File: //usr/share/doc/nvidia-driver-580-server/html/kernel_open.html
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta name="generator" content=
"HTML Tidy for Linux/x86 (vers 1 September 2005), see www.w3.org">
<meta http-equiv="Content-Type" content=
"text/html; charset=us-ascii">
<title>Chapter 45. Open Linux Kernel Modules</title>
<meta name="generator" content="DocBook XSL Stylesheets V1.68.1">
<link rel="start" href="index.html" title=
"NVIDIA Accelerated Linux Graphics Driver README and Installation Guide">
<link rel="up" href="installationandconfiguration.html" title=
"Part I. Installation and Configuration Instructions">
<link rel="prev" href="gsp.html" title=
"Chapter 44. GSP Firmware">
<link rel="next" href="addtlresources.html" title=
"Chapter 46. NVIDIA Contact Info and Additional Resources">
</head>
<body>
<div class="navheader">
<table width="100%" summary="Navigation header">
<tr>
<th colspan="3" align="center">Chapter 45. Open Linux
Kernel Modules</th>
</tr>
<tr>
<td width="20%" align="left"><a accesskey="p" href=
"gsp.html">Prev</a> </td>
<th width="60%" align="center">Part I. Installation and
Configuration Instructions</th>
<td width="20%" align="right"> <a accesskey="n" href=
"addtlresources.html">Next</a></td>
</tr>
</table>
<hr></div>
<div class="chapter" lang="en">
<div class="titlepage">
<div>
<div>
<h2 class="title"><a name="kernel_open" id=
"kernel_open"></a>Chapter 45. Open Linux Kernel
Modules</h2>
</div>
</div>
</div>
<h3>Introduction</h3>
<p>The NVIDIA Linux GPU Driver contains several kernel modules:
nvidia.ko, nvidia-modeset.ko, nvidia-uvm.ko, nvidia-drm.ko, and
nvidia-peermem.ko.</p>
<p>Two "flavors" of these kernel modules are provided:</p>
<div class="itemizedlist">
<ul type="disc">
<li>
<p>Open, i.e. source-published, kernel modules that are dual
licensed MIT/GPLv2. With every driver release, the source code to
the open kernel modules is published on <a href=
"https://github.com/NVIDIA/open-gpu-kernel-modules" target=
"_top">https://github.com/NVIDIA/open-gpu-kernel-modules</a> and a
tarball is provided on <a href=
"https://download.nvidia.com/XFree86/" target=
"_top">https://download.nvidia.com/XFree86/</a>.</p>
</li>
<li>
<p>Proprietary. This is the flavor that NVIDIA has historically
shipped.</p>
</li>
</ul>
</div>
<p>We recommend the use of open kernel modules on all GPUs that
support it.</p>
<p>Though the kernel modules in the two flavors are different, they
are based on the same underlying source code. The two flavors are
mutually exclusive: they cannot be used within the kernel at the
same time, and they should not be installed on the filesystem at
the same time.</p>
<p>The user space components of the NVIDIA Linux GPU driver are
identical and behave in the same way, regardless of which flavor of
kernel module is used.</p>
<h3>Supported Features</h3>
<p>The proprietary flavor supports the GPU architectures Maxwell,
Pascal, Volta, Turing, and later GPUs until Blackwell. Blackwell
and later are only supported by the open kernel modules.</p>
<p>The open flavor of kernel modules supports Turing and later
GPUs. The open kernel modules cannot support GPUs before Turing,
because the open kernel modules depend on the GPU System Processor
(GSP) first introduced in Turing.</p>
<p>The following features will only work with the open kernel
modules flavor of the driver:</p>
<div class="itemizedlist">
<ul type="disc">
<li>
<p>NVIDIA Confidential Computing</p>
</li>
<li>
<p>Magnum IO GPUDirect Storage (GDS)</p>
</li>
<li>
<p>Heterogeneous Memory Management (HMM)</p>
</li>
<li>
<p>CPU affinity for GPU fault handlers</p>
</li>
<li>
<p>DMABUF support for CUDA allocations</p>
</li>
</ul>
</div>
<h3>Known Issues</h3>
<p>The following are some known limitations of the open kernel
