4:06 a.m.
Hi,
I'm using F36 on a 10 year old laptop running an nividia GPU (GeForce
GT 560M). I'm using the nvidia drivers from rpmfusion (390xx).
Yesterday, I ran dnf upgrade and got the new 6.0.5 kernel. I lost the
native screen resolution (1920x1080) and have a 1024x768 instead.
I then removed the nvidia driver with dnf but it did not improve.
Maybe what is provided below may help. In particular, it seems that
nouveau is blacklisted but nvidia is not installed anymore...
Thanks for your help,
F
$ lspci |grep "VGA"
01:00.0 VGA compatible controller: NVIDIA Corporation GF116M [GeForce
GT 560M] (rev a1)
$ cat /proc/cmdline
BOOT_IMAGE=(hd0,msdos1)/vmlinuz-6.0.5-200.fc36.x86_64
root=UUID=4aab7630-84b2-4eae-9c5e-804ff8657b7f ro rhgb quiet
$ cat /etc/sysconfig/grub
GRUB_TIMEOUT=5
GRUB_DISTRIBUTOR="$(sed 's, release .*$,,g' /etc/system-release)"
GRUB_DEFAULT=saved
GRUB_DISABLE_SUBMENU=true
GRUB_TERMINAL_OUTPUT="console"
GRUB_CMDLINE_LINUX="rhgb quiet rd.driver.blacklist=nouveau
modprobe.blacklist=nouveau nvidia-drm.modeset=1"
GRUB_DISABLE_RECOVERY="true"
GRUB_ENABLE_BLSCFG=true
$ journalctl -b 0
oct. 31 09:28:49 vouise kernel: video: probe of PNP0A08:00 failed with error -22
oct. 31 09:28:49 vouise kernel: ACPI: video: [Firmware Bug]:
ACPI(MXM3) defines _DOD but not _DOS
oct. 31 09:28:49 vouise kernel: ACPI: video: Video Device [MXM3]
(multi-head: yes rom: no post: no)
oct. 31 09:28:49 vouise kernel: acpi device:1a: registered as cooling_device8
oct. 31 09:28:49 vouise kernel: input: Video Bus as
/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:18/LNXVIDEO:00/input/input18
5:37 a.m.
I tried to reinstall the nvidia drivers from rpmfusion:
$ sudo dnf install xorg-x11-drv-nvidia-390xx akmod-nvidia-390xx
I still have the same issue on the screen resolution and here is the
output of the commands:
$ lspci|grep VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GF116M [GeForce
GT 560M] (rev a1)
$ cat /proc/cmdline
BOOT_IMAGE=(hd0,msdos1)/vmlinuz-6.0.5-200.fc36.x86_64
root=UUID=4aab7630-84b2-4eae-9c5e-804ff8657b7f ro rhgb quiet
rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
nvidia-drm.modeset=1
$ cat /etc/sysconfig/grub
GRUB_TIMEOUT=5
GRUB_DISTRIBUTOR="$(sed 's, release .*$,,g' /etc/system-release)"
GRUB_DEFAULT=saved
GRUB_DISABLE_SUBMENU=true
GRUB_TERMINAL_OUTPUT="console"
GRUB_CMDLINE_LINUX="rd.driver.blacklist=nouveau
modprobe.blacklist=nouveau nvidia-drm.modeset=1 rhgb quiet
rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
nvidia-drm.modeset=1"
GRUB_DISABLE_RECOVERY="true"
GRUB_ENABLE_BLSCFG=true
9:53 a.m.
On Mon, 31 Oct 2022 11:37:43 +0100
Frédéric <ufospoke(a)gmail.com> wrote:
I tried to reinstall the nvidia drivers from rpmfusion:
$ sudo dnf install xorg-x11-drv-nvidia-390xx akmod-nvidia-390xx
I still have the same issue on the screen resolution and here is the
output of the commands:
$ lspci|grep VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GF116M [GeForce
GT 560M] (rev a1)
$ cat /proc/cmdline
BOOT_IMAGE=(hd0,msdos1)/vmlinuz-6.0.5-200.fc36.x86_64
root=UUID=4aab7630-84b2-4eae-9c5e-804ff8657b7f ro rhgb quiet
rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
nvidia-drm.modeset=1
$ cat /etc/sysconfig/grub
GRUB_TIMEOUT=5
GRUB_DISTRIBUTOR="$(sed 's, release .*$,,g' /etc/system-release)"
GRUB_DEFAULT=saved
GRUB_DISABLE_SUBMENU=true
GRUB_TERMINAL_OUTPUT="console"
GRUB_CMDLINE_LINUX="rd.driver.blacklist=nouveau
modprobe.blacklist=nouveau nvidia-drm.modeset=1 rhgb quiet
rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
nvidia-drm.modeset=1"
GRUB_DISABLE_RECOVERY="true"
GRUB_ENABLE_BLSCFG=true
Is there anything in the journal?
journalctl -r
to look at messages in reverse order from the current time or
journalctl -b 0
to pick up the last boot messages. If it is failing to set a
resolution that it previously had no trouble with, there should be some
kind of warning or error, I think.
If you remove the rpmfusion nvidia drivers, and then enable nouveau,
and reboot, does the resolution work.
It shouldn't matter, but are you booting into text or gui, and what is
your desktop?
