In fresh install of 12.3 RC 1 with GeForce 6150 I tried to install nvidia 304.64. It first complained about missing version.h file. I was able to correct that by creating a link but then nvidia-installer gives a general error complaining about kernel sources or files not being correctly configured. I had to reinstall and am using the illegitimate nouveau driver currently. [What has happened to nouveau? It’s now harder to get rid of than roaches in New Orleans. Are nouveau devs proud of this?] The first question is does anyone else have nvidia 304.64 installed in 12.3 and if so how did you do that?
I can attempt to install nvidia 304.64 again in order to provide output for troubleshooting but have currently seen only posts about ‘modern nvidia 310.x’. Hence the title question. Is it that nvidia 304.64 hasn’t been updated to work with kernel 3.7.x-desktop? Or missing packages dependencies? I’m not a novice and am reasonably sure I more than covered needed packages. But, again, I’m willing to try again to provide output.
Aside: Is it coincidence that ‘nouveau’ and 'nvidia (legacy) 304.64 have gone to feces at the same time.>:)
rotfl!: Or is this some vast MS$,nVidia conspiracy to force the purchase of new hardware that a working stiff in South East Louisiana can’t afford thereby creating a group of 2nd class citizens leading to chaos and ultimately the collapse of civilization as we know it?rotfl!
We too have a board that requires nVidia 304.64 -or- better to get quality graphics. However, here in this latest release, RC1, we can not get a command line execution. ‘e’ at Grub2 and Ctrl-Alt-F1 will NOT get us a command line, therefore we can not “try” to install the nVidia Driver.
However, the nouveau Driver is working great for us. This is the best version yet. We have even been able to run our VMware with NO complaints. -But- even though it is better it is still not as smooth and clean as the Native Driver. However, (another one of those) with the nVidia drivers of late, 295.xx and later, we have been experiencing "Freezes". Our screen would freeze and the only out was a Power Off. With nouveau we do not experience that -and- as stated above, in this newest version of nouveau we are getting lots better graphics operations.
We told our wife and friends when we purchased this machine back in Jan of '07 that the only 'Big' update it would need would be a new Video card. Looks like that day may have come for both of us cause we haven't seen -or- heard of any activity on a newer nVidia driver for our cards. In fact we think that they have ceased updates on them.
On 02/16/2013 02:56 AM, chucktr wrote:
>
> Hello Ben,
>
> We too have a board that requires nVidia 304.64 -or- better to get
> quality graphics. However, here in this latest release, RC1, we can not
> get a command line execution. ‘e’ at Grub2 and Ctrl-Alt-F1 will NOT get
> us a command line, therefore we can not “try” to install the nVidia
> Driver.
Press ‘e’ in the grub menu does get an edit window. Use the arrow keys to
position at the end of the line that starts with ‘linux’, add a ‘3’, and press
F10. The system will boot to a command-line prompt with X not running.
> However, the nouveau Driver is working great for us. This is the
> best version yet. We have even been able to run our VMware with NO
> complaints. -But- even though it is better it is still not as smooth
> and clean as the Native Driver. However, (another one of
> those) with the nVidia drivers of late, 295.xx and later, we have
> been experiencing “Freezes”. Our screen would freeze and the only out
> was a Power Off. With nouveau we do not experience that -and- as stated
> above, in this newest version of nouveau we are getting lots better
> graphics operations.
>
> We told our wife and friends when we purchased this machine back in
> Jan of '07 that the only ‘Big’ update it would need would be a new Video
> card. Looks like that day may have come for both of us cause we haven’t
> seen -or- heard of any activity on a newer nVidia driver for our cards.
> In fact we think that they have ceased updates on them.
Yes, nVidia has stopped supporting older hardware with the newest drivers;
however, as I will continue to need using 304.60 for my hardware, I will
continue to publish patches for newer kernels at http://www.lwfinger.com/nvidia_patches/.
