NVIDIA beta, Linux-CK & Bumblebee

Hi all,
I have been wondering about this "issue" -- for a long time I used -bumblebee versions of (lib32-)nvidia-utils and the nvidia kernel driver, modified for linux-ck, not knowing they had disappeared from AUR.
Now I'd like to be able to continue using linux-ck and use the NVIDIA beta driver along with Bumblebee. Is there any way to achieve that, apart from maintaining my own version of Bumblebee since the default depends on nvidia?
Thanks for any insight. Clay

It's required if you want to build kernel modules for example.
If you don't want it to show up when doing "pacman -Qqd" just install it explicity.
edit: And take a look at the nvidia-beta PKGBUILD to see why it was installed.
Last edited by loafer (2012-11-12 20:22:15)

Similar Messages

[Solved] nvidia-beta & linux headers

Just installed nvidia-beta 310.14 & appears to be working fine.
However, when i run:
sudo pacman -R $(pacman -Qqdt)
it offers to remove linux-headers 3.6.6-1.
This seems like a very bad idea. As far as i know, they are in use:
$ uname -a
Linux archlinux 3.6.6-1-ARCH #1 SMP PREEMPT Mon Nov 5 11:57:22 CET 2012 x86_64 GNU/Linux
Would someone kindly explain ?
Last edited by flipper T (2012-11-12 20:29:38)

It's required if you want to build kernel modules for example.
If you don't want it to show up when doing "pacman -Qqd" just install it explicity.
edit: And take a look at the nvidia-beta PKGBUILD to see why it was installed.
Last edited by loafer (2012-11-12 20:22:15)

The "nvidia-beta-ck" aur package and the "linux-ck" package issue?

I use the linux-ck unofficail repo to install the BFQ scheduler enabled kernel. Today it upgraded to the 3.8.8 and it updated fine but the 319 nvidia beta drivers weren't installed after the update. So I reinstalled the nvidia-beta aur package and everything was fine. Is there a way for it to automatically hook the nvidia drivers into the kernel -- say with dkms or will I need to re-install the drivers everytime the kernel updates.

I think there's an nvidia-all package in the AUR no? You can probably modify that to work with the beta package.

Nvidia-beta-dkms + linux-git = no CUDA

hello everyone!
I recently made the switch from the official repo-kernel to the git version (driver issues) and am now forced to use the nvidia-beta-dkms package. but now my current cuda code, which used to be working just fine does still execute, but at runtime no cuda device is detected. I have installed the latest the official cuda package. what is going wrong?
thanks in advance for any help!

I think there's an nvidia-all package in the AUR no? You can probably modify that to work with the beta package.

Request for developers to include the Dell Quadro FX 3450 in their beta linux flashplayer for 10.1

request for developers to include the Dell Quadro FX 3450 in their beta linux flashplayer for 10.1. i'm guessing it wouldn't be hard to do. its the same chip that is used in the Nvidia Geforce 6800. It has NV42GL core processor. please don't leave this device behind, any acceleration that can be done to the existing flash player would definitely count, as great.
thank you

Hi it turns out that the SVN version is the most up to date which does not need to be patched as it has the right version numbers. I have created a new PKGBUILD.
# Maintainer: kso <keansum AT gmail DOT com>
pkgname=freechart-svn
pkgver=r3169
pkgrel=1
pkgdesc="Free powerful charting library based on wxWidgets."
arch=('x86_64' 'i686')
url="http://wxcode.sourceforge.net/components/freechart/"
license=('custom:"wxWindows"')
depends=('wxgtk')
makedepends=('subversion')
source=('svn+http://svn.code.sf.net/p/wxcode/code/trunk/wxCode/components/freechart/')
md5sums=('SKIP')
_svntrunk=http://svn.code.sf.net/p/wxcode/code/trunk/wxCode/components/freechart/
_svnmod=freechart
pkgver() {
cd "$_svnmod"
local ver="$(svnversion)"
printf "r%s" "${ver//[[:alpha:]]}"
build() {
cd "$srcdir"
msg "Connecting to SVN server...."
if [[ -d "$_svnmod/.svn" ]]; then
(cd "$_svnmod" && svn up -r "$pkgver")
else
svn co "$_svntrunk" --config-dir ./ -r "$pkgver" "$_svnmod"
fi
msg "SVN checkout done or server timeout"
msg "Starting build..."
rm -rf "$srcdir/$_svnmod-build"
svn export "$srcdir/$_svnmod" "$srcdir/$_svnmod-build"
cd "$srcdir/$_svnmod-build"
# BUILD HERE
./configure --prefix=/usr
make
package() {
cd "$srcdir/$_svnmod-build"
make DESTDIR="$pkgdir/" install
# install LICENSE
install -D -m644 $srcdir/$_svnmod/license.txt "${pkgdir}/usr/share/licenses/${pkgname}/LICENSE"
I have uploaded onto the AUR. I have checked using namcap. Please let me know if there are any mistakes. Many thanks!

