View previous topic :: View next topic |
Author |
Message |
Aradayn n00b
Joined: 16 Jul 2017 Posts: 19
|
Posted: Fri Oct 12, 2018 5:16 am Post subject: |
|
|
Here you are:
https://pastebin.com/rNT6ugE2
Thank you for your time, I really appreciate it. This was working in the past, with an earlier version of this kernel configuration. I guess my problem has nothing to do with the latest driver anyway. I've tried running through the Nvidia driver wiki page again, but no luck, exactly the same results. |
|
Back to top |
|
|
Cthulhu666 Apprentice
Joined: 02 Apr 2004 Posts: 195
|
Posted: Fri Oct 12, 2018 8:18 am Post subject: |
|
|
No problem, only happy to help.
I'll take a look at the log when time allows it.
If you still have the old kernel config, you could try diff'ing the old and the new and see if anything you have removed stands out (PCI, ACPI, etc...). _________________ A certified Gentoo lover!
Got 4 Gentoo boxes (2 desktops, 1 laptop and a server/MythTV HTPC).
Bought a computer bundled with windows? Don't want to use windows? Demand a refund! |
|
Back to top |
|
|
krinn Watchman
Joined: 02 May 2003 Posts: 7470
|
Posted: Fri Oct 12, 2018 9:54 am Post subject: |
|
|
Your log is showing at second 100 that a new usb device was add, showing the boot has been complete (at 8rd second) and a new usb device has been plug, a sign everything is working.
And your log show nvidia-drivers version 396.54 in use.
I don't think showing anyone a kernel.log of a working nvidia drivers and kernel would help anyone. |
|
Back to top |
|
|
Aradayn n00b
Joined: 16 Jul 2017 Posts: 19
|
Posted: Sat Oct 13, 2018 3:31 am Post subject: |
|
|
Should I move this to a different thread? Like I said earlier, my problem appears to not actually be related to the latest drivers.
And should I keep this in Kernel & Hardware? I kind of think so since my video card is not showing up when I run lspci. |
|
Back to top |
|
|
Aradayn n00b
Joined: 16 Jul 2017 Posts: 19
|
Posted: Mon Oct 15, 2018 3:37 am Post subject: |
|
|
My problem was unrelated. (I was disabling some PCI devices in the local script, and since a hardware configuration change this included my graphics card.)
Additionally, the latest drivers work fine for me without any issue. |
|
Back to top |
|
|
i4dnf Apprentice
Joined: 18 Sep 2005 Posts: 271 Location: Bucharest, Romania
|
Posted: Thu Oct 25, 2018 5:29 am Post subject: |
|
|
Follow-up on this:
The problem appears to be the nvidia module hanging if loaded too early in the boot process.
Blacklisting the nvidia modules so that they only get loaded later, when X starts, is a usable workaround for now, at least in my case.
(Only blacklisting 'nvidia' doesn't work, the other nvidia-* modules need to be blacklisted too, as they pull 'nvidia' if they get loaded)
There's also what seems to be a related bug report: #667362 _________________ "The only difference between me and a madman is that I am not MAD" (SALVATOR DALI) |
|
Back to top |
|
|
ct85711 Veteran
Joined: 27 Sep 2005 Posts: 1791
|
Posted: Thu Oct 25, 2018 6:18 am Post subject: |
|
|
Just curious, but did you try adding haveged to your startup. My thinking is that this may be related similar to the slow startup due to insufficient entropy. In this case, the system isn't timing out the module loading (hence the hanging part). |
|
Back to top |
|
|
i4dnf Apprentice
Joined: 18 Sep 2005 Posts: 271 Location: Bucharest, Romania
|
Posted: Thu Oct 25, 2018 7:34 am Post subject: |
|
|
No, I didn't, but the problem still manifested on a 4.19.0 kernel with RANDOM_TRUST_CPU=y which prevents the entropy related delay, so I don't think that has any relevance. _________________ "The only difference between me and a madman is that I am not MAD" (SALVATOR DALI) |
|
Back to top |
|
|
pickd.mask n00b
Joined: 02 Aug 2011 Posts: 28
|
Posted: Wed Oct 31, 2018 9:42 pm Post subject: |
|
|
Problem still exists with nvidia-drivers-410.73 and 410.66 with both 4.18.16 and 4.19.0 kernels.
Blacklisting nvidia modules doesn't help.
I get same stuff as mentioned here: https://bugs.gentoo.org/667362#c0
Quote: | timeout 'nvidia-udev.sh add'
slow: 'nvidia-udev.sh add'
timeout: killing 'nvidia-udev.sh add'
slow: 'nvidia-udev.sh add' |
And sometimes udevd says that "specified group 'kvm' unknown".
Applying the patch from https://devtalk.nvidia.com/default/topic/1043346/linux/nvidia-driver-v410-73-fails-to-build-functional-modules/ doesn't help too, so I guess, I have to wait until devs from Nvidia will fix it.
