Do kernel updates correlate with wayland functioning (window behavior)?

Issues & Assistance
1

I often see some window oddness like increased or decreased state which makes windows of programs look not like they intended to look. Scalability issues, shortly. Also Issues with fullscreen mode.

Why or how it happens?

Is it really because Wayland too much rely on kernel comparing with X11?

inxi -F output:

System:
  Host: pc Kernel: 6.17.9-1-cachyos-bore arch: x86_64 bits: 64
  Desktop: GNOME v: 49.2 Distro: CachyOS
Machine:
  Type: Laptop System: LENOVO product: 83LT v: Legion R7000P ADR10
    serial: <superuser required>
  Mobo: LENOVO model: LNVNB161216 v: SDK0T76479 WIN
    serial: <superuser required> Firmware: UEFI vendor: LENOVO v: RLCN20WW
    date: 04/14/2025
Battery:
  ID-1: BAT0 charge: 64.5 Wh (79%) condition: 81.7/80 Wh (102.1%)
CPU:
  Info: 16-core model: AMD Ryzen 9 8945HX with Radeon Graphics bits: 64
    type: MT MCP cache: L2: 16 MiB
  Speed (MHz): avg: 3195 min/max: 428/5463 cores: 1: 3195 2: 3195 3: 3195
    4: 3195 5: 3195 6: 3195 7: 3195 8: 3195 9: 3195 10: 3195 11: 3195 12: 3195
    13: 3195 14: 3195 15: 3195 16: 3195 17: 3195 18: 3195 19: 3195 20: 3195
    21: 3195 22: 3195 23: 3195 24: 3195 25: 3195 26: 3195 27: 3195 28: 3195
    29: 3195 30: 3195 31: 3195 32: 3195
Graphics:
  Device-1: NVIDIA GB206M [GeForce RTX 5070 Max-Q / Mobile] driver: nvidia
    v: 580.105.08
  Device-2: Syntek Integrated Camera driver: uvcvideo type: USB
  Display: wayland server: X.Org v: 24.1.9 with: Xwayland v: 24.1.9
    compositor: gnome-shell driver: X: loaded: amdgpu,nvidia
    unloaded: modesetting dri: radeonsi
    gpu: nv_platform,nvidia,nvidia-nvswitch resolution: 2560x1600~240Hz
  API: EGL v: 1.5 drivers: nvidia,swrast
    platforms: gbm,wayland,x11,surfaceless,device
  API: OpenGL v: 4.6.0 compat-v: 4.5 vendor: nvidia mesa v: 580.105.08
    renderer: NVIDIA GeForce RTX 5070 Laptop GPU/PCIe/SSE2
  API: Vulkan v: 1.4.328 drivers: nvidia surfaces: N/A
  Info: Tools: api: eglinfo, glxinfo, vulkaninfo
    gpu: nvidia-settings,nvidia-smi x11: xdpyinfo, xprop, xrandr
Audio:
  Device-1: NVIDIA GB206 High Definition Audio driver: snd_hda_intel
  Device-2: Advanced Micro Devices [AMD] Audio Coprocessor
    driver: snd_rpl_pci_acp6x
  Device-3: Advanced Micro Devices [AMD] Family 17h/19h/1ah HD Audio
    driver: snd_hda_intel
  API: ALSA v: k6.17.9-1-cachyos-bore status: kernel-api
  Server-1: PipeWire v: 1.4.9 status: active
Network:
  Device-1: MEDIATEK driver: mt7925e
  IF: wlan0 state: down mac: 62:29:cc:db:87:16
  Device-2: Realtek RTL8111/8168/8211/8411 PCI Express Gigabit Ethernet
    driver: r8169
  IF: enp5s0 state: up speed: 1000 Mbps duplex: full mac: c8:53:09:bb:f6:87
  IF-ID-1: virbr0 state: down mac: 52:54:00:e6:d5:5c
Bluetooth:
  Device-1: Foxconn / Hon Hai Wireless_Device driver: N/A type: USB
Drives:
  Local Storage: total: 2.75 TiB used: 838.8 GiB (29.8%)
  ID-1: /dev/nvme0n1 vendor: Kingston model: SNV2S2000G size: 1.82 TiB
  ID-2: /dev/nvme1n1 vendor: Lenovo model: UMIS RPJYJ1T24MML1AWY
    size: 953.87 GiB
Partition:
  ID-1: / size: 933.58 GiB used: 836.17 GiB (89.6%) fs: btrfs dev: /dev/dm-0
  ID-2: /boot size: 20.26 GiB used: 2.63 GiB (13.0%) fs: vfat
    dev: /dev/nvme1n1p2
  ID-3: /home size: 933.58 GiB used: 836.17 GiB (89.6%) fs: btrfs
    dev: /dev/dm-0
  ID-4: /var/log size: 933.58 GiB used: 836.17 GiB (89.6%) fs: btrfs
    dev: /dev/dm-0
  ID-5: /var/tmp size: 933.58 GiB used: 836.17 GiB (89.6%) fs: btrfs
    dev: /dev/dm-0
Swap:
  ID-1: swap-1 type: zram size: 62.58 GiB used: 0 KiB (0.0%) dev: /dev/zram0
Sensors:
  System Temperatures: cpu: 65.8 C mobo: 48.5 C
  Fan Speeds (rpm): N/A
Info:
  Memory: total: 64 GiB note: est. available: 62.58 GiB used: 5.38 GiB (8.6%)
  Processes: 582 Uptime: 29m Shell: fish inxi: 3.3.40

