Why Blacklist nouveau?
Hello, after update thumbleweed to kernel 6.1.8 X won't start I'm looking for the reason and find out that the kernel module 'nouveau' was not loaded, after "modprobe nouveau" X starts as expected. I found a new file: '/lib/modprobe.d/50-nvidia-default.conf' with "blacklist nouveau" in first line. Rename this file to *.bak and after reboot X starts normally. regards tp
Nice, can you proceed and post the output of `rpm -qf /lib/modprobe.d/50-nvidia-default.conf` here? On Tue, Jan 31, 2023 at 10:50 AM tponline@e.mail.de <tponline@e.mail.de> wrote:
Hello,
after update thumbleweed to kernel 6.1.8 X won't start
I'm looking for the reason and find out that the kernel module 'nouveau' was not loaded, after "modprobe nouveau" X starts as expected.
I found a new file: '/lib/modprobe.d/50-nvidia-default.conf' with "blacklist nouveau" in first line.
Rename this file to *.bak and after reboot X starts normally.
regards tp
-- Regards, Andrei Dziahel
On Tue, Jan 31, 2023 at 10:54:50AM +0100, Andrei Dziahel wrote:
Nice, can you proceed and post the output of `rpm -qf /lib/modprobe.d/50-nvidia-default.conf` here?
On Tue, Jan 31, 2023 at 10:50 AM tponline@e.mail.de <tponline@e.mail.de> wrote:
Hello,
after update thumbleweed to kernel 6.1.8 X won't start
I'm looking for the reason and find out that the kernel module 'nouveau' was not loaded, after "modprobe nouveau" X starts as expected.
I found a new file: '/lib/modprobe.d/50-nvidia-default.conf' with "blacklist nouveau" in first line.
Rename this file to *.bak and after reboot X starts normally.
regards tp
Probably same as bug https://bugzilla.suse.com/show_bug.cgi?id=1207520 Ciao, Marcus
On 1/31/23 10:50, tponline@e.mail.de wrote:
Hello,
after update thumbleweed to kernel 6.1.8 X won't start
I'm looking for the reason and find out that the kernel module 'nouveau' was not loaded, after "modprobe nouveau" X starts as expected.
I found a new file: '/lib/modprobe.d/50-nvidia-default.conf' with "blacklist nouveau" in first line.
Rename this file to *.bak and after reboot X starts normally.
regards tp
You have the binary NVIDIA drivers installed (nvidia-driver-G06-kmp-default package) and it blacklists nouveau.
On Tue, 31 Jan 2023 at 10:01, Michael Pujos <pujos.michael@gmail.com> wrote:
You have the binary NVIDIA drivers installed (nvidia-driver-G06-kmp-default package) and it blacklists nouveau.
Or _have previously had_ them, but later removed. Also, or G04 or G05 versions. But agreed. Installing the binary drivers blacklists Nouveau. The proprietary driver must blacklist the built-in Nouveau driver, or the built-in driver in the kernel "grabs" the hardware and the binary driver can't get the complete access it needs in order to function. -- Liam Proven ~ Profile: https://about.me/liamproven Email: lproven@cix.co.uk ~ gMail/gTalk/FB: lproven@gmail.com Twitter/LinkedIn: lproven ~ Skype: liamproven UK: (+44) 7939-087884 ~ Czech [+ WhatsApp/Telegram/Signal]: (+420) 702-829-053
On Tue, Jan 31, 2023 at 10:50:26AM +0100, tponline@e.mail.de wrote:
Hello,
after update thumbleweed to kernel 6.1.8 X won't start
I'm looking for the reason and find out that the kernel module 'nouveau' was not loaded, after "modprobe nouveau" X starts as expected.
I found a new file: '/lib/modprobe.d/50-nvidia-default.conf' with "blacklist nouveau" in first line.
Rename this file to *.bak and after reboot X starts normally.
