Nvidia Display 27.21.14
Aline Braunbeck
Hi, i have an issue with HDMI, when I install Ubuntu 18.04. it does not detect HDMI, after I update additional driver. I select 460.
currently, i only display HDMI and not detect the default laptop monitor
nvidia-bug-report.log.gz (1.1 MB)
I had the same issue (2nd monitor not detected after updates) on my linux (Ububtu 20.04) machine with nvidia installed and tried almost every suggestions listed above.
After a few waste of hours, I found a solution to my case.
Looking at your bug report I can see that you have a Notebook with an RTX 2060. Does Ubuntu come pre-installed with the Laptop? If not please check with the vendor if it supports Ubuntu and multiple display configurations.
Because from the log itself everything looks ok, the driver registers the different (virtual) DP and HDMI outputs correctly, but it seems only the internal (eDP) connection to the Laptop screen is used.
Acer ConceptD Driver: 530.30.02 CUDA Version 12.1, NVIDIA GeForce RTX 3080 running Xorg. I step away from my desk for more than 15 minutes and I come back to a frozen system. Getting extremely frustrated.
I am still on 515.105.01 and can confirm this bug exists on this version, but only when i play on Steam.
So i was going to upgrade to 530 as @amrits said that this bug will be fixed, but it seems like it isnt yet, since someone after that post reported they are still experiencing these issues.
Would also love any news/to help fix this issue if I can. I seem to be having the exact symptoms of richard.decal, in that my Lenovo P15 Gen 2 running Ubuntu 20.04.6 has recently started experiencing external monitors connected via lenovo dock freezing. I was running 535, downgraded to 525 and still see the nvidia-modeset: ERROR: GPU:0: Idling display engine timed out error. It seems to have lessened/stopped the freezing, but there are other performance bugs present in this version that I would like to avoid.
On suspend/resume, the brightness should be maintained by the kernel, i.e. the amdgpu driver so I suspect a bug there. Maybe check if this contains some help:
-/issues/1438
For a workaround, you could copy and modify the systemd-backlight units to save/restore brightness on suspend/resume.
This saves me some effort. Was going to upgrade. I have been thinking that everyone was happy here. I have 3090 and tried multiple distros and unable to solve this. Just end up powering the machine off at night. Would like to get ths resolved. Just get nvidia wait one resume. using Opensuse Tumbleweed here currently and 525.89.02
A: The Display Mode Selector tool should only be used if the default GPU configuration is insufficient for the desired use case. In addition, consult your system OEM to determine if the desired mode is supported with your hardware.
A: The Display Mode Selector tool is a special tool for NVIDIA L40S, NVIDIA L40, NVIDIA RTX 6000 Ada, NVIDIA A40, NVIDIA RTX A5000, NVIDIA RTX A5500, and NVIDIA RTX A6000 only. It should not be used with any other GPU.
A: A system (workstation or server) needs to be specifically qualified for any non-default mode. It is critical to check with the system OEM to confirm that the desired display mode is supported. Misuse of the tool can permanently damage the GPU without the option to return to the default configuration.
I have an HP Envy 17.3 inch laptop with Intel(R) Iris(R) Xe Graphics and NVIDIA GeForce RTX 3050 4GB Laptop GPU. I cannot get the Laptop display, or any associated outside monitor, to use the Nvidia display card. I am running Windows 11 Pro, All drivers are up-to-date, both from Windows and from HP.
From the Nvidia Control Panel, Manage 3d Settings tab, I have selected the High Performance Nvidia processor as my preferred graphics processor. I have also selected the GeForce RTX 3050 processor for PhysX. However, the laptop will not use the Nvidia processor for the laptop screen or my external monitor.
Within Windows Display options, at either the "Advanced Display" selection or the "Graphics" selection, I see no option for me to select a preferred Vermoegenet.de option. There does not seem to be any options to select preferences for the displays within the Windows Display options.
RElatedly, I also went to the Device Manager, and disabled the Intel Iris GPU, thinking that the monitors would then default to the Nvidia GPU. Nope Instead, the external monitor reported receiving no signal, and the laptop display just used the basic onboard graphics.
I have seen many posts out there about people having problems with Nvidia drivers, so it seems I'm far from alone. Besides, there have already been a lot of useful answers/suggestions provided, so I don't want to add yet another question on the subject itself (I will take the time to read and try out the different suggestions once I solve my present problem).
