# 560 release feedback & discussion

**URL:** <https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830>\
**Category:** Linux\
**Created:** [July 23, 2024, 2:23pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830 "2024-07-23T14:23:27Z")\
**Posts on this page:** 20\
**Page:** 23

<div class="post-metadata">

**Author:** ![edu4rdshl](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/edu4rdshl/32/364158_2.png) [@edu4rdshl](https://forums.developer.nvidia.com/u/edu4rdshl)\
**Post date:** [September 23, 2024, 6:09pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/458 "2024-09-23T18:09:50Z")

</div>

The issue is not limited to wayland/xwayland, it also affects X11, so I’m pretty sure it’s a problem with the driver and how they manage the memory. The issue may have been existed before, but it’s ridiculous right now, you can make a single app use up to ~2 GB VRAM if you keep resizing it, and the worst part is that the memory doesn’t get free-ed.

---

<div class="post-metadata">

**Author:** ![shelter](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/shelter/32/405595_2.png) [@shelter](https://forums.developer.nvidia.com/u/shelter)\
**Post date:** [September 23, 2024, 6:14pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/459 "2024-09-23T18:14:05Z")

</div>

It’s the same with 550 dev in X11 so it isn’t new. It just became more apparent with Wayland

Edit:  
I had to change back to 550 dev and X11 because of the current issues with 555/560

---

<div class="post-metadata">

**Author:** ![omarhanykasban706](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@omarhanykasban706](https://forums.developer.nvidia.com/u/omarhanykasban706)\
**Post date:** [September 23, 2024, 8:58pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/460 "2024-09-23T20:58:48Z")

</div>

what did you do nvidia  
dvi port is broken on `550.120`  
it was working on `550.107.02`

---

<div class="post-metadata">

**Author:** ![PcChip](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/pcchip/32/364244_2.png) [@PcChip](https://forums.developer.nvidia.com/u/PcChip)\
**Post date:** [September 23, 2024, 11:20pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/461 "2024-09-23T23:20:05Z")

</div>

I wanted to submit feedback about the 560 drivers, on CachyOS (arch) with KDE Plasma

When using the Open version, there are random microstutters when dragging windows around and scrolling, and in games there are tons of microstutters (and normal stutters too!) - I tried with multiple different versions of Proton / Proton-GE / Cachy-Proton just to make sure it wasn’t that.

When I switched back to the Closed version and disabled GSP firmware, all these issues went away. Dragging and scrolling on the desktop is smooth as glass. Games are like butter.

(The closed version does have the annoying downside that sometimes windows flicker through the foreground window, so for example I can see background windows through a fullscreen youtube video, but that seems related to DSC at 240hz on my LG OLED. When I drop it to144hz that goes away. That happens on every driver version from 535-545-550-555-560 on both X11 and Wayland, both kubuntu and arch, so that problem is unrelated to this 560 complaint I’m making)

I am writing this because I understand that in the future, you will force the open version with GSP Firmware enabled. I would ask that you please iron out these issues I’ve described before forcing that on everyone!

---

<div class="post-metadata">

**Author:** ![phoenix91140](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@phoenix91140](https://forums.developer.nvidia.com/u/phoenix91140)\
**Post date:** [September 24, 2024, 5:00am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/462 "2024-09-24T05:00:48Z")

</div>

Fully operational here on OpenSUSE Tumbleweed with kernel 6.11

---

<div class="post-metadata">

**Author:** ![kuba10j](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@kuba10j](https://forums.developer.nvidia.com/u/kuba10j)\
**Post date:** [September 24, 2024, 8:09am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/464 "2024-09-24T08:09:03Z")

</div>

But nvidia open modules mentioned here are not “nouveau” and are mostly made by NVIDIA themselves. AFAIK Moreover they mostly come from the same codebase as proprietary drivers.

