3

I recently installed Ubuntu 16.04 on my personal notebook (Dell Inspiron 5548) and everytime I try to logout from any user account, I get a Kernel BUG ("Oops" message) related to amdgpu.

I tried both upgrading from 14 LTS and full installation using an 16.04 image on usb drive. I also have been using Ubuntu 15 in my computer for quite a while without any similar problems.

I have read that it might be related to the new amd graphic cards replacing fglrx and that only brand new graphic cards would be supported. However, I have installed Ubuntu 16 on Dell Studio 1458 (can't remember the graphic card now, but it is also Radeon) and it worked like a charm.

The BUG report follows below. Does any one knows some fix for this?

EDIT: my graphics card is AMD Radeon™ HD R7 M265.

May  3 10:57:58 ubuntu-5548 kernel: [  329.916153] [drm] PCIE GART of 2048M enabled (table at 0x0000000000040000).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155529] [drm] ring test on 0 succeeded in 15 usecs
May  3 10:57:58 ubuntu-5548 kernel: [  330.155722] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 1 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155736] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 2 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155747] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 3 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155757] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 4 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155766] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 5 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155775] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 6 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155784] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 7 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155793] [drm:gfx_v8_0_ring_test_ring [amdgpu]] *ERROR* amdgpu: cp failed to lock ring 8 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155821] [drm] ring test on 9 succeeded in 7 usecs
May  3 10:57:58 ubuntu-5548 kernel: [  330.155837] [drm:sdma_v2_4_ring_test_ring [amdgpu]] *ERROR* amdgpu: dma failed to lock ring 10 (-2).
May  3 10:57:58 ubuntu-5548 kernel: [  330.155844] [drm:amdgpu_resume [amdgpu]] *ERROR* resume 5 failed -2
May  3 10:57:58 ubuntu-5548 kernel: [  330.155852] [drm:amdgpu_resume_kms [amdgpu]] *ERROR* amdgpu_resume failed (-2).
May  3 10:57:58 ubuntu-5548 acpid: client 984[0:0] has disconnected
May  3 10:57:58 ubuntu-5548 acpid: client connected from 3312[0:0]
May  3 10:57:58 ubuntu-5548 acpid: 1 client rule loaded
May  3 10:57:59 ubuntu-5548 kernel: [  330.329604] BUG: unable to handle kernel NULL pointer dereference at 0000000000000248
May  3 10:57:59 ubuntu-5548 kernel: [  330.329631] IP: [<ffffffffc0348ea2>] amdgpu_vm_grab_id+0x122/0x310 [amdgpu]
May  3 10:57:59 ubuntu-5548 kernel: [  330.329667] PGD 0 
May  3 10:57:59 ubuntu-5548 kernel: [  330.329674] Oops: 0000 [#1] SMP 
May  3 10:57:59 ubuntu-5548 kernel: [  330.329686] Modules linked in: drbg ansi_cprng ctr ccm rfcomm bnep rtsx_usb_sdmmc rtsx_usb_ms memstick rtsx_usb nls_iso8859_1 intel_rapl x86_pkg_temp_thermal intel_powerclamp arc4 coretemp kvm_intel kvm iwlmvm mac80211 irqbypass crct10dif_pclmul dell_wmi crc32_pclmul dell_laptop sparse_keymap dcdbas snd_hda_codec_hdmi dell_smm_hwmon aesni_intel aes_x86_64 lrw gf128mul glue_helper iwlwifi ablk_helper cryptd dell_led btusb btrtl btbcm input_leds btintel serio_raw snd_hda_codec_realtek bluetooth hid_multitouch snd_hda_codec_generic joydev cfg80211 snd_soc_rt5640 snd_hda_intel snd_soc_ssm4567 snd_soc_rl6231 snd_hda_codec lpc_ich elan_i2c snd_soc_core snd_hda_core snd_hwdep snd_compress ac97_bus snd_pcm_dmaengine snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi snd_seq snd_seq_device snd_timer snd dw_dmac dw_dmac_core soundcore dell_rbtn snd_soc_sst_acpi shpchp mei_me i2c_designware_platform 8250_dw mei spi_pxa2xx_platform i2c_designware_core acpi_pad mac_hid uvcvideo videobuf2_vmalloc videobuf2_memops videobuf2_v4l2 videobuf2_core v4l2_common videodev media parport_pc ppdev lp parport autofs4 hid_generic usbhid amdkfd amd_iommu_v2 amdgpu i915 ttm psmouse i2c_algo_bit drm_kms_helper syscopyarea ahci sysfillrect libahci sysimgblt fb_sys_fops r8169 drm mii wmi video i2c_hid hid sdhci_acpi sdhci fjes