I did a quick search on your device and linux kernel and didn't find
anything, so it shouldn't be related to the kernel update. But, what
happens if you boot an older kernel?
11:06 a.m.
On 31/10/2022 14:53, stan via users wrote:
On Mon, 31 Oct 2022 11:37:43 +0100
Frédéric <ufospoke(a)gmail.com> wrote:
> I tried to reinstall the nvidia drivers from rpmfusion:
> $ sudo dnf install xorg-x11-drv-nvidia-390xx akmod-nvidia-390xx
>
> I still have the same issue on the screen resolution and here is the
> output of the commands:
>
> $ lspci|grep VGA
> 01:00.0 VGA compatible controller: NVIDIA Corporation GF116M [GeForce
> GT 560M] (rev a1)
>
> $ cat /proc/cmdline
> BOOT_IMAGE=(hd0,msdos1)/vmlinuz-6.0.5-200.fc36.x86_64
> root=UUID=4aab7630-84b2-4eae-9c5e-804ff8657b7f ro rhgb quiet
> rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
> nvidia-drm.modeset=1
>
> $ cat /etc/sysconfig/grub
> GRUB_TIMEOUT=5
> GRUB_DISTRIBUTOR="$(sed 's, release .*$,,g' /etc/system-release)"
> GRUB_DEFAULT=saved
> GRUB_DISABLE_SUBMENU=true
> GRUB_TERMINAL_OUTPUT="console"
> GRUB_CMDLINE_LINUX="rd.driver.blacklist=nouveau
> modprobe.blacklist=nouveau nvidia-drm.modeset=1 rhgb quiet
> rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
> nvidia-drm.modeset=1"
> GRUB_DISABLE_RECOVERY="true"
>
>
>
> GRUB_ENABLE_BLSCFG=true
Is there anything in the journal?
journalctl -r
to look at messages in reverse order from the current time or
journalctl -b 0
to pick up the last boot messages. If it is failing to set a
resolution that it previously had no trouble with, there should be some
kind of warning or error, I think.
If you remove the rpmfusion nvidia drivers, and then enable nouveau,
and reboot, does the resolution work.
It shouldn't matter, but are you booting into text or gui, and what is
your desktop?
I did a quick search on your device and linux kernel and didn't find
anything, so it shouldn't be related to the kernel update. But, what
happens if you boot an older kernel?
I'm on F36 with the rpmfusion 470xx driver, and saw the same lower
screen resolution on reboot with the 6.0.5 kernel. The previous 5.19.16
kernel still booted as before.
A new version of 470xx in rpmfusion nonfree testing was announced around
midnight and I now have its kmod and akmod installed. atop showed
several cycles of akmods activity. After they appeared to have
finished, a 'systemctl reboot' has delivered a system that seems to be
ok.
There were reports from dracut-shutdown of 'failure to perform cleanup',
and akmods-shutdown.service took longer than might have been expected
given the earlier akmods activity.
Wait for a 390xx update?
John P
12:28 p.m.
I'm on F36 with the rpmfusion 470xx driver, and saw the same
lower
screen resolution on reboot with the 6.0.5 kernel. The previous 5.19.16
kernel still booted as before.
A new version of 470xx in rpmfusion nonfree testing was announced around
midnight and I now have its kmod and akmod installed. atop showed
several cycles of akmods activity. After they appeared to have
finished, a 'systemctl reboot' has delivered a system that seems to be
ok.
Wait for a 390xx update?
That would be nice but why doesn't it work with nouveau?
Thanks,
F
12:26 p.m.
Thanks for your message.
Is there anything in the journal?
nothing that seems related to the issue. If you look at the first
message of this thread, I posted en excerpt from the journal.
If you remove the rpmfusion nvidia drivers, and then enable nouveau,
and reboot, does the resolution work.
That's what I did but it didn't work.
It shouldn't matter, but are you booting into text or gui, and
what is
your desktop?
I'm using plasma but it fails already in kdm.
I did a quick search on your device and linux kernel and didn't
find
anything, so it shouldn't be related to the kernel update. But, what
happens if you boot an older kernel?
now it also fails with old kernel...
Thanks,
F
5:29 p.m.
On 1/11/22 04:26, Frédéric wrote:
Thanks for your message.
> Is there anything in the journal?
nothing that seems related to the issue. If you look at the first
message of this thread, I posted en excerpt from the journal.
> If you remove the rpmfusion nvidia drivers, and then enable nouveau,
> and reboot, does the resolution work.
That's what I did but it didn't work.
> It shouldn't matter, but are you booting into text or gui, and what is
> your desktop?
I'm using plasma but it fails already in kdm.
> I did a quick search on your device and linux kernel and didn't find
> anything, so it shouldn't be related to the kernel update. But, what
> happens if you boot an older kernel?
now it also fails with old kernel...