I'm posting an answer to my question as I managed to solve the problem (thanks to the AskUbuntu community). This may not be a completely generic answer, but I will explain the steps which I followed in case it helps someone else. Here's how I did it:
Here I followed the instructions provided by Marmayogi in his comment to my question, except for the installation command of the Nvidia driver which was incorrect as it has changed (see the answer by Kerry Kobashi in this post for details):
After rebooting the GUI was in a normal resolution (except GRUB which was still in low resolution, but it's less of a problem). To check that everything went well, I issued the following command: lsmod grep nvidia, and there was output, so the installation was successful.
Here I am want to stress out that you need to check which version of the Nvidia driver you need (i.e. which is compatible with your graphics card). You can check on the Nvidia website which products are supported by each driver (the latest one is usually the best if you have a recent graphics card). Not doing so and blindly following instructions in another post which did not specify this is what led me to having all these problems in the first place (unfortunately, this point is too often ignored in a lot of posts I've seen on the subject). Also when choosing the version, I read somewhere that you only need to write the main version (don't bother about the number after the point, so if latest driver is 418.56, just write 418).
Here is a detailed procedure to blacklist Nouveau followed by the installation of nvidia driver. You may read this as a continuation of @johnwolf1987 answer since blacklisting Nouveau was not covered over there.
Made changes from the file after pressing E including Add nouveau.modeset=0 at the end on a text starting with Linux. Also replaced quiet splash with nomodesetand reboot using F10 but still same problem.
The above command will autoinstall the drivers suited to your machine and update them. I did this to eliminate the hassle of finding out the appropriate driver version for my system and let Ubuntu handle that.
I had a similar problem. I blacklisted the Nouveu driver, then could not get a display. I used an ssh client from another machine on my network, and ran the installer fr my NVidia driver from there. one reboot later, and I was in!
I've created bug report @bugs.launchpad.net so you can add your info too.
I prefer to install nvidia driver from default Ubuntu app 'Software & Updates'-> Additional Drivers->choose Nvidia->apply. Reboot. You can check your driver version by nvidia-smi command in terminal. For me it's 390.116
I had the same problem and nothing helped me. The only thing i did and it works i described above.
seems the problem is in Kernel v4.18.x, v4.20.x (that's i tested for my hardware on different OS's and Kernels. Everything still work okay on Kernel v4.15.x, so if you want you can use it instead of 5.0.6)
Just want to add some comments based on the above answers, I believe the most important thing in my case is, you have to make sure to install the correct nvidia driver, never blindly follow other answers, they are old posts.
I hope this message finds you well. I recently decided to switch from Fedora 38 to Manjaro due to some challenges I encountered with the Fedora 39 update. However, I am currently facing difficulties in getting the Nvidia graphics driver to work properly on my system.
I used this command for installation:
sudo mhwd -a pci nonfree 0300
There was no change after that. My external monitors were still not working. Then I uninstalled the drivers again and made a fresh installation using this command:
sudo mhwd -i pci video-nvidia
During this installation, external monitors worked, but the main screen of the laptop did not work. Just a black screen comes up.
But, for the last couple of weeks - maybe since EIS 16.2.13 was released - every time I reboot the PC (which is at least once each day), after Windows starts up, EIS asks again for internet access permission for NVDisplay.Container.exe (Nvidia display driver telemetry).
I keep checking the Remember rule option and denying access but it asks again on the next reboot like no rule was created, and firewall makes another identical rule for this executable as it already exists in the EIS firewall rule list (see attached picture). Again, all these rules are completely identical, are concerning the same executable and have the blanket OUT-Deny settings for ANY local and remote ports and hosts.
I was only able to boot into Fedora using the nomodeset parameter in grub (I assume because of the missing Nvidia drivers). I ran dnf update and then installed the latest Nvidia drivers from the Nvidia CUDA repository (Later I also tried the RPM Fusion method with the same results).
After rebooting, I tried starting Fedora using the kernel 6.1.14 without the nomodeset parameter and I was able to get to the decryption screen and enter my password. The loading animation starts but the screen goes black before I can see the login screen.
My suggestion is to remove what was installed from the cuda repo and instead enable the rpmfusion-nonfree-nvidia-driver repo and install from there with sudo dnf install akmod-nvidia xorg-x11-drv-nvidia-cuda which will give you both the nvidia drivers and cuda as well.
c80f0f1006