Proprietary drivers are not going away any time soon. For more info see: [Clarification on GPU support for Maxwell/Pascal archs and binary/OS relationship · Issue #19 · NVIDIA/open-gpu-kernel-modules · GitHub](https://github.com/NVIDIA/open-gpu-kernel-modules/issues/19#issuecomment-1124360887)

---

<div class="post-metadata">

**Author:** ![edu4rdshl](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/edu4rdshl/32/364158_2.png) [@edu4rdshl](https://forums.developer.nvidia.com/u/edu4rdshl)\
**Post date:** [September 24, 2024, 9:24pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/465 "2024-09-24T21:24:26Z")

</div>

If you would at least understand the minimum of what you’re talking, you will not be saying that nonsense. Nvidia had planned that for a long time ago, GSP is there to be used, Nvidia did that on purpose. The drivers are still developed by Nvidia, please keep your bs at home.

---

<div class="post-metadata">

**Author:** ![kelvie](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@kelvie](https://forums.developer.nvidia.com/u/kelvie)\
**Post date:** [September 25, 2024, 6:23am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/466 "2024-09-25T06:23:26Z")

</div>

Anyone know what the last version of the nvidia driver and egl-wayland doesn’t have the vram leak on wayland compositors (NOT including xwayland, I think that’s been a problem forever), but still has explicit sync?

---

<div class="post-metadata">

**Author:** ![shelter](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/shelter/32/405595_2.png) [@shelter](https://forums.developer.nvidia.com/u/shelter)\
**Post date:** [September 25, 2024, 7:04am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/467 "2024-09-25T07:04:54Z")

</div>

I don’t think it’s very new, it’s just that it became more obvious when more people started using wayland. I could be wrong tho’.

---

<div class="post-metadata">

**Author:** ![edu4rdshl](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/edu4rdshl/32/364158_2.png) [@edu4rdshl](https://forums.developer.nvidia.com/u/edu4rdshl)\
**Post date:** [September 25, 2024, 8:05am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/468 "2024-09-25T08:05:58Z")

</div>

Doesn’t exist, but iirc 555 does a “better job” on that.

---

<div class="post-metadata">

**Author:** ![mark.c.dixon](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@mark.c.dixon](https://forums.developer.nvidia.com/u/mark.c.dixon)\
**Post date:** [September 25, 2024, 2:41pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/469 "2024-09-25T14:41:46Z")

</div>

These symlinks have vanished in 560 (el9 aarch64), it would be nice if they came back please:

/usr/lib64/libnvidia-cfg.so  
/usr/lib64/libnvidia-fbc.so  
/usr/lib64/libnvidia-ml.so  
/usr/lib64/libnvidia-nvvm.so  
/usr/lib64/libnvidia-ptxjitcompiler.so

Used to be in nvidia-driver-devel in 555 and earlier.

Missing /usr/lib64/libnvidia-ml.so means the Slurm batch scheduler doesn’t detect the gpus in the system.

---

<div class="post-metadata">

**Author:** ![opisalwaysgreat](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@opisalwaysgreat](https://forums.developer.nvidia.com/u/opisalwaysgreat)\
**Post date:** [September 25, 2024, 7:14pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/470 "2024-09-25T19:14:01Z")

</div>

Interesting, I’m also experiencing the `NVRM: nvAssertFailedNoLog: Assertion failed: pEventNotificationList->pendingEventNotifyCount == 0 @ event_notification.c:289` spam when running a game, in my case Cyberpunk. Are we the only two?  
However, in my case it makes Cyberpunk freeze a couple of seconds after its launch. Will be posting a log tomorrow.  
Running 560.35.03, on a RTX 4070 Laptop.

---

<div class="post-metadata">

**Author:** ![jNines](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/jnines/32/376188_2.png) [@jNines](https://forums.developer.nvidia.com/u/jNines)\
**Post date:** [September 25, 2024, 7:28pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/471 "2024-09-25T19:28:41Z")

</div>

@langdalepl also has it. I’m guessing it’s something specific to certain engines, as other haven’t. Both of you in Cyberpunk, myself in HFW.

Don’t have the issue when using the closed module.