Thanks,
You might be best waiting for a 370 upgrade as John suggested. I have
akmod-nvidia-520 installed and kmod-nvidia-520 and kmod-nvidia-6.0.5
installed having had the 6.0.5 kernel installed several days ago and my
system is running running kde at the set 3840x2160 resolution. When I
had issues with the nvidia driver loading because of not being
secureboot signed I was getting a message displayed that the nvidia
driver couldn't be found and it was defaulting back to the nouveau
driver even though the nouveau driver was blacklisted in
/etc/default/grub. After uninstalling the nvidia drivers, correcting the
secureboot signing issue, and reinstalling akmod-nvidia and kmod-nvidia
my system is still running at 3840x2160 resolution with the nvidia
drivers and the install of those drivers has not blacklisted nouveau in
/etc/default/grub.
regards,
Steve
F
_______________________________________________
users mailing list -- users(a)lists.fedoraproject.org
To unsubscribe send an email to users-leave(a)lists.fedoraproject.org
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
4:51 p.m.
On 31 Oct 2022, at 09:06, Frédéric <ufospoke(a)gmail.com> wrote:
Hi,
I'm using F36 on a 10 year old laptop running an nividia GPU (GeForce
GT 560M). I'm using the nvidia drivers from rpmfusion (390xx).
Yesterday, I ran dnf upgrade and got the new 6.0.5 kernel. I lost the
native screen resolution (1920x1080) and have a 1024x768 instead.
Did you try rebooting and using the previous kernel?
I then removed the nvidia driver with dnf but it did not improve.
It's not that simple as the rpmfusion drivers install source code that is built
by akmod. akmod makes more RPMs that are have the binary .ko files in them
that is installed on first boot with the new kernel.
I found that to wipe the nvidia driver and then reinstall it takes a number of steps.
This is from memory, so check they make sense on your system.
I have all these RPMs installed to support my RX3060:
$ rpm -qa | grep nvidia
xorg-x11-drv-nvidia-cuda-libs-520.56.06-1.fc36.x86_64
nvidia-gpu-firmware-20221012-141.fc36.noarch
nvidia-persistenced-520.56.06-1.fc36.x86_64
xorg-x11-drv-nvidia-kmodsrc-520.56.06-1.fc36.x86_64
xorg-x11-drv-nvidia-libs-520.56.06-1.fc36.i686
xorg-x11-drv-nvidia-libs-520.56.06-1.fc36.x86_64
akmod-nvidia-520.56.06-1.fc36.x86_64
nvidia-settings-520.56.06-1.fc36.x86_64
xorg-x11-drv-nvidia-520.56.06-1.fc36.x86_64
xorg-x11-drv-nvidia-power-520.56.06-1.fc36.x86_64
xorg-x11-drv-nvidia-cuda-libs-520.56.06-1.fc36.i686
xorg-x11-drv-nvidia-cuda-520.56.06-1.fc36.x86_64
kmod-nvidia-5.19.16-200.fc36.x86_64-520.56.06-1.fc36.x86_64
kmod-nvidia-5.19.14-200.fc36.x86_64-520.56.06-1.fc36.x86_64
And
$ rpm -qa |grep akmod
akmods-0.5.7-8.fc36.noarch
akmod-nvidia-520.56.06-1.fc36.x86_64
I removed all of them when I get into any issues.
I also rm -rf /var/cache/akmods to remove any of the built intermediates.
Then I installed fresh.
Make sure that any akmods services have completed.
Then reboot and wait for any akmods builds to finish and install the new nvidia drivers.
This can take a while - ssh in from another system if you are curious what is going on.
Barry
Maybe what is provided below may help. In particular, it seems that
nouveau is blacklisted but nvidia is not installed anymore...
Thanks for your help,
F
$ lspci |grep "VGA"
01:00.0 VGA compatible controller: NVIDIA Corporation GF116M [GeForce
GT 560M] (rev a1)
$ cat /proc/cmdline
BOOT_IMAGE=(hd0,msdos1)/vmlinuz-6.0.5-200.fc36.x86_64
root=UUID=4aab7630-84b2-4eae-9c5e-804ff8657b7f ro rhgb quiet
$ cat /etc/sysconfig/grub
GRUB_TIMEOUT=5
GRUB_DISTRIBUTOR="$(sed 's, release .*$,,g' /etc/system-release)"
GRUB_DEFAULT=saved
GRUB_DISABLE_SUBMENU=true
GRUB_TERMINAL_OUTPUT="console"
GRUB_CMDLINE_LINUX="rhgb quiet rd.driver.blacklist=nouveau
modprobe.blacklist=nouveau nvidia-drm.modeset=1"
GRUB_DISABLE_RECOVERY="true"
GRUB_ENABLE_BLSCFG=true
$ journalctl -b 0
oct. 31 09:28:49 vouise kernel: video: probe of PNP0A08:00 failed with error -22
oct. 31 09:28:49 vouise kernel: ACPI: video: [Firmware Bug]:
ACPI(MXM3) defines _DOD but not _DOS
oct. 31 09:28:49 vouise kernel: ACPI: video: Video Device [MXM3]
(multi-head: yes rom: no post: no)
oct. 31 09:28:49 vouise kernel: acpi device:1a: registered as cooling_device8
oct. 31 09:28:49 vouise kernel: input: Video Bus as
/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:18/LNXVIDEO:00/input/input18
_______________________________________________
users mailing list -- users(a)lists.fedoraproject.org
To unsubscribe send an email to users-leave(a)lists.fedoraproject.org
Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Do not reply to spam, report it: https://pagure.io/fedora-infrastructure/new_issue
2:23 a.m.
On 1/11/22 09:37, Joe Zeff wrote:
On 10/31/2022 03:51 PM, Barry Scott wrote:
> I found that to wipe the nvidia driver and then reinstall it takes a
> number of steps.