I got nvidia 310.32 x86_64 to install using a sym link:
in root at level 3:
quote
cd /lib/modules/<$KERNEL-VERSION>/build/include/
ln -s generated/uapi/linux/
unquote
where <$KERNEL-VERSION> came from “uname -r”, eg 3.7.6-1.2-desktop
Guess we were to succinct -or- not enough in our first reply. We realize that 'e' is not really an editor -but- it will show the lines of Grub2 -and- we know that we can go down to the appropriate line and add the '3'. However, all we are getting is a black screen. To top that off, it turns out the problem in Grub2 is two fold... #1 our lack of understanding. Grub legacy was much easier -but- we are trying to learn the new. jdmcdaniel3 has some interesting scripts to deal with this new Grub2 but we haven't taken the time -yet- to read and understand them. And #2 the Grub2 problem is in our 12.2 version. We have a triple boot setup and at the Grub2 screen we arrow down to the openSuSE 12.3 entry and hit 'e'. All we get is a black screen. Then, since we received your reply we also tried to use the 'e' key... which we have before... on the first line which would get us to the boot line for openSuSE 12.2. It also only produced a black screen. So, something has changed... either by us -or- one of the updates. Till that is fixed we will not be able to boot into mode 3 via Grub.
The other problem is… Ctrl-Alt-F1 in openSuSE 12.3 will not get us a command line screen either. It does nothing. Like we never even hit any keys.
We see that you have some patches for 304.60 but do you have any for 304.64 -or- will your ‘60’ patches work on the ‘64’??
On 02/18/2013 12:36 PM, chucktr wrote:
>
> Hello lwfinger,
>
> Guess we were to succinct -or- not enough in our first reply. We
> realize that ‘e’ is not really an editor -but- it will show the lines of
> Grub2 -and- we know that we can go down to the appropriate line and add
> the ‘3’. However, all we are getting is a black screen. To top that
> off, it turns out the problem in Grub2 is two fold… #1 our lack of
> understanding. Grub legacy was much easier -but- we are trying to learn
> the new. jdmcdaniel3 has some interesting scripts to deal with this new
> Grub2 but we haven’t taken the time -yet- to read and understand them.
> And #2 the Grub2 problem is in our 12.2 version. We have a triple boot
> setup and at the Grub2 screen we arrow down to the openSuSE 12.3 entry
> and hit ‘e’. All we get is a black screen. Then, since we received
> your reply we also tried to use the ‘e’ key… which we have before…
> on the first line which would get us to the boot line for openSuSE 12.2.
> It also only produced a black screen. So, something has changed…
> either by us -or- one of the updates. Till that is fixed we will not be
> able to boot into mode 3 via Grub.
>
> The other problem is… Ctrl-Alt-F1 in openSuSE 12.3 will not get
> us a command line screen either. It does nothing. Like we never even
> hit any keys.
Something is messed up on your system. Here, the ‘e’ key does exactly as it
should, and Ctrl-Alt-FX keys do exactly as they did in 12.2. The only difference
is that 12.3 is my primary boot, and my 12.2 partition is secondary.
> We see that you have some patches for 304.60 but do you have any for
> 304.64 -or- will your ‘60’ patches work on the ‘64’??
As one might suspect the “black/black” -or- as we would interpret it, black on black, produces an unreadable screen. The “fix” was to change the parameter to “white/black”. We tried grey and gray -but- it complained that gray did not exist. Anywho, that took care of the boot up allowing us to enter a ‘3’ at the end of the linux line and get into Command Line Mode.
However, it did NOT fix the Ctrl-Alt-F1 situation. Yep, we understand that some -or- even all of you can get a response with that key combo -but- we can NOT. Absolutely nothing happens. It is like we never pressed the keys. Now then, we vaguely remember, -or- we are imagining, that we read somewhere that there is a setting that Enables/Disables those key combos. Are we dreaming or is that a fact??
Thanks,
Chuck
P.S.
Larry, we have not tried your patches for nVidia… yet… but will soon.
On 2013-02-19 17:46, chucktr wrote:
> However, it did NOT fix the Ctrl-Alt-F1 situation. Yep, we understand
> that some -or- even all of you can get a response with that key combo
> -but- we can NOT. Absolutely nothing happens. It is like we never
> pressed the keys. Now then, we vaguely remember, -or- we are imagining,
> that we read somewhere that there is a setting that Enables/Disables
> those key combos. Are we dreaming or is that a fact??
