nvidia – Ubuntu 20.10 update from 20.04 will only boot from recovery using llvmpipe for graphics

Tried to perform an update from 20.04 LTS to 20.10. All was going well until the final reboot “into” 20.10, and my computer fails to get to the desktop when booting normally. It flashes between a purple screen and a black screen and never loads the gnome desktop. It then seems to basically hang.

Naturally I then tried recovery mode from grub, and weirdly, everything boots when I go in via this route, choosing “resume”. However there’s still some graphical anomalies during boot, (dialogue box appears with a basic theme on the UEFI background, and then disappears), and I noticed that in – Settings – About, it’s listing llvmpipe as the graphics, rather than the Intel iGPU when I go in this route. Weirdly, it also offers to (and defaults to) boot using the 5.8.0.26 kernel (which I didn’t manually install) above the listing for 5.8.0.25 (which is what I believe 20.10 ships with.

My box is an HP ex-chromebox (Haswell Celeron 2995U. 1.4GHz) which ubuntu has breathed new life into even if it’s a rather weak machine. Thought I’d get used to Ubuntu on something I already had first even if it is weak! Ubuntu opens up so much more than it could originally do so I’d like to get it back, especially as it was really impressive so far.

I have used a discrete GPU on this box though an rather hacky EGPU connected to the WiFi pci-express port (an Nvidia GTX 960) but I usually use the iGPU unless it needs the extra oomph for something – what I think happened during the update, is that the update disabled the proprietary Nvidia drivers and the Nvidia dGPU/iGPU switching, and this has messed up the X.org server settings meaning it doesn’t know what to do for graphics when it boots. Recovery works by switching to the lowest form of compatible graphics modes. Yes I wish I had not made life hard for myself.

I thought purging the Nvidia drivers would fix it, but now I’ve far enough down this rabbit hole that I’m lost. I started trying sudo X -configure and get an error:

FATAL: Module fbcon not found in directory

I don’t know enough about X.org.conf to have a clue what to edit in there.

I then thought I’d try a repair install from an Ubuntu live-USB, and run into a mount point error. At this point I’m ready to give up, but I’d really rather not have to start over. I figure the other thing worth trying is to reinstall the nvidia drivers and see if I can undo what I’ve done that way, but I thought I should ask before I cause any more damage!

I do thankfully have a backup of my home directory if I have to resort to that, but any better ideas as to how I can reset this mess without basically wiping and starting again?

Seems like Ubuntu 18.04 Doesn’t recognise NVIDIA Drivers

I am a beginner in Ubuntu.

Just last week I have installed Ubuntu18. I have a dual boot machine on new Laptop with Secure boot switched-on (Windows 10 and Ubuntu 18.04.5).

I have NVIDIA GeForce GTX 1650 in my laptop, so i tried to install Nvidia drivers by following the steps mentioned here: Installing GPU drivers for Data Science

After performing the steps exactly the same, you can see that I have installed NVIDIA Driver :
Installed Recommended Driver

I have then rebooted and I can see that nvidia-smi runs perfectly fine (earlier i would get error)
Output of nvidia-smi command

Then I ran sudo prime-select nvidia
Getting output that nvidia is already set

enter image description here

Again, I rebooted
But still the Ubuntu Settings mention that I am using intel drivers.
enter image description here

How do I fix this?

nvidia – Why is Firefox’s window on X11 crashing on every display configuration change?

I am using Firefox on Ubuntu 20.10 (recently updated from 20.04 where this issue was also occurring). Every time the display configuration changes or something goes full screen (I plug in a monitor, the lock screen starts, I put the laptop to sleep), the window that Firefox is in suddenly moves behind all other windows and desktop icons (as if it was part of the desktop), and it loses its contents (everything being displayed in the window gets garbled or disappears). I have to press Alt + F4 and restart Firefox. I am running gnome 3.38.1, but it happened in 3.36 as well.

It might have started after I updated my proprietary nivida drivers, but I’m not sure.

Any ideas? Let me know if I can provide other helpful information.

mining pools – CryptoDredge 0.23.0 – NVIDIA GPU Miner

CryptoDredge

Download

Overview
CryptoDredge is an easy-to-use and highly optimized cryptocurrency mining software. It takes full advantage of modern NVIDIA graphics cards through unique optimization techniques. We have also paid great attention to stable energy consumption. These benefits, along with very low development fees, make our product one of the best public miners available.

