inttf NVIDIA 340.108 / 390.154 Patcher for Kernel 6.2/6.1/6.0/5.19/5.18/5.17/5.16/5.15/5.14/5.13 [BASH Script] - Comment Page: 8

This is a BASH script, which download NVIDIA installer, extract it, patch it and make new patched installer package. This is very quickly tested alpha version so if you have any problems please let me know or if this works as it should you can also let me know. I use here currently ARCH Linux Patches + my own patches. Currently this works with NVIDIA 340.108 / 390.154 / 418.113 / 435.21 drivers and latest 6.2, 6.1, 6.0, 5.19, 5.18, 5.17, 5.16, 5.15, 5.14, 5.13, 5.12, 5.11, 5.10 and 5.9 kernels. This should work with any distribution. You just need install wget and patch packages. Check video version of guide: Watch More Videos and Subscribe to Youtube...

257 comments on “inttf NVIDIA 340.108 / 390.154 Patcher for Kernel 6.2/6.1/6.0/5.19/5.18/5.17/5.16/5.15/5.14/5.13 [BASH Script] - Comment Page: 8

1 6 7 8 9
    1. Hi. I was trying to install the patched drivers for my Geforce gt 120M.
      I got this error:

      I can confirm autoconf.h and auto.conf exist in the kernel sources.

    2. Love the script. Can you apply some patches for the NVIDIA 418.113 driver for kernels 5.12, 5.15, 5.19. I use Garuda Linux and the build is failing against these kernels.

      • Thanks, have to check what it will need to patch also 418.113 for new kernels.

    3. On OpenSuSE 15.4 and 340.108 the patch ran successfully, .. as did the installation script, .. but the video hung (the main monitor is on the MB), .. could not reboot, .. had to reset.

      On startup, nothing different; last lines in the nvidia log are:

      -> Installing both new and classic TLS OpenGL libraries.
      -> Installing both new and classic TLS 32bit OpenGL libraries.

      Only change in modprobe.d is a new version of nvidia.conf with blacklist noveau. The script generates install for 5.19 and back to 5.09.

      Would OpenSuSE be different?


    4. Built and installed on OpenSuSE 15.4 (Kernel 5.14), no errors in the log, .. but no video on the card after reboot:

      lspci |grep -E “VGA|3D” :

      03:00.0 3D controller: NVIDIA Corporation G98 [Quadro NVS 450] (rev a1)
      04:00.0 VGA compatible controller: NVIDIA Corporation G98 [Quadro NVS 450] (rev a1)

      Is there any special about OpenSuSE?


      • Hi Lee,

        Thanks for reporting, I have to check this, maybe I can reproduce this.

        OpenSuSE is sometimes tricky, because they use some LTS style kernel versions and just backport newest features to old kernel.

    5. Hey JR. Is 6.0 RC5 not working on 515.xx with the 6.0 patcher?
      I tried upgrading my desktop (GTX1060) from RC3 to RC5 this AM using both the stock .run from NVIDIA
      as well as your patched ones and both failed. I’ve noticed their autobuild systems failing a crapload lately also.
      I was using the MAINLINE DAILY builds from the 12th and 13th, all previous builds are failed on AMD64
      I can get them to install, and they run with NOUVEAU but fail with stock, or patched NVIDIA drivers.
      (I got RC3 working fine of course tho and I notice your patch only says for up to RC4, that’s why I ask about RC5.)

      I don’t understand what’s up with 6.0, as stated before, Linux Torvalds has stated that “It’s just a number” and
      there’s no *major* difference but if so, how come all the autobuilds are failing, why don’t NVIDIA’s most recent
      drivers work and how come you have said that it will probably be a while, if ever, for 340.108 to be patched?
      Obviously there’s some major differences with 6.0 and it’s more than “just a number”.

      – MENTD

      • Hey MENTD,

        Tested this with 6.0 RC5 on Fedora 36 with my custom build. This patched version still installs normally, so have to say that this is Ubuntu problem. I can check this when I have a little bit more time. Hopefully NVIDIA publish soon working version(s) for Kernel 6.0.

        • Okay well knowing that, I will have another go at it tomorrow AM as well & report back. 8)

          If I recall, the daily builds have always been kinda flaky.

          • Success! 8)
            Now running RC5 daily 20220915 with your 515.65.01 patch on my GTX 1060 desktop! 8)
            My laptop is still running 5.19.8-xanmod1 on 340.108 with your patch.

