NVIDIA Driver NVIDIA-Linux-x86_64-340.107 Kernel bigger than 5.1.12-1

ulrich.klaas · June 27, 2019, 5:20pm

Hello,

i am running Suse LEAP 15.1 on a Lenovo T510 Notebook.
I have installed NVIDIA-Linux-x86_64-340.107 with patches for Kernel 5.0 and 5.1.
Everything works fine until Kernel 5.1.11-2.
From kernel 5.1.12-1 until now 5.1.15-7 the state is: I can compile and install the
driver without errors but after reboot i have only a black screen.
Ich cannot schwitch to virtual consoles.
The notebook is still reachable via ssh.

Is there a new patch existing ?

kind regards

Ulli
nvidia-installer.tar.gz (21.5 KB)

vatbier · July 21, 2019, 10:18pm

Ulrich,

I can’t help you but on my 2007 Dell inspiron 1720 laptop on Kubuntu 19.04 (KDE plasma 5.15.4) I have been able to install NVIDIA-Linux-x86_64-340.107 with kernel 5.0 and 5.1 patches from nonfree/nvidia-340xx-kmod.git - nvidia-340xx-kmod (I had to fix one of the patches myself) with ubuntu’s mainline kernel 5.1.16.
I also had to extract the *custom.run to disable compiling of uvm (dkms.conf and Makefile).

ulrich.klaas · July 22, 2019, 6:33pm

Hallo,
thank you for yor answer.

Can you tell me what you have changed in dkms.conf and in the Makefile (which Makefile).
I removed everything that contains “uvm” but can no more compile the driver.

kind regards
Ulli

vatbier · July 27, 2019, 7:01pm

My apologies, Ulrich, for not answering sooner but the last days were very busy and too hot with the heat wave.

To not load nouveau I created /etc/modprobe.d/blacklist-nouveau.conf :
blacklist nouveau
options nouveau modeset=0

In one of the 2 patch files I changed “+++ b/kernel/nv_time.h” to “+++ kernel/nv_time.h” so it would show up alright in the kernel folder (the --apply-patch parameter of the nvidia .run file should get fixed).

bash NVIDIA-Linux-x86_64-340.107.run --apply-patch buildfix_kernel_5.0_fix_nv-time.patch
./NVIDIA-Linux-x86_64-340.107-custom.run --list | more
bash NVIDIA-Linux-x86_64-340.107-custom.run --apply-patch xf86-video-nvidia-legacy-02-fix-linux-5.1.patch

extract the custom.run

In file .manifest I had to add nv_time.h

In the kernel folder I did not change the Makefile (it has “nv_uvm_interface.c” in it but no need to remove that as that is not the uvm module).
In the dkms.conf in the kernel folder I commented out the lines for uvm to:

#BUILT_MODULE_NAME[1]=“${PACKAGE_NAME}-uvm”
#BUILT_MODULE_LOCATION[1]=“uvm/”
#DEST_MODULE_LOCATION[1]=“/kernel/drivers/video”
#MAKE[0]+=“; make -C uvm module KERNEL_UNAME=${kernelver} KBUILD_EXTMOD=${dkms_tree}/${PACKAGE_NAME}/${PACKAGE_VERSION}/build/uvm”
#CLEAN+=“; make -C uvm clean”
#BUILT_MODULE_NAME[1]=“${PACKAGE_NAME}-uvm”
#BUILT_MODULE_LOCATION[1]=“uvm/”
#DEST_MODULE_LOCATION[1]=“/kernel/drivers/video”
#MAKE[0]+=“; make -C uvm module KERNEL_UNAME=${kernelver} KBUILD_EXTMOD=${dkms_tree}/${PACKAGE_NAME}/${PACKAGE_VERSION}/build/uvm”
#CLEAN+=“; make -C uvm clean”

In the uvm folder I moved the files dkms.conf.fragment and Makefile to unusable files (not sure if this was really necessary):
mv Makefile donotbuilduvmMakefile
mv dkms.conf.fragment donotusedkms_conf_fragment

running then “sh ./nvidia-installer” (in a root shell of course, not with X running) when it’s building it gives an error “unable to open ‘kernel/uvm/Makefile’ for copying” but it asks “would you like to continue installation anyway?”
and the nvidia kernel module then got installed.