This is anoter duplicate of https://bugzilla.suse.com/show_bug.cgi?id=1207520 Should be fixed now with next TW update. Check for update of - pesign-obs-integration package. pesign-obs-integration.changes: ------------------------------------------------------------------- Mon Jan 23 14:16:22 UTC 2023 - Callum Farmer <gmbr3@opensuse.org> - Add dependency-generators.patch to support copying source files and macros to the re-package build (jsc#PED-2658) [...] Hopefully we'll also get changes for nvidia-open-driver-G06-signed in. See https://build.opensuse.org/request/show/1062105 nvidia-open-driver-G06-signed.changes: ------------------------------------------------------------------- Mon Jan 30 17:14:28 UTC 2023 - Stefan Dirsch <sndirsch@suse.com> - no longer use %hardcode_pci_list macro on TW, since the issue with generating wrong PCI ID modaliases for Supplements has been fixed now in pesign-obs-integration package for TW (boo#1207520) which is sitting on top of ------------------------------------------------------------------- Sun Jan 29 16:16:33 UTC 2023 - Stefan Dirsch <sndirsch@suse.com> - introduce %hardcode_pci_list macro for hardcoding PCI ID modaliases for Supplements; make it default to *temporarily* workaround boo#1207520 * TW: enable only officially supported headless GPUs! * Leap/sle: enable all Turing and Ampere GPUs and set the appropriate modprobe.d option, so we can at least begin testing on sle15-sp5 ... - adds pci-table.patch, generati-pci-table.sh Thanks, Stefan Public Key available ------------------------------------------------------ Stefan Dirsch (Res. & Dev.) SUSE Software Solutions Germany GmbH Tel: 0911-740 53 0 Frankenstraße 146 FAX: 0911-740 53 479 D-90461 Nürnberg http://www.suse.de Germany ---------------------------------------------------------------- Geschäftsführer: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman (HRB 36809, AG Nürnberg) ----------------------------------------------------------------
On 2023-01-31 08:04, Stefan Dirsch wrote:
On Tue, Jan 31, 2023 at 10:50:26AM +0100, tponline@e.mail.de wrote:
Hello,
after update thumbleweed to kernel 6.1.8 X won't start
I'm looking for the reason and find out that the kernel module 'nouveau' was not loaded, after "modprobe nouveau" X starts as expected.
I found a new file: '/lib/modprobe.d/50-nvidia-default.conf' with "blacklist nouveau" in first line.
Rename this file to *.bak and after reboot X starts normally. This is anoter duplicate of
https://bugzilla.suse.com/show_bug.cgi?id=1207520
Should be fixed now with next TW update. Check for update of
- pesign-obs-integration
package.
Hi Stefan, Do you have an idea when the above package will roll out? I just upgraded again to the latest TW (20230202) and on first boot, I still have to kill Xorg to get SDDM. Not a big deal but I wanted to be sure that I had not missed anything. :) # rpm -q pesign-obs-integration pesign-obs-integration-10.2+git20220504.8690743-5.1.x86_64 Thanks a lot! --- Pablo Sanchez - Blueoak Database Engineering Ph: 819.459.1926 iNum: 883.5100.0990.1054
On Sat, Feb 04, 2023 at 05:58:51PM -0500, Pablo Sanchez wrote:
On 2023-01-31 08:04, Stefan Dirsch wrote:
On Tue, Jan 31, 2023 at 10:50:26AM +0100, tponline@e.mail.de wrote:
Hello,
after update thumbleweed to kernel 6.1.8 X won't start
I'm looking for the reason and find out that the kernel module 'nouveau' was not loaded, after "modprobe nouveau" X starts as expected.
I found a new file: '/lib/modprobe.d/50-nvidia-default.conf' with "blacklist nouveau" in first line.
Rename this file to *.bak and after reboot X starts normally. This is anoter duplicate of
https://bugzilla.suse.com/show_bug.cgi?id=1207520
Should be fixed now with next TW update. Check for update of
- pesign-obs-integration
package.
Hi Stefan,
Do you have an idea when the above package will roll out?
Both packages have been checked into TW - pesign-obs-integration and nvidia-open-driver-G06-signed.
I just upgraded again to the latest TW (20230202) and on first boot, I still have to kill Xorg to get SDDM.
Not sure why it would help to kill the Xserver once to fix the issue. The topic of this thread is that it was mistakenly installed the nvidia-open-driver-G06-signed-kmp-default package on *all* sysmtems with nvidia cards, which installs a driver, which currently doesn't support more than a handful nVidia GPUs out-of-the-box and then resulted in no graphical desktop at all (because blacklisting 'nouveau' driver).
