Crash on desktop logout: nouveau ... AMD-Vi: Event logged [IO_PAGE_FAULT

Hardware is a Lenovo Legion 7 16ACHg6 (with Geforce RTX 3070) connected to Dell WD19 dock, running Ubuntu 24.04.

Boot with dock connected, log in - so far it works ok. Now logout - laptop screen and external monitors all go black - kernel shows

Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: drm: ttm_dma 0x00000000b1f4b5e2: pages NULL
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6c6440 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6c6460 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6d5c40 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6d5c60 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6ccd40 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: drm: [systemd-logind[1485]/00000001:fenceNonStallIntr] [BLOCK]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: drm: ttm_dma 0x00000000b1f4b5e2: pages NULL
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6ccd60 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6c2e40 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6c1340 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6cb540 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0xfb6c1360 flags=0x0000]
Feb 02 21:13:25 ubuntu2404 kernel: [drm:nouveau_framebuffer_new [nouveau]] offset=0 stride=6720 h=1050 gobs_in_block=16 bw=105 bh=9 gob_size=512 bl_size=7741440 size=8388608
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: [drm:drm_mode_addfb2_ioctl] [FB:134]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: drm: [systemd-logind[1485]/00000001:fenceNonStallIntr] [ALLOW]
Feb 02 21:13:25 ubuntu2404 kernel: nouveau 0000:01:00.0: drm: ttm_dma 0x00000000a0af9800: pages NULL

I tested with Ubuntu kernel v6.8.0-51 and mainline v6.13.0 with same result.

Kernel log: dmesg-logout.gz

Possibly related #225

Yeah... I'd say it's the same issue. Have you checked if the same workaround fixes it for you?

iommu=pt does not fix it.

what about iommu=off?

For some reason, I'm getting this immediately after logging in now and external monitors immediately fail.

Full dmesg: dmesg-ubuntu24-6.14.0-rc4-iommu-off.gz

With iommu=off, nouveau reports a DMA error:

[    9.583218] pci 0000:01:00.0: optimus capabilities: enabled, status dynamic power, hda bios codec supported
[    9.583226] VGA switcheroo: detected Optimus DSM method \_SB_.PCI0.GPP0.PEGP handle
[    9.583228] nouveau: detected PR support, will not use DSM
[    9.583284] nouveau 0000:01:00.0: enabling device (0000 -> 0003)
[    9.583596] nouveau 0000:01:00.0: NVIDIA GA104 (b74000a1)
[    9.699131] nouveau 0000:01:00.0: bios: version 94.04.3f.00.c4
[    9.704541] ------------[ cut here ]------------
[    9.704543] nouveau 0000:01:00.0: DMA addr 0x000000011a538000+4096 overflow (mask ffffffff, bus limit 0).
[    9.704550] WARNING: CPU: 3 PID: 587 at kernel/dma/direct.h:103 dma_map_page_attrs+0x31b/0x3f0
[    9.704555] Modules linked in: amdgpu(+) nouveau(+) mxm_wmi drm_gpuvm binfmt_misc i2c_algo_bit drm_ttm_helper ttm drm_panel_backlight_quirks intel_rapl_msr intel_rapl_comm
on kvm_amd kvm btusb snd_hda_codec_realtek btrtl polyval_clmulni snd_hda_codec_generic btintel polyval_generic snd_usb_audio(+) snd_hda_scodec_component ee1004 uvcvideo btbcm
 snd_hda_codec_hdmi ghash_clmulni_intel wmi_bmof snd_usbmidi_lib videobuf2_vmalloc btmtk aesni_intel snd_hda_intel snd_ump videobuf2_memops crypto_simd snd_intel_dspcfg snd_s
eq_midi uvc cryptd drm_suballoc_helper rapl snd_hda_codec snd_seq_midi_event videobuf2_v4l2 bluetooth cec video snd_hwdep k10temp snd_rawmidi videodev rc_core snd_hda_core i2
c_piix4 snd_seq videobuf2_common amdxcp i2c_smbus snd_pcm snd_seq_device drm_exec drm_buddy snd_timer gpu_sched snd mc drm_display_helper soundcore ccp joydev input_leds seri
o_raw wmi mac_hid sch_fq_codel msr parport_pc ppdev lp parport efi_pstore nfnetlink dmi_sysfs ip_tables x_tables autofs4 btrfs blake2b_generic xor raid6_pq
[    9.704633]  hid_microsoft ff_memless hid_cmedia r8152 mii usbhid nvme ucsi_acpi hid_multitouch nvme_core hid_generic r8169 typec_ucsi ahci libahci realtek nvme_auth typec
 i2c_hid_acpi i2c_hid hid uas usb_storage
[    9.704652] CPU: 3 UID: 0 PID: 587 Comm: (udev-worker) Not tainted 6.14.0-rc4-00266-gdf87d843c6eb #207
[    9.704654] Hardware name: LENOVO 82N6/LNVNB161216, BIOS GKCN65WW 01/16/2024
[    9.704655] RIP: 0010:dma_map_page_attrs+0x31b/0x3f0
[    9.704658] Code: c0 e8 f9 92 a0 00 4c 8b 45 c0 48 8d 4d c8 48 c7 c7 38 4b 2d 95 48 89 c6 41 50 4c 8b 4d b8 4d 89 f8 48 8b 55 b0 e8 d5 6d f0 ff <0f> 0b 4c 8b 55 a8 5a 48 c
7 c1 ff ff ff ff e9 d1 fd ff ff 49 8b 44
[    9.704659] RSP: 0018:ffffc08181b676e0 EFLAGS: 00010246
[    9.704661] RAX: 0000000000000000 RBX: ffff9923828c90c8 RCX: 0000000000000000
[    9.704663] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000
[    9.704664] RBP: ffffc08181b67740 R08: 0000000000000000 R09: 0000000000000000
[    9.704665] R10: 0000000000000000 R11: 0000000000000000 R12: ffffe1f404694e00
[    9.704666] R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000001000
[    9.704667] FS:  000075cd53c358c0(0000) GS:ffff992991b80000(0000) knlGS:0000000000000000
[    9.704668] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    9.704670] CR2: 00007fffbda33258 CR3: 00000001062e8000 CR4: 0000000000f50ef0
[    9.704671] PKRU: 55555554
[    9.704672] Call Trace:
[    9.704673]  <TASK>
[    9.704675]  ? show_regs.cold+0x19/0x24
[    9.704677]  ? dma_map_page_attrs+0x31b/0x3f0
[    9.704679]  ? __warn.cold+0xce/0x188
[    9.704681]  ? dma_map_page_attrs+0x31b/0x3f0
[    9.704684]  ? report_bug+0x110/0x160
[    9.704686]  ? __wake_up_klogd.part.0+0x40/0x70
[    9.704689]  ? handle_bug+0x6a/0xb0