Nvidia beta driver link down

The link for the nvidia beta driver has been down for the last couple of days everytime I try to download it. http://aur.archlinux.org/packages.php?I … hans=&SeB=
I have found another site ftp://download.nvidia.com/XFree86/Linux-x86/185.18.10/ which has two links. They look as if they are in the right format for a package to use in Arch, but will they actually work? Also, which one would I need? There are two - pkg0 and pkg1.run.
Very respectfully!

Okay I think found the answer to my question from the "readme" file.
What is the significance of the pkg# suffix on the .run file?
The pkg# suffix is used to distinguish between .run files containing the same driver, but different sets of precompiled kernel interfaces. If a distribution releases a new kernel after an NVIDIA driver is released, the current NVIDIA driver can be repackaged to include a precompiled kernel interface for that newer kernel (in addition to all the precompiled kernel interfaces that were included in the previous package of the driver).
.run files with the same version number, but different pkg numbers, only differ in what precompiled kernel interfaces are included. Additionally, .run files with higher pkg numbers will contain everything the .run files with lower pkg numbers contain.
So I am assuming it doesnt matter. Is my assumption correct.
very respectfully!

Nvidia-Beta drivers crash with Kernel 3.18.2

After the upgrade to kernel 3.18.2 I recompiled the NVIDIA-Beta drivers to discover that crashes in the kernel. Has anyone got any ideas on how to rectify this issue or have had similar problems?
Log is at the following link - http://pastebin.com/frfMi9qJ
And below is a tiny snippet
[ 15.933352] ------------[ cut here ]------------
[ 15.933363] WARNING: CPU: 0 PID: 663 at drivers/gpu/drm/drm_ioctl.c:143 drm_setversion+0x17e/0x190 [drm]()
[ 15.933393] No drm_driver.set_busid() implementation provided by nvidia_frontend_exit_module [nvidia]. Use drm_dev_set_unique() to set the unique name explicitly.
[ 15.933394] Modules linked in:
[ 15.933395] arc4 ecb md4 md5 hmac nls_utf8 cifs dns_resolver fscache tda18271 tda10048 cfg80211 rfkill snd_hda_codec_hdmi xpad ff_memless led_class mousedev joydev nvidia(PO) coretemp hwmon intel_rapl x86_pkg_temp_thermal intel_powerclamp kvm_intel kvm crct10dif_pclmul saa7164(+) iTCO_wdt crc32_pclmul crc32c_intel iTCO_vendor_support ghash_clmulni_intel mxm_wmi evdev aesni_intel mac_hid tveeprom aes_x86_64 lrw dvb_core gf128mul alx glue_helper v4l2_common ablk_helper videodev cryptd drm psmouse tpm_tis media nuvoton_cir mdio tpm serio_raw rc_core battery video processor snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel snd_hda_controller snd_hda_codec snd_hwdep snd_pcm snd_timer mei_me snd i2c_i801 mei i2c_core soundcore lpc_ich shpchp wmi pcspkr button sch_fq_codel vboxdrv(O) sg ext4 crc16
[ 15.933420] mbcache jbd2 uvesafb hid_generic usbhid hid sr_mod cdrom sd_mod atkbd libps2 ahci libahci libata ehci_pci xhci_pci xhci_hcd ehci_hcd scsi_mod usbcore usb_common i8042 serio
[ 15.933429] CPU: 0 PID: 663 Comm: Xorg.bin Tainted: P O 3.18.2-2-ARCH #1
[ 15.933430] Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./Z87 Killer, BIOS P1.00 09/16/2013
[ 15.933431] 0000000000000000 00000000aa63387f ffff8807eb967c98 ffffffff8154f134
[ 15.933432] 0000000000000000 ffff8807eb967cf0 ffff8807eb967cd8 ffffffff81072bc1
[ 15.933433] ffff8807eb967f1c ffff8807eb967df0 ffff8807eb990000 ffff8807f295a780
[ 15.933435] Call Trace:
[ 15.933439] [<ffffffff8154f134>] dump_stack+0x4e/0x71
[ 15.933442] [<ffffffff81072bc1>] warn_slowpath_common+0x81/0xa0
[ 15.933444] [<ffffffff81072c35>] warn_slowpath_fmt+0x55/0x70
[ 15.933449] [<ffffffffa03f037e>] drm_setversion+0x17e/0x190 [drm]
[ 15.933451] [<ffffffff811de6b2>] ? path_openat+0xc2/0x6e0
[ 15.933455] [<ffffffffa03efe0c>] drm_ioctl+0x2cc/0x680 [drm]
[ 15.933458] [<ffffffffa03f0200>] ? drm_noop+0x40/0x40 [drm]
[ 15.933460] [<ffffffff811dfd59>] ? do_filp_open+0x49/0xd0
[ 15.933462] [<ffffffff811e23e0>] do_vfs_ioctl+0x2d0/0x4b0
[ 15.933463] [<ffffffff811df196>] ? final_putname+0x26/0x50
[ 15.933465] [<ffffffff811e2641>] SyS_ioctl+0x81/0xa0
[ 15.933468] [<ffffffff81554ca9>] system_call_fastpath+0x12/0x17
[ 15.933469] ---[ end trace 1f206c9a64b932eb ]---
[ 17.257683] ------------[ cut here ]------------

