I was not actually testing the Xorg version I now see.
Using the login "gnome on Xorg" option I can see by the different display that I am using it and it makes no difference to the problem.
The good thing I am fairly sure this is just a display issue and other memory is not getting damaged so should not cause anything really bad or strange to happen.