Drm.ko is missing for CUDA 6.5 / Ubuntu 14.04 / AWS EC2 instance of GP2 g2.2xlarge

Question

To install CUDA 6.5 on Ubuntu 14.04.1 LTS on an AWS EC2 g2.2xlarge instance, do I install via a .deb file or a .run file

.sudo ./cuda_6.5.14_linux_64.run --kernel-source-path=/usr/src/linux-headers-3.13.0-34-generic 

I always get the same error in skipping drm.ko. Code compilation seems successful. Below was a magazine. (I rebooted before installation)

Compilation of the kernel module is complete.

Unable to determine if secure boot is enabled: no such file or directory

Error loading kernel module: no such file or directory

Kernel Messages:

[3.595939] type = 1400 audit (1408809902.911: 5): apparmor = "STATUS"

operation = "profile_replace" profile = "unconfined"

name = "/usr/lib/NetworkManager/nm-dhcp-client.action" pid = 492

Comm = "apparmor_parser"

[3.595942] type = 1400 audit (1408809902.911: 6): apparmor = "STATUS"

operation = "profile_replace" profile = "unconfined"

name = "/ usr / lib / connman / scripts / dhclient- script" pid = 492

Comm = "apparmor_parser"

[3.596140] type = 1400 audit (1408809902.915: 7): apparmor = "STATUS"

operation = "profile_replace" profile = "unconfined"

operation = "profile_replace" profile = "unconfined"

name = "/ usr / lib / connman / scripts / dhclient- script" pid = 492

Comm = "apparmor_parser"

[4.696067] init: fault-tolerant core process (833) killed by the TERM signal

[4.793261] type = 1400 audit (1408809904.107: 8): apparmor = "STATUS"

operation = "profile_replace" profile = "unconfined" name = "/ sbin / dhclient"

pid = 952 comm = "apparmor_parser"

[4.793267] type = 1400 audit (1408809904.107: 9): apparmor = "STATUS"

operation = "profile_replace" profile = "unconfined"

name = "/usr/lib/NetworkManager/nm-dhcp-client.action" pid = 952

Comm = "apparmor_parser"

[5.036249] init: the main plymouth-upstart-bridge process is over, packs

[6.589233] init: udev-fallback-graphics (1203) main process terminated

with status 1

[136.367014] nvidia: "NVIDIA" kernel modulation license.

[136.367019] Disabling debug lock due to kernel corruption

[136.370281] nvidia: module verification error: signature and / or

the required key is missing - kernel damage

[136.370383] nvidia: Unknown character drm_open (err 0)

[136.370393] nvidia: Unknown character drm_poll (err 0)

[136.370404] nvidia: Unknown character drm_pci_init (err 0)

[136.370449] nvidia: Unknown character drm_gem_prime_handle_to_fd (err 0)

[136.370462] nvidia: Unknown character drm_gem_private_object_init (err 0)

[136.370474] nvidia: Unknown character drm_gem_mmap (err 0)

[136.370478] nvidia: Unknown character drm_ioctl (err 0)

[136.370486] nvidia: Unknown character drm_gem_object_free (err 0)

[136.370496] nvidia: Unknown drm_read character (err 0)

[136.370509] nvidia: Unknown character drm_gem_handle_create (err 0)

[136.370515] nvidia: Unknown character drm_prime_pages_to_sg (err 0)

[136.370550] nvidia: Unknown character drm_pci_exit (err 0)

[136.370563] nvidia: Unknown character drm_release (err 0)

[136.370565] nvidia: Unknown character drm_gem_prime_export (err 0)

Driver installation cannot find the kernel source. please make sure the kernel source packages are installed and configured correctly.

Answer 1

The error was caused by the absence of the drm module required by the NVIDIA driver. By default, Ubuntu AMI installs the minimum common Linux kernel (linux-image-virtual), which does not include the drm module. To fix this, install the full linux-image-generic generic kernel. Installing linux-image-extra-virtual will work as it is just a transitional package for linux-image-generic. I would suggest installing linux-generic to include both headers and images. Summarizing:

 sudo apt-get install linux-generic 

The AWS forum has a similar question

Answer 2

In fact, right after the new start of the GPU instance, apt-get upgrade wanted to leave 4 packages as linux-virtual , linux-image-virtual . I still installed them so that I would not update anything else. (The new installation does not have previous nvidia or any nouveau drivers.)

The fact is that linux-image-virtual is a lean build without drm.ko Just do

 sudo apt-get install linux-image-extra-virtual 

which contains drm.ko

Then proceed with the installation of CUDA using the .deb or .run .