Not a big deal but I wanted to be sure that I had not missed anything. :)
# rpm -q pesign-obs-integration pesign-obs-integration-10.2+git20220504.8690743-5.1.x86_64
Actually pesign-obs-integration doesn't matter here, since it's only relevant for building nvidia-open-driver-G06-signed-kmp-default package. Looks like you have a completely different issue, if killing Xserver once at boot helps. Maybe it's a race - driver not loaded completely, when Xserver starts ... Thanks, Stefan Public Key available ------------------------------------------------------ Stefan Dirsch (Res. & Dev.) SUSE Software Solutions Germany GmbH Tel: 0911-740 53 0 Frankenstraße 146 FAX: 0911-740 53 479 D-90461 Nürnberg http://www.suse.de Germany ---------------------------------------------------------------- Geschäftsführer: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman (HRB 36809, AG Nürnberg) ----------------------------------------------------------------
On 2023-02-06 06:58, Stefan Dirsch wrote:
Looks like you have a completely different issue, if killing Xserver once at boot helps. Maybe it's a race - driver not loaded completely, when Xserver starts ...
Hi Stefan, Thank you for the explanation. I'll look into creating a reproducible test case and open a bug. Cheers! --- Pablo Sanchez - Blueoak Database Engineering Ph: 819.459.1926 iNum: 883.5100.0990.1054
Hi, after returning from FOSDEM I updated my workhorse to TW 20230205 and ran into exactly the same issue. Afte entering the LUKS password the screen remains black, killing the X-Server does not help, neither renaming the 50-nvidia-... file. Booting into init 3 as well as booting into an older snapshot works neither - always black screen. The kernel crashes with a version mismatch (client has version 470.161.x, kernel module 525.85.05) in snapshot 20230205. Currently searching a solution.... Cheers Axel
* Axel Braun <axel.braun@gmx.de> [02-06-23 12:13]:
Hi, after returning from FOSDEM I updated my workhorse to TW 20230205 and ran into exactly the same issue. Afte entering the LUKS password the screen remains black, killing the X-Server does not help, neither renaming the 50-nvidia-... file. Booting into init 3 as well as booting into an older snapshot works neither - always black screen. The kernel crashes with a version mismatch (client has version 470.161.x, kernel module 525.85.05) in snapshot 20230205. Currently searching a solution....
change or boot to a text screen, change to root do rpm -qa *nvidia* and remove inappropriate nvidia drivers zypper -v rm <wrong><nvidia><drivers> then lock out the incorrect drivers, zypper al *G06* (works for me and undoubtedly for you seeing 470.161.x) then install correct drivers, probably: zypper -v dup *nvidia* keep the lock until the package correction comes mainstream and you should be good to go. -- (paka)Patrick Shanahan Plainfield, Indiana, USA @ptilopteri http://en.opensuse.org openSUSE Community Member facebook/ptilopteri Photos: http://wahoo.no-ip.org/piwigo paka @ IRCnet oftc
Hello Patrick, Am Montag, 6. Februar 2023, 18:44:27 CET schrieb Patrick Shanahan:
* Axel Braun <axel.braun@gmx.de> [02-06-23 12:13]:
Hi, after returning from FOSDEM I updated my workhorse to TW 20230205 and ran into exactly the same issue. Afte entering the LUKS password the screen remains black, killing the X-Server does not help, neither renaming the 50-nvidia-... file. Booting into init 3 as well as booting into an older snapshot works neither - always black screen. The kernel crashes with a version mismatch (client has version 470.161.x, kernel module 525.85.05) in snapshot 20230205. Currently searching a solution....
change or boot to a text screen, change to root do
As written, booting into a text screen did not work - the issue as described abvoe existed for all snapshots containing the 6.1.8 kernel (for which the G06 driver was build in the latest update). That was not my expectation, but I assume it has to do with the way that btrfs takes snapshots) Going back to an older snapshot (6.1.4 kernel) worked, and allowed a zypper dup without G06 driver Thanks Axel
participants (10)
-
Andrei Dziahel -
Axel Braun -
Axel Braun -
Liam Proven -
Marcus Meissner -
Michael Pujos -
Pablo Sanchez -
Patrick Shanahan -
Stefan Dirsch -
tponline＠e.mail.de