Musings » nVidia

Inverting Large Images Using CUDA

fpmurphy — Sun, 18 Oct 2009 17:10:45 +0000

This is post demonstrates how to invert a very large image, stored as a vector, using nVidia's CUDA programming environment and a GeForce 260 GTX graphics card.

Project Plymouth

fpmurphy — Sun, 13 Sep 2009 21:25:08 +0000

The aim of the Plymouth project is to answer the oft-stated user requirement that their computer startup be flicker-free, seamless, shiny and logged. In this post I examine the current state of the project and show you how to customize your startup and shutdown experience.

Fedora 11 nVidia Twinview Support

fpmurphy — Thu, 11 Jun 2009 17:07:00 +0000

Fedora 11 (Leonidas) ships with the nouveau nVidia graphics driver preloaded by default if a nVidia graphics card is detected at install time. Previous versions of Fedora used the older X.Org nv driver.

The nouveau project aims at producing Open Source 3D drivers for nVidia graphics cards. According to the nouveau project Wiki

2D-support is in fairly good shape with EXA acceleration, Xv and Randr12 (think of dual-head, rotations, etc.). Randr12 should work for all cards up to, and including, Geforce 9000 series, although some issues with Geforce 8/9 laptops may still exist, for such issues bug reports should be submitted. Randr12 is now the default. Any 3D functionality that might exist is still unsupported, do not ask for instructions to try it. Also, VT switching while X is running is considered lucky.”

Well, I certainly quickly ran into the VT switching issue! It worked but not consistently.

Unfortunately the nouveau driver currently does not support nVidia TwinView functionality and I suspect that it will be a long time before it does if ever!

To use TwinView with Fedora 11, you have to load the correct nVidia drivers from rpmfusion.org. I described how to do this in detail in a previous post so I will not repeat that information here.

You also need to modify your grub.conf file to include the nopat kernel boot option as shown below. title Fedora (2.6.29.4-167.fc11.x86_64) root (hd0,1) kernel /vmlinuz-2.6.29.4-167.fc11.x86_64 ro root=/dev/mapper/vg_ultra-lv_root rhgb quiet nopat initrd /initrd-2.6.29.4-167.fc11.x86_64.img

The nopat option is needed for this particular kernel (2.6.29.4) as it appears to still have broken PAT functionality.

For those readers who are unaware of what PAT is, here is a brief explanation. Traditionally page caching was controlled by a CPU feature called Memory Type Range Registers (MTRR). A CPU has a finite and limited set of MTRRs each of which control part of the physical address space. To overcome this limitation and provide a more flexible architecture, Intel and other x86 CPU vendors added a set of bits to page table entries to control how a CPU does page caching. These bits are called the Page Attribute Table (PAT). Incidentally, the 2.6.26 kernel was the first Linux kernel to support PATs.

Unless you rebuild your initial ramdisk (initrd), the nouveau driver will remain loaded in the kernel. I prefer not to have the nouveau driver loaded in my kernel if I am

Fedora 10 Dual Head Graphics Card

fpmurphy — Sat, 28 Mar 2009 17:50:00 +0000

This post details how to install and configure two monitors on a single EVGA GeForce 9500 GT graphics (video) card under Fedora 10.

When I was researching dual head graphics cards for use with Fedora, I came across a lot of forum and blog entries on the subject but most of them contained obsolete or incorrect information – especially about TwinView, Xinerama and the X Window System configuration file. Hopefully this post will provide readers with some useful up-to-date information on the subject.

This particular graphics card (EVGA P/N 01G-P3-N959-TR) was of interest to me for a number of reasons but primarly because the card contains two heads each with DVI-I connectors which would allow me to move away from VGA cabling altogether since my two HP w1907 19″ LCD monitors also have DVI-I connectors. In addition two heads on one graphics card enabled me to free up a PCI slot by eliminating the need for a second graphics card in my workstation.

Using TwinView was also of interest to me because it uses a single X screen, i.e. the driver conceals all information about the two separate monitors from the X server, and both monitors share the same frame buffer. This means that I can continue to use functionality such as accelerated OpenGL without a problem.

There are many graphics cards available which contain the nVidia 9500GT GPU but, to date, few of these cards come with dual DVI-I heads. In addition nVidia fully supports their graphics cards on GNU/Linux and Solaris platforms. Another feature of modern nVidia GPUs which is of great interest to me is their support for the CUDA (Compute Unified Device Architecture) architecture which I wish to experiment with a part of a project which I am currently working on for a customer.

You can use a single monitor with this card out-of-the-box using the default nv driver which comes standard with Fedora. However, to use two monitors with this card, you need to download and install the correct package for your specific kernel, i.e. kmod-nvidia-2.6.27.19-170.2.35.fc10.x86_64-180.29-1.fc10.1.x86_64 in my particular case as my present kernel is 2.6.27.19-170.2.35. Selecting this package will cause three other packages to be also downloaded and installed on your system unless they already exist. These packages are livna-config-display-*, xorg-x11-drv-nvidia-*, and xorg-x11-drv-nvidia-libs-*.

These packages not not available in the regular Fedora repositories. You need to add the rpmfusion repositories to /etc/yum.conf or to