I am not sure when this disappeared. Probably during upgrade from 24.04 to 24.10.
On my system, there is no option to set the Night Light from "Sunset to Sunrise".
What can be the cause? And how can I restore old functionality?
This is an almost 6 years old notebook originally installed with Ubuntu 18.10 and after that upgraded every six months to the latest "normal" Ubuntu release.
I haven't met with this problem before. After yesterday's upgrade from 24.04 to 24.10, I started having occasional screen flickers (once or several times successively for almost a time period of a second the screen goes black for a few milliseconds).
At the time of the flicker I have the following in system's syslog:
2024-11-17T19:12:36.160592+03:00 _myhostname_ kernel: i915 0000:00:02.0: [drm] *ERROR* CPU pipe A FIFO underrun
I am currently using the nvidia-driver-560 driver, but I tried the others (X.Org X server and 535-server), with the same result.
I don't know if it a coincidence or not, however after upgrading to 24.10 it started taking tooooo much time (almost two minutes with peak CPU usage) to display the Additional Drivers tab in Software & Updates:
The same for the below command:
# time ubuntu-drivers devices
== /sys/devices/pci0000:00/0000:00:01.0/0000:01:00.0 ==
modalias : pci:v000010DEd00001C8Dsv00001D05sd00001042bc03sc00i00
vendor : NVIDIA Corporation
model : GP107M [GeForce GTX 1050 Mobile]
driver : nvidia-driver-560 - distro non-free recommended
driver : nvidia-driver-535-server - distro non-free
driver : xserver-xorg-video-nouveau - distro free builtin
real 1m52.061s
user 0m36.741s
sys 1m15.267s
I tried to set first i915.modeset=0 and then i915.modeset=1 in /etc/default/grub followed with
$ sudo update-grub
and system reboot, but it didn't help either.
Choosing Wayland during login may solve the problem. However, I am not going to use Wayland, because Wayland is not mature enough and has many problems that Xorg does not.¹
Choosing NVIDIA (Performance Mode) seems to solve the problem. However, I wasn't able to test for an extended period, because of another problem mentioned here.
Disabling NVIDIA and choosing onboard graphics adapter Intel® UHD Graphics 630 does not solve the problem.
¹ I can create a long list of the problems and missing functionalities of Wayland; but this is not in the scope of this question.
System Details Report
Report details
- Date generated: 2024-11-17 19:58:33
Hardware Information:
- Hardware Model: MONSTER ABRA A5 V13.2
- Memory: 24.0 GiB
- Processor: Intel® Core™ i7-8750H × 12
- Graphics: Intel® UHD Graphics 630 (CFL GT2)
- Graphics 1: NVIDIA GeForce GTX 1050
- Disk Capacity: 756.2 GB
Software Information:
- Firmware Version: N.1.10
- OS Name: Ubuntu 24.10
- OS Build: (null)
- OS Type: 64-bit
- GNOME Version: 47
- Windowing System: X11
- Kernel Version: Linux 6.11.0-9-generic
After today's update of openssh-server (1:9.6p1-3ubuntu13.5 ⇒ 1:9.6p1-3ubuntu13.7) any custom port definition on systemctl edit ssh.socket configuration like this will not cause ssh.socket to listen on the IPv4 of the system:
ListenStream=2222
init will listen only on the IPv6 of the system.
Assume that I had a quite old message in my Inbox (or any other folder) and just deleted it. It will go into my Trash folder.
How can I sort my Trash folder so that the message I deleted right now appears on the top or bottom of the message list in the Trash folder? A sort criterion something like "Date Deleted" will be need, but there is none.
Please, feel free to advise me to migrate this question to another more appropriate StackExchange site.
After upgrading to 23.10 (from 23.04) I noticed that on my system there are more than 220 Noto fonts. This is kind of insane. In today's world of UTF standard why do we have so many fonts for some languages? How, can I safely get rid of them?
Please, note that his question is different than one of my previous question about fonts.