Removing them all can be done in one step:
sudo dnf remove *nvidia*
I've also found that sudo dnf remove akmod-nvidia
kmod-nvidia also
removes all the nvidia files and all the associated dependencies.
regards,
Steve
HTH, HAND.
_______________________________________________
users mailing list -- users(a)lists.fedoraproject.org
To unsubscribe send an email to users-leave(a)lists.fedoraproject.org
Fedora Code of Conduct:
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives:
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.org
Do not reply to spam, report it:
https://pagure.io/fedora-infrastructure/new_issue
3:55 a.m.
On 01/11/2022 07:23, Stephen Morris wrote:
On 1/11/22 09:37, Joe Zeff wrote:
> On 10/31/2022 03:51 PM, Barry Scott wrote:
>> I found that to wipe the nvidia driver and then reinstall it takes a
>> number of steps.
>
> Removing them all can be done in one step:
>
> sudo dnf remove *nvidia*
I've also found that sudo dnf remove akmod-nvidia kmod-nvidia also
removes all the nvidia files and all the associated dependencies.
regards,
Steve
>
> HTH, HAND.
I haven't wanted to use nouveau recently, although IIRC it will now be
activated as a fallback. But in the past I had the impression that only
the last-installed driver would behave properly. If you want nouveau,
try reinstalling it.
John P
2:12 p.m.
Thank you all for your answers. However, I completely removed the
nvidia driver so I guess this can be excluded from the issue.
I still have the wrong screen resolution (1024x768) instead of full HD
(1920x1080) and do not know what to do. I update the system everyday
but nothing changes.
I don't know what to do. If anyone has an idea, that would help.
Thanks,
F
2:47 p.m.
On 11/6/22 12:12, Frédéric wrote:
Thank you all for your answers. However, I completely removed the
nvidia driver so I guess this can be excluded from the issue.
I still have the wrong screen resolution (1024x768) instead of full HD
(1920x1080) and do not know what to do. I update the system everyday
but nothing changes.
I don't know what to do. If anyone has an idea, that would help.
The nouveau driver might not be able to support your device properly.
You most likely do need the proprietary driver installed.
3:15 p.m.
Frédéric composed on 2022-11-06 21:12 (UTC+0100):
Thank you all for your answers. However, I completely removed the
nvidia driver so I guess this can be excluded from the issue.
I still have the wrong screen resolution (1024x768) instead of full HD
(1920x1080) and do not know what to do. I update the system everyday
but nothing changes.
I don't know what to do. If anyone has an idea, that would help.
Completely purging NVidia is easier said than done. How do you know you succeeded?
Did your rebuild your initrds after removing NVidia's drivers?
Is nvidia-gpu-firmware installed?
Please fpaste /var/log/Xorg.0.log and give us the resulting URI!
grep veau /etc/modprobe.d/* should produce null output.
lsmod | grep veau should resemble:
nouveau 2416640 2
mxm_wmi 16384 1 nouveau
video 61440 1 nouveau
drm_ttm_helper 16384 1 nouveau
ttm 90112 2 drm_ttm_helper,nouveau
drm_display_helper 172032 1 nouveau
wmi 32768 3 wmi_bmof,mxm_wmi,nouveau
inxi -GSaz output, as I included in a reply here a week ago, can also be quite
informative:
<https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
Mesa is also important:
# rpm -qa | grep mesa
mesa-libGLU-9.0.1-6.fc36.x86_64
mesa-libglapi-22.1.7-1.fc36.x86_64
mesa-libgbm-22.1.7-1.fc36.x86_64
mesa-libGL-22.1.7-1.fc36.x86_64
mesa-libEGL-22.1.7-1.fc36.x86_64
mesa-filesystem-22.1.7-1.fc36.x86_64
mesa-dri-drivers-22.1.7-1.fc36.x86_64
mesa-vulkan-drivers-22.1.7-1.fc36.x86_64
--
Evolution as taught in public schools is, like religion,
based on faith, not based on science.
Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!
Felix Miata
11:54 p.m.
Hi,
Thanks for your answer.
Completely purging NVidia is easier said than done. How do you know
you succeeded?
I do not know. I just removed all packages containing nvidia in their
name with dnf remove "*nvidia*"
Did your rebuild your initrds after removing NVidia's drivers?
I do not know what that means.
Is nvidia-gpu-firmware installed?
no:
$ dnf list --installed|grep nvidia
does not return anything
Please fpaste /var/log/Xorg.0.log and give us the resulting URI!
https://paste.centos.org/view/2af15415
grep veau /etc/modprobe.d/* should produce null output.
$ grep veau /etc/modprobe.d/*
/etc/modprobe.d/blacklist-nvidia-nouveau.conf:blacklist nouveau
/etc/modprobe.d/blacklist-nvidia-nouveau.conf:options nouveau modeset=0
This seems wrong, no?
$ lsmod | grep veau
does not return anything
inxi -GSaz output, as I included in a reply here a week ago, can also
be quite
informative:
<https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
It was written pinxi and I couldn't find it :=) Thanks for rewritting
it. inxi, I have it.