            The trick for the desktop is there are two header files:

            I had to use the ALL one, NOT the GENERIC one. I had to do the same thing with RC3 but had forgotten.
            So yeah, all is still good with your patcher for the latest RC5 daily on Ubuntu on 515.65.01.

            Hope 340.108 is gonna be possible! 8) Good work and good luck!


      NVIDIA *finally* released a new 515 driver and it’s working on KERNEL 6.0 *without* need of patching!

      I updated my GTX 1060 6GB desktop PC (i5-3470 QUAD CORE w/ 16GB RAM) running UBUNTU v22.04.1 LTS to
      NVIDIA v515.76 this AM using the normal .run file procedure without a hitch! 8)

      So for everyone on 515, go snag the new driver now and enjoy a working, patchless driver! 8)

      Hopefully our resident genius JR will get 340.108 working on KERNEL 6.0 soon! 8)

      • Strange. 6.0 is just a bitch.
        Although RC3 and RC5 installed, RC6 is giving me issues and won’t work.
        It installs, but fails with NVIDIA 515.76 or the 515.15 patched one.
        At least, for me it does. Half the builds from mainline are failed.
        I am using the daily one from 09/22/22.
        It appears to be a path error.
        Hopefully all these issues are gone by the time 6.0 gets out of RC mode.

      • You’re totally right, I will test it soon and update script. Thanks!

          • Just pushed new version, 340, 390 and 470 should work with kernel 6.0.2.

            Will run some extra tests, but installers build normally.

            • Script errors out –
              [email protected]:~/NVIDIA$ ./
              bash: ./ /brn/bash: bad interpreter: No such file or directory

              But it seems to be a simple fix, I think you used # to comment instead of /* */ ?
              I just nuked the first line:

              Seems to be working great now but you prob. wanna update it for the others. 8)

              • I’m stupid it isn’t a comment error, I just don’t have /brn/bash heheh…
                It’s late…

                • I mean it was a typo it should be /bin/bash
                  heheheh u can nuke all my half asleep posts

    7. Thanks for patches and gathering them in one place!

    8. I can’t seem to get it to build the kernel.
      It just errors out. I’m on Ubuntu 20.04.5 LTS.

      I have GCC-11, it seems to be saying it needs GCC-12 but I don’t understand why?
      The 6.0 KERNEL runs fine on the NOUVEAU drivers and the 340.108 is ancient, so what is requiring GCC-12?
      I usually have to do EXPORT CC=/usr/bin/gcc-11 and then IGNORE CC VERSION MISMATCH to get previous KERNELS working,
      but it has always worked fine after that. (All PRE-6.0 KERNELS work great.)

      • Odd. APT downloaded some kernel source modules other than the header source yesterday and it seems to be working now.
        Thank you!!!
        Now on 340.108 running KERNEL 6.0.2!

    9. Hi! So, 340.108 driver is not need to patch for 5.19, 6.00 kernel?

      • Hi,

        I will update this, 340.108 needs patched version and both 5.19 and 6.0 patches are available with script.

    10. Howdy
      On my old AUS X73S laptop with NVIDIA GT520M the patch isn’t working with kernel 6.03 and 390.154:
      the driver compiles OK but XORG doesn’t start !
      On XORG.log :
      263.575] (EE) Screen(s) found, but none have a usable configuration.
      [ 263.575] (EE)
      Fatal server error:
      [ 263.575] (EE) no screens found(EE)

      dmesg sys :
      [ 213.761217] nvidia-nvlink: Nvlink Core is being initialized, major device number 246
      [ 213.761917] nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=none
      [ 213.762354] NVRM: loading NVIDIA UNIX x86_64 Kernel Module 390.154 Wed Jun 22 04:50:54 UTC 2022 (using threaded interrupts)
      [ 213.766552] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 390.154 Wed Jun 22 04:42:33 UTC 2022
      [ 213.768542] [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
      [ 213.768547] [drm] Initialized nvidia-drm 0.0.0 20160202 for 0000:01:00.0 on minor 1
      [ 263.563727] resource sanity check: requesting [mem 0x000e0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000e7fff window]
      [ 263.563738] caller _nv027866rm+0x58/0x90 [nvidia] mapping multiple BARs
      [ 263.578699] NVRM: failed to copy vbios to system memory.
      [ 263.578972] NVRM: RmInitAdapter failed! (0x30:0xffff:663)
      [ 263.579011] NVRM: rm_init_adapter failed for device bearing minor number 0

      • Hi Bob,

        Is your system working normally with 5.19 kernel + NVIDIA 390.154?