What I may say that window issues noticeable with just looking at steam and games, and other programs in some places too. Maybe next time I get it, I share pics of odd looking windows.

Such issues can occure with booting another kernel or after update, so I revert to last btrfs snapshot in such case.

For now I use “Gnome on Xorg” (pkgver 49.2). Thanks to this: Reddit - The heart of the internet

2

That shouldn’t be a thing. I have never seen anything you describe and i run cachy on multiple devices, desktops and laptops. None with nvidia tho and seeing the still ongoing issues i would not be surprised if that is the cause, but really can’t comment on that.

Are you using desktop scaling?

3

no. only 100%

4

Hmm well in that case i can only say: fuck nvidia! :stuck_out_tongue:
But really it’s a kind of cop-out as i really can’t tell since i have no access to nvidia hardware.

Only thing i can really tell you that i haven’t seen such things, and i’m pretty anal about stuff program windows and usually notice it instantly and eventually get annoyed enough to look into it. Also game a lot with no fullscreen issues. The way you describe it it sounds like it happens to a lot of programs which makes it pretty weird, is that really the case or is it just limited to specific programs? Electron apps maybe for example?

Well gotta admit i don’t have anything to go on, definitely not normal. And tbh even just saying nvidia isn’t fair as i never seen any posts about your issues linked to nvidia.
If anything i’m thinking in the direction of your screen not always being correctly detected/read, which would be strange since it’s a laptop but then again when it comes to brands following hw specifications it’s also no suprise if it would the issue, might want to look into that direction and stuff like EDID, but that’s probably going to be a bitch.

5

I can tell you several things.

First: CachyOS dropped support for NVidia proprietary drivers when they failed to build with DKMS. In particular the 580 driver. If you got it installed via the repo it isn’t properly installed and will have oddities. Machines I had that needed the 580 driver I, sadly, had to install Ubuntu 24.04 on.

Second: Wayland is a looooooong way from “ready.” Switch to an X11 based desktop like Mate and let others endure the river of bugs and needless frustration that is Wayland. The X11 emulator support built into Wayland is a sad sad feeling. Most application developers aren’t updating for Wayland, just waiting for it to be recorded as a catastrophic failure like Ubuntu’s Phone, UpStart, Unity desktop, and Snap.

Third: The 580 driver under Ubuntu 24.04 has issues with mismatched monitors. Same resolution, just different sizes. Dragging window from 24 inch monitor to 22 or smaller can cause issues.