modules versus the proprietary kernel modules with GSP firmware
mode disabled:</p>
<div class="itemizedlist">
<ul type="disc">
<li>
<p>GPU initialization is slower. One possible mitigation is to use
nvidia-persistenced to initialize the GPU(s) in advance, before
running applications that use the GPU.</p>
</li>
<li>
<p>Enter and exit latencies for power-saving modes like S3, S4 and
Run Time D3 (RTD3) can be longer due to additional GSP state being
restored.</p>
</li>
<li>
<p>GPU power consumption can be marginally impacted in some
scenarios.</p>
</li>
<li>
<p>Run Time D3 (RTD3) is only supported on Ampere and above
GPUs.</p>
</li>
</ul>
</div>
<h3>Installation</h3>
<p>Because the two flavors of kernel modules are mutually
exclusive, one or the other must be chosen at install time. By
default, installation will choose which flavor of kernel modules to
install, based on the GPUs detected in the system. If a pre-Turing
GPU is detected, installation will default to the proprietary
flavor of kernel modules. Otherwise, installation will default to
the open flavor of kernel modules.</p>
<p>This default can be overridden with the "--kernel-module-type"
.run file option, or its short form "-M". Use "-M=proprietary" to
force installation of the proprietary flavor of kernel modules, or
"-M=open" to force installation of the open flavor of kernel
modules.</p>
<p>E.g.,</p>
<pre class="screen">
sh ./NVIDIA-Linux-[...].run -m=kernel-open
</pre>
<p></p>
<p>As a convenience, the open kernel modules distributed in the
.run file are pre-compiled.</p>
<p>Advanced users, who want to instead build the open kernel
modules from source, should do the following:</p>
<div class="itemizedlist">
<ul type="disc">
<li>
<p>Uninstall any existing driver with `nvidia-uninstall`.</p>
</li>
<li>
<p>Install from the .run file with "--no-kernel-modules" option, to
install everything except the kernel modules.</p>
</li>
<li>
<p>Fetch, build, and install the open kernel module source from
<a href="https://github.com/NVIDIA/open-gpu-kernel-modules" target=
"_top">https://github.com/NVIDIA/open-gpu-kernel-modules</a>. See
<a href=
"https://github.com/NVIDIA/open-gpu-kernel-modules/blob/main/README.md"
target=
"_top">https://github.com/NVIDIA/open-gpu-kernel-modules/blob/main/README.md</a>
for details on building.</p>
</li>
</ul>
</div>
<p>Note that you must use the same version of the .run file and the
open kernel module source from <a href=
"https://github.com/NVIDIA/open-gpu-kernel-modules" target=
"_top">https://github.com/NVIDIA/open-gpu-kernel-modules</a></p>
<p>You can determine which flavor of kernel modules is installed
using either `modinfo` or looking at
/proc/driver/nvidia/version.</p>
<p>E.g., the proprietary flavor will report:</p>
<pre class="screen">
# modinfo nvidia | grep license
license: NVIDIA
# cat /proc/driver/nvidia/version
NVRM version: NVIDIA UNIX x86_64 Kernel Module [...]
</pre>
<p>The open flavor will report:</p>
<pre class="screen">
# modinfo nvidia | grep license
license: Dual MIT/GPL
# cat /proc/driver/nvidia/version
NVRM version: NVIDIA UNIX Open Kernel Module for x86_64 [...]
</pre>
<p></p>
</div>
<div class="navfooter">
<hr>
<table width="100%" summary="Navigation footer">
<tr>
<td width="40%" align="left"><a accesskey="p" href=
"gsp.html">Prev</a> </td>
<td width="20%" align="center"><a accesskey="u" href=
"installationandconfiguration.html">Up</a></td>
<td width="40%" align="right"> <a accesskey="n" href=
"addtlresources.html">Next</a></td>
</tr>
<tr>
<td width="40%" align="left" valign="top">Chapter 44. GSP
Firmware </td>
<td width="20%" align="center"><a accesskey="h" href=
"index.html">Home</a></td>
<td width="40%" align="right" valign="top">
Chapter 46. NVIDIA Contact Info and Additional
Resources</td>
</tr>
</table>
</div>
</body>
</html>