        • yes it works with 390.154 but NVIDIA 390.157 is released .

    11. dpkg e: Sub-process /usr/bin/dpkg returned an error code (1)

      your solution didn’t work.
      pop os-gt540m (390.154)

      • Hard to say why you got this error message. Could you post full error message?

    12. Hi, Just want to say thanks for publishing this script, it enabled me to give a new lease of life to my 2013 MBP (now running Pop_OS).

      Thanks you


      • Excellent! You’re welcome!

    13. Arghhh. I followed your guide and created, but when I tried to execute it in run level 3 it said

      “gzip: stdin: not in gzip format” and quit.

      What should I try?
      I’m on fedora 36 with lspci | grep VGA showing

      01:00.0 VGA compatible controller: NVIDIA Corporation GK208B [GeForce GT 710] (rev a1)

      Thanks for any hints…

    14. OK, I followed your clear instructions, but could not get your script
      to work. When I try to execute it (from anywhere) it quits, saying

      “gzip: stdin: not in gzip format
      Execution failed”

      What does this mean? I’m trying to get kernel-6.0.7 to work on Fedora 36.
      The output of lspci | grep VGA is
      01:00.0 VGA compatible controller: NVIDIA Corporation GK208B [GeForce GT 710] (rev a1)

      The output of uname -a is
      Linux 500g 5.19.16-602.inttf.fc36.x86_64 #1 SMP PREEMPT_DYNAMIC Mon Oct 17 16:51:38 EEST 2022 x86_64 x86_64 x86_64 GNU/Linux

      I see I had posted a similar “gzip: stdin: not in gzip format” error last June but a
      day or so later said I’d made a dumb mistake. But I do not remember what “dumb” mistake I’d made. Can you help me refresh my memory?

      I see that if I say ./ –info
      it says Compression: gzip, but if I say
      ./ –info
      it says Compression: xz
      This latter driver (which of course does not work for kernel-6.0.7) otherwise behaves
      normally, producing its usual series of dots. Is any of this relevant?

      Any thoughts would be most welcome.

      • Hi Peter,

        I just tested this on Fedora 36 with kernel 6.0.7, it works normally on my Fedora system.

        Actually this is NVIDIA’s own script which compress patched version using gzip instead xz. For me both works normally.

        Could you post output of following command:

        dnf list installed gzip xz

        I’m actually not sure what happened earlier with 5.17 -> 5.18 kernel. You can always start start from scratch, delete whole directory with installers.

        • Hi JR, dnf list installed gzip xz shows

          gzip.x86_64 1.11-3.fc36 @anaconda
          xz.x86_64 5.2.5-9.fc36 @anaconda

          I’m still getting (now 11/23/22), trying patched driver, freshly installed in an otherwise empty NVIDIA directory:

          Verifying archive integrity… OK
          Uncompressing NVIDIA driver 390.154 patched for kernel 6.0+
          gzip: stdin: not in gzip format
          Extraction failed.

          So I’m still puzzled, alas. What could I be doing wrong?

          I’m on fedora 36, uname -a shows Linux 500g 5.19.16-602.inttf.fc36.x86_64 #1 SMP PREEMPT_DYNAMIC Mon Oct 17 16:51:38 EEST 2022 x86_64 x86_64 x86_64 GNU/Linux

          nvidia-installer -v | grep versiom shows nvidia-installer: version 390.154 ([email protected]) Wed Jun 22

          and lspci | grep VGA shows
          01:00.0 VGA compatible controller: NVIDIA Corporation GK208B [GeForce GT 710] (rev a1)

          Cheers, and hoping your TG goes well.

        • Oh: I tried running JR’s patched driver ( from the fedora 36 live-server (“Try Fedora”), doing the six steps above, and it seemed to run correctly, with no “not in gzip format” error. So I’m thinking that I must have installed some package(s) prior to trying the patched driver—package(s) that somehow caused the “not in gzip format” error. What could such packages be? Will I have to start all over with a fresh install of fedora from its iso?

1 6 7 8 9

Leave a Reply to Alexander Procenko Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.