6

If I had to guess - display resolution detection issues? Nvidia apparently got some in the 580 driver series. Maybe check in display settings if everything seems right. Failing that, sudo cachyos-bugreport.sh could be helpful.

Interesting, doesn’t seem to be that way from looking at packages; linux-cachyos-nvidia has the same driver version as its open counterpart. Probably outdated?

Your experience may vary, of course, but this is plain wrong: a lot of applications simply don’t need to be updated due using a toolkit like GTK/Qt/SDL/Electron and not caring what the underlying windowing system is, the rest (Firefox/Chrome/Wine/Proton) are very much updating or already there. I even have a case of a program not supporting X11, although that’s admittedly more of an exception than a rule.

7

I prefer Gnome. I specifically switched from KDE, because KDE has MS Windows paradigm, and as other reasons because of which I don’t want to spend time on MS Windows paradigm desktops (imitating full functionality but not so completely being). I prefer something like MacOS.

8

It occurs with amdgpu enabled too (UMA in UEFI disabled, now I have only discrete graphics, I don’t have crashes to login screen with also crash of all my programs as before, so nvdia doesn’t lead to crash of all my programs and exiting to login screen).

I don’t say anything based on nvidia yet, don’t be just a guesser without a base for that in my case, we haven’t such. Just please require more info if you need.

What I may say that window issues noticeable with just looking at steam and games, and other programs in some places too. Maybe next time I get it, I share pics of odd looking windows.

Such issues can occure with booting another kernel or after update, so I revert to last btrfs snapshot in such case.

9

I have package linux-cachyos-bore-nvidia-open

So my driver is “open”

10

I’ve been a software developer for over 40 years. Most applications are not updating for Wayland.

Qt is buggy as all get-out.

Electron has a supported life matching that of fruit flies.

GTK defaults to the X11 emulation under Wayland. Your application has to force Wayland at runtime

To force a GTK application to use the Wayland backend, set the GDK_BACKEND environment variable to wayland before launching the application

Unless you’ve ported to SDL3 your SDL application isn’t “just using” Wayland.

The vast majority of applications are currently using the X11 emulation. It will be this way for at least the next decade. That is making the massive assumption Wayland isn’t taken out to the woods and given two behind the ear like it deserves.

Having spent many months under the hood of CopperSpice creating LsCs before abandoning the Qt base completely, I can tell you they can’t easily rip out the default of XKB without breaking everything. That’s why I started BasisDoctrina. None of the legacy frameworks can successfully make the jump. SDL3 broke from SDL2 so they made the jump.

Right now the desktops that DO NOT use Wayland are growing in popularity.

X11 may eventually be replaced, but it won’t be Wayland. That experiment shouldn’t have gotten out of the lab.

11

That’s cool and all, but makes confidently stated mistakes both in your reply and linked project’s documentation that much more puzzling; the part about inherited core affinity in spawned threads by default is a quirk of some platform/toolkit combination at best (and that’s just one example).

Doesn’t exactly match my experience in recent times as a user, neither under Wayland nor X11, but fair, may be a setup difference. Plasma 5 was admittedly unusable with Wayland and Qt API can be quite a legacy mess.

I’m not sure what that has to do with Wayland support? I have my gripes with Electron, but I really don’t see how this is relevant here - aside from meaning that people will get windowing-related updates faster. The important part is that it’s pretty common to use Electron-based apps.

This one is just verifiably false, both for GTK3 and GTK4. No, seriously, check for yourself if you don’t trust me. Say what you want about Gnome, but they aren’t wild enough to add an option to build Mutter without XWayland when the end result would not even run their own apps.

Bonus: under Flatpak+Wayland applications that support Wayland don’t have access to the X11 socket. As you might expect, this includes GTK3/4 apps, meaning these wouldn’t be able to use XWayland in such situation even if they wanted.

SDL2 already had a Wayland backend - even if it was not used by default, it worked quite well towards the SDL3 release. Now sdl2-compat implements SDL2 API on top of SDL3 and seems to default to Wayland if possible.