[    9.704692]  ? exc_invalid_op+0x18/0x80
[    9.704694]  ? asm_exc_invalid_op+0x1b/0x20
[    9.704700]  ? dma_map_page_attrs+0x31b/0x3f0
[    9.704705]  nvkm_fb_ctor+0xbe/0xf0 [nouveau]
[    9.704759]  gf100_fb_new_+0x87/0xc0 [nouveau]
[    9.704810]  ga102_fb_new+0x5f/0x70 [nouveau]
[    9.704838]  nvkm_device_ctor+0xd84/0x3bb0 [nouveau]
[    9.704885]  nvkm_device_pci_new+0x117/0x2d0 [nouveau]
[    9.704919]  nouveau_drm_probe+0x56/0x200 [nouveau]
[    9.704966]  local_pci_probe+0x42/0x90
[    9.704970]  pci_device_probe+0xea/0x2b0
[    9.704974]  really_probe+0xf0/0x360
[    9.704976]  ? pm_runtime_barrier+0x55/0x90
[    9.704980]  __driver_probe_device+0x8c/0x140
[    9.704983]  driver_probe_device+0x24/0xd0
[    9.704985]  __driver_attach+0xdd/0x210
[    9.704988]  ? __device_attach_driver+0x150/0x150
[    9.704990]  bus_for_each_dev+0x8f/0xf0
[    9.704993]  driver_attach+0x1e/0x30
[    9.704995]  bus_add_driver+0x14d/0x240
[    9.704998]  driver_register+0x73/0xe0
[    9.705001]  __pci_register_driver+0x7d/0x90
[    9.705003]  nouveau_drm_init+0x1b9/0x1000 [nouveau]
[    9.705036]  ? 0xffffffffc0fa0000
[    9.705038]  do_one_initcall+0x76/0x3e0
[    9.705044]  do_init_module+0x97/0x270
[    9.705047]  load_module+0x1f18/0x2530
[    9.705057]  init_module_from_file+0x97/0xe0
[    9.705059]  ? init_module_from_file+0x97/0xe0
[    9.705066]  idempotent_init_module+0x135/0x350
[    9.705073]  __x64_sys_finit_module+0x6b/0xd0
[    9.705076]  x64_sys_call+0x2f5/0xfb0
[    9.705078]  do_syscall_64+0x87/0x140
[    9.705080]  ? _raw_spin_unlock+0x22/0x50
[    9.705083]  ? irqentry_exit_to_user_mode+0xc2/0x210
[    9.705086]  ? irqentry_exit+0x6f/0xa0
[    9.705087]  ? common_interrupt+0x64/0xe0
[    9.705090]  entry_SYSCALL_64_after_hwframe+0x4b/0x53
[    9.705092] RIP: 0033:0x75cd53b2725d
[    9.705094] Code: ff c3 66 2e 0f 1f 84 00 00 00 00 00 90 f3 0f 1e fa 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c
3 48 8b 0d 8b bb 0d 00 f7 d8 64 89 01 48
[    9.705096] RSP: 002b:00007ffde5ba32b8 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
[    9.705098] RAX: ffffffffffffffda RBX: 0000579dbfc299f0 RCX: 000075cd53b2725d
[    9.705099] RDX: 0000000000000000 RSI: 000075cd53cbe07d RDI: 000000000000002b
[    9.705100] RBP: 00007ffde5ba3370 R08: 0000000000000040 R09: 00007ffde5ba3300
[    9.705101] R10: 000075cd53c03b20 R11: 0000000000000246 R12: 000075cd53cbe07d
[    9.705102] R13: 0000000000020000 R14: 0000579dbfbd9360 R15: 0000579dbfbe87b0
[    9.705107]  </TASK>
[    9.705108] irq event stamp: 130369
[    9.705110] hardirqs last  enabled at (130375): [<ffffffff93cc9fb5>] __up_console_sem+0x75/0x90
[    9.705112] hardirqs last disabled at (130380): [<ffffffff93cc9f9a>] __up_console_sem+0x5a/0x90
[    9.705114] softirqs last  enabled at (125292): [<ffffffff93c139d8>] __irq_exit_rcu+0xb8/0xf0
[    9.705116] softirqs last disabled at (125267): [<ffffffff93c139d8>] __irq_exit_rcu+0xb8/0xf0
[    9.705118] ---[ end trace 0000000000000000 ]---
[    9.725149] nouveau 0000:01:00.0: fb: 8192 MiB GDDR6
[    9.772426] nouveau 0000:01:00.0: drm: VRAM: 8192 MiB
[    9.772429] nouveau 0000:01:00.0: drm: GART: 536870912 MiB
[    9.772432] nouveau 0000:01:00.0: drm: BIT table 'A' not found
[    9.772434] nouveau 0000:01:00.0: drm: BIT table 'L' not found
[    9.772435] nouveau 0000:01:00.0: drm: TMDS table version 2.0
[    9.775216] nouveau 0000:01:00.0: drm: MM: using COPY for buffer copies

(the full dmesg also contains a later lock error which was previously reported at #371 )