After trying to switch from Intel driver to the Nvidia graphics driver on my notebook computer, using Software Updates option from the gnome-control-center, it now boots only to command prompt with no attempt to display a graphical login screen. I tried all the "terminals", using Ctrl+Alt+F1, F2, and so on, but none has the GUI login prompt.
I tried some answers in AskUbuntu, like reinstalling ubuntu-desktop, using the tasksel command to (re)install desktop, removing and reinstalling NVIDIA drivers, trying to switch back to Intel driver, using systemctl set-default graphical.target, but none of them helped.
I have done similar graphics driver switches (Intel ⟷ Nvidia) in the previous versions of Ubuntu for various reasons and never had this problem. My system was upgraded to 22.04 from 21.10 in the end of April. I am using this very system since 18.10 and upgraded every six months with no Ubuntu re-installation so far. The system is fully updated.
The startx command opens a tiny white terminal at the top left corner of the screen. I can run the command gnome-shell to bring up GNOME, but it is very slow and some GUI tools like gnome-terminal or gnome-control-center just crash.
I also noticed that the directory /etc/X11 and its sub-directories of are all empty now. It may be a problem with Xorg.
The command hostnamectl, when executed with no argument, will dump a bunch of information to the terminal. One of that is the Operating System line as shown in the sample output below:
$ hostnamectl
Static hostname: xxxxxx
Icon name: computer-laptop
Chassis: laptop
Machine ID: e8..43
Boot ID: 7b..2a
Operating System: Ubuntu 21.10
Kernel: Linux 5.13.0-28-generic
Architecture: x86-64
Hardware Vendor: XXXXXX
Hardware Model: XXXXXX
However, the output "Ubuntu 21.10" is underlined like a hyperlink and when you Ctrl+click, it will open a web browser with the following link address: https://ubuntu.com/.
How is this accomplished? What are the terminal commands (escape sequences?) that I need to output from a text based program or script to create such a hyperlink on the terminal?
I know that there are many calculator applications for GNOME. But, my question is for the specific gnome-calculator application that comes by default in Ubuntu. Currently, I am using its version 3.38.1 on Ubuntu 20.10.
Most conventional calculators allow to repeat the last operation by repetitively pressing the Enter or = key like this:
- 3+4Enter will display 7
- 3+4EnterEnter will display 11
- 3+4EnterEnterEnter will display 15
- and so on.
Is there a method or workaround to do this in gnome-calculator? Currently, when I press Enter or = key the second time, the calculator displays the original expression and this is something I do not need. I also found nothing special in Keyboard Shortcuts on this.
I have been using Viber for Desktop (the official .deb package downloaded from https://download.cdn.viber.com/cdn/desktop/Linux/viber.deb ) for at least 4 normal Ubuntu versions (18.10 through 20.04) with no problem. A few days ago I upgraded from Ubuntu 20.04 to 20.10 and now the /opt/viber/Viber executable crashes on startup with the following message on terminal:
Segmentation fault (core dumped)
System log has the following:
Oct 27 19:04:42 xxx kernel: [ 3782.065767] Viber[25573]: segfault at 0 ip 00007ff5e4804bc5 sp 00007fff2e7dc3b0 error 4 in iHD_drv_video.so[7ff5e46eb000+348000]
Oct 27 19:04:42 xxx kernel: [ 3782.065771] Code: 01 48 83 f8 42 75 e6 31 db 66 41 89 9f 58 0c 00 00 49 8b 5f 28 8b 83 f0 1d 00 00 85 c0 0f 84 ba 01 00 00 48 8b b3 e8 1d 00 00 <81> 3e 00 00 01 00 48 89 75 c8 0f 85 a3 01 00 00 44 8d 70 f0 4d 63
A crash dump on system's /var/crash directory is also generated.
I have reinstalled the latest version from Viber's website.
I have deleted the ~/.ViberPC and ~/.cache/Viber Media S.à r.l directories from the home directory, but the problem stays the same.
Any other user having the same problem?
