Bug 619114 – x11-drivers/nvidia-drivers-340.102 does not work with 4.9.* kernels, failed to register procfs!

• Status: RESOLVED DUPLICATE of bug 602844
• Product: Gentoo Linux

[Description Octavian 2017-05-20 20:12:37 UTC]

Created attachment 473548 [details, diff]
patch needs to be applied after running NVIDIA-Linux-x86_64-340.102.run --extract-only

=x11-drivers/nvidia-drivers-340.102 does not work with 4.9.* kernels. 
It compiles fine but nvidia module loading fails with the errors listed at the end of the post.

At the provided URL, there is a path and a description on how to apply it. I have tested it and it works.

The tricky part is that NVIDIA-Linux-x86_64-340.102.run (NVIDIA script) needs to be run with --extract-only, then the attached patch applied, then the NVIDIA-Linux-x86_64-340.102.run needs to be run again.

Best regards,
Octavian


May 20 21:50:35 ferb kernel: ------------[ cut here ]------------
May 20 21:50:35 ferb kernel: WARNING: CPU: 1 PID: 16772 at fs/proc/generic.c:345 0xffffffff803e8d8d
May 20 21:50:35 ferb kernel: proc_dir_entry 'driver/nvidia' already registered
May 20 21:50:35 ferb kernel: Modules linked in: nvidia(PO+) zram zsmalloc autofs4 ipv6 snd_hda_codec_realtek snd_hda_codec_generic binfmt_misc snd_usb_audio snd_usbmidi_lib snd_hwdep snd_rawmidi snd_seq_device vboxpci(O) snd_hda_intel vboxnetadp(O) drm i2c_i801 snd_hda_codec i2c_smbus vboxnetflt(O) snd_hda_core snd_pcm vboxdrv(O) snd_timer snd acpi_cpufreq virtio_pci virtio_blk virtio_balloon virtio_ring virtio cbc sha512_generic sha256_generic sha1_generic scsi_transport_iscsi fuse jfs btrfs xor zlib_deflate raid6_pq usbhid xhci_pci xhci_hcd ohci_pci ohci_hcd usb_storage scsi_transport_fc uhci_hcd ehci_pci ehci_hcd usbcore usb_common [last unloaded: pcspkr]
May 20 21:50:35 ferb kernel: CPU: 1 PID: 16772 Comm: modprobe Tainted: P        W  O    4.9.28-gentoo #1
May 20 21:50:35 ferb kernel: Hardware name: FUJITSU SIEMENS CELSIUS W370                  /D2817-A1, BIOS 6.00 R1.11.2817.A1               01/15/2009
May 20 21:50:35 ferb kernel:  0000000000000000 ffffffff806314fd ffffc900042a3c00 0000000000000000
May 20 21:50:35 ferb kernel:  ffffffff80290a54 ffff8801c5078f40 ffffc900042a3c50 ffff8801b5887385
May 20 21:50:35 ferb kernel:  ffff8801c6c0a938 ffff8801b5887300 ffffffffa0ccbce8 ffffffff80290aba
May 20 21:50:35 ferb kernel: Call Trace:
May 20 21:50:35 ferb kernel:  [<ffffffff806314fd>] ? 0xffffffff806314fd
May 20 21:50:35 ferb kernel:  [<ffffffff80290a54>] ? 0xffffffff80290a54
May 20 21:50:35 ferb kernel:  [<ffffffff80290aba>] ? 0xffffffff80290aba
May 20 21:50:35 ferb kernel:  [<ffffffff808f16de>] ? 0xffffffff808f16de
May 20 21:50:35 ferb kernel:  [<ffffffff803e8d8d>] ? 0xffffffff803e8d8d
May 20 21:50:35 ferb kernel:  [<ffffffff803e8eab>] ? 0xffffffff803e8eab
May 20 21:50:35 ferb kernel:  [<ffffffffa08d5657>] ? 0xffffffffa08d5657
May 20 21:50:35 ferb kernel:  [<ffffffffa002628b>] ? 0xffffffffa002628b
May 20 21:50:35 ferb kernel:  [<ffffffffa0026783>] ? 0xffffffffa0026783
May 20 21:50:35 ferb kernel:  [<ffffffffa00267d0>] ? 0xffffffffa00267d0
May 20 21:50:35 ferb kernel:  [<ffffffff802003fa>] ? 0xffffffff802003fa
May 20 21:50:35 ferb kernel:  [<ffffffff80378f8b>] ? 0xffffffff80378f8b
May 20 21:50:35 ferb kernel:  [<ffffffff80333b91>] ? 0xffffffff80333b91
May 20 21:50:35 ferb kernel:  [<ffffffff8030208b>] ? 0xffffffff8030208b
May 20 21:50:35 ferb kernel:  [<ffffffff802ffc60>] ? 0xffffffff802ffc60
May 20 21:50:35 ferb kernel:  [<ffffffff8038cef0>] ? 0xffffffff8038cef0
May 20 21:50:35 ferb kernel:  [<ffffffff803024b3>] ? 0xffffffff803024b3
May 20 21:50:35 ferb kernel:  [<ffffffff808f1a60>] ? 0xffffffff808f1a60
May 20 21:50:35 ferb kernel: ---[ end trace cdabddabd4921b4e ]---
May 20 21:50:35 ferb kernel: NVRM: failed to register procfs!
May 20 21:50:35 ferb kernel: NVRM: request_mem_region failed for 16M @ 0xbd000000. This can\x0aNVRM: occur when a driver such as rivatv is loaded and claims\x0aNVRM: ownership of the device's registers.
May 20 21:50:35 ferb kernel: nvidia: probe of 0000:01:00.0 failed with error -1
May 20 21:50:35 ferb kernel: NVRM: The NVIDIA probe routine failed for 1 device(s).
May 20 21:50:35 ferb kernel: NVRM: None of the NVIDIA graphics adapters were initialized!
May 20 21:50:35 ferb kernel: [drm] Module unloaded

[Comment 1 Jeroen Roovers (RETIRED) 2017-05-21 20:17:08 UTC]

I suppose you already know where to store that patch.

[Comment 2 Octavian 2017-05-23 16:47:30 UTC]

Is this resolved? Since the *102 version still has the issue. 
The patch I applied myself a bit crude: while emerge was active I have modified that file by hand....

Are you sure is fixed upstream? 

Best regards,
Octavian

[Comment 3 Jeroen Roovers (RETIRED) 2017-06-03 13:50:15 UTC]

*** This bug has been marked as a duplicate of bug 602844 ***