| View previous topic :: View next topic |
| Author |
Message |
FictionPimp
n00b
Joined: 24 Aug 2004
Posts: 43
|
 Posted: Sat Dec 17, 2005 3:37 am Post subject: |
 |
|
| Tweakin wrote: | Well, after trying everyones suggestions and messing around myself for a day, I wasn't able to get it working. I completely reinstalled Gentoo (it was a fresh install) again and did not use genkernel. X is working fine now, and I had no problems. I did everything the same way, minus genkernel, so that leads me to believe there is something being included in genkernels kernel that the nvidia-kernel isn't playing nicely with.
Wish I could have gotten to the bottom of it, i hate leaving things open like that, but the system needs to be built and dropped into place tomorrow. |
From the sound of it I would be willing to bet genkernel installed nvidiafb. With that enabled the nvidia module wont load and gives all sorts of fun errors. |
|
| Back to top |
|
 |
depontius
Veteran
Joined: 05 May 2004
Posts: 1925
|
| Posted: Mon Dec 19, 2005 1:39 pm Post subject: |
|
|
I use genkernel, but that hasn't been related to my problems, because I'm still using it and am working.
Oops... Are you doing bootsplash?
Bootsplash requires framebuffer, and there is the interaction between the GPL framebuffer for nVidia and the nVidia driver. About the time I had my nVidia troubles moving off of 2.6.12, I also had framebuffer problems. So at the moment, I'm back to a straight text console, no framebuffer at all. To begin, you might try that. Beyond that, you might try using the straight svga framebuffer, not the nVidia-specific one.
In my case, somewhere between 2.6.12 and 2.6.14, the framebuffer began using scan frequencies too high for my monitor. I haven't bothered to bug it out, since text mode gets me working, and the framebuffer is just a nice thing.
_________________
.sigs waste space and bandwidth |
|
| Back to top |
|
|
iMike
Apprentice
Joined: 01 Apr 2005
Posts: 193
Location: Stockholm, Sweden
|
| Posted: Mon Dec 19, 2005 5:01 pm Post subject: |
|
|
The suggestion to use
| Code: |
/sbin/NVmakedevices.sh
|
worked for me. Great! Thanks a lot. Also, thanks for the script,
Very convenient.  |
|
| Back to top |
|
|
lorenct
n00b
Joined: 20 May 2004
Posts: 31
|
| Posted: Tue Dec 20, 2005 5:20 am Post subject: |
|
|
| depontius wrote: | I use genkernel, but that hasn't been related to my problems, because I'm still using it and am working.
Oops... Are you doing bootsplash?
Bootsplash requires framebuffer, and there is the interaction between the GPL framebuffer for nVidia and the nVidia driver. About the time I had my nVidia troubles moving off of 2.6.12, I also had framebuffer problems. So at the moment, I'm back to a straight text console, no framebuffer at all. To begin, you might try that. Beyond that, you might try using the straight svga framebuffer, not the nVidia-specific one.
In my case, somewhere between 2.6.12 and 2.6.14, the framebuffer began using scan frequencies too high for my monitor. I haven't bothered to bug it out, since text mode gets me working, and the framebuffer is just a nice thing. |
Interesting... I saw the same nvidia / linux kernel problems moving from 2.6.12-r10 to 2.6.14-r4.
I had to go back to 2.6.12-r10 because I started getting the following error during boot:
| Code: |
Dec 13 18:56:19 kayak PCI: Failed to allocate mem resource #6:20000@f0000000 for 0000:01:00.0
|
After that started to happen (actually as early as trying to update to 2.6.13-r5) X would no longer start.
But I never had any problems with bootsplash. That still happens correctly with the 2.6.14-r4 kernel, but the memory allocation problem occurs and X bombs out saying that it cannot find a usable screen even though screens exist.
I even tried upgrading xorg-x11 to 6.8.99.15 to see if actually X had the problem and it still will not start.
Maybe I will try to remove bootsplash from the kernel and see what happens.