---

<div class="post-metadata">

**Author:** ![noahblaker333](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@noahblaker333](https://forums.developer.nvidia.com/u/noahblaker333)\
**Post date:** [September 25, 2024, 9:51pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/472 "2024-09-25T21:51:06Z")

</div>

I hate to harp on this more but I’m again asking for some acknowledgement of the HDR freeze issue in Plasma. There is an open pull for Plasma that disables HDR on Nvidia cards specifically because of this issue.

> **[backends/drm: block the colorspace capability by default on NVidia (!6497) ·...](https://invent.kde.org/plasma/kwin/-/merge_requests/6497)**
>
> There have been many reports that enabling HDR on NVidia can lead to KWin's atomic commits being rejected on login, which means KWin can't display anything. To prevent...

---

<div class="post-metadata">

**Author:** ![kelvie](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@kelvie](https://forums.developer.nvidia.com/u/kelvie)\
**Post date:** [September 26, 2024, 4:46am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/473 "2024-09-26T04:46:49Z")

</div>

Started another thread here for the VRAM / wayland issue: [Multiple wayland compositors not freeing VRAM after resizing windows - #2 by kelvie](https://forums.developer.nvidia.com/t/multiple-wayland-compositors-not-freeing-vram-after-resizing-windows/307939/2)

It appears to affect more than just the 560 driver after more testing

---

<div class="post-metadata">

**Author:** ![thesword53](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@thesword53](https://forums.developer.nvidia.com/u/thesword53)\
**Post date:** [September 26, 2024, 8:37pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/474 "2024-09-26T20:37:20Z")

</div>

Am I the only person who doesn’t have any freeze issue with `kwin` 6.1 and HDR enabled ?  
However I experience freezes with HDR on Windows: [Black screen and system freezes when I enable DSC, 10 bits HDR and VRR - #4 by MarkusHoHo](https://forums.developer.nvidia.com/t/black-screen-and-system-freezes-when-i-enable-dsc-10-bits-hdr-and-vrr/307921/4)

---

<div class="post-metadata">

**Author:** ![thesourcehim](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@thesourcehim](https://forums.developer.nvidia.com/u/thesourcehim)\
**Post date:** [September 27, 2024, 6:27am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/475 "2024-09-27T06:27:03Z")

</div>

GravityMark OpenGL test can not be started:

```auto
E: 59.73 ms: GL::open(): glXChooseFBConfig(): failed
E: 61.19 ms: GLWindow::create_visual(): can't open OpenGL
E: 61.33 ms: GLWindow::create_context(): OpenGL is not opened
E: 61.35 ms: Window::create(): can't create context
E: 61.36 ms: GravityMark::create(): can't create Window for OpenGL platform

```

Runs fine on Intel iGPU.

---

<div class="post-metadata">

**Author:** ![asubstanced](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@asubstanced](https://forums.developer.nvidia.com/u/asubstanced)\
**Post date:** [September 27, 2024, 7:42am UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/476 "2024-09-27T07:42:44Z")

</div>

No freezes here. Same DE. Just limited colour space issue reported previously.

---

<div class="post-metadata">

**Author:** ![JamesJones](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/jamesjones/32/10056_2.png) [@JamesJones](https://forums.developer.nvidia.com/u/JamesJones)\
**Post date:** [September 27, 2024, 7:56pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/477 "2024-09-27T19:56:31Z")

</div>

Please see the latest update here for a potential workaround:

[Workaround for Excessive VRAM usage](https://github.com/NVIDIA/egl-wayland/issues/126#issuecomment-2379945259)

---

<div class="post-metadata">

**Author:** ![BlueGoliath](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@BlueGoliath](https://forums.developer.nvidia.com/u/BlueGoliath)\
**Post date:** [September 27, 2024, 11:03pm UTC](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830/479 "2024-09-27T23:03:09Z")

</div>

I already posted a screenshot showing XWayland is where the memory leak is, atleast on Gnome. Some code is clearly reserving 10% of VRAM.

[Previous page](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830.md?page=22)

[Next page](https://forums.developer.nvidia.com/t/560-release-feedback-discussion/300830.md?page=24)
