/[gentoo]/xml/htdocs/doc/en/nvidia-guide.xml
Gentoo

Contents of /xml/htdocs/doc/en/nvidia-guide.xml

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1.41 - (show annotations) (download) (as text)
Sun Jul 22 07:32:10 2007 UTC (7 years, 5 months ago) by nightmorph
Branch: MAIN
Changes since 1.40: +105 -123 lines
File MIME type: application/xml
update nvidia references as nvidia-legacy-drivers is deprecated and will be removed. nvidia-drivers now contains every available version. see Cardoe's message to gentoo-dev for more information.

1 <?xml version='1.0' encoding="UTF-8"?>
2 <!-- $Header: /var/cvsroot/gentoo/xml/htdocs/doc/en/nvidia-guide.xml,v 1.40 2007/04/16 05:39:54 nightmorph Exp $ -->
3 <!DOCTYPE guide SYSTEM "/dtd/guide.dtd">
4
5 <guide link="/doc/en/nvidia-guide.xml">
6 <title>Gentoo Linux nVidia Guide</title>
7
8 <author title="Author">
9 <mail link="swift@gentoo.org">Sven Vermeulen</mail>
10 </author>
11 <author title="Author">
12 <mail link="nightmorph@gentoo.org">Joshua Saddler</mail>
13 </author>
14 <author title="Editor">
15 <mail link="curtis119@gentoo.org">M Curtis Napier</mail>
16 </author>
17 <author title="Editor">
18 <mail link="wolf31o2@gentoo.org">Chris Gianelloni</mail>
19 </author>
20
21 <abstract>
22 Many Gentooists have an nVidia chipset on their system. nVidia provides specific
23 Linux drivers to boost the performance of your card. This guide informs you how
24 to install and configure these drivers.
25 </abstract>
26
27 <!-- The content of this document is licensed under the CC-BY-SA license -->
28 <!-- See http://creativecommons.org/licenses/by-sa/2.5 -->
29 <license/>
30
31 <version>1.32</version>
32 <date>2007-07-22</date>
33
34 <chapter>
35 <title>Introduction</title>
36 <section>
37 <body>
38
39 <p>
40 The nVidia drivers in the tree are released by nVidia and are built against the
41 Linux kernel. They contain a binary blob that does the heavy lifting for talking
42 to the card. The drivers consist of two parts, a kernel module, and an X11
43 driver. Both parts are included in a single package. Due to the way nVidia has
44 been packaging their drivers, you will need to make some choices before you
45 install the drivers.
46 </p>
47
48 <p>
49 Currently, there are two driver packages in the tree, though this will be
50 reduced to just one package in the near future. The first,
51 <c>nvidia-drivers</c>, contains the latest drivers from nVidia with support for
52 <e>all</e> cards, with several versions available depending on how old your
53 card is. It uses an eclass to detect what kind of card you're running so that it
54 installs the proper version. The second package, <c>nvidia-legacy-drivers</c>,
55 supports older cards. It is deprecated in favor of the all-inclusive
56 <c>nvidia-drivers</c> package, which contains the same driver version(s). <b>Do
57 not install <c>nvidia-legacy-drivers</c></b>; it's only kept in the tree to give
58 its current users time to switch over to the new package. You should only use
59 <c>nvidia-drivers</c>, as explained in the next chapter.
60 </p>
61
62 <note>
63 Previously, Gentoo provided separate ebuilds for the nVidia kernel module
64 (<c>nvidia-kernel</c>) and the X11 GLX libraries (<c>nvidia-glx</c>). These
65 ebuilds have since been removed from the Portage tree in favor of
66 <c>nvidia-drivers</c>. If you use <c>nvidia-kernel</c> and <c>nvidia-glx</c>,
67 then you should migrate to the newer packages.
68 </note>
69
70 </body>
71 </section>
72 </chapter>
73
74 <chapter>
75 <title>Driver compatibility</title>
76 <section>
77 <title>Which version?</title>
78 <body>
79
80 <p>
81 The <c>nvidia-drivers</c> package supports the full range of available nVidia
82 cards. Multiple versions are available for installation, depending on the
83 card(s) you have.
84 </p>
85
86 <p>
87 Newer cards such as the GeForce 8, 7, 6, and FX 5 series should use the newer
88 drivers from the 100.x series.