_________________
--
That's my story and I am sticking to it... |
|
| Back to top |
|
|
yamakawa
Guru
Joined: 28 Jul 2003
Posts: 340
|
| Posted: Fri Dec 23, 2005 5:37 pm Post subject: |
|
|
In my case, I fixed it by
| Code: | | # rc-update del coldplug boot (or maybe default) |
It seemed apparent to me that right after the bootstrap ran coldplug, the message on the screen went wrong (before launching X). |
|
| Back to top |
|
|
yamakawa
Guru
Joined: 28 Jul 2003
Posts: 340
|
| Posted: Fri Dec 23, 2005 8:50 pm Post subject: |
|
|
Man, after updating nvidia-kernel & glx to 1.0.8178, I'm f**ked up! 
The workround is here. |
|
| Back to top |
|
|
lorenct
n00b
Joined: 20 May 2004
Posts: 31
|
| Posted: Wed Dec 28, 2005 8:19 pm Post subject: |
|
|
Interestingly enough, I changed my grub.conf entry in /boot/grub/grub.conf to:
| Code: |
# Menu listing - GenToo Linux kernel-genkernel-x86-2.6.14-gentoo-r4
title GenToo Linux kernel-genkernel-x86-2.6.14-gentoo-r4
root (hd0,0)
kernel (hd0,0)/boot/kernel-genkernel-x86-2.6.14-gentoo-r4 root=/dev/i2o/hda3 init=/linuxrc CONSOLE=/dev/tty1 splash=silent udev
|
Originally I had:
| Code: |
# Menu listing - GenToo Linux kernel-genkernel-x86-2.6.14-gentoo-r4 [bootsplash]
title GenToo Linux kernel-genkernel-x86-2.6.14-gentoo-r4 [bootsplash]
root (hd0,0)
kernel (hd0,0)/boot/kernel-genkernel-x86-2.6.14-gentoo-r4 root=/dev/i2o/hda3 udev video=vesafb:1280x1024-32@75 splash=silent,theme:livecd-2005.0 quiet CONSOLE=/dev/tty1
initrd (hd0,0)/boot/splash-livecd-2005.0-1280x1024
|
Which included the bootsplash method for a nice graphical startup screen.
But removing it while still receiving the error message:
| Code: |
PCI: Failed to allocate mem resource #6:20000@f0000000 for 0000:01:00.0
|
in my /var/log/messages file, at least the system boots the 2.6.14-r4 kernel and X starts without a problem.
| Code: |
Linux kayak 2.6.14-gentoo-r4 #1 SMP PREEMPT Fri Dec 16 21:07:40 CST 2005 i686 Pentium III (Katmai) GenuineIntel GNU/Linux
|
Not sure if this is a bootsplash/kernel/X11 compatibility issue or not, but hopefully this will help anyone else who was stuck at 2.6.12-r10 and could not run X after boot with the newer 2.6.13-r5 or newer kernels.
_________________
--
That's my story and I am sticking to it... |
|
| Back to top |
|
|
weedy
Apprentice
Joined: 24 Jun 2005
Posts: 225
|
| Posted: Thu Dec 29, 2005 3:27 am Post subject: |
|
|
| Tweakin wrote: | | paul555 wrote: | | I had the same problem i solved it by emerging the version of the nvidia-kernel and nvidia-glx |
I gave that a try, it increased the error output to this...
| Code: | Error: API mismatch: the NVIDIA kernel module is version 7.1.0, but
this X module is version 1.0.7676. Please be sure that your kernel
module and all NVIDIA driver files have the same driver version.
(EE) NVIDIA(0): Failed to initialize the NVIDIA kernel module! Please ensure
(EE) NVIDIA(0): that there is a supported NVIDIA GPU in this system, and
(EE) NVIDIA(0): that the NVIDIA device files have been created properly.
(EE) NVIDIA(0): Please consult the NVIDIA README for details.
(EE) NVIDIA(0): *** Aborting ***
(EE) Screen(s) found, but none have a usable configuration. |
|
you didnt reload the nvidia module, i just had this error the other day when a emerge -avuDN world updated NVIDIA kernel & NVIDIA glx. in short you made 500x mroe work for your self. |
|
| Back to top |
|
|
Lunpa
Tux's lil' helper
Joined: 27 Jan 2004
Posts: 115
Location: mars
|
| Posted: Thu Dec 29, 2005 7:28 am Post subject: |
|
|
I've been having similar problems...