# inxi -C --vs
inxi 3.3.22-00 (2022-10-08)
CPU:
Info: quad core model: Intel Core i7-2760QM bits: 64 type: MT MCP cache:
L2: 1024 KiB
Speed (MHz): avg: 907 min/max: 800/3500 cores: 1: 1661 2: 800 3: 800
4: 798 5: 800 6: 798 7: 800 8: 800
# inxi -GSaz --zl --hostname
System:
Host: vouise Kernel: 6.0.5-200.fc36.x86_64 arch: x86_64 bits: 64
compiler: gcc v: 2.37-36.fc36
parameters: BOOT_IMAGE=(hd0,msdos1)/vmlinuz-6.0.5-200.fc36.x86_64
root=UUID=4aab7630-84b2-4eae-9c5e-804ff8657b7f ro rhgb quiet
Console: pty pts/2 wm: kwin_x11 DM: 1: KDM 2: XDM Distro: Fedora release
36 (Thirty Six)
Graphics:
Device-1: NVIDIA GF116M [GeForce GT 560M] vendor: CLEVO/KAPOK driver: N/A
alternate: nouveau non-free: series: 390.xx+ status: legacy-active
(EOL~late 2022) arch: Fermi code: GF1xx process: 40/28nm built: 2010-16
pcie: gen: 1 speed: 2.5 GT/s lanes: 16 bus-ID: 01:00.0 chip-ID: 10de:1251
class-ID: 0300
Display: x11 server: X.Org v: 1.20.14 with: Xwayland v: 22.1.3
compositor: kwin_x11 driver: X: loaded: nouveau,vesa
unloaded: fbdev,modesetting alternate: nv dri: swrast gpu: N/A
display-ID: :0 screens: 1
Screen-1: 0 s-res: 1024x768 s-dpi: 96 s-size: 271x203mm (10.67x7.99")
s-diag: 339mm (13.33")
Monitor-1: default res: 1024x768 size: N/A modes: N/A
OpenGL: renderer: llvmpipe (LLVM 14.0.0 256 bits) v: 4.5 Mesa 22.1.7
direct render: Yes
# rpm -qa|grep mesa
mesa-libGLU-9.0.1-6.fc36.x86_64
mesa-libGLU-devel-9.0.1-6.fc36.x86_64
mesa-libGLU-9.0.1-6.fc36.i686
mesa-libgbm-22.1.7-1.fc36.x86_64
mesa-libglapi-22.1.7-1.fc36.x86_64
mesa-libEGL-22.1.7-1.fc36.x86_64
mesa-libGL-22.1.7-1.fc36.x86_64
mesa-filesystem-22.1.7-1.fc36.x86_64
mesa-dri-drivers-22.1.7-1.fc36.x86_64
mesa-filesystem-22.1.7-1.fc36.i686
mesa-libEGL-devel-22.1.7-1.fc36.x86_64
mesa-libOpenCL-22.1.7-1.fc36.x86_64
mesa-libGL-devel-22.1.7-1.fc36.x86_64
mesa-libOSMesa-22.1.7-1.fc36.x86_64
mesa-libxatracker-22.1.7-1.fc36.x86_64
mesa-vulkan-drivers-22.1.7-1.fc36.x86_64
mesa-libglapi-22.1.7-1.fc36.i686
mesa-libGL-22.1.7-1.fc36.i686
mesa-libgbm-22.1.7-1.fc36.i686
mesa-libEGL-22.1.7-1.fc36.i686
mesa-dri-drivers-22.1.7-1.fc36.i686
mesa-libOSMesa-22.1.7-1.fc36.i686
mesa-vulkan-drivers-22.1.7-1.fc36.i686
thanks a lot!
F
12:25 a.m.
Frédéric composed on 2022-11-07 06:54 (UTC+0100):
> Completely purging NVidia is easier said than done. How do you
know you succeeded?
I do not know. I just removed all packages containing nvidia in
their
name with dnf remove "*nvidia*"
rpm/dnf only remove packages. Their post scripts usually are incapable of
completing a purge. The NVidia packages are supposed to be provided with
installation instructions that detail how to completely purge. When those are not
found and followed, what's been happening to you is expected.
> Did your rebuild your initrds after removing NVidia's
drivers?
I do not know what that means.
Boot initialization involves loading a kernel and initrd (initramfs) via Grub.
Initrds are custom generated normally with installation of every kernel. Adding
proprietary drivers such as NVidia's usually requires that they be re-generated.
These are two of my currently installed ones:
# ls -Ggh /boot/ini*5.1*
-rw------- 1 15M Sep 11 01:43 /boot/initramfs-5.18.19-200.fc36.x86_64.img
-rw------- 1 15M Oct 17 23:26 /boot/initramfs-5.19.15-201.fc36.x86_64.img
> Is nvidia-gpu-firmware installed?
no:
$ dnf list --installed|grep nvidia
does not return anything
I don't know whether your GPU requires it, so go ahead and install it as last step
after completing other instructions here ready to reboot to test. IIRC, this will
cause regeneration of at least your current initrd, if not all, so that you need
not do it manually. If it doesn't happen automatically, then run:
dracut -f
> Please fpaste /var/log/Xorg.0.log and give us the resulting URI!
Among other problems noted there:
(EE) open /dev/dri/card0: No such file or directory
This is a typical result of failure to load a required kernel module, in your
case, nouveau, which provides KMS needed for X to run competently on NVidia GPUs
using only FOSS software.