Then I did a update-initramfs -u -k on the 5.1.16, I don’t know what command for that is used in SUSE.

mhell · October 4, 2019, 11:38am

Dear vatbier

I am using SuSe TW Linux 5.2.11-1-default #1 SMP Thu Aug 29 06:41:19 UTC 2019 (6385110) i686 i686 i386 GNU/Linux means 32 bit and tried above mentioned patches.
Nouveau is blacklisted and not loaded during install.

The 340.107 Driver then compiled w/ --no-unified-memory and did not include DKMS in menu-install.
The machine starts in 5.2 but screen stays black, Xorg.0.log stating “no screen found”.
Am I missing any configuration ?

Same patched driver runs w/ 4.20.6-1-default #1 SMP Thu Jan 31 07:37:50 UTC 2019 (463cfd2) i686 i686 i386 GNU/Linux w/o problem.

See below the Xorg.0.log

Thanks, Freddy

PS 340.107 patch for linux kernel 5.2 - Linux - NVIDIA Developer Forums maybe same question

[ 13.597] (–) Log file renamed from “/var/log/Xorg.pid-1830.log” to “/var/log/Xorg.0.log”
[ 13.598]
X.Org X Server 1.20.5
X Protocol Version 11, Revision 0
[ 13.598] Build Operating System: openSUSE SUSE LINUX
[ 13.598] Current Operating System: Linux chimera3.homesite 5.2.11-1-default #1 SMP Thu Aug 29 06:41:19 UTC 2019 (6385110) i686
[ 13.598] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-5.2.11-1-default root=UUID=3a2095f0-d479-4bcd-933e-899906ff64f6 resume=/dev/sda1 splash=silent quiet nomodeset showopts
[ 13.598] Build Date: 30 August 2019 12:00:00PM
[ 13.598]
[ 13.598] Current version of pixman: 0.36.0
[ 13.598] Before reporting problems, check http://wiki.x.org
to make sure that you have the latest version.
[ 13.598] Markers: (–) probed, () from config file, (==) default setting,
(++) from command line, (!!) notice, (II) informational,
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 13.598] (==) Log file: “/var/log/Xorg.0.log”, Time: Fri Oct 4 12:49:00 2019
[ 13.611] (==) Using config file: “/etc/X11/xorg.conf”
[ 13.611] (==) Using config directory: “/etc/X11/xorg.conf.d”
[ 13.611] (==) Using system config directory “/usr/share/X11/xorg.conf.d”
[ 13.617] (==) ServerLayout “Layout0”
[ 13.617] () |–>Screen “Screen0” (0)
[ 13.617] () | |–>Monitor “Monitor0”
[ 13.618] () | |–>Device “Device0”
[ 13.618] () |–>Input Device “Keyboard0”
[ 13.618] () |–>Input Device “Mouse0”
[ 13.618] (==) Automatically adding devices
[ 13.618] (==) Automatically enabling devices
[ 13.618] (==) Automatically adding GPU devices
[ 13.618] (==) Max clients allowed: 256, resource mask: 0x1fffff
[ 13.628] (WW) The directory “/usr/share/fonts/misc/sgi” does not exist.
[ 13.628] Entry deleted from font path.
[ 13.629] (==) FontPath set to:
/usr/share/fonts/misc:unscaled,
/usr/share/fonts/Type1/,
/usr/share/fonts/100dpi:unscaled,
/usr/share/fonts/75dpi:unscaled,
/usr/share/fonts/ghostscript/,
/usr/share/fonts/cyrillic:unscaled,
/usr/share/fonts/truetype/,
built-ins
[ 13.629] (==) ModulePath set to “/usr/lib/xorg/modules”
[ 13.629] (WW) Ignoring unrecognized extension “XFree86-DGA”
[ 13.629] (WW) Hotplugging is on, devices using drivers ‘kbd’, ‘mouse’ or ‘vmmouse’ will be disabled.
[ 13.629] (WW) Disabling Keyboard0
[ 13.629] (WW) Disabling Mouse0
[ 13.629] (II) Loader magic: 0x6966a0
[ 13.629] (II) Module ABI versions:
[ 13.629] X.Org ANSI C Emulation: 0.4
[ 13.629] X.Org Video Driver: 24.0
[ 13.629] X.Org XInput driver : 24.1
[ 13.629] X.Org Server Extension : 10.0
[ 13.630] (++) using VT number 7