A while back, some hardware failed, and forced me to upgrade my processor, and so to avoid potential problems with a stage 1 install that had some unreasonable optimisation flags, and general curiosity, I thought I'd try out debian testing.
In debian, I found that X would freeze up, without any aparent pattern (I've observed this for 2 long months, over 3 total installs of debian testing/etch). So, over xmas, I thought I'd go back to gentoo, because it had always worked with the system, even with some wierd (and quite mis-guided) cflags.
The problems remained, and with some insight this hardware has given me, I've found a version that works rather well (eg, I can type this long winded anectdote with all the nvidia goodness I love and know).
Anyhow, to get to the point, I think there is a bad regression in the later nvidia drivers, atleast for users with two screens.
What I did was downgrade to my best guess of what was out when I had them working last (sometime late september 2005)...
and (for some reason) suprisingly, it works quite well!
heres what I'm using:
| Code: |
media-video/nvidia-kernel-1.0.7676-r1]
media-video/nvidia-glx-1.0.7676-r1
x11-base/xorg-x11-6.8.2-r4
|
note, that none of these are the latest and greatest, but they seem to work like a charm thus far.
I wish I knew what was wrong with it before with newer drivers, but I've been unable to pinpoint it.
Also, I'm using a geforce fx5200 and I don't use twinview mode with two monitors (just normal xinerama, or I'm pretty sure I'm using it).
I hope this helps anyone. I'll porbably take back much of what I've said in a few hours, depending on if the crashes are truely gone. in the mean time, its time to get the tv card working ^_^
_________________
I wanna learn how to play the accordian trumpet... |
|
| Back to top |
|
|
Lunpa
Tux's lil' helper
Joined: 27 Jan 2004
Posts: 115
Location: mars
|
| Posted: Fri Dec 30, 2005 8:46 am Post subject: |
|
|
yeah, so I figure everything I know is wrong now, as after 3 days it works damn nice, I reboot to make sure something seemingly unrelated is working ok (apache), and 2 hours later, x freezes to a halt, and all I can do is move my mouse, but only in one screen.
I've seen this before, in debian, so I kill X remotely from ssh to experiment, but unlike in debian, I can safely restart X again, without a reboot.
either way, what the hell? what if I was working on something?
_________________
I wanna learn how to play the accordian trumpet... |
|
| Back to top |
|
|
whtwtr
n00b
Joined: 12 Jun 2005
Posts: 53
Location: Munich
|
| Posted: Tue Jan 03, 2006 7:46 pm Post subject: Same issue here |
|
|
Hi guys, sorry to add to this long thread but I am having the same issue, just upgraded to 2.6.14-gentoo-r5 using genkernal as always. Then re-emerged nvidia-kernel and nvidia-glx. I am using the unmasked "latest" versions 1.0.8174, rebooted into my -r5 kernel, started X and ??? I get the dreaded API mismatch error.
So I tried everything that this great thread has suggested but I am still getting the error I have tried unmerging nvidia-* rebooting and re-emerging... still get the error??? Device nodes are present... I tried "eselect opengl set nvidia" returned with no error but still the API blah blah! opengl-update nvidia... nothing!
should I just commit myself to the local loonie bin, or can someone offer a little insight. 
Is there a way to really clean out the nvidia driver files so you a user can start from a clean slate? (without having to reinstall gentoo  )
any help is greatly appreciated...
Cheers!
BTW, my arch is amd64 |
|
| Back to top |
|
|
OPelerin
Apprentice
Joined: 17 Jul 2004
Posts: 287
Location: Belgium
|
| Posted: Wed Jan 04, 2006 10:58 pm Post subject: |
|
|
Still having same issue. Reverting to 2.6.12 now
_________________
Olivier PELERIN |
|
| Back to top |
|
|
val
Apprentice
Joined: 30 Jun 2004
Posts: 234
|
| Posted: Sat Jan 07, 2006 10:51 pm Post subject: |
|
|
| iMike wrote: | The suggestion to use
| Code: |
/sbin/NVmakedevices.sh
|
worked for me. Great! Thanks a lot. Also, thanks for the script,
Very convenient. |
Same! Thank you! In fact, the init script was the only way it would work. For some reason, once xorg fails to load for me, I can never successfully restart it - I have to reboot - which meant I needed the init script.
