Linux install without driver

How to install Linux on a computer without a video card?

It is fully functional, except — it has no video card. So, is it possible to install Linux or FreeBSD on this machine? I’m not new to Linux, so any Linux distribution is suitable.

I would suggest making the HDD an external USB drive intermediately (or putting it into another PC while unplugging its HDDs) and then install your distro there. Make sure to then set up your ssh-server and put it back into your old PC. Alternatively: if you know that the PC will boot from USB first, you can do the same procedure, but install the system on a flash drive first, then install it on the PC.

There’s an answer to that over here serverfault.com/questions/21255/headless-linux-install (the unix.stackexchange.com question above doesn’t have an accepted answer and seems light on detail).

wiki.debian.org/DebianInstaller/Remote. Try googling «installing debian over ssh». This was the first hit.

Though if I was in your place, I’d just buy a video card. A basic video card is not very expensive, and will make the process much easier. Well worth it, unless your time is of no value.

3 Answers 3

As far as I know, installing on a headless computer is something which can be done using a variety of tricks. The first of the following suggestions uses accessibility features to compensate for the lack of screen. The second one is the easy way to pretend you know the exact sequence of keys to press. The third and fourth are one of many possibilities for installing over a remote connection.

• Debian has accessibility features directly in the installer, and it beeps when the installer is ready. After that beep, you simply have to press s and Enter to get speech synthesis, which you can use if your old machine has an audio jack. See https://wiki.debian.org/accessibility#Debian_installer_accessibility for more details.
• Install in a virtual machine at the same time as on your old machine, and be sure to press the exact same keys.
• Install via SSH, as noted by @FaheemMitha in the comments. If you’re willing to try out something new, NixOs is pretty easy to install over SSH from an existing system. It then is just a matter of booting off a Live USB where you have already configured ssh (add your public key to the ~/.ssh/authorized_keys, with chmod 700 for the folder and 600 for the file), and log in from another computer. Don’t forget to set up SSH in the freshly installed system, before rebooting 🙂 .
• If you have a (couple of) USB to serial dongles, you can always try to pretend you found a time machine, and run the installer over serial (Debian supports that, I think, as well as a lot of other distributions).

Finally, you can install the operating system on another computer, as Fiximan suggested. An easy way to do this without removing the hard disk drive is to install in a virtual machine (QEMU comes to mind) using a raw disk image, and then simply dd that disk image onto the harddisk (which can easily be done by blindly typing the command on the headless computer from a live USB: press Ctrl + Alt + F1 , log-in (check the username and password beforehand), and type zcat /path/to/img.gz | dd of=/dev/sda . And hope that /dev/sda hasn’t been mapped to your harddisk and not to the USB key 🙂 .

Источник

How to install Ubuntu without a USB drive?

I installed Fedora 34 about a week ago and found out that most of the apps used .deb which Fedora doesn’t support, so I was thinking about switching to Ubuntu, BUT I lost my USB drive, so does anyone know how to install Ubuntu 21.04 without USB drive? My laptop uses UEFI, so unetbootin won’t work.

You can alter grub so it boots a ISO from a drive on the system (I use it in QA-testing on a device without working USB ports). Note: I didn’t QA-test Ubuntu 21.04 on that device, and I know changes have been occurring in recent ISOs with live boots, so I can’t guarantee it’ll work, but I still believe it would

How did you install Fedora? If you have an optical drive, then you could always burn the Ubuntu installation ISO to a DVD. It’s much slower, but it’ll give you an opportunity to relive the glorious 90s 😉

2 Answers 2

Booting ISO from Fedora GRUB2 Menu

The method on this page should work for booting an Ubuntu ISO from the Fedora GRUB bootloader 20.04 booting .iso from GRUB menu No need for a USB here.

To reiterate:

Add the following menuentry to /etc/grub.d/40_custom:

Where (hdX,Y) is the disk and partition the ISO is on, for example /dev/sda3 would be (hd0,3) . [path] is the path to the folder the ISO file is in, and [isoname] is the name of the ISO, for example /isos/ubuntu-20.04.2-desktop-amd64.iso is used if the Ubuntu ISO is located in a folder named isos on the root partition. rmmod tpm is only needed when booting in UEFI mode.

Источник

How to install a device driver on Linux

One of the most daunting challenges for people switching from a familiar Windows or MacOS system to Linux is installing and configuring a driver. This is understandable, as Windows and MacOS have mechanisms that make this process user-friendly. For example, when you plug in a new piece of hardware, Windows automatically detects it and shows a pop-up window asking if you want to continue with the driver’s installation. You can also download a driver from the internet, then just double-click it to run a wizard or import the driver through Device Manager.