ok, that worked.
but my x server will not start after downgrading linux kernel, nvidia and nvidia-utils. oO
looks like the known problem with the kernel and nvidia freeze. are there more packages that i have to downgrade?

[solved] nvidia: requires linux 3.2

pacman -Syu wrote:...
warning: nvidia: ignoring package upgrade (285.05.09-3 => 290.10-2)
warning: nvidia-utils: ignoring package upgrade (285.05.09-1 => 290.10-1)
resolving dependencies...
looking for inter-conflicts...
error: failed to prepare transaction (could not satisfy dependencies)
:: nvidia: requires linux<3.2
I've downgraded nvidia because nvidia-290.10 doesn't work at all for me.
Q: Now, what options do I have?
To cut a long story short...
A: use pacman -Syud
Thanks again, I've learned a lot these days.
Last edited by rtfreedman (2012-01-25 16:49:42)

rtfreedman wrote:
Hm, I don't get pacman to upgrade - ignoring nvidia
I tried 'pacman -Syu --force' to no avail
error: failed to prepare transaction (could not satisfy dependencies)
:: nvidia: requires linux<3.2
I was going to upgrade the kernel, going to init 1 and compiling the nvidia.ko from source.
How do I do it?
I got the same problem, and my solution was:
1) Use pacman -U to install nvidia-290.10-2 and nvidia-utils-290.10-1 (downloaded directly from a mirror). This will force a kernel upgrade.
2) Use pacman -Syu to make entire system up-to-date
3) Use pacman -R nvidia to remove this package
4) Through ABS get the PKGBUILD of nvidia-utils version 290.10 and change the version to 285.05.09
4a) You must also change the md5sum number, for i686 it's 2d469a90abef50320f548cfa8085e3a0
4b) Use makepkg to build the package, and install both created packages (nvidia-utils and opencl-nvidia)
5) Through ABS get the PKGBUILD of nvidia version 290.10 and change the version to 285.05.09 and pkgrel from 1 to 3.
5a) You must also change the md5sum number, for i686 it's 2d469a90abef50320f548cfa8085e3a0
5b) Use makepkg to build the package, and install created package nvidia
6) Edit /etc/pacman.conf add opencl-nvidia to the ignore list
7) Reboot and enjoy
Hope this works for you too.
Ps: Just for the record, I really hate nvidia for not supporting my graphic card anymore...

J2sdkee-1_4-beta-linux.sh installation

We have executed the above file as follows
sh j2sdkee-1_4-beta-linux.sh
We got the following error
ERROR: could not initialize interface awt - exception: java.lang.InternalError:Can't connect to X11 window server using ':0.0' as the value of the DISPLAY variable.
We are using Redhat LINUX 7.2 version.
Could anyone help us?
Thanks&Rgds...Raveen.S

the DISPLAY environment variable is used by linux/unix programs to figure out which X server to connect to when displaying graphics. There're a couple of long explanations at the bottom of
http://www.linux.org/docs/ldp/howto/HOWTO-INDEX/howtos.html. Google gave me a more concise explanation at
http://www.yale.edu/ceo/Documentation/UnixManual/node23.html. If you're running the program on the same machine as your monitor is hooked up to, you'll very likely want DISPLAY to be localhost:0.0
If you're running your program remotely, you'll want to replace localhost with the name of the machine to which your monitor is connected.
As for the original question on this thread. I'm guessing that you probably don't have an X server running. The HOWTO-INDEX link above will probably explain what's going on better than I could.

J2sdkee-1_4-beta-linux installation pb

I install j2re-1_4_2-beta-linux-i586 first
then I wanted to install j2sdkee-1_4-beta-linux
When I run j2sdkee-1_4-beta-linux.sh
I get :
Cannot find Java(TM) 2 Platform, Standard Edition
Version 1.3 or higher is the Java(TM) 2 Platform, Standard Edition
is required to run this installation program
I suppose some installation step is missing.
But didn't find anything in doc.
Andre Legendre

First of all, you'll have to install full JDK in order to run J2EE 1.4 Application Server - JRE will not be enough. If installation still reports the same message after you install JDK, provide -javahome <jkd_path> parameter when you invoke installer.