> grep veau /etc/modprobe.d/* should produce null output.
$ grep veau /etc/modprobe.d/*
/etc/modprobe.d/blacklist-nvidia-nouveau.conf:blacklist nouveau
/etc/modprobe.d/blacklist-nvidia-nouveau.conf:options nouveau modeset=0
This seems wrong, no?
This is wrong, the foundational and possibly only problem. Remove both files. They
prevent loading of the necessary nouveau kernel module.
$ lsmod | grep veau
does not return anything
Because nouveau could not be loaded.
> inxi -GSaz output, as I included in a reply here a week ago, can
also be quite
> informative:
>
<https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
It was written pinxi and I couldn't find it :=) Thanks for
rewritting
it. inxi, I have it.
pinxi is the devel version of inxi.
--
Evolution as taught in public schools is, like religion,
based on faith, not based on science.
Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!
Felix Miata
6:26 a.m.
> $ grep veau /etc/modprobe.d/*
> /etc/modprobe.d/blacklist-nvidia-nouveau.conf:blacklist nouveau
> /etc/modprobe.d/blacklist-nvidia-nouveau.conf:options nouveau modeset=0
> This seems wrong, no?
This is wrong, the foundational and possibly only problem. Remove both files. They
prevent loading of the necessary nouveau kernel module.
removing this file alone fixed the issue after reboot!
Thank you so much.
Afterwards, I installed package nvidia-gpu-firmware but it did not
change anything. Maybe, as you suggest, it is not needed for my GPU.
Thanks to you, I was able to leave the PC in better shape to my
children just before leaving home for a business trip.
I will then have to try to reinstall the nvidia proprietary driver as
it is much more efficient to play Minecraft (my son). I'll try that on
Friday.
Thanks
F
12:48 a.m.
Frédéric composed on 2022-11-07 06:54 (UTC+0100):
...
Graphics:
Device-1: NVIDIA GF116M [GeForce GT 560M] vendor: CLEVO/KAPOK driver: N/A
alternate: nouveau non-free: series: 390.xx+ status: legacy-active
(EOL~late 2022) arch: Fermi
"EOL" suggests the applicable 390.xx
driver has had, or is about to have, support
terminated for Fermi architecture (GF116M), possibly as of the 6.0 kernel.
--
Evolution as taught in public schools is, like religion,
based on faith, not based on science.
Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!
Felix Miata
4:42 p.m.
On 06/11/2022 20:12, Frédéric wrote:
Thank you all for your answers. However, I completely removed the
nvidia driver so I guess this can be excluded from the issue.
I still have the wrong screen resolution (1024x768) instead of full HD
(1920x1080) and do not know what to do. I update the system everyday
but nothing changes.
I don't know what to do. If anyone has an idea, that would help.
Thanks,
F
I don't use the 390xx driver, but try this:
sudo dnf --enablerepo=rpmfusion-nonfree-updates-testing \
install akmod-nvidia-390xx xorg-x11-drv-nvidia-390xx \
xorg-x11-drv-nvidia-drv-390xx
HTH
John P
4:58 p.m.
On 06/11/2022 22:42, John Pilkington wrote:
On 06/11/2022 20:12, Frédéric wrote:
> Thank you all for your answers. However, I completely removed the
> nvidia driver so I guess this can be excluded from the issue.
> I still have the wrong screen resolution (1024x768) instead of full HD
> (1920x1080) and do not know what to do. I update the system everyday
> but nothing changes.
> I don't know what to do. If anyone has an idea, that would help.
> Thanks,
> F
I don't use the 390xx driver, but try this:
sudo dnf --enablerepo=rpmfusion-nonfree-updates-testing \
install akmod-nvidia-390xx xorg-x11-drv-nvidia-390xx \
xorg-x11-drv-nvidia-drv-390xx
Sorry, a typo. The last one, which you may not need, should have been
xorg-x11-drv-nvidia-cuda-390xx
HTH
John P
12:25 a.m.
I don't use the 390xx driver, but try this:
sudo dnf --enablerepo=rpmfusion-nonfree-updates-testing \
install akmod-nvidia-390xx xorg-x11-drv-nvidia-390xx \
xorg-x11-drv-nvidia-cuda-390xx
That's the driver I normally use but I tried multiple times to install
and uninstall without any success.
Thanks
F
1:12 a.m.
Hi
On Mon, 07 Nov 2022 07:25:18 +0100 Fr��d��ric wrote:
> I don't use the 390xx driver, but try this:
> sudo dnf --enablerepo=rpmfusion-nonfree-updates-testing \
> install akmod-nvidia-390xx xorg-x11-drv-nvidia-390xx \
> xorg-x11-drv-nvidia-cuda-390xx
That's the driver I normally use but I tried multiple times to
install
and uninstall without any success.
As John Pilkington said in another thread (monitor/display problem):
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
390xx went into the testing repo on 3 Nov
dnf confirm:
dnf --enablerepo=rpmfusion-nonfree-updates-testing repoquery --changelog
akmod-nvidia-390xx
...
Changelog for akmod-nvidia-390xx-3:390.154-3.fc36.x86_64
* Thu Nov 03 2022 Leigh Scott <leigh123linux(a)gmail.com> - 3:390.154-3
- Patch for 6.0 kernel
- Clean up old patches
Have you tried this 390.154-3 version ?