89 </p>
90
91 <p>
92 Older cards such as the GeForce 3 or GeForce 4
93 series require the 96xx drivers. For these cards, you should mask
94 <c>>=x11-drivers/nvidia-drivers-1.0.9700</c> in your
95 <path>/etc/portage/package.mask</path> file. This will prevent newer versions of
96 the driver which are incompatible with your card from being installed.
97 </p>
98
99 <p>
100 Old NV2x-based cards (such as TNT, TNT2, GeForce, and GeForce 2) require the
101 older 71xx drivers (such as <c>nvidia-drivers-1.0.7185</c>). For these cards,
102 you should mask <c>>=nvidia-drivers-1.0.8700</c> in
103 <path>/etc/portage/package.mask</path>.
104 </p>
105
106 <p>
107 You can check for driver compatibility for your card at to determine which
108 driver supports it at
109 <uri>http://us.download.nvidia.com/XFree86/Linux-x86_64/100.14.11/README/appendix-a.html</uri>.
110 </p>
111
112 <impo>
113 The <c>nvidia-legacy-drivers</c> package is deprecated and should not be
114 installed on a fresh system. Although it is a 71xx driver, the corresponding
115 71xx driver version found in <c>nvidia-drivers</c> should be used instead.
116 </impo>
117
118 </body>
119 </section>
120 </chapter>
121
122 <chapter>
123 <title>Configuring your Card</title>
124 <section>
125 <title>Kernel Configuration</title>
126 <body>
127
128 <p>
129 As mentioned above, the nVidia kernel driver installs and runs against your
130 current kernel. It builds as a module, so it makes sense that your kernel must
131 support the loading of kernel modules. If you used <c>genkernel all</c> to
132 configure the kernel for you, then you're all set. If not, double check your
133 kernel configuration so that this support is enabled:
134 </p>
135
136 <pre caption="Enabling the Loading of Kernel Modules">
137 Loadable module support ---&gt;
138 [*] Enable loadable module support
139 </pre>
140
141 <p>
142 You also need to enable <e>Memory Type Range Register</e> in your kernel:
143 </p>
144
145 <pre caption="Enabling MTRR">
146 Processor and Features ---&gt;
147 [*] MTRR (Memory Type Range Register) support
148 </pre>
149
150 <p> Also, if you have an AGP graphics card, you can optionally enable
151 <c>agpgart</c> support to your kernel, either compiled in or as a module. If
152 you do not use the in-kernel agpgart, then the drivers will use their own
153 <c>agpgart</c> implementation, called <c>NvAGP</c>. On certain systems, this
154 performs better than the in-kernel agpgart, and on others, it performs worse.
155 You will need to evaluate this on your own system to get the best performance.
156 If you are unsure what to do, use the in-kernel agpgart:
157 </p>
158
159 <pre caption="Enabling agpgart">
160 Device Drivers ---&gt;
161 Character devices ---&gt;
162 &lt;*&gt; /dev/agpgart (AGP Support)
163 </pre>
164
165 <note>
166 On amd64, the IOMMU controls the agpgart setting.
167 </note>
168
169 </body>
170 </section>
171 <section>
172 <title>Arch-specific notes</title>
173 <body>
174
175 <impo>
176 For x86 and AMD64 processors, the in-kernel driver conflicts with the binary
177 driver provided by nVidia. If you will be compiling your kernel for these CPUs,
178 you must completely remove support for the in-kernel driver as shown:
179 </impo>
180
181 <pre caption="Remove the in-kernel driver">
182 Device Drivers ---&gt;
183 Graphics Support ---&gt;
184 &lt; &gt; nVidia Framebuffer Support
185 &lt; &gt; nVidia Riva support
186 </pre>
187
188 <p>
189 A good framebuffer alternative is <c>VESA</c>:
190 </p>
191
192 <pre caption="Enable VESA support">
193 Device Drivers ---&gt;
194 Graphics Support ---&gt;
195 &lt;*&gt; VESA VGA graphics support
196 </pre>
197
198 <p>
199 Then, under "VESA driver type" select either <c>vesafb</c> or
200 <c>vesafb-tng</c>. If you are using an AMD64 processor, you should select
201 <c>vesafb</c> rather than <c>vesafb-tng</c>:
202 </p>
203
204 <pre caption="Select framebuffer type">
205 (X) vesafb
206 ( ) vesafb-tng
207 </pre>
208
209 <p>
210 For more information, you can read up
211 <path>/usr/src/linux/Documentation/fb/vesafb.txt</path> if you are using
212 <c>vesafb</c> or look for your framebuffer documentation under
213 <path>/usr/src/linux/Documentation/fb/</path>.