Update
I contacted viber.com. First, they answered by proposing a complete removal and re-install. However, this was something that I have already done. After informing them so, they requested the crash dump, which I sent to them.
Currently Viber for Ubuntu is in version 13.3.1 and has not been updated since July 2020. So, we are stuck to this version until an update is released.
For the time being, the temporary workaround is to retry starting Viber after crash, until it does not crash. It seems that the crash is pretty random, with a chance of 10% of the times Viber starting without a crash. So, I wrote the following Bash function and added it to the end of my ~/.bash_aliases file:
function vib
{
while ! /usr/bin/pgrep -i Viber ; do
nohup /opt/viber/Viber >/tmp/vib.out 2>/tmp/vib.err &
sleep 3
done
}
To start Viber, I simply type vib into a terminal and wait until it starts.
Update (Oct 2021)
This problem seems to have been solved in Viber version 16.1.0.37:
I am using Ubuntu 20.04.1. Recently (I haven't noticed exactly when), the graphic login screen of Ubuntu started showing the login names (the first field in the /etc/passwd file) instead of the full user names as set in the 5th (comment) field in the /etc/passwd file.
A copy of my Ubuntu user definition from /etc/passwd is given below:
fedon:x:1000:1000:Fedon Kadifeli:/home/fedon:/bin/bash
For this user, I am expecting the full name Fedon Kadifeli to be displayed on the graphical login user selection screen at the right of the account picture. However, the login name fedon is displayed instead. However, after I select the user and switch to the graphical password entry screen, the full user name of the selected login name is displayed (as expected).
How can I set Ubuntu to display the full names in the login screen again?
Notes:
I have a test (virtual) system with exactly the same Ubuntu version and it displays the user names correctly in the login screen. I need to find the difference between these the two systems. The user entries in
/etc/passwdare exactly the same on both systems. I have checked Settings -> Users on both systems and the user definitions are also the same.I tried to change the entries in
/org/gnome/login-screen/withdconf-editor. However, they have no effect. I guess the changes are made for the current user, whereas a system-wide configuration should be needed for the login screen!I have also looked at: https://help.gnome.org/admin/system-admin-guide/stable/login.html.en , but I found nothing useful.
May be important:
There is an "accessibility" icon (the little guy with open arms) on the log in screen. I have not enabled any accessibility option; all options are in their defaults which is off. However, when I enable the Large Text option, the text becomes larger and the login names change to full user names. Even after I disable the Large Text option back to its original off position, the text reverts back to its normal (smaller) size, but the full user names stay on the screen (exactly as I want them). However, this state is not preserved after reboot.
I have a feeling that this problem may be related to another one observed on my system: After system boot, many processes belonging to `gdm` user are created
I have a project with several files that are opened automatically when I launch the atom editor.
Recently, the atom editor (version 1.48) starts with two empty files with strange names:
ATOM_DISABLE_SHELLING_OUT_FOR_ENVIRONMENT=false
and
atom
It tries to save the first file in my home directory and the second file to the /usr/bin directory (which is not possible due to permissions, of course).
So, I reverted to atom version 1.47 (snap revision 252) where this problem does not occur.
Is there something wrong with revision 257 of atom snap?
Update: Same problem with version 1.49.0 too!
Update: Same problem with version 1.50.0 too!
tl;dr
The summary of all the following is roughly this:
- Remove or move the directory
/usr/share/help/C/gnome-help. - Run the command
sudo apt reinstall gnome-user-docs gnome-getting-started-docs ubuntu-docs. - The missing directory and its contents must come back.
This works OK in one of my 20.04 system, but not on another 20.04 system. Both systems are up-to-date.
I want to find out what makes this difference.
After upgrading from 19.10 to 20.04 the GNOME Help application (yelp) cannot find GNOME help files. For example, when I am in the GNOME Files application (nautilus) and press F1 I get the following error message in the yelp window:
Document Not Found
The URI ‘help:gnome-help/files’ does not point to a valid page.