[ 13.630] (II) systemd-logind: logind integration requires -keeptty and -keeptty was not provided, disabling logind integration
[ 13.634] (–) PCI:*(1@0:0:0) 10de:06e1:1acc:853d rev 161, Mem @ 0xfd000000/16777216, 0xd0000000/268435456, 0xfa000000/33554432, I/O @ 0x0000cc00/128, BIOS @ 0x???/131072
[ 13.634] (–) PCI: (2@0:0:0) 14f1:8880:0070:f038 rev 4, Mem @ 0xfe800000/2097152
[ 13.634] (II) LoadModule: “glx”
[ 13.642] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
[ 13.731] (II) Module glx: vendor=“NVIDIA Corporation”
[ 13.731] compiled for 4.0.2, module version = 1.0.0
[ 13.731] Module class: X.Org Server Extension
[ 13.731] (II) NVIDIA GLX Module 340.107 Thu May 24 21:47:46 PDT 2018
[ 13.731] (II) LoadModule: “nvidia”
[ 13.735] (II) Loading /usr/lib/xorg/modules/drivers/nvidia_drv.so
[ 13.750] (II) Module nvidia: vendor=“NVIDIA Corporation”
[ 13.750] compiled for 4.0.2, module version = 1.0.0
[ 13.750] Module class: X.Org Video Driver
[ 13.753] (II) NVIDIA dlloader X Driver 340.107 Thu May 24 21:24:47 PDT 2018
[ 13.753] (II) NVIDIA Unified Driver for all Supported NVIDIA GPUs
[ 13.801] (II) Loading sub module “fb”
[ 13.801] (II) LoadModule: “fb”
[ 13.801] (II) Loading /usr/lib/xorg/modules/libfb.so
[ 13.803] (II) Module fb: vendor=“X.Org Foundation”
[ 13.803] compiled for 1.20.5, module version = 1.0.0
[ 13.803] ABI class: X.Org ANSI C Emulation, version 0.4
[ 13.803] (WW) Unresolved symbol: fbGetGCPrivateKey
[ 13.803] (II) Loading sub module “wfb”
[ 13.803] (II) LoadModule: “wfb”
[ 13.803] (II) Loading /usr/lib/xorg/modules/libwfb.so
[ 13.804] (II) Module wfb: vendor=“X.Org Foundation”
[ 13.804] compiled for 1.20.5, module version = 1.0.0
[ 13.804] ABI class: X.Org ANSI C Emulation, version 0.4
[ 13.804] (II) Loading sub module “ramdac”
[ 13.805] (II) LoadModule: “ramdac”
[ 13.805] (II) Module “ramdac” already built-in
[ 13.811] (EE) NVIDIA: Failed to initialize the NVIDIA kernel module. Please see the
[ 13.811] (EE) NVIDIA: system’s kernel log for additional error messages and
[ 13.811] (EE) NVIDIA: consult the NVIDIA README for details.
[ 13.811] (EE) No devices detected.
[ 13.811] (EE)
Fatal server error:
[ 13.811] (EE) no screens found(EE)
[ 13.811] (EE)
Please consult the The X.Org Foundation support
at http://wiki.x.org
for help.
[ 13.811] (EE) Please also check the log file at “/var/log/Xorg.0.log” for additional information.
[ 13.811] (EE)
[ 13.911] (EE) Server terminated with error (1). Closing log file.