214 </p>
215
216 </body>
217 </section>
218 <section>
219 <title>Continuing with Kernel Configuration</title>
220 <body>
221
222 <p>
223 The <c>nvidia-drivers</c> ebuild automatically discovers your kernel version
224 based on the <path>/usr/src/linux</path> symlink. Please ensure that you have
225 this symlink pointing to the correct sources and that your kernel is correctly
226 configured. Please refer to the "Configuring the Kernel" section of the <uri
227 link="/doc/en/handbook/">Installation Handbook</uri> for details on configuring
228 your kernel.
229 </p>
230
231 <p>
232 First, you'll need to emerge <c>eselect</c> (if you don't already have it). If
233 you are using <c>gentoo-sources-2.6.18-r4</c>, your kernel listing might look
234 something like this:
235 </p>
236
237 <pre caption="Check your /usr/src/linux symlink">
238 # <i>eselect kernel list</i>
239 Available kernel symlink targets:
240 [1] linux-2.6.18-gentoo-r4 *
241 [2] linux-2.6.20-gentoo-r7
242 [3] linux-2.6.20-gentoo-r8
243 <comment>(Verify that the right kernel is marked with an asterisk</comment>
244 </pre>
245
246 <p>
247 In the above output, you'll notice that the <c>linux-2.6.18-gentoo-r4</c> kernel
248 is marked with an asterisk (<b>*</b>) to show that it is the symlinked kernel.
249 </p>
250
251 <p>
252 If the symlink is not pointing to the correct sources, you must update the link
253 by selecting the number of your desired kernel sources, as in the example
254 above.
255 </p>
256
257 <pre caption="Create/Update /usr/src/linux symlink">
258 <comment>(Select the correct kernel)</comment>
259 # <i>eselect kernel set 1</i>
260 </pre>
261
262 </body>
263 </section>
264 <section>
265 <title>Installing the Appropriate Drivers</title>
266 <body>
267
268 <p>
269 Now it's time to install the drivers.
270 </p>
271
272 <pre caption="Installing the nVidia drivers">
273 # <i>emerge nvidia-drivers</i>
274 </pre>
275
276 <note>
277 The newer 100.x drivers can be installed with the <c>gtk</c> USE flag set. This
278 will install <c>media-video/nvidia-settings</c>, a handy graphical tool for
279 monitoring and configuring several aspects of your nVidia card.
280 </note>
281
282 <impo>
283 Every time you <uri link="/doc/en/kernel-upgrade.xml">compile a new kernel</uri>
284 or recompile the current one, you will need to run <c>emerge nvidia-drivers</c>
285 to reinstall the nVidia modules. An easier way to keep track of modules
286 installed by ebuilds (such as <c>nvidia-drivers</c>) is to install
287 <c>sys-kernel/module-rebuild</c>. Once you've installed it, simply run
288 <c>module-rebuild populate</c> to populate its database with a list of packages
289 to be rebuilt. Once you've finished compiling or recompiling a kernel, just run
290 <c>module-rebuild rebuild</c> to rebuild the drivers for your new kernel.
291 </impo>
292
293 <p>
294 Once the installation has finished, run <c>modprobe nvidia</c> to load the
295 kernel module into memory. If this is an upgrade, you should remove the
296 previous module first.
297 </p>
298
299 <pre caption="Loading the kernel module">
300 # <i>lsmod | grep nvidia &amp;&amp; rmmod nvidia</i>
301 # <i>modprobe nvidia</i>
302 </pre>
303
304 <p>
305 To prevent you having to manually load the module on every bootup, you probably
306 want to have this done automatically each time you boot your system, so edit
307 <path>/etc/modules.autoload.d/kernel-2.6</path> and add <c>nvidia</c> to it.
308 Don't forget to run <c>update-modules</c> afterwards.
309 </p>
310
311 <impo>
312 If you compiled <c>agpgart</c> as a module, you will need to add it to
313 <path>/etc/modules.autoload.d/kernel-2.6</path>.
314 </impo>
315
316 <pre caption="Running update-modules">
317 # <i>update-modules</i>
318 </pre>
319
320 </body>
321 </section>
322 <section>
323 <title>Configuring the X Server</title>
324 <body>
325
326 <p>
327 Once the appropriate drivers are installed you need to configure your X Server
328 to use the <c>nvidia</c> driver instead of the default <c>nv</c> driver.