May  3 10:57:59 ubuntu-5548 kernel: [  330.330101] CPU: 0 PID: 163 Comm: gfx Not tainted 4.4.0-21-generic #37-Ubuntu
May  3 10:57:59 ubuntu-5548 kernel: [  330.330120] Hardware name: Dell Inc. Inspiron 5548/0YDTG3, BIOS A04 05/15/2015
May  3 10:57:59 ubuntu-5548 kernel: [  330.330140] task: ffff8804460e44c0 ti: ffff8804448dc000 task.ti: ffff8804448dc000
May  3 10:57:59 ubuntu-5548 kernel: [  330.330160] RIP: 0010:[<ffffffffc0348ea2>]  [<ffffffffc0348ea2>] amdgpu_vm_grab_id+0x122/0x310 [amdgpu]
May  3 10:57:59 ubuntu-5548 kernel: [  330.330197] RSP: 0018:ffff8804448dfce0  EFLAGS: 00010246
May  3 10:57:59 ubuntu-5548 kernel: [  330.330211] RAX: 0000000000000000 RBX: ffff880445530000 RCX: ffff88008eefd400
May  3 10:57:59 ubuntu-5548 kernel: [  330.330230] RDX: ffffffff81ef3cc0 RSI: ffff880445532d78 RDI: ffff880449cc3000
May  3 10:57:59 ubuntu-5548 kernel: [  330.330249] RBP: ffff8804448dfdb0 R08: ffff88008eefd400 R09: 000000018080004d
May  3 10:57:59 ubuntu-5548 kernel: [  330.330268] R10: ffff8803f30ee020 R11: 0000000000000004 R12: ffff880445532d78
May  3 10:57:59 ubuntu-5548 kernel: [  330.330286] R13: ffff880449cc3000 R14: ffff880445530838 R15: 0000000000000001
May  3 10:57:59 ubuntu-5548 kernel: [  330.330305] FS:  0000000000000000(0000) GS:ffff88045ec00000(0000) knlGS:0000000000000000
May  3 10:57:59 ubuntu-5548 kernel: [  330.330327] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
May  3 10:57:59 ubuntu-5548 kernel: [  330.330342] CR2: 0000000000000248 CR3: 0000000002e0a000 CR4: 00000000003406f0
May  3 10:57:59 ubuntu-5548 kernel: [  330.330361] Stack:
May  3 10:57:59 ubuntu-5548 kernel: [  330.330367]  ffff88042a99dc48 0000000000000000 ffff88008eefd400 0000000000000000
May  3 10:57:59 ubuntu-5548 kernel: [  330.330390]  0000000000000000 0000000000000000 0000000000000000 0000000000000000
May  3 10:57:59 ubuntu-5548 kernel: [  330.330412]  0000000000000000 0000000000000000 0000000000000000 0000000000000000
May  3 10:57:59 ubuntu-5548 kernel: [  330.330435] Call Trace:
May  3 10:57:59 ubuntu-5548 kernel: [  330.330454]  [<ffffffffc034ace0>] amdgpu_ib_schedule+0x90/0x390 [amdgpu]
May  3 10:57:59 ubuntu-5548 kernel: [  330.330487]  [<ffffffffc03873b6>] amdgpu_sched_run_job+0x36/0x140 [amdgpu]
May  3 10:57:59 ubuntu-5548 kernel: [  330.330519]  [<ffffffffc0386bcf>] amd_sched_main+0x23f/0x400 [amdgpu]
May  3 10:57:59 ubuntu-5548 kernel: [  330.330538]  [<ffffffff810c3a10>] ? wake_atomic_t_function+0x60/0x60
May  3 10:57:59 ubuntu-5548 kernel: [  330.330567]  [<ffffffffc0386990>] ? amd_sched_entity_wakeup+0x70/0x70 [amdgpu]
May  3 10:57:59 ubuntu-5548 kernel: [  330.330587]  [<ffffffff810a0528>] kthread+0xd8/0xf0
May  3 10:57:59 ubuntu-5548 kernel: [  330.330601]  [<ffffffff810a0450>] ? kthread_create_on_node+0x1e0/0x1e0
May  3 10:57:59 ubuntu-5548 kernel: [  330.330621]  [<ffffffff8182488f>] ret_from_fork+0x3f/0x70
May  3 10:57:59 ubuntu-5548 kernel: [  330.330636]  [<ffffffff810a0450>] ? kthread_create_on_node+0x1e0/0x1e0
May  3 10:57:59 ubuntu-5548 kernel: [  330.330653] Code: c0 44 89 bc 85 48 ff ff ff 41 83 c7 01 44 39 bb 1c 09 00 00 76 4f 49 83 c6 10 4d 8b 6e f0 4d 85 ed 74 66 4c 89 ef e8 fe 2e ff ff <8b> b8 48 02 00 00 48 8b b4 fd 50 ff ff ff 48 85 f6 74 b2 41 8b 
May  3 10:57:59 ubuntu-5548 kernel: [  330.330757] RIP  [<ffffffffc0348ea2>] amdgpu_vm_grab_id+0x122/0x310 [amdgpu]
May  3 10:57:59 ubuntu-5548 kernel: [  330.330788]  RSP <ffff8804448dfce0>
May  3 10:57:59 ubuntu-5548 kernel: [  330.330797] CR2: 0000000000000248
May  3 10:57:59 ubuntu-5548 kernel: [  330.337194] ---[ end trace f4393c5763eacaf5 ]---