--
francis
6:32 a.m.
As John Pilkington said in another thread (monitor/display problem):
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
390xx went into the testing repo on 3 Nov
Have you tried this 390.154-3 version ?
Not yet. I have never used testing. Thank you for your explaination on
how to do it. If it is still in testing on Friday when I go back home,
I will test it. Otherwise, the normal update may be enough.
Thanks
F
4:34 a.m.
On 07/11/2022 12:32, Frédéric wrote:
> As John Pilkington said in another thread (monitor/display
problem):
>
>
https://lists.fedoraproject.org/archives/list/users@lists.fedoraproject.o...
>
> 390xx went into the testing repo on 3 Nov
>
> Have you tried this 390.154-3 version ?
Not yet. I have never used testing. Thank you for your explaination on
how to do it. If it is still in testing on Friday when I go back home,
I will test it. Otherwise, the normal update may be enough.
Thanks
F
In fact I did my installation of the 470xx update from 'testing' using
the dnfdragora GUI, not from the command line; but by Friday the move
to the 'stable' repo will probably have happened.
John
5:17 a.m.
Hi,
Back home. This morning, I just upgraded my system with dnf upgrade
(so I still did not installed the nvidia proprietary driver). I
obtained the new kerner 6.0.7. I rebooted and here am I again with my
wrong screen resolution...
# grep veau /etc/modprobe.d/*
returns nothing
I rebuilt initrds although the file was already there for 6.0.7:
# dracut -f
# lsmod | grep veau
returns nothing
# inxi -GSaz --zl --hostname
https://paste.centos.org/view/0496bee0
# inxi -C --vs
https://paste.centos.org/view/97d00d79
/var/log/Xorg.0.log:
https://paste.centos.org/view/355ead1b
I see strange things:
[ 23.494] (EE) Failed to load module "nv" (module does not exist, 0)
[ 23.628] (EE) [drm] Failed to open DRM device for pci:0000:01:00.0: -19
[ 23.628] (EE) open /dev/dri/card0: No such file or directory
[ 23.628] (WW) Falling back to old probe method for modesetting
[ 23.628] (EE) open /dev/dri/card0: No such file or directory
[ 23.634] (EE) open /dev/fb0: No such file or directory
[ 23.634] (EE) Screen 0 deleted because of no matching config section.
[ 23.634] (II) UnloadModule: "modesetting"
[ 23.634] (EE) Screen 0 deleted because of no matching config section.
but also
[ 23.492] (II) LoadModule: "nouveau"
[ 23.492] (II) Loading /usr/lib64/xorg/modules/drivers/nouveau_drv.so
which seems to indicate that nouveau is used.
Thanks for your help!
F
6:03 a.m.
On 11/11/2022 11:17, Frédéric wrote:
Hi,
Back home. This morning, I just upgraded my system with dnf upgrade
(so I still did not installed the nvidia proprietary driver). I
obtained the new kerner 6.0.7. I rebooted and here am I again with my
wrong screen resolution...
# grep veau /etc/modprobe.d/*
returns nothing
I rebuilt initrds although the file was already there for 6.0.7:
# dracut -f
# lsmod | grep veau
returns nothing
# inxi -GSaz --zl --hostname
https://paste.centos.org/view/0496bee0
# inxi -C --vs
https://paste.centos.org/view/97d00d79
/var/log/Xorg.0.log:
https://paste.centos.org/view/355ead1b
I see strange things:
[ 23.494] (EE) Failed to load module "nv" (module does not exist, 0)
[ 23.628] (EE) [drm] Failed to open DRM device for pci:0000:01:00.0: -19
[ 23.628] (EE) open /dev/dri/card0: No such file or directory
[ 23.628] (WW) Falling back to old probe method for modesetting
[ 23.628] (EE) open /dev/dri/card0: No such file or directory
[ 23.634] (EE) open /dev/fb0: No such file or directory
[ 23.634] (EE) Screen 0 deleted because of no matching config section.
[ 23.634] (II) UnloadModule: "modesetting"
[ 23.634] (EE) Screen 0 deleted because of no matching config section.
but also
[ 23.492] (II) LoadModule: "nouveau"
[ 23.492] (II) Loading /usr/lib64/xorg/modules/drivers/nouveau_drv.so
which seems to indicate that nouveau is used.
Thanks for your help!
F
Today's email from rpmfusion does not show any pushes to the nonfree
stable repo. I suggest:
Boot the 5.16 kernel and install akmod470xx
or reinstall nouveau
or dnf --enablerepo epmfusion-nonfree-updates-testing install akmod470xx
or wait until the push-to-stable happens and update
These are ideas, probably not the exact commands.
John
9:12 a.m.
Frédéric composed on 2022-11-11 12:17 (UTC+0100):
# inxi -GSaz --zl --hostname
https://paste.centos.org/view/0496bee0
from above:
rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
causes all the following fatal errors:
> # lsmod | grep veau
returns nothing
...