IF they're about to fix it, of course.
Sometimes I think that Linus was right about Nvidia |
|
Back to top |
|
|
pickd.mask n00b
Joined: 02 Aug 2011 Posts: 28
|
Posted: Fri Nov 09, 2018 6:24 pm Post subject: |
|
|
Here's what I did to solve it. Works for me.
Also thank you all guys for posting about your issues and thoughts, it helped me to find out what was going on.
Link https://bugs.gentoo.org/670340#c8
I'll copypaste it here, just in case
Quote: | So it appears that I was able to circumwent that issue by rethinking all comments about blacklisting modules coming from wise people.
At last I noticed that IF I blacklist all modules to prevent them from loading (by udev, from what I know), I can actually load modules manually via modprobe and somehow that works perfectly. NOTE: I couldn't load or remove nvidia modules if I haven't blacklisted them.
After that the solution was simple. Probably it's not the best way, maybe it's plain dumb way, but it works for me.
So here's what I did:
1. Added the in /etc/modprobe.d/blacklist.conf following lines:
Code: | blacklist nvidia
blacklist nvidia_drm
blacklist nvidia_modeset |
(basically, I just blacklisted all nvidia modules that usually are loaded, which you can see by typing 'lsmod | grep -i nvidia' when your DE works)
2. Created file /etc/local.d/nvidia-udev-workaround.start
Added the following lines in it:
Code: | #!/bin/sh
echo "NVIDIA WORKAROUND IN PROGRESS";
modprobe nvidia_drm; |
3. Made that script executable by:
Code: | chmod +x /etc/local.d/nvidia-udev-workaround.start |
4. Made sure that local appears in default runlevel:
Code: | rc-update show default |
If there's no "local", in order to try that workaround, you should add it by Code: | rc-update add local default |
Then reboot.
Works for me with 410.73 and 415.13 nvidia-drivers, with 4.18.17 and 4.19.1 kernels. |
|
|
Back to top |
|
|
Utsuho Reiuji Apprentice
Joined: 03 Apr 2013 Posts: 179
|
Posted: Sat Nov 10, 2018 6:07 pm Post subject: |
|
|
how do you prevent xdm to start before local was invoked? Your fix also works for me, but I have to restart xdm before I get to use my DE.
Edit: On a second look, I noticed that xdm actually doesn't start automatically using this fix. Still better than no desktop... |
|
Back to top |
|
|
pickd.mask n00b
Joined: 02 Aug 2011 Posts: 28
|
Posted: Sat Nov 10, 2018 10:45 pm Post subject: |
|
|
Utsuho Reiuji wrote: | how do you prevent xdm to start before local was invoked? Your fix also works for me, but I have to restart xdm before I get to use my DE.
Edit: On a second look, I noticed that xdm actually doesn't start automatically using this fix. Still better than no desktop... |
On my system everything autostarts just fine. I didn't make any particular setup related to boot order or something. Maybe you should try to add line with "sleep 10;" to delay modprobe. Just a wild guess.
Here's some info about my runlevels:
Code: | ivan@pc ~ $ rc-update show boot
binfmt | boot
bootmisc | boot
consolefont | boot
fsck | boot
hostname | boot
hwclock | boot
keymaps | boot
localmount | boot
loopback | boot
modules | boot
mtab | boot
opentmpfiles-setup | boot
procfs | boot
root | boot
swap | boot
sysctl | boot
syslog-ng | boot
termencoding | boot
urandom | boot
ivan@pc ~ $ rc-update show default
cronie | default
cupsd | default
dbus | default
elogind | default
iptables | default
local | default
net.eth0 | default
netmount | default
openvpn | default
xdm | default |
Also I have #rc_parallel="NO" (commented) in /etc/rc.conf. Nothing else about parallel boot is present. That might be the case.
It also worth mentioning that I use kde and sddm.
Maybe something could be helpful. |
|
Back to top |
|
|
Utsuho Reiuji Apprentice
Joined: 03 Apr 2013 Posts: 179
|
Posted: Sun Nov 11, 2018 2:13 am Post subject: |
|
|
Thanks for your reply, pickd.mask, this helps a bit. I need to finish this massive update 1st to test things, but I'll try using rc_parallel="NO" just in case.
Not sure if that was the cause, but I noticed that /etc/init.d/local had
Code: | depend()
{
after *
keyword -timeout
}
| in it. Maybe uncommenting that "after *" bit will help. I also added "after local" to /etc/init.d/xdm in the same section.
