vBulletin Search Engine Optimization
| |||||||
| Register | FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| Hey folks, have a problem with my X-Server. I recently installed on my 64-bit-system (amd64 and asus mb) gentoo (kernel x86_64 2.6.13-gentoo-r3). Now the os ist running and I like to run some nice graphic applications. My graphic-card is an RIVA TNT2. What have I done till now? - I installed the nvidia-kernel (6629), nvidia-settings (6629) and the nvidia-glx (6629) - The output of "lsmod" tells me that the modul ist loaded after booting. In the directory "/lib/modules/2.6.13-gentoo-r3/video" the module nvidia.ko exist - I called "opengl-update nvidia" - The relevant section in my "xorg.conf" looks like: Section "Device" Identifier "Graphic1" Screen 0 Driver "nvidia" BoardName "Riva TNT2" VendorName "nVidia" BusID "PCI:1:0:0" EndSection - The broken start of the X-Server tells: NVIDIA(0):Failed to initialize the NVIDIA kernel module! NVIDIA(0): *** Aborting *** Screens found, but none have a usable configuration Strange to me is the output of the cmd "cat /proc/driver/nvidia/agp/status": Status: Disabled and "cat /proc/driver/nvidia/cards/0": Model: RIVA TNT2 Model 64/Model 64 Pro IRQ: 193 Video BIOS: ??.??.??.??.?? Card Type: AGP Can anybody help fixing the problem with my TNT2 card? Thanks Christian |
| |||
| christian.wenz@arcor.de wrote: > graphic applications. My graphic-card is an RIVA TNT2. 1) TNT2 is not supported by new NVidia drivers. See your /var/log/messages. Get older driver by masking new ones. 2) On UDEV system loading the nvidia.o driver does not make /dev/nv* nodes. This puzzled me for quite a long time. They are created later, during X startup, so missing devices does not necessarily mean broken driver. HTH -- Pawel Kraszewski |
| |||
| On Thu, 13 Oct 2005 20:28:33 +0200, Pawel Kraszewski <daemon@pro.wp.pl> wrote: >christian.wenz@arcor.de wrote: > >> graphic applications. My graphic-card is an RIVA TNT2. > >1) TNT2 is not supported by new NVidia drivers. See your /var/log/messages. >Get older driver by masking new ones. > In the output of "dmesg" I find these lines: nvidia: module license NVIDIA taints kernel NVRM: loading NVIDIA Linux x86_64 NVIDIA Kernel Module 1.0-6629 It looks fine to me. But ok, I try it with an older one. I am not so familiar with portage and emerge. Could you tell me how to install an older driver? >2) On UDEV system loading the nvidia.o driver does not make /dev/nv* nodes. >This puzzled me for quite a long time. They are created later, during X >startup, so missing devices does not necessarily mean broken driver. > Thanks for this info. |
| |||
| |
| |||
| christian.wenz@arcor.de wrote: > In the output of "dmesg" I find these lines: > nvidia: module license NVIDIA taints kernel > NVRM: loading NVIDIA Linux x86_64 NVIDIA Kernel Module 1.0-6629 > It looks fine to me. But ok, I try it with an older one. I am not so > familiar with portage and emerge. Could you tell me how to install an > older driver? So, the driver loads correctly. Putting it that way - it already IS a very old driver. Current is 1.0.7676-r1 and doesn't work with anything below GeForce. The higest driver version supporting TNT2 is 1.0-7174. To force it, add >media-video/nvidia-kernel-1.0.7176 >media-video/nvidia-glx-1.0.7176 to your /etc/portage/package.mask and re-emerge nvidia-kernel and nvidia-glx As I see you have 64-bit platform, like me. Why do you have so old driver? What version of Gentoo do you have? Did you 'emerge sync' recently? -- Pawel Kraszewski |
| |||