The same thing occurs in other GNOME applications, like gedit, gnome-calculator etc.
yelp can display the man pages OK. The problem is only with the "help" pages. Is there any package that needs to be installed?
I tried to install package gnome-user-guide, but it didn't help. Packages:
gnome-getting-started-docs
gnome-user-docs
ubuntu-docs
xorg-docs-core
are already installed and up-to-date.
It seems* that the files needed are the ones under /usr/share/help/C/gnome-help/ and the package gnome-user-docs is the one that provides these (https://packages.ubuntu.com/focal/all/gnome-user-docs/filelist). However, even removing, purging and re-installing this package does not bring these files to my system.
*Note: dpkg -S /usr/share/help/C/gnome-help/files.page output is this:
gnome-user-docs: /usr/share/help/C/gnome-help/files.page
Also ls -lisad /usr/share/help/ /usr/share/help/C output is:
8651957 4 drwxr-xr-x 59 root root 4096 May 14 19:55 /usr/share/help/
8651631 4 drwxr-xr-x 31 root root 4096 May 17 17:39 /usr/share/help/C
and permissions and ownership of the files within these directories is -rw-r--r-- root root.
Similarly, other GNOME packages (for example, gnome-calculator) when (re)installed, they do not populate the relevant directory under /usr/share/help/C/.
I tested by copying /usr/share/help/C/gnome-help/ directory and its contents from another system and then reinstalling the gnome-user-docs package. The re-installation deleted the necessary files which I just copied there! Before doing the re-installation, but after copying these files manually, yelp was able to display its home screen and the relevant help screens OK. After the re-installation, the above mentioned problem occurs.
I have tried with various locales with the installation of gnome-user-docs package. Every time the same thing occurs. The files that I copied manually under /usr/share/help/C/ are being deleted:
# LANG=en_US.UTF-8 apt --reinstall install gnome-user-docs
# LANG=en_GB.UTF-8 apt --reinstall install gnome-user-docs
# LANG=C.UTF-8 apt --reinstall install gnome-user-docs
My locale output is this:
LANG=C.UTF-8
LANGUAGE=
LC_CTYPE="C.UTF-8"
LC_NUMERIC=en_GB.UTF-8
LC_TIME=en_GB.UTF-8
LC_COLLATE="C.UTF-8"
LC_MONETARY=en_GB.UTF-8
LC_MESSAGES="C.UTF-8"
LC_PAPER=en_GB.UTF-8
LC_NAME=en_GB.UTF-8
LC_ADDRESS=en_GB.UTF-8
LC_TELEPHONE=en_GB.UTF-8
LC_MEASUREMENT=en_GB.UTF-8
LC_IDENTIFICATION=en_GB.UTF-8
LC_ALL=
(Note that I tested with LANG=en_US.UTF-8 also; result was the same.)
The output of check-language-support --show-installed is:
firefox-locale-en gimp-help-en hunspell-en-au hunspell-en-ca hunspell-en-gb hunspell-en-us hunspell-en-za hyphen-en-ca hyphen-en-gb hyphen-en-us language-pack-en language-pack-gnome-en libreoffice-help-en-gb libreoffice-help-en-us libreoffice-l10n-en-gb libreoffice-l10n-en-za mythes-en-au mythes-en-us poppler-data thunderbird-locale-en thunderbird-locale-en-gb thunderbird-locale-en-us wamerican wbritish
Currently installed packages:
# apt --installed list | grep -i "docs"
gnome-getting-started-docs/focal,focal,now 3.36.1-0ubuntu1 all [installed]
gnome-user-docs/focal,focal,now 3.36.1-0ubuntu1 all [installed]
ubuntu-docs/focal,focal,now 20.04.2 all [installed]
xorg-docs-core/focal,focal,now 1:1.7.1-1.1 all [installed,automatic]
More Information
(1)
The output of sudo apt-cache depends yelp is like this:
yelp
Depends: libc6
Depends: libglib2.0-0
Depends: libgtk-3-0
Depends: libwebkit2gtk-4.0-37
Depends: libyelp0
|Depends: dconf-gsettings-backend
Depends: <gsettings-backend>
dconf-gsettings-backend
Depends: python3-distro
Depends: yelp-xsl
Depends: man-db
man-db:i386
Recommends: docbook-xml
Note: This is the same output as on another system which does not have this problem.