No, I have read a similar report somewhere, related to systemd starting
something in a different order.
You should report in Bugzilla.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4, with Evergreen, x86_64 “Celadon” (Minas Tirith))
Hey, hey, hey… we have a new driver for our nVidia GeForce 7 series cards. 304.84. Just downloaded it -but- have to go to work so won’t be able to check it out till later. Anyway, just wanted to let the rest of you know … there is a new one!! And we just installed Larry’s fix for the ole 304.64. Thanks Larry and here’s hoping that the new one won’t require any fixes.
Following the excellent instructions in this thread, I installed the 304.60 driver that I download here: NVIDIA DRIVERS 304.60Certified (32 bit system)
Unfortunately this causes a hang on boot, even with nomodeset in the boot command line.
The dying words of the relevant Xorg.0.log are:
25.452] Initializing built-in extension DRI2
25.452] (II) "glx" will be loaded by default.
25.452] (II) LoadModule: "dri2"
25.452] (II) Module "dri2" already built-in
25.452] (II) LoadModule: "glamoregl"
25.476] (II) Loading /usr/lib/xorg/modules/libglamoregl.so
26.078] (EE) Failed to load /usr/lib/xorg/modules/libglamoregl.so: /usr/lib/xorg/modules/libglamoregl.so: undefined symbol: _glapi_tls_C$
26.078] (II) UnloadModule: "glamoregl"
26.078] (II) Unloading glamoregl
26.078] (EE) Failed to load module "glamoregl" (loader failed, 7)
26.078] (II) LoadModule: "glx"
26.079] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
26.111] (II) Module glx: vendor="NVIDIA Corporation"
26.111] compiled for 4.0.2, module version = 1.0.0
26.111] Module class: X.Org Server Extension
26.111] (II) NVIDIA GLX Module 304.60 Sun Oct 14 20:34:01 PDT 2012
26.111] Loading extension GLX
26.111] (II) LoadModule: "nvidia"
26.114] (II) Loading /usr/lib/xorg/modules/drivers/nvidia_drv.so
26.168] (II) Module nvidia: vendor="NVIDIA Corporation"
26.168] compiled for 4.0.2, module version = 1.0.0
26.168] Module class: X.Org Video Driver
26.185] (II) NVIDIA dlloader X Driver 304.60 Sun Oct 14 20:14:37 PDT 2012
26.185] (II) NVIDIA Unified Driver for all Supported NVIDIA GPUs
26.186] (++) using VT number 7
26.186] (EE) No devices detected.
26.186]
Fatal server error:
26.187] no screens found
26.187] (EE)
Any advice? I sure do miss my high-performance graphics!
On 03/06/2013 11:46 AM, saillwhite wrote:
>
> Unfortunately this causes a hang on boot, even with nomodeset in the
> boot command line.
It should not surprise you that nomodeset has no effect. It prevents nouveau
from loading, but you are already doing that by installing the nVidia driver.
> The dying words of the relevant Xorg.0.log are:
>
>> 25.452] Initializing built-in extension DRI2
>> 25.452] (II) “glx” will be loaded by default.
>> 25.452] (II) LoadModule: “dri2”
>> 25.452] (II) Module “dri2” already built-in
>> 25.452] (II) LoadModule: “glamoregl”
>> 25.476] (II) Loading /usr/lib/xorg/modules/libglamoregl.so
>> 26.078] (EE) Failed to load /usr/lib/xorg/modules/libglamoregl.so:
>> /usr/lib/xorg/modules/libglamoregl.so: undefined symbol: _glapi_tls_C$
>> 26.078] (II) UnloadModule: “glamoregl”
>> 26.078] (II) Unloading glamoregl
>> 26.078] (EE) Failed to load module “glamoregl” (loader failed, 7)
This error is not a problem. It happens even on my system where the nVidia
driver works.