This process isn’t as easy on a Linux operating system. For one reason, Linux is an open source operating system, so there are hundreds of Linux distribution variations. This means it’s impossible to create one how-to guide that works for all Linux distros. Each Linux operating system handles the driver installation process a different way.

Second, most default Linux drivers are open source and integrated into the system, which makes installing any drivers that are not included quite complicated, even though most hardware devices can be automatically detected. Third, license policies vary among the different Linux distributions. For example, Fedora prohibits including drivers that are proprietary, legally encumbered, or that violate US laws. And Ubuntu asks users to avoid using proprietary or closed hardware.

To learn more about how Linux drivers work, I recommend reading An Introduction to Device Drivers in the book Linux Device Drivers.

Two approaches to finding drivers

1. User interfaces

If you are new to Linux and coming from the Windows or MacOS world, you’ll be glad to know that Linux offers ways to see whether a driver is available through wizard-like programs. Ubuntu offers the Additional Drivers option. Other Linux distributions provide helper programs, like Package Manager for GNOME, that you can check for available drivers.

2. Command line

What if you can’t find a driver through your nice user interface application? Or you only have access through the shell with no graphic interface whatsoever? Maybe you’ve even decided to expand your skills by using a console. You have two options:

Check if a driver is already installed

Before jumping further into installing a driver in Linux, let’s look at some commands that will determine whether the driver is already available on your system.

The lspci command shows detailed information about all PCI buses and devices on the system:

Or with grep:

$ lscpci | grep SOME_DRIVER_KEYWORD

For example, you can type lspci | grep SAMSUNG if you want to know if a Samsung driver is installed.

The dmesg command shows all device drivers recognized by the kernel:

Or with grep:

$ dmesg | grep SOME_DRIVER_KEYWORD

Any driver that’s recognized will show in the results.

If nothing is recognized by the dmesg or lscpi commands, try these two commands to see if the driver is at least loaded on the disk:

Tip: As with lspci or dmesg, append | grep to either command above to filter the results.

If a driver is recognized by those commands but not by lscpi or dmesg, it means the driver is on the disk but not in the kernel. In this case, load the module with the modprobe command:

$ sudo modprobe MODULE_NAME

Run as this command as sudo since this module must be installed as a root user.

Add the repository and install

There are different ways to add the repository through yum, dnf, and apt-get; describing them all is beyond the scope of this article. To make it simple, this example will use apt-get, but the idea is similar for the other options.

1. Delete the existing repository, if it exists.

$ sudo apt-get purge NAME_OF_DRIVER*

where NAME_OF_DRIVER is the probable name of your driver. You can also add pattern match to your regular expression to filter further.

2. Add the repository to the repolist, which should be specified in the driver guide.

$ sudo add-apt-repository REPOLIST_OF_DRIVER

where REPOLIST_OF_DRIVER should be specified from the driver documentation (e.g., epel-list).

3. Update the repository list.

4. Install the package.

$ sudo apt-get install NAME_OF_DRIVER

5. Check the installation.

Run the lscpi command (as above) to check that the driver was installed successfully.

For more information

Источник

Ubuntu 14.04 how to install cuda 6.5 without installing nvidia driver

I am working on workstation with CPU core I7 4790 and two GPUs GTX 760 4 GB ram/1152 core the system’s memory is 16 GB RAM I have Ubuntu 14.04.1 LTS after many tries and reinstalling Ubuntu many times finally i have correctly installed nvidia driver 340.46 using 3 shell commands ppa xorg-edgers now i want to install Cuda 6.5 toolkit but in the manual they say that the cuda toolkit installer will also install nvidia Graphics driver how to prevent the installer from reinstall the graphics driver how to install cuda 6.5 toolkit without reinstalling my graphics driver because i faced many problems until i installed 340.46 correctly and after successfully installing cuda toolkit 6.5 how to upgrade my graphics driver without disturbing cuda toolkit and reinstall everything from the beginning

1 Answer 1

If you use the runfile installer method, the toolkit installer will prompt you individually for each of the 3 components:

So use the runfile installer method, and simply select «no» to the first prompt, if you don’t want to install the driver. You can still install the toolkit and samples. You can download the runfile installer from here

and after successfully installing cuda toolkit 6.5 how to upgrade my graphics driver without disturbing cuda toolkit and reinstall everything from the beginning

The driver can always be installed separately by downloading a driver installer package from www.nvidia.com.

If you use a package manager method, it should be possible to install the toolkit without the driver by installing the cuda-toolkit-X.Y package, where X.Y is the CUDA version. Refer to the package manager section of the install guide

Источник