Edit: that worked. |
|
Back to top |
|
|
pickd.mask n00b
Joined: 02 Aug 2011 Posts: 28
|
Posted: Mon Nov 12, 2018 2:09 am Post subject: |
|
|
Utsuho Reiuji wrote: |
Edit: that worked. |
That's great
BTW, I checked my /etc/init.d/local and that line was uncommented by default, I never touched it. |
|
Back to top |
|
|
buratino2015 n00b
Joined: 16 Apr 2015 Posts: 14
|
Posted: Mon Nov 26, 2018 6:02 pm Post subject: |
|
|
I have the same issue on the kernel 4.14.83 and nvidia-drivers 410.78. Rolled back to the 396.54. Videocard: GTX-1070. |
|
Back to top |
|
|
Maf Guru
Joined: 15 May 2005 Posts: 309
|
Posted: Sun Dec 02, 2018 11:08 am Post subject: |
|
|
Same here with nvidia-drivers-415.18 with Linux 4.14.83, GTX 970. Had to roll back to 396.54. |
|
Back to top |
|
|
Muso Veteran
Joined: 22 Oct 2002 Posts: 1052 Location: The Holy city of Honolulu
|
Posted: Sun Dec 02, 2018 8:12 pm Post subject: |
|
|
Maf wrote: | Same here with nvidia-drivers-415.18 with Linux 4.14.83, GTX 970. Had to roll back to 396.54. |
With gentoo-sources 4.19.6 and the 415.18 nvidia-drivers, I have no issues at all.
My video card is a GTX 950 _________________ "You can lead a horticulture but you can't make her think" ~ Dorothy Parker
2021 is the year of the Linux Desktop! |
|
Back to top |
|
|
pvh1987 n00b
Joined: 06 Jun 2012 Posts: 56
|
Posted: Mon Dec 03, 2018 8:42 pm Post subject: |
|
|
Thanks a lot! This workaround solved a big problem I had today, with a totally broken system after updating all packages with emerge. Turned out the culprit was the buggy nvidia-drivers.
If you are using CUDA, you might want to add
Code: |
blacklist nvidia_uvm
|
to /etc/modprobe.d/blacklist.conf as well. I do not know if this is necessary, but I would not take any chances. However, it does not seem to load when nvidia_drm is loaded, so I added it to /etc/local.d/nvidia-udev-workaround.start as well:
Code: |
modprobe nvidia_uvm;
|
Finally, after a whole day of troubleshooting, my workstation works again. I don't know how buggy the driver is when it comes to gaming... maybe a performance boost would be in order after all this work, but probably too much to ask for |
|
Back to top |
|
|
mrbassie l33t
Joined: 31 May 2013 Posts: 772 Location: over here
|
Posted: Tue Dec 04, 2018 8:54 am Post subject: |
|
|
pvh1987 wrote: | I don't know how buggy the driver is when it comes to gaming... |
I use bumblebee so I didn't have any of the probs in the thread, however I masked > 396.54 as the 4.10 driver rendered some textures incorrectly in games. |
|
Back to top |
|
|
Kaorukun n00b
Joined: 27 Aug 2014 Posts: 5
|
Posted: Wed Dec 05, 2018 4:59 am Post subject: |
|
|
pickd.mask wrote: | Here's what I did to solve it. Works for me.
Also thank you all guys for posting about your issues and thoughts, it helped me to find out what was going on.
Link https://bugs.gentoo.org/670340#c8
I'll copypaste it here, just in case
|
That worked for me too.
Now using x11-drivers/nvidia-drivers-415.18 and sys-kernel/gentoo-sources-4.19.6 |
|
Back to top |
|
|
uraes Tux's lil' helper
Joined: 28 Nov 2002 Posts: 135 Location: Estonia
|
Posted: Sun Dec 30, 2018 10:42 am Post subject: |
|
|
Uh, got hit by this... Defenitely graziest bug for many passed years Three days without working PC... and counting. Even tried stable/unstable/older versions of udev,nvidia-drivers, xorg... recompiled all packages (1116 items)...
So far : blacklisting modules helps - no high load anymore, sddm starts.. somehow (meaning: I see mouse cursor frame on black screen. ATM I'm not even sure is this bug/nvidia(related) or I have messed up my configs). Going on with experimentations ...
kernel: 4.19.12
nvidia-drivers: 415.18
Edit: As complaining is biggest helper in IT industry.. after my post I found another thread suggesting about mess with symlinks. Probably, as meanwhile I tried to switch to nouveau, I got broken symlinks under /usr/lib64/ .. so I did
Code: |
$ cd /usr/lib64/
$ ls -l libGLESv*
$ rm libGLESv*
$ emerge nvidia-drivers
|
|
|
Back to top |
|
|
Stolz Moderator
Joined: 19 Oct 2003 Posts: 3028 Location: Hong Kong
|
Posted: Thu Dec 12, 2019 12:04 pm Post subject: |
|
|
One year later and I'm also having the same issue with current drivers
- sys-kernel/gentoo-sources-5.4.2
- drivers/nvidia-drivers-440.36
- sys-fs/eudev-3.2.9
The workaround to blacklist the modules works. Thanks! |
|
Back to top |
|
|
|