329 </p>
330
331 <p>
332 Open <path>/etc/X11/xorg.conf</path> with your favorite editor (such as
333 <c>nano</c> or <c>vim</c>) and go to the <c>Device</c> section. In that
334 section, change the <c>Driver</c> line:
335 </p>
336
337 <pre caption="Changing nv to nvidia in the X Server configuration">
338 Section "Device"
339 Identifier "nVidia Inc. GeForce2"
340 <i>Driver "nvidia"</i>
341 VideoRam 65536
342 EndSection
343 </pre>
344
345 <p>
346 Then go to the <c>Module</c> section and make sure the <c>glx</c> module gets
347 loaded while the <c>dri</c> module doesn't:
348 </p>
349
350 <pre caption="Updating the Module section">
351 Section "Module"
352 <comment>(...)</comment>
353 <i># Load "dri"
354 Load "glx"</i>
355 <comment>(...)</comment>
356 EndSection
357 </pre>
358
359 <p>
360 Next, in section <c>Screen</c>, make sure that either the <c>DefaultDepth</c>
361 directive is set to 16 or 24, or that you only have <c>Display</c> subsections
362 with <c>Depth</c> settings of 16 or 24. Without it, the nVidia GLX extensions
363 will not start.
364 </p>
365
366 <pre caption="Updating the Screen section">
367 Section "Screen"
368 <comment>(...)</comment>
369 <i>DefaultDepth 16</i>
370 Subsection "Display"
371 <comment>(...)</comment>
372 EndSection
373 </pre>
374
375 <p>
376 Run <c>eselect</c> so that the X Server uses the nVidia GLX libraries:
377 </p>
378
379 <pre caption="Running eselect">
380 # <i>eselect opengl set nvidia</i>
381 </pre>
382
383 </body>
384 </section>
385 <section>
386 <title>Adding your Users to the video Group</title>
387 <body>
388
389 <p>
390 You have to add your user to the <c>video</c> group so he has access to the
391 nVidia device files:
392 </p>
393
394 <pre caption="Adding your user to the video group">
395 # <i>gpasswd -a youruser video</i>
396 </pre>
397
398 <p>
399 This might not be totally necessary if you aren't using <c>udev</c> but it
400 doesn't hurt either and makes your system future-proof.
401 </p>
402
403 </body>
404 </section>
405 <section>
406 <title>Testing your Card</title>
407 <body>
408
409 <p>
410 To test your nVidia card, fire up X and run <c>glxinfo</c>, which is part of the
411 <c>mesa-progs</c> package. It should say that direct rendering is activated:
412 </p>
413
414 <pre caption="Checking the direct rendering status">
415 $ <i>glxinfo | grep direct</i>
416 direct rendering: Yes
417 </pre>
418
419 <p>
420 To monitor your FPS, run <c>glxgears</c>.
421 </p>
422
423 </body>
424 </section>
425 <section>
426 <title>Enabling nvidia Support</title>
427 <body>
428
429 <p>
430 Some tools, such as <c>mplayer</c> and <c>xine-lib</c>, use a local USE flag
431 called <c>xvmc</c> which enables XvMCNVIDIA support, useful when watching high
432 resolution movies. Add in <c>xvmc</c> in your USE variable in
433 <path>/etc/make.conf</path> or add it as USE flag to <c>media-video/mplayer</c>
434 and/or <c>media-libs/xine-lib</c> in <path>/etc/portage/package.use</path>.
435 </p>
436
437 <p>
438 There are also some applications that use the <c>nvidia</c> USE flag, so you
439 may want to add it to <path>/etc/make.conf</path>.
440 </p>
441
442 <p>
443 Then, run <c>emerge -uD --newuse world</c> to rebuild the applications that
444 benefit from the USE flag change.
445 </p>
446
447 </body>
448 </section>
449 <section>
450 <title>Using NVidia Settings Tool</title>
451 <body>
452
453 <p>
454 nVidia also provides you with a settings tool. This tool allows you to monitor
455 and change graphical settings without restarting the X server and is available
456 through Portage as <c>media-video/nvidia-settings</c>. As mentioned earlier, it
457 will be pulled in automatically if you install
458 <c>>=media-video/nvidia-drivers-100.14.09</c> with the <c>gtk</c> USE flag set
459 in <path>/etc/make.conf</path> or in <path>/etc/portage/package.use</path>.