Devfee
Development fee is 1% (2% for MTP)

Supported algorithms
Allium
Argon2d (250/4096 / DYN / NIM)
BCD
BitCore
Chukwa (Argon2-512)
Chukwa WRKZ (Argon2-256)
CryptoNightConceal
CryptoNightFastV2 (Masari and Stellite)
CryptoNightGPU
CryptoNightHaven
CryptoNightHeavy
CryptoNightSaber (Bittube)
CryptoNightTurtle
CryptoNightUPX
CryptoNightZLS
Cuckaroo29 (Grin)
CuckooCycle (Aeternity)
HMQ1725
Lyra2REv3
Lyra2vc0ban
Lyra2z
Lyra2zz (LAPO)
MTP (see the MTP Algorithm item)
MTP-TCR
NeoScrypt
Phi2
Pipe
Skunkhash
Tribus
X16R
X16RT
X16RV2
X16S
X17
X21S
X22i

System requirements
NVIDIA GPUs with Compute Capability 5.0 or higher.
MTP requires a graphics card with at least 5 GB of memory.
Latest GeForce driver
2 GB of RAM (4 GB recommended). Some algorithms, such as NeoScrypt, require virtual memory (paging file) the same size as the entire GPU memory.
Internet connection

Windows
Windows 7 / 8.1 / 10 (64-bit)
Visual C ++ Redistributable for Visual Studio 2015

mining pools – dstm’s ZCash / Equihash Nvidia Miner v0.6.2

dstm’s ZCash Cuda miner

I wrote this originally for my own purposes. It works pretty well – here it is.
Compatible with devices with Compute Capability 5.0 or higher, i.e. it should work fine on Maxwell / Pascal GPUs.
Supports every server using the stratum protocol. If you have problems with certain pools, please. report them here.
Contains 2% devfee.
Suggestions are welcome.

Download

Performance:
1080Ti 725-740 Sol / s
1070 440-450 Sol / s
stock settings

UI:
An example of mining on a rig with 8 GPUs.
GPU0-7: 1070Ti

Sol / s: decisions per second
Avg: average number of decisions per second
I / s: the number of iterations per second performed by the GPU
S / W: efficiency – average Sol / s per watt
W: power consumption in watts
last 3 columns:

Version 0.6.2

fixed handshake bugs

  • fixed device selection bug introduced in 0.6.1
  • improve device initialization in large systems

nvidia – Ubuntu 20.04 Second monitor is not working

I tried to connect my laptop with the second monitor, however, it displays on the second monitor that No HDMI signal from your device. When I checked the Displays setting, there is no monitor too.
I use ASUS Gaming. This is the configuration of Ubuntu 20.04
enter image description here

When I checked xrandr in the terminal:
enter image description here

I also updated to the kernel 5.6 but nothing changes
Here is the setting of the software updates, there are no additional drives
enter image description here

I hope anyone can help me because I need to work on a big screen.
Thanks in advance.

nvidia – Cannot connect second monitor via HDMI cable in Ubuntu 20.04

I just installed ubuntu 20.04 on my laptop. However, when I tried to connect my external monitor via HDMI cable the monitor is not detecting any signal.
I am using nvidia-driver-435. My GPU is “NVIDIA Corporation TU106GLM (Quadro RTX 3000 Mobile / Max-Q)”. I am using a HP G6-15 laptop. xrandr is not detecting the display

Screen 0: minimum 320 x 200, current 1920 x 1080, maximum 16384 x 16384
eDP-1 connected primary 1920×1080+0+0 (normal left inverted right x axis y axis) 344mm x 193mm
1920×1080 60.03*+ 60.01 59.97 59.96 59.93 40.02
1680×1050 60.00 59.95 59.88
1600×1024 60.17
1400×1050 59.98 60.00
1600×900 59.99 59.94 59.95 59.82
1280×1024 59.95 60.02
1440×900 59.89
1400×900 59.96 59.88
1280×960 60.00 59.99
1440×810 60.00 59.97
1368×768 59.88 59.85
1360×768 59.80 59.96
1280×800 59.99 59.97 59.81 59.91
1152×864 59.97 60.00
1280×720 60.00 59.99 59.86 59.74
1024×768 60.04 60.00 59.95
960×720 60.00
928×696 60.05
896×672 60.01
1024×576 59.95 59.96 59.90 59.82
960×600 59.93 60.00
960×540 59.96 59.99 59.63 59.82
800×600 60.00 60.32 59.96 56.25
840×525 60.01 59.88
864×486 59.92 59.57
800×512 60.17
700×525 59.98
800×450 59.95 59.82
640×512 60.02
720×450 59.89
700×450 59.96 59.88
640×480 60.00 59.94 59.94
720×405 59.51 58.99
720×400 59.97
684×384 59.88 59.85
680×384 59.80 59.96
640×400 59.88 59.98 59.96
576×432 60.06
640×360 59.86 59.83 59.84 59.32
640×350 59.84
512×384 60.00
512×288 60.00 59.92
480×270 59.63 59.82
400×300 60.32 56.34
432×243 59.92 59.57
320×240 60.05
360×202 59.51 59.13
320×180 59.84 59.32
DP-1 disconnected (normal left inverted right x axis y axis)
DP-2 disconnected (normal left inverted right x axis y axis)
HDMI-1 disconnected (normal left inverted right x axis y axis)