(2)
The checksums for the yelp package are OK:
$ debsums yelp | wc -l
230
$ debsums yelp | grep 'OK$' | wc -l
230
$ debsums -a yelp | grep 'OK$' | wc -l
230
However, the checksums for the gnome-user-docs package show inconsistent results. On the system with this problem:
$ debsums gnome-user-docs | wc -l
4
$ debsums gnome-user-docs | grep 'OK$' | wc -l
4
$ debsums -a gnome-user-docs | grep 'OK$' | wc -l
4
$ debsums gnome-user-docs
/usr/share/doc/gnome-user-docs/NEWS.gz OK
/usr/share/doc/gnome-user-docs/README OK
/usr/share/doc/gnome-user-docs/changelog.Debian.gz OK
/usr/share/doc/gnome-user-docs/copyright OK
Whereas on a system without this problem:
$ debsums gnome-user-docs | wc -l
465
and the difference is in the /usr/share/help/C/gnome-help/ and /usr/share/help/C/system-admin-guide/ files: The files I am talking about throughout this post.
My system was upgraded from 18.10 to 19.04, then to 19.10 and finally to 20.04. I have been using it for more than 15 months as my primary computer. This is a typical default Ubuntu installation with some extra manual "apt install"s.
I noticed that there are a lot of fonts and font variations (mostly Asian) that I haven't installed explicitly. Selecting a font in an application (e.g. in LibreOffice) is cumbersome because of the long font drop-down list box. Also, I think that having so many unneeded fonts will make applications slower and is a waste of system resources.
I would like to remove most of the seemingly unnecessary fonts from my system. However, I am not sure which ones are necessary for proper system and core applications operation. Are there any pointers on how to decide on this?
I am using Ubuntu 20.04 desktop. I have only the default themes installed on my system. Currently, I am using Adwaita (default) which does not have arrows at the two ends of vertical or horizontal window scrollbars.
I have tested the other themes; only HighContrast and HighContrastInverse seem to provide the arrows I need. However, these themes are not appropriate for me. Adwaita (default) is just fine, except for the missing arrows in scrollbars.
I have tried a few solutions provided on the Internet (like https://www.slsmk.com/adding-scrollbars-and-buttons-to-desktop-theme/ ) that try to create a file like ~/.config/gtk-3.0/gtk.css and put the necessary CSS declarations, but these do not work.
What is the recommended method to locally modify a theme for GTK2 applications and add these scrollbar arrows?
Update:
Actually, it turns out that the above-mentioned action does work for GTK3 applications. But, my application that needs these arrows uses GTK2.
I tried locations like ~/.config/gtk-2.0/gtkrc, ~/.gtkrc, and ~/.gtkrc-2.0, but it didn't help.
Editing directly the file /usr/share/themes/Adwaita/gtk-2.0/main.rc and commenting out the lines:
GtkScrollbar::has-backward-stepper = 0
GtkScrollbar::has-forward-stepper = 0
seems to help, but I want to do it on user basis (by settings their values to 1).