[ 23.628] (EE) [drm] Failed to open DRM device for
pci:0000:01:00.0: -19
[ 23.628] (EE) open /dev/dri/card0: No such file or directory
[ 23.628] (EE) open /dev/dri/card0: No such file or directory
[ 23.634] (EE) open /dev/fb0: No such file or directory
[ 23.634] (II) UnloadModule: "modesetting"
Those kernel command line options need to be removed from /etc/default/grub
and /boot/grub2/grub.cfg regenerated to exclude them.
nvidia-drm.modeset=1
Is probably unnecessary. You might be best off removing it too. None of my
NVidia GPUs require it.
To test, strike the E key at the Grub menu and remove them all before proceeding
with boot.
--
Evolution as taught in public schools is, like religion,
based on faith, not based on science.
Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!
Felix Miata
3:01 p.m.
Hi,
rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
causes all the following fatal errors:
Those kernel command line options need to be removed from /etc/default/grub
and /boot/grub2/grub.cfg regenerated to exclude them.
You are right: I edited with e and removed those 2 commands and it works.
So I edited /etc/default/grub, removed the commands ran grub2-mkconfig
-o /boot/grub2/grub.cfg.
And it works!
nvidia-drm.modeset=1
Is probably unnecessary. You might be best off removing it too. None of my
NVidia GPUs require it.
You're right, I also removed it.
After that, I was able to reinstall the nvidia proprietary driver
(390xx) and again, it works!
Thanks a lot!
F
6:13 p.m.
Frédéric composed on 2022-10-31 10:06 (UTC+0100):
I'm using F36 on a 10 year old laptop running an nividia GPU
(GeForce
GT 560M). I'm using the nvidia drivers from rpmfusion (390xx).
Yesterday, I ran dnf upgrade and got the new 6.0.5 kernel. I lost
the
native screen resolution (1920x1080) and have a 1024x768 instead.
I then removed the nvidia driver with dnf but it did not improve.
I don't think NVidia drivers can be eradicated so easily. NVidia provides
instructions for their installation, which include instructions for their removal.
If removal instructions aren't followed, the library tainting they cause and/or
.conf files they create can remain, thus thwarting any attempt to use any
competent FOSS driver. This is an example of what should be seen running F36 on an
old NVidia GPU using only FOSS:
# pinxi -C --vs
pinxi 3.3.22-13 (2022-10-27)
CPU:
Info: quad core model: AMD Phenom II X4 965 bits: 64 type: MCP cache:
L2: 2 MiB
Speed (MHz): avg: 3423 min/max: N/A cores: 1: 3423 2: 3423 3: 3423 4: 3423
# pinxi -GSaz --zl --hostname
System:
Host: gb970 Kernel: 5.19.17-200.fc36.x86_64 arch: x86_64 bits: 64
compiler: gcc v: 2.37-36.fc36 parameters: ro root=LABEL=<filter> audit=0
ipv6.disable=1 net.ifnames=0 audit=0 plymouth.enable=0 noresume
mitigations=auto consoleblank=0 selinux=0 vga=791 video=1440x900@60 5
Desktop: Trinity v: R14.0.12 tk: Qt v: 3.5.0 info: kicker wm: Twin v: 3.0
vt: 7 dm: 1: TDM 2: XDM Distro: Fedora release 36 (Thirty Six)
Graphics:
Device-1: NVIDIA GF108 [GeForce GT 630] vendor: Gigabyte driver: nouveau
v: kernel non-free: series: 390.xx+ status: legacy-active (EOL~late 2022)
arch: Fermi code: GF1xx process: 40/28nm built: 2010-16 pcie: gen: 1
speed: 2.5 GT/s lanes: 16 ports: active: DVI-I-1,HDMI-A-1 empty: VGA-1
bus-ID: 01:00.0 chip-ID: 10de:0f00 class-ID: 0300 temp: 50.0 C
Display: x11 server: X.Org v: 1.20.14 driver: X: loaded: modesetting
unloaded: fbdev,vesa dri: nouveau gpu: nouveau display-ID: :0 screens: 1
Screen-1: 0 s-res: 3600x1200 s-dpi: 120 s-size: 761x253mm (29.96x9.96")
s-diag: 802mm (31.57")
Monitor-1: DVI-I-1 pos: primary,left model: NEC EA243WM serial: <filter>
built: 2011 res: 1920x1200 hz: 60 dpi: 94 gamma: 1.2
size: 519x324mm (20.43x12.76") diag: 612mm (24.1") ratio: 16:10 modes:
max: 1920x1200 min: 640x480
Monitor-2: HDMI-A-1 mapped: HDMI-1 pos: right model: Dell P2213
serial: <filter> built: 2012 res: 1680x1050 hz: 60 dpi: 90 gamma: 1.2
size: 473x296mm (18.62x11.65") diag: 558mm (22") ratio: 16:10 modes:
max: 1680x1050 min: 720x400
API: OpenGL v: 4.3 Mesa 22.1.7 renderer: NVC1 direct render: Yes
Note that the X display driver is modesetting, not the reverse-engineered,
"experimental", optional nouveau provided by xorg-x11-drv-nouveau.
--
Evolution as taught in public schools is, like religion,
based on faith, not based on science.
Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!
Felix Miata
531
days inactive
542
days old
28 comments
9 participants
participants (9)
-
Barry Scott -
Felix Miata -
Francis.Montagnac＠inria.fr -
Frédéric -
Joe Zeff -
John Pilkington -
Samuel Sieb -
stan -
Stephen Morris