460 </p>
461
462 </body>
463 </section>
464 </chapter>
465
466 <chapter>
467 <title>Troubleshooting</title>
468 <section>
469 <title>Getting 2D to work on machines with 4Gb or more memory</title>
470 <body>
471
472 <p>
473 If you are having troubles with the nVidia 2D acceleration it is likely that
474 you are unable to set up a write-combining range with MTRR. To verify, check
475 the contents of <path>/proc/mtrr</path>:
476 </p>
477
478 <pre caption="Checking if you have write-combining enabled">
479 # <i>cat /proc/mtrr</i>
480 </pre>
481
482 <p>
483 Every line should contain "write-back" or "write-combining". If you see a line
484 with "uncachable" in it you will need to change a BIOS setting to fix this.
485 </p>
486
487 <p>
488 Reboot and enter the BIOS, then find the MTRR settings (probably under "CPU
489 Settings"). Change the setting from "continuous" to "discrete" and boot back
490 into Linux. You will now find out that there is no "uncachable" entry anymore
491 and 2D acceleration now works without any glitches.
492 </p>
493
494 </body>
495 </section>
496 <section>
497 <title>
498 When I attempt to load the kernel module I receive a "no such device"
499 </title>
500 <body>
501
502 <p>
503 This usually occurs when you don't have a matching video card. Make sure that
504 you have an nVidia-powered graphical card (you can double-check this using
505 <c>lspci</c>).
506 </p>
507
508 <p>
509 If you are confident that you have an nVidia card, check your BIOS and see if
510 the directive <e>Assign IRQ to VGA</e> is set.
511 </p>
512
513 </body>
514 </section>
515 </chapter>
516
517 <chapter>
518 <title>Expert Configuration</title>
519 <section>
520 <title>Documentation</title>
521 <body>
522
523 <p>
524 The nVidia driver package also comes with comprehensive documentation. This is
525 installed into <c>/usr/share/doc</c> and can be viewed with the following
526 command:
527 </p>
528
529 <pre caption="Viewing the NVIDIA documentation">
530 $ <i>less /usr/share/doc/nvidia-drivers-*/README.gz</i>
531 </pre>
532
533 </body>
534 </section>
535 <section>
536 <title>Kernel module parameters</title>
537 <body>
538
539 <p>
540 The <c>nvidia</c> kernel module accepts a number of parameters (options) which
541 you can use to tweak the behaviour of the driver. Most of these are mentioned in
542 the documentation. To add or change the values of these parameters, edit the
543 file <c>/etc/modules.d/nvidia</c>. Remember to run <c>update-modules</c> after
544 modifying this file, and bear in mind that you will need to reload the
545 <c>nvidia</c> module before the new settings take effect.
546 </p>
547
548 <pre caption="Adjusting nvidia options">
549 <comment>(Edit /etc/modules.d/nvidia in your favourite editor)</comment>
550 # <i>nano -w /etc/modules.d/nvidia</i>
551 <comment>(Update module information)</comment>
552 # <i>update-modules</i>
553 <comment>(Unload the nvidia module...)</comment>
554 # <i>modprobe -r nvidia</i>
555 <comment>(...and load it once again)</comment>
556 # <i>modprobe nvidia</i>
557 </pre>
558
559 </body>
560 </section>
561 <section>
562 <title>Advanced X configuration</title>
563 <body>
564
565 <p>
566 The GLX layer also has a plethora of options which can be configured. These
567 control the configuration of TV out, dual displays, monitor frequency detection,
568 etc. Again, all of the available options are detailed in the documentation.
569 </p>
570
571 <p>
572 If you wish to use any of these options, you need to list them in the relevant
573 Device section of your X config file (usually <c>/etc/X11/xorg.conf</c>). For
574 example, suppose I wanted to disable the splash logo:
575 </p>
576
577 <pre caption="Advanced nvidia configuration in the X configuration">
578 Section "Device"
579 Identifier "nVidia Inc. GeForce2"
580 Driver "nvidia"
581 <i>Option "NoLogo" "true"</i>
582 VideoRam 65536
583 EndSection
584 </pre>
585
586 </body>
587 </section>
588 </chapter>
589 </guide>

  ViewVC Help
Powered by ViewVC 1.1.20