I have an Ubuntu 20.04 server installed on a single 8GB drive. The default installation has some "snap"s installed also. So, there are some "squashfs" file systems also reported by the df command:
# df -mT
Filesystem Type 1M-blocks Used Available Use% Mounted on
/dev/root ext4 7877 1837 6025 24% /
devtmpfs devtmpfs 465 0 465 0% /dev
tmpfs tmpfs 477 0 477 0% /dev/shm
tmpfs tmpfs 96 1 95 1% /run
tmpfs tmpfs 5 0 5 0% /run/lock
tmpfs tmpfs 477 0 477 0% /sys/fs/cgroup
/dev/loop0 squashfs 18 18 0 100% /snap/amazon-ssm-agent/1566
/dev/loop1 squashfs 94 94 0 100% /snap/core/9066
/dev/loop2 squashfs 55 55 0 100% /snap/core18/1705
/dev/loop3 squashfs 69 69 0 100% /snap/lxd/14804
/dev/loop4 squashfs 70 70 0 100% /snap/lxd/14890
/dev/loop5 squashfs 55 55 0 100% /snap/core18/1754
tmpfs tmpfs 96 0 96 0% /run/user/1000
As you see, there is only 1837MB of data stored in the (only) disk.
Now, I am trying the list that disk usage for each directory present under root (/) using the following command:
# du -smc /* 2>/dev/null
0 /bin
48 /boot
0 /dev
8 /etc
1 /home
0 /lib
0 /lib32
0 /lib64
0 /libx32
1 /lost+found
1 /media
1 /mnt
1 /opt
0 /proc
1 /root
1 /run
0 /sbin
1116 /snap
1 /srv
0 /sys
1 /tmp
1166 /usr
601 /var
2938 total
The output shows a large amount of disk space used by the /snap directory, which of course is not true.
What is the correct way to count the size of files residing only on "real disk" filesystems? Adding the option -x to du does not make me feel comfortable, because in the future I may have another "real disk" filesystem mounted under /home for example and I do want that to be counted in du's output.
The ubuntu-support-status command is no longer provided by the update-manager-core package. The closest I found is ubuntu-security-status:
# dpkg --search "ubuntu-*-status*"
update-manager-core: /usr/bin/ubuntu-security-status
There are no man pages for these commands. So, I cannot compare their usage.
Any ideas?
By removing the parameters quiet and splash from GRUB_CMDLINE_LINUX_DEFAULT line at the /etc/default/grub file, I am able to display the messages displayed by system startup scripts (services) to the screen during Ubuntu boot.
However, these messages are scrolling very fast and it is practically impossible to catch any FAILED messages. The file /var/log/boot.log used to hold these messages. For example:
# grep -a FAILED /var/log/boot.log
[FAILED] Failed to start Raise network interfaces.
...
However, this file is no longer updated (since April 2019) on my system, which suggests that starting with Ubuntu 19.04 it is deprecated. Where can I find that content on Ubuntu 19.10? Is there a way to capture the data displayed by init scripts on the console during boot to a file?
Note
The same functionality (i.e., the exact console text output) is not provided by journalctl. For example:
# journalctl -b 0 | grep Raise
Jan 13 12:01:25 ... systemd[1]: Starting Raise network interfaces...
Jan 13 12:01:25 ... systemd[1]: Started Raise network interfaces.
I have the following lines repeating every minute in my /var/log/syslog file.
Oct 9 19:19:42 my_machine dbus-daemon[2138]: [session uid=1000 pid=2138] Activating via systemd: service name='org.freedesktop.Tracker1.Miner.Extract' unit='tracker-extract.service' requested by ':1.66' (uid=1000 pid=2555 comm="/usr/lib/tracker/tracker-miner-fs " label="unconfined")
Oct 9 19:19:42 my_machine systemd[2118]: Starting Tracker metadata extractor...
Oct 9 19:19:42 my_machine dbus-daemon[2138]: [session uid=1000 pid=2138] Successfully activated service 'org.freedesktop.Tracker1.Miner.Extract'
Oct 9 19:19:42 my_machine systemd[2118]: Started Tracker metadata extractor.
Oct 9 19:19:53 my_machine systemd[2118]: tracker-extract.service: Succeeded.
Can somebody explain what is going on.
My system is Ubuntu 19.04 with latest updates. This problem seems to have started around April 21st. Probably after upgrading 18.10 to 19.04.
How can I change the period from 1 minute to (say) 1 hour?
Note: Even after I followed the explanation in another question (as my question was identified as a possible duplicate of it) nothing changed of what I am detailing below.