_________________
My log saw what happened to Laura Palmer. Go ahead. Ask it. |
|
| Back to top |
|
|
OPelerin
Apprentice
Joined: 17 Jul 2004
Posts: 287
Location: Belgium
|
| Posted: Tue Jan 10, 2006 10:14 pm Post subject: |
|
|
just emerged media-video/nvidia-kernel-1.0.8178-r1
This fixed my issue. No needs to use /etc/init.d/nvidia start or what so ever quick fixes. This works well.
I would encourage to upgrade to this version.
_________________
Olivier PELERIN |
|
| Back to top |
|
|
jamessnell
n00b
Joined: 24 Oct 2005
Posts: 19
Location: Calgary, Canada
|
| Posted: Fri Jan 20, 2006 5:27 am Post subject: |
|
|
| FranZ wrote: | Try | Code: | | /sbin/NVmakedevices.sh |
. It did the job for me.
Cheerio |
Excellent! Thank you!
- for some illusive reason, the card did initially work for me, then after I did something, I'm not sure what, I stopped working and I got the message: | Code: | | NVIDIA(0): Failed to initialize the NVIDIA kernel module! |
My xorg.conf never changed, infact, I think the only thing I changed was I made an rc script to automatically load the nvidia modules.
Anyway, whatever, I'm happy now as that fixed it for me too. Thank you! |
|
| Back to top |
|
|
KamikazeMicrowave
n00b
Joined: 20 Jan 2006
Posts: 71
|
| Posted: Thu Feb 23, 2006 3:21 am Post subject: nvidia wont create on boot |
|
|
| Like the subject says, I dont have any nvidia devices under /dev. I have to manually add it with the NVmakedevices.sh script. |
|
| Back to top |
|
|
avieth
Veteran
Joined: 17 Sep 2004
Posts: 1945
Location: Canada
|
| Posted: Thu Feb 23, 2006 3:49 am Post subject: |
|
|
EDIT: Woops, disregard this.
Last edited by avieth on Thu Feb 23, 2006 4:10 am; edited 1 time in total |
|
| Back to top |
|
|
yabbadabbadont
Advocate
Joined: 14 Mar 2003
Posts: 4790
Location: 2 exits past crazy
|
| Posted: Thu Feb 23, 2006 4:01 am Post subject: |
|
|
| Make sure that you have nvidia in your /etc/modules.autoload.d/kernel-2.6 file. Also, udev will not create the devices for the nvidia binary driver. Once you have created the devices using the script provided with the driver, make sure that you have RC_DEVICE_TARBALL="yes" in /etc/conf.d/rc. That way the dev directory will be saved and restored when rebooting. |
|
| Back to top |
|
|
curtis119
Bodhisattva
Joined: 10 Mar 2003
Posts: 2158
Location: Toledo, OH, USA, North America, Earth, SOL System, Milky Way, The Universe, The Cosmos, and Beyond.
|
|
| Back to top |
|
|
hcintra
n00b
Joined: 27 Jan 2005
Posts: 51
|
| Posted: Thu May 04, 2006 5:21 pm Post subject: Sugestion |
|
|
(I am cross posting this here as it was one of the topics that I reach after searching for: API mistacht nvidia
I had the API mismatch problem after upgrading nvidia-kernel and nvidia-glx. After lots of forum reading and trying some solutions (the above included), even resorting to downgrading nvidia to version 8178 once, I found out that my driver definition in the xorg.conf had been altered to "nvidia".
I am positive on the fact that I had set it to "nv" during my original installation of X. Go figure...
My sugestion: check your Graphic Devide Section on xorg.conf and make sure driver is set to "nv". |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
Desktop Environments