>> 26.078] (II) LoadModule: “glx”
>> 26.079] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
>> 26.111] (II) Module glx: vendor=“NVIDIA Corporation”
>> 26.111] compiled for 4.0.2, module version = 1.0.0
>> 26.111] Module class: X.Org Server Extension
>> 26.111] (II) NVIDIA GLX Module 304.60 Sun Oct 14 20:34:01 PDT
>> 2012
>> 26.111] Loading extension GLX
>> 26.111] (II) LoadModule: “nvidia”
>> 26.114] (II) Loading /usr/lib/xorg/modules/drivers/nvidia_drv.so
>> 26.168] (II) Module nvidia: vendor=“NVIDIA Corporation”
>> 26.168] compiled for 4.0.2, module version = 1.0.0
>> 26.168] Module class: X.Org Video Driver
>> 26.185] (II) NVIDIA dlloader X Driver 304.60 Sun Oct 14 20:14:37
>> PDT 2012
>> 26.185] (II) NVIDIA Unified Driver for all Supported NVIDIA GPUs
>> 26.186] (++) using VT number 7
>>
>> 26.186] (EE) No devices detected.
>> 26.186]
>> Fatal server error:
>> 26.187] no screens found
>> 26.187] (EE)
>>
>
> Any advice? I sure do miss my high-performance graphics!
Check the files in /etc/X11/xorg.conf.d/ for correct info.
About those files in /etc/X11/xorg.conf.d - they contain comments only. The /etc/X11/xorg.conf file that was generated by the driver install has valid-looking (but rather generic) monitor, device and screen sections:
Thanks namlak - actually have made progress. Realized looking at failsafe log that I have TWO graphics cards and the default one is Intel. Added the BusId of the NVIDIA driver to xorg.conf, like so:
Also used YaST to install driver that is newly in repository.
Now my Xorg.0.log fails with:
27.019] (II) Module glx: vendor="NVIDIA Corporation"
27.019] compiled for 4.0.2, module version = 1.0.0
27.019] Module class: X.Org Server Extension
27.019] (II) NVIDIA GLX Module 310.32 Mon Jan 14 14:59:59 PST 2013
27.019] Loading extension GLX
27.019] (II) LoadModule: "nvidia"
27.019] (II) Loading /usr/lib/xorg/modules/updates/drivers/nvidia_drv.so
27.073] (II) Module nvidia: vendor="NVIDIA Corporation"
27.073] compiled for 4.0.2, module version = 1.0.0
27.073] Module class: X.Org Video Driver
27.090] (II) NVIDIA dlloader X Driver 310.32 Mon Jan 14 14:40:21 PST 2013
27.090] (II) NVIDIA Unified Driver for all Supported NVIDIA GPUs
27.091] (++) using VT number 7
27.092] (II) Loading sub module "wfb"
27.092] (II) LoadModule: "wfb"
27.092] (II) Loading /usr/lib/xorg/modules/libwfb.so
27.107] (II) Module wfb: vendor="X.Org Foundation"
27.107] compiled for 1.13.2, module version = 1.0.0
27.107] ABI class: X.Org ANSI C Emulation, version 0.4
27.107] (II) Loading sub module "ramdac"
27.107] (II) LoadModule: "ramdac"
27.107] (II) Module "ramdac" already built-in
27.127] (**) NVIDIA(0): Depth 24, (--) framebuffer bpp 32
27.127] (==) NVIDIA(0): RGB weight 888
27.127] (==) NVIDIA(0): Default visual is TrueColor
27.127] (==) NVIDIA(0): Using gamma correction (1.0, 1.0, 1.0)
27.127] (**) NVIDIA(0): Enabling 2D acceleration
27.128] (EE) NVIDIA(0): Failed to initialize the NVIDIA kernel module. Please see the
27.128] (EE) NVIDIA(0): system's kernel log for additional error messages and
27.128] (EE) NVIDIA(0): consult the NVIDIA README for details.
27.128] (EE) NVIDIA(0): *** Aborting ***
27.128] (EE) NVIDIA(0): Failing initialization of X screen 0
27.128] (II) UnloadModule: "nvidia"
27.128] (II) UnloadSubModule: "wfb"
27.128] (EE) Screen(s) found, but none have a usable configuration.
27.128]
I consider this progress! Probably just need to clean things up a bit and try again.