[SOLVED]Problem updating nvidia-beta

If I update nvidia-utils-beta it tells me I need nvidia-beta
If I try updating nvidia-beta it gives me the following error:
error: target not found: nvidia-utils-beta=275.09.07
==> ERROR: 'pacman' failed to install missing dependencies.
==> ERROR: Makepkg was unable to build nvidia-beta.
Last edited by sbfreak (2011-06-15 10:48:43)

That would take me to the stable branch would it not ? I just want to keep using nvidia-beta to have the latest driver,beta or not.

Configuring nvidia proprietary drivers in bumblebee

I have arch linux 64bit and installed bumblebee and nvidia 302.17-1.
How can I access settings of these drivers?
I"m getting this error when I try it the usual way:
#nvidia-settings
You do not appear to be using the NVIDIA X driver. Please edit your X configuration file (just run `nvidia-xconfig` as root), and restart the X server.
(even after running nvidia-xconfig as root)
Last edited by ebshankar (2012-06-24 14:21:47)

Durden wrote:
Doesn't look like you have them installed. try:
pacman -S nvidia
and see if they are already installed correctly
Its already installed
Gusar wrote:Everything you want to run on the nvidia card needs to be started with optirun. This of course includes nvidia-settings.
tried:
#optirun nvidia-settings
also, it gives the exact same error.

How to add the flash plugin in firefox beta linux x86_64

Someone give me the solution in a live chat session so I share it here.
If you are testing firefox under linux (x86_64 in my case) and the flash plugin is installed but not detected by firefox, it's probably because you are testing it in a single user environment.
For example, you have downloaded and unarchived firefox beta in your home directory :
/home/toto/downloads/firefox
And flash plugin was already installed in :
/usr/lib64/mozilla/plugins/libflashplayer.so
All you have to do is to create a personal plugin directory :
mkdir -p ~/.mozilla/plugins
And then, copy the flash plugin :
cp -a /usr/lib64/mozilla/plugins/libflashplayer.so ~/.mozilla/plugins/

A symbolic link to that file should be sufficient.

New Nvidia beta driver available: 197.13

Nvidia has a new set of beta drivers out (197.13) that fixes problems with the 196.75 set, plus adds some new capabilities. Maybe it can help solve some of the PE8 GPU bugs we've been reading about in this forum...
To get it, go here: http://www.nvidia.com/Download/Find.aspx?lang=en-us and do a search.
Please report any improvements/problems to us all! I will be trying it tonight!
Included in this release:
Adds support for Next Generation ION.
Adds support for GeForce GT 320, GeForce GT 330, and GeForce GT 340.
Upgrades PhysX System Software to version 9.10.0129.
Upgrades HD Audio driver to version 1.0.9.1 (for supported GPUs).
Increases performance in several PC games from v196.21 WHQL. The following are examples of measured improvements. Results will vary depending on your GPU and system configuration:
        * Up to 13% performance increase in Crysis: Warhead with a single GPU.
        * Up to 30% performance increase in Crysis: Warhead with SLI technology.
        * Up to 13% performance increase in H.A.W.X. with single GPU.
        * Up to 15% performance increase in H.A.W.X. with SLI technology.
        * Up to 30% performance increase in Left 4 Dead with single GPU.
        * Up to 28% performance increase in
Left 4 Dead with SLI technology.
Adds SLI and multi-GPU support for top new gaming titles including Assassin Creed II, Battlefield: Bad Company 2, Command and Conquer 4: Tiberium Twilight, Grand Theft Auto IV: Episodes from Liberty City, Mass Effect 2, Napoleon: Total War, and Zombie Driver.
Enhanced SLI support for Metro 2033, Supreme Commander 2, Unigine, and World of Warcraft.
Adds override anti-aliasing support for Mass Effect 2.
Includes numerous bug fixes. Refer to the release notes on the documentation tab for information about the key bug fixes in this release.
Users without US English operating systems can select their language and download the International driver here.

Thanks for this info...
I guess the usage of such drivers might cause some problems since this are just BETA drivers

MOVED: MSI GT70 Nvidia gtx870 linux driver

This topic has been moved to Linux/Unix.
https://forum-en.msi.com/index.php?topic=182029.0

This topic has been moved to Linux/Unix.
https://forum-en.msi.com/index.php?topic=182029.0

NVIDIA beta, Linux-CK & Bumblebee

Similar Messages

Maybe you are looking for