Arch Linux User Repository

Basically, take the sha512sum of the kernel-6.3.patch file and replace with it the last checksum in the PKGBUILD file

@SoftExpert I cloned this repo, replaced your PKGBUILD and created your patch file. After running makepkg it fails:

==> Making package: nvidia-470xx-utils 470.199.02-1 (Wed 05 Jul 2023 12:15:07 PM CEST)
==> Checking runtime dependencies...
==> Checking buildtime dependencies...
==> Retrieving sources...
  -> Found nvidia-drm-outputclass.conf
  -> Found nvidia-470xx-utils.sysusers
  -> Found nvidia-470xx.rules
  -> Found NVIDIA-Linux-x86_64-470.199.02.run
  -> Found kernel-6.4.patch
==> Validating source files with sha512sums...
    nvidia-drm-outputclass.conf ... Passed
    nvidia-470xx-utils.sysusers ... Passed
    nvidia-470xx.rules ... Passed
    NVIDIA-Linux-x86_64-470.199.02.run ... Passed
    kernel-6.4.patch ... FAILED
==> ERROR: One or more files did not pass the validity check!

@simplecard: You are missing the fix provided by the patch for the kernel 6.4. Please review the comments from the past

@SoftExpert My apologies. It was actually error 2.

CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_pmm_sysmem_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_perf_events_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_perf_module_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_get_rm_ptes_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_fault_buffer_flush_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_mmu_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_peer_identity_mappings_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_va_block_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_range_group_tree_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_thread_context_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-uvm/uvm_rb_tree_test.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-modeset/nvidia-modeset-linux.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-modeset/nv-kthread-q.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-drv.o
  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-utils.o
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-drv.c: In function ‘nv_drm_update_drm_driver_features’:
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-drv.c:876:18: error: ‘struct drm_driver’ has no member named ‘dumb_destroy’
  876 |     nv_drm_driver.dumb_destroy     = nv_drm_dumb_destroy;
      |                  ^
make[2]: *** [scripts/Makefile.build:252: /var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-drv.o] Error 1
make[2]: *** Waiting for unfinished jobs....
make[1]: *** [Makefile:2026: /var/lib/dkms/nvidia/470.182.03/build] Error 2
make: *** [Makefile:80: modules] Error 2

@giniu: the solution is described in the comments posted a few days ago. Just browse the comments. To summarize: - use the latest version of the driver 470.199.02 - apply the patch for kernel 6.4

Hello, part of log:

  CC [M]  /var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-modeset.o
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-drv.c: In function ‘nv_drm_update_drm_driver_features’:
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-drv.c:876:18: error: ‘struct drm_driver’ has no member named ‘dumb_destroy’
  876 |     nv_drm_driver.dumb_destroy     = nv_drm_dumb_destroy;
      |                  ^
make[2]: *** [scripts/Makefile.build:252: /var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-drv.o] Błąd 1
make[2]: *** Oczekiwanie na niezakończone zadania....
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-crtc.c: In function ‘plane_req_config_update’:
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-crtc.c:89:9: warning: unused variable ‘ret’ [-Wunused-variable]
   89 |     int ret = 0;
      |         ^~~
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-crtc.c: In function ‘nv_drm_plane_atomic_set_property’:
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-crtc.c:371:32: warning: unused variable ‘nv_drm_plane_state’ [-Wunused-variable]
  371 |     struct nv_drm_plane_state *nv_drm_plane_state =
      |                                ^~~~~~~~~~~~~~~~~~
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-modeset.c: In function ‘__will_generate_flip_event’:
/var/lib/dkms/nvidia/470.182.03/build/nvidia-drm/nvidia-drm-modeset.c:96:23: warning: unused variable ‘primary_plane’ [-Wunused-variable]
   96 |     struct drm_plane *primary_plane = crtc->primary;
      |                       ^~~~~~~~~~~~~

I think this is same or similar to https://forums.developer.nvidia.com/t/error-compiling-with-recent-mainline-kernel-linux-6-4-rc1/252732 and there is patch at https://gist.github.com/joanbm/77f0650d45747b9a4dc8e330ade2bf5c

I've tested the patch and it works.

@simplecard: can you provide the relevant part of the build log where the error is being recorded ? Error 10 is quite vague ...

I'm on kernel 6.4.1-arch2-1 and have a GTX 650, this package fails to build. Error 10. Is there a way around this? My integrated GPU is very weak and I'm starting to experience lag without this package..

@wviana: Well, we are in the process of selecting new maintainers / contributors; it will take a short while until that is cleared up (the standard procedure). Since this driver serves old(ish) hardware, the interest for maintaining it is winding down slowly. Only you know if you can afford to wait or not. The package is basically, a recipe to correctly install the driver. If you need / must / want to use kernel 6.4 / experiment, the solution described in the comments will help you install the newest driver. Otherwise you can wait for the package to be officially updated.

edit: added missing word.

@SoftExpert Oh, ok, thanks. At some time the maintainer will add this patch to the package won't they? So is it ok just wait until then?