I was earlier using ubuntu 18.04 on the same computer and the second monitor was working. I freshly installed 20.04 after erasing the previous version after deleting everything. I switched to lightdm switched back to gdm3 but nothing worked so far.In some answers I have found to edit

sudo nano /lib/modprobe.d/nvidia-kms.conf
and comment the options nvidia-drm modeset=1.
But it is not present on my computer. I can onlye find the nvidia-graphics-drivers.conf file on /lib/modprobe.d

I don’t know what is wrong. Please let me know if you need more information.

editing – Edit host file – Nvidia shield pro

I am trying to edit my host file in nvidia shield pro device. Its not rooted and I get read only error when I try to update the host file using adb push. I am a newbie in this. I need to do some url redirecting.

I am not trying to block anything. I have a local Apache server set up and want to redirect the traffic from the shield to my local server for some custom urls.

18.04 – Nvidia driver installs on Ubuntu, but fails to load

I am running Ubuntu 18.04.3 with an NVIDIA GeForce GTX 780 Ti graphics card, and I want to install and use the official NVIDIA driver. I have previously been successful with this, and have done many installations in this way for years. But with my current system, after installing the recommended driver, it doesn’t load correctly.

Here is what I have done so far specifically:

  1. First, I tried installing the driver via apt: sudo apt install nvidia-driver-390. But after rebooting, I could tell that NVIDIA driver wasn’t actually running, for three reasons: (i) the screen was in a low resolution (it is in a higher resolution when the NVIDIA driver is running), (ii) When I looked in Ubuntu Software Centre -> Additional Drivers, it showed that one of the open-source drivers was selected, (iii) When I run nvidia-smi, it prints the following: NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.
  1. Then, I tried to disable the Nouveau driver, using the instructions here: https://linuxconfig.org/how-to-disable-nouveau-nvidia-driver-on-ubuntu-18-04-bionic-beaver-linux. Following this, I manually selected the NVIDIA proprietary driver in Ubuntu Software Centre -> Additional Drivers. I then rebooted, but the driver still was not loaded (low resolution, nvidia-smi error).

  2. I then read that this may be something to do with secure boot being enabled. So, I disabled secure boot at the BIOS. However, this did not change anything.

  3. I then tried installing a different NVIDIA driver (nvidia-driver-435), using the instructions here: https://linuxconfig.org/how-to-install-the-nvidia-drivers-on-ubuntu-18-04-bionic-beaver-linux. This seemed to install the driver correctly, but resulted in a few error messages along the way. I have pasted the full printout here (https://pastebin.com/mgBKXczC), but the two error messages are:
    ERROR: Cannot create report: (Errno 17) File exists: '/var/crash/nvidia-kernel-source-435.0.crash' and Error! Bad return status for module build on kernel: 5.4.0-48-generic (x86_64).
    However, this did not seem to load the driver correctly (again, low resolution, and nvidia-smi error). And when I take a look at Ubuntu Software Centre -> Additional Drivers, I see that actually, the nvidia-driver-435 I installed is actually some open source version, rather than the proprietary version:
    (click here for a screenshot).
    So, all I can do is select the proprietary nvidia-driver-390 from before, which did not work either, as I have already described.

I am completely stuck now as to why this is happening, and what else I should try. Can anybody help please?

nvidia – Ubuntu 20.04.1 Keeps Disabling One of My Displays

I recently asked a very similar question in relation to 18.04.5, and upgraded to 20.04.1 and am experiencing the same issue.

I am curious about why Ubuntu keeps disabling one of my displays.

Currently I am running 4 displays, 3 Dell monitors and a Samsung TV.

If the computer locks and turns off the display connections and the TV shuts itself off after x amount of time , or if I manually turn the TV off, whenever I turn it back on, whether before waking Ubuntu, or after, the screen stays disabled and I have to go into display settings and turn it back on. Which in turn causes me to lose the resolution setting that I had it on.

This is on a fresh install + update + installing NVIDIA drivers.

The difference now is that I also lose audio sometimes and need to enable the monitor, disable it, then enable it again to get my sound back.

This is quite frustrating.

I am running an NVIDIA 1070 and AMD Ryzen 5 2600, the three monitors are on DisplayPort while the TV is running over HDMI.

I am using the recommended drivers for NVIDIA (450.66) using ubuntu-drivers devices and ubuntu-drivers autoinstall

Short of this issue, everything else works great.

Any ideas or help would be greatly appreciated.