Qt made the jump - however inelegantly, GTK made the jump, Electron is well usable there, SDL did not need an API break just for Wayland support (you could debate if redesign of HiDPI handling in SDL3 counts, but that’s hardly Wayland-specific) - and these (+Wine/Proton) cover a whole lot of applications.

Simply untrue for an out-of-box on a Cachy install of Plasma/Gnome, that’s for sure. I personally also have only a few applications that use XWayland (notably Steam, Audacity, rare native game that touches X11 functions directly).

I’m really curious if you have a citation for this. Seems especially odd when you can still run an X11 session of your usual desktop of choice (and MATE, XFCE, Cinnamon, all are working towards Wayland).

12

No, that’s fact. When you create a thread on Linux said thread has affinity for the parent core.

affinity-001
affinity-001956Ă—279 31.2 KB

They will get shiny new bugs faster.

Having just spent a whole lot of time with Scintilla GTK and having exchanged multiple messages on the same mailing list for Geany with the developer, my statement is correct. I have not used GTK4, but GTK, like Qt and CopperSpice has a legacy platform issue.

XKB cannot be fallen back to. They test for it first. If found, that’s what is used. Where Geany has been having issues they tell people to set the environment variable forcing Wayland.

You need to check via Flatseal.

Let us not forget MacOS is really FreeBSD under the hood. Hacked for sure, but FreeBsd, the Unix nobody used . . . until now.

Last but not least, both Canonical and Google are pouring resources into Fuchsia. Basically because nobody likes Wayland. Adding insult to injury Wayland really fails with the Vulkan stack

The Vulkan stack was a much better all around idea than Wayland.

We shall see. I know my friend doing research dev for Meta says they are ready to kick NXP and all things Wayland to the curb. Just non-stop failures due to a needless layer.

At any rate, it’s time for supper. Tomorrow I actually have to get some work done.

13

I will only say: fuck Wayland!:wink:

14

Same thing just about everybody on the qt-interest mailing list is saying while complaining about the abysmal Wayland support in Qt (which KDE is written with and a primary reason many distros have dropped or are dropping the KDE desktop.)

kde-quote
kde-quote708Ă—357 18.3 KB

With RedHat dropping KDE from the commercially viable product line one has to wonder how much longer the yet-another-red-hat distros will continue to endure the pain.

Check out all of the liberated screenshots in this thread. X11Libre

They even have it packaged for CachyOS as well as all other Arch based distros.

Qtc keeps pointing to this discussion.

You can put lipstick on a pig, but it can’t be your prom date.

Wayland is a failed experiment.

15

I would like to see your criticism of Wayland structure compared to X11 and why Wayland enthusiasts like to mention that X11 deeply became a legacy software and Wayland is the only one way to go further, something like that.

No cliche please.

16

https://www.reddit.com/r/wayland/comments/1osv85i/wayland_protocol_development_is_it_really_as/

Rather than cut & paste or retype all of that…You are looking for the answer that starts

reddit-001
reddit-001765Ă—259 14.1 KB

17

Insert “why not both” meme :stuck_out_tongue:

18

This is not funny, stupid

19

Not really?

Wayland cares a lot more about the window-manager/compositor and relies on it .. the kernel, especially if we are not counting long deprecated ones, does not really matter so much.


Oh my .. look .. a thread does not require real assistance so its not really paid much mind … and all of a sudden its open season for xlibre shilling.

We still didnt get anything besides platitudes and outright fabrications.

Even an eager would-be-believer asks for real information as to why we think X is some sort of golden martyr .. to a response that is just a link to a reddit thread .. where its the same individual still just saying things and not giving any supporting evidence. Not even making realistic claims like “its slower because the API does such and such”. Just “wayland bad, heres a bunch of other stuff I believe but that is not really related like resumes and nvidia, and by the way I have an X bridge to sell you.”

:roll_eyes:

20

Closed.

The Wayland, X11 debate is a broken record. Even worse with Xlibre in the mix here.

Move on.