| On Fri, 14 Oct 2005 11:17:15 +0200, Pawel Kraszewski <Pawel_Kraszewski@wp.pl> wrote: >christian.wenz@arcor.de wrote: > >> In the output of "dmesg" I find these lines: >> nvidia: module license NVIDIA taints kernel >> NVRM: loading NVIDIA Linux x86_64 NVIDIA Kernel Module 1.0-6629 >> It looks fine to me. But ok, I try it with an older one. I am not so >> familiar with portage and emerge. Could you tell me how to install an >> older driver? > >So, the driver loads correctly. > >Putting it that way - it already IS a very old driver. Current is >1.0.7676-r1 and doesn't work with anything below GeForce. > >The higest driver version supporting TNT2 is 1.0-7174. To force it, add > >>media-video/nvidia-kernel-1.0.7176 >>media-video/nvidia-glx-1.0.7176 > >to your /etc/portage/package.mask > >and re-emerge nvidia-kernel and nvidia-glx I have done this with no success. But it seems that the X-Server takes a longer time to abort. Maybe the right way, but anything else is wrong. Do you have some more tricks and hints? > > >As I see you have 64-bit platform, like me. Why do you have so old driver? >What version of Gentoo do you have? Did you 'emerge sync' recently? I get the latest stage2-install-file three days ago and installed gentoo described in the handbook. The kernel version is 2.6.13-gentoo-r3. Means r3 the version of Gentoo? |
| |||
| Christian Wenz wrote: > I have done this with no success. But it seems that the X-Server takes > a longer time to abort. Maybe the right way, but anything else is > wrong. Do you have some more tricks and hints? Run a plain X (type capital 'X' and press enter) Then, if you get no NVidia logo and no "hatchy" pattern, post your "/var/log/Xorg.0.log" and "/etc/X11/xorg.conf" > I get the latest stage2-install-file three days ago and installed I usually go from S1 > gentoo described in the handbook. The kernel version is > 2.6.13-gentoo-r3. Means r3 the version of Gentoo? No, this means 3rd Gentoo revision (let's say "update") of the 2.6.13 kernel. It has nothing to do with the actual Gentoo version. The easiest way to check actual version of Gentoo is to look, where does /etc/make.profile point to: ls -l /etc/make.profile last two parts of the shown path are architecture and gentoo version. Well, for more direct (interactive) help get me at skype (p.kraszewski) or at jabber (pkraszewski@jabber.org) from 7pm up. HTH -- Pawel Kraszewski |
| ||||
| On Fri, 14 Oct 2005 15:45:33 +0200, Pawe? Kraszewski <Daemon@pro.wp.pl> wrote: >Christian Wenz wrote: > >> I have done this with no success. But it seems that the X-Server takes >> a longer time to abort. Maybe the right way, but anything else is >> wrong. Do you have some more tricks and hints? > >Run a plain X (type capital 'X' and press enter) > >Then, if you get no NVidia logo and no "hatchy" pattern, post your >"/var/log/Xorg.0.log" and "/etc/X11/xorg.conf" > >> I get the latest stage2-install-file three days ago and installed > >I usually go from S1 > >> gentoo described in the handbook. The kernel version is >> 2.6.13-gentoo-r3. Means r3 the version of Gentoo? > >No, this means 3rd Gentoo revision (let's say "update") of the 2.6.13 >kernel. It has nothing to do with the actual Gentoo version. > >The easiest way to check actual version of Gentoo is to look, where >does /etc/make.profile point to: > >ls -l /etc/make.profile > >last two parts of the shown path are architecture and gentoo version. > >Well, for more direct (interactive) help get me at skype (p.kraszewski) or >at jabber (pkraszewski@jabber.org) from 7pm up. > >HTH So many thanks to pawel for his effort helping me with my RIVA TNT2 and the X-Server, which doesn't started. Unfortunately all attempts were in vain. After a sleepless night I tried out to create the nodes for the nvidia chip manually, I read this tip at http://www.athemis.de/content/linux/...e.html#nvidia2. Making this changes the X-Server starts and the nvidia-module could be found :-). So, to everybody who has a problem with nvidia and uses udev make sure your nodes are created. After creating the nodes manually udev remvoves und creates them by itself. Greetings from a happy TNT2-user |