3 Answers 3

7

Your symptoms are the same as the bug reported here:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1579374

You may log into launchpad and click the "it also affects me" link in the bug above if you want to bring more attention from the maintainers.

Until the bug is fixed in Ubuntu kernel packages, try to use the 4.6 upstream kernel packages as suggested by Renê Barbosa. You may download them here:

http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.6-rc7-wily/

Please download and install the following files:

  • linux-headers-4.6.0-040600rc7_4.6.0-040600rc7.201605081830_all.deb
  • linux-headers-4.6.0-040600rc7-generic_4.6.0-040600rc7.201605081830_amd64.deb
  • linux-image-4.6.0-040600rc7-generic_4.6.0-040600rc7.201605081830_amd64.deb

(Sorry, I cannot post direct links to the files because of askubuntu reputation restrictions)

1
  • Perfect! It took me some time to make the full installation of 16.04 again. After it, I updated / upgraded the system and installed the files you listed. It is working fine now, no bugs, no bug black screen anymore (so far, so good). Thank you very much!
    – fontanads
    May 20, 2016 at 23:00
0

Same issue here. Using the same graphics card.

EDIT: I got it fixed by installing the latest 4.6 kernel packages from the mainline repository. It looks like something needs to be backported to the default 4.4 kernel of Ubuntu 16.04.

1
  • Thank you very much. Following Paulo Matias advice, inspired by your answer, we found a solution that works for me at the moment.
    – fontanads
    May 20, 2016 at 23:03
0

I see you upgraded kernel and solved your problem. I have same specs and I upgraded to kernel 4.6 and now DRI_PRIME=1 glxgears works fine.

... but running games with DRI_PRIME=1 command brings me same (poor) result as running with integrated card. Is it normal or I am missing something?

2
  • Hi Bernardo. I'm sorry, I do not run any games on my notebook. Anyway, I'm not aware about how this flag DRI_PRIME works. Are there any tests or verifications I could do in my system that could help you compare the performance?
    – fontanads
    Jun 1, 2016 at 22:41
  • Hello. I don't know... DRI_PRIME=1 activates discrete card. I've searched and reached a conclusion that the open source driver from AMD is worse than Intel driver so the performance on games are indeed worse with discrete card in this case. Maybe when amdgpu PRO (closed source) comes, the performance of my discrete card will receive a boost. Since you don't run games, you can just stay with amdgpu.
    – bernardo
    Jun 15, 2016 at 15:35

Not the answer you're looking for? Browse other questions tagged .