I know there are myriad questions and how-to's about installing and using NVIDIA graphics card on Ubuntu, but my case is a bit different (probably due to using Ubuntu 19.04).
My laptop has a GeForce GTX 1050 NVIDIA graphics card besides the "default" Intel graphics card. (I am using the laptop's own display and nothing is connected to the HDMI port.)
$ lspci -k | grep -A 2 -i "VGA"
00:02.0 VGA compatible controller: Intel Corporation UHD Graphics 630 (Mobile)
Subsystem: Tongfang Hongkong Limited UHD Graphics 630 (Mobile)
Kernel driver in use: i915
--
01:00.0 VGA compatible controller: NVIDIA Corporation GP107M [GeForce GTX 1050 Mobile] (rev a1)
Subsystem: Tongfang Hongkong Limited GP107M [GeForce GTX 1050 Mobile]
Kernel driver in use: nvidia
I already have installed the latest recommended NVIDIA driver (probably several weeks ago!):
$ sudo ubuntu-drivers devices
== /sys/devices/pci0000:00/0000:00:01.0/0000:01:00.0 ==
modalias : pci:v000010DEd00001C8Dsv00001D05sd00001042bc03sc00i00
vendor : NVIDIA Corporation
model : GP107M [GeForce GTX 1050 Mobile]
driver : nvidia-driver-390 - distro non-free
driver : nvidia-driver-418 - distro non-free recommended
driver : xserver-xorg-video-nouveau - distro free builtin
$ sudo apt-get install nvidia-driver-418
Reading package lists... Done
Building dependency tree
Reading state information... Done
nvidia-driver-418 is already the newest version (418.56-0ubuntu1).
0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded.
$ prime-select query
nvidia
However, system Settings | Details | About displays that the following graphics processor is active:
Intel® UHD Graphics 630 (Coffeelake 3x8 GT2)
I get the following output from nvidia-smi command:
$ nvidia-smi
Wed May 29 19:17:55 2019
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 418.56 Driver Version: 418.56 CUDA Version: 10.1 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce GTX 1050 Off | 00000000:01:00.0 Off | N/A |
| N/A 33C P8 N/A / N/A | 2MiB / 4040MiB | 0% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+
The GPU seems idle.
And the nvidia-settings command just displays a simple window like this:
Considering all these, I believe that currently the Intel graphics card, and not the NVIDIA graphics card is active in my system.
Purging and reinstalling the NVIDIA driver does not help.
Certainly I am missing something. But what?
GSmartControl and any other command line tool (like fdisk, smartctl, cat /sys/block/sd*/queue/hw_sector_size, cat /sys/block/sd*/queue/physical_block_size) I had used report the same for both of my disks:
Sector Size: 512 bytes logical/physical
This is a default Ubuntu 18.10 (later upgraded to 19.04) installation. However, the stat -f command on both disks reports:
Block size: 4096 Fundamental block size: 4096
Both of my disks are SSDs and AFAIK SSD disks require a sector size of 4K. Is this OK or am I missing something? Does the information returned by stat (=4K) ensure that the OS will always send IOs to the disk in multiple of 4K and these blocks will never cross 4K boundaries (IO blocks will always be aligned to 4K)?
Please note the following output (sdb2 is my root partition, sda is my /home disk):
# fdisk -l /dev/sd?
Disk /dev/sda: 465.8 GiB, 500107862016 bytes, 976773168 sectors
Disk model: SanDisk SDSSDH35
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk /dev/sdb: 238.5 GiB, 256060514304 bytes, 500118192 sectors
Disk model: ADATA SU800NS38
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: xxxx....
Device Start End Sectors Size Type
/dev/sdb1 2048 1050623 1048576 512M EFI System
/dev/sdb2 1050624 500117503 499066880 238G Linux filesystem
# df / /home
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sdb2 244568380 17799136 214276188 8% /
/dev/sda 479670976 129685112 325550152 29% /home