From mboxrd@z Thu Jan 1 00:00:00 1970
From: bugzilla-daemon-CC+yJ3UmIYqDUpFQwHEjaQ@public.gmane.org
Subject: [Bug 100423] New: MMIO read of 00000000 FAULT at 022554 [
IBUS ]
Date: Mon, 27 Mar 2017 20:42:44 +0000
Message-ID:
Bug ID
100423
Summary
MMIO read of 00000000 FAULT at 022554 [ IBUS ]
Product
xorg
Version
unspecified
Hardware
Other
OS
All
Status
NEW
Severity
normal
Priority
medium
Component
Driver/nouveau
Assignee
nouveau@lists.freedesktop.org
Reporter
bruno.n.pagani@gmail.com
QA Contact
xorg-team@lists.x.org
Created attachment 130492 =
[details]
dmesg output
Hi there,
I=E2=80=99ve got a new laptop since a couple months, and I=E2=80=99m starti=
ng to configure it
intensively and report bug. It=E2=80=99s an HP zBook Studio G3, with a Core=
i7-6700HQ
with Intel HD graphics and a dedicated NVIDIA Corporation GM107GLM [Quadro
M1000M] card. Amongst the issue I find in dmesg, there is those two lines t=
hat
appear in red:
[ 16.246988] nouveau 0000:01:00.0: priv: HUB0: 614900 00800000 (18408200)
[ 16.248921] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 02=
2554
[ IBUS ]
I=E2=80=99ve attached the full dmesg.
I have similar issue on dell precision 5520 with=20 Intel(R) Core(TM) i7-7820HQ CPU @ 2.90GHz Has nvidia maxwell (NV117) and 4k touch screen I see these : nouveau 0000:01:00.0: fb: 4096 MiB GDDR5=20 nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 10ac08 [ IBUS ]= =20 nouveau 0000:01:00.0: priv: HUB0: 10ecc0 ffffffff (1f40822c)=20 vga_switcheroo: enabled [TTM] Zone kernel: Available graphics memory: 16334986 kiB=20 [TTM] Zone dma32: Available graphics memory: 2097152 kiB=20 [TTM] Initializing pool allocator=20 [TTM] Initializing DMA pool allocator=20 nouveau 0000:01:00.0: DRM: VRAM: 4096 MiB=20 nouveau 0000:01:00.0: DRM: GART: 1048576 MiB=20 nouveau 0000:01:00.0: DRM: Pointer to TMDS table invalid=20 nouveau 0000:01:00.0: DRM: DCB version 4.0=20 nouveau 0000:01:00.0: DRM: Pointer to flat panel table invalid
What | Removed | Added |
---|---|---|
OS | All | Linux (All) |
Hardware | Other | x86-64 (AMD64) |
CC | arthurborsboom@gmail.com |
Same for: Intel i7-4710HQ + GM107M [GeForce GTX 850M] [ 1.950230] nouveau: detected PR support, will not use DSM [ 1.950384] nouveau 0000:01:00.0: NVIDIA GM107 (117110a2) [ 1.979948] nouveau 0000:01:00.0: bios: version 82.07.47.00.06 [ 2.071839] nouveau 0000:01:00.0: fb: 2048 MiB DDR3 [ 2.071853] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 02= 2554 [ IBUS ] [ 2.071913] nouveau 0000:01:00.0: priv: HUB0: 10ecc0 ffffffff (1840822c) [ 3.157435] nouveau 0000:01:00.0: DRM: VRAM: 2048 MiB [ 3.157436] nouveau 0000:01:00.0: DRM: GART: 1048576 MiB [ 3.157438] nouveau 0000:01:00.0: DRM: Pointer to TMDS table invalid [ 3.157438] nouveau 0000:01:00.0: DRM: DCB version 4.0 [ 3.157439] nouveau 0000:01:00.0: DRM: Pointer to flat panel table inval= id [ 3.227691] nouveau 0000:01:00.0: hwmon_device_register() is deprecated. Please convert the driver to use hwmon_device_register_with_info(). [ 3.286804] nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies [ 3.286815] [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on minor 1
Almost same for i7-6700, GM206 [GTX 950] No touch screen, no hybrid graphics. Pure desktop. Intel 530 on chip but disabled in BIOS. Only one Asus Graphic Card. Problem started on Kernel 4.10.x and is going on 4.11.0. Kernel 4.9.x is FA= ULT free. Different distributions Fedora 25, Gentoo. ------------------------------------ dmesg | grep -i nouveau [ 1.688288] fb: switching to nouveaufb from EFI VGA [ 1.689206] nouveau 0000:01:00.0: NVIDIA GM206 (126020a1) [ 1.787066] nouveau 0000:01:00.0: bios: version 84.06.3d.00.6f [ 1.787209] nouveau 0000:01:00.0: disp: dcb 15 type 8 unknown [ 1.787571] nouveau 0000:01:00.0: fb: 2048 MiB GDDR5 [ 1.787585] nouveau 0000:01:00.0: bus: MMIO write of 80000136 FAULT at 10eb14 [ IBUS ] [ 1.794508] nouveau 0000:01:00.0: DRM: VRAM: 2048 MiB [ 1.794509] nouveau 0000:01:00.0: DRM: GART: 1048576 MiB [ 1.794512] nouveau 0000:01:00.0: DRM: TMDS table version 2.0 [ 1.794513] nouveau 0000:01:00.0: DRM: DCB version 4.1 [ 1.794515] nouveau 0000:01:00.0: DRM: DCB outp 00: 01000f02 00020030 [ 1.794517] nouveau 0000:01:00.0: DRM: DCB outp 01: 02000f00 00000000 [ 1.794518] nouveau 0000:01:00.0: DRM: DCB outp 02: 02811f66 04400010 [ 1.794520] nouveau 0000:01:00.0: DRM: DCB outp 03: 02011f62 00020010 [ 1.794522] nouveau 0000:01:00.0: DRM: DCB outp 04: 02022f72 00020020 [ 1.794523] nouveau 0000:01:00.0: DRM: DCB outp 05: 04033f82 00020030 [ 1.794525] nouveau 0000:01:00.0: DRM: DCB outp 15: 01df4ff8 00000000 [ 1.794526] nouveau 0000:01:00.0: DRM: DCB conn 00: 00001030 [ 1.794528] nouveau 0000:01:00.0: DRM: DCB conn 01: 00010146 [ 1.794529] nouveau 0000:01:00.0: DRM: DCB conn 02: 00020261 [ 1.794531] nouveau 0000:01:00.0: DRM: DCB conn 03: 02000331 [ 1.794532] nouveau 0000:01:00.0: DRM: DCB conn 04: 00000470 [ 1.794534] nouveau 0000:01:00.0: DRM: Pointer to flat panel table inval= id [ 1.828756] nouveau 0000:01:00.0: DRM: unknown connector type 70 [ 1.828775] nouveau 0000:01:00.0: DRM: failed to create encoder 1/8/0: -= 19 [ 1.828778] nouveau 0000:01:00.0: DRM: Unknown-1 has no encoders, removi= ng [ 1.934421] nouveau 0000:01:00.0: hwmon_device_register() is deprecated. Please convert the driver to use hwmon_device_register_with_info(). [ 1.993646] nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies [ 2.072744] nouveau 0000:01:00.0: priv: GPC0: 419df4 00000000 (1c40820e) [ 2.072763] nouveau 0000:01:00.0: priv: GPC1: 419df4 00000000 (1c40820e) [ 2.174707] nouveau 0000:01:00.0: DRM: allocated 1600x1200 fb: 0x60000, = bo ffff9b09ca6eb800 [ 2.176838] fbcon: nouveaufb (fb0) is primary device [ 2.375660] nouveau 0000:01:00.0: fb0: nouveaufb frame buffer device [ 2.389828] [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on minor 0 Almost same, because my FAULT is at 10eb14=20 ------------------------------------ 01:00.0 VGA compatible controller: NVIDIA Corporation GM206 [GeForce GTX 95= 0] (rev a1) (prog-if 00 [VGA controller]) Subsystem: ASUSTeK Computer Inc. Device 8555 Flags: bus master, fast devsel, latency 0, IRQ 134 Memory at de000000 (32-bit, non-prefetchable) [size=3D16M] Memory at 2fe0000000 (64-bit, prefetchable) [size=3D256M] Memory at 2ff0000000 (64-bit, prefetchable) [size=3D32M] I/O ports at e000 [size=3D128] Expansion ROM at 000c0000 [disabled] [size=3D128K] Capabilities: [60] Power Management version 3 Capabilities: [68] MSI: Enable+ Count=3D1/1 Maskable- 64bit+ Capabilities: [78] Express Legacy Endpoint, MSI 00 Capabilities: [100] Virtual Channel Capabilities: [258] L1 PM Substates Capabilities: [128] Power Budgeting <?> Capabilities: [420] Advanced Error Reporting Capabilities: [600] Vendor Specific Information: ID=3D0001 Rev=3D1 = Len=3D024 <?> Capabilities: [900] #19 Kernel driver in use: nouveau Kernel modules: nouveau ------------------------------------ xorg-x11-drv-nouveau-1.0.14-2.fc25.x86_64 libdrm-2.4.79-1.fc25.x86_64 and: xf86-video-nouveau-1.0.13 libdrm-2.4.75 Should I open a new bug with "FAULT at 10eb14" or is it this same= issue?
@gene, Johnny B. Goode: From what I know, the =E2=80=9Cat= =E2=80=9D value matters, because they have meanings. Though from one generation to another this could be different values for the same register. Also, @Johnny, in your case it=E2=80=99s a write instead of a read, so = I would definitively open another issue.
I've bisected the issue. It happens with this commit: fb/gf100-: modify constructors to allow more customisation
Looks like 22554 is no longer available on Maxwell. See where = ltc reads it from for the new location. Needs to be split out of ramgk104 somehow.
(In reply to Ilia Mirkin from comment #6) > Looks like 22554 is no longer available on Maxwe= ll. See where ltc reads it > from for the new location. Needs to be split out of ramgk104 somehow.<= /span > I have WIP patches in my Fermi reclocking code (I'll try and push before I leave for xmas break) that begin to address this.
There is a new one now (10ac08, in addition to 022554): ``` [ 10.667586] nouveau 0000:01:00.0: enabling device (0006 -> 0007) [ 10.667788] nouveau 0000:01:00.0: NVIDIA GM107 (117310a2) [ 10.695951] nouveau 0000:01:00.0: bios: version 82.07.a9.00.0c [ 10.849391] nouveau 0000:01:00.0: fb: 4096 MiB GDDR5 [ 10.849404] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 02= 2554 [ IBUS ] [ 10.851426] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 10= ac08 [ IBUS ] [ 11.937159] nouveau 0000:01:00.0: DRM: VRAM: 4096 MiB [ 11.937160] nouveau 0000:01:00.0: DRM: GART: 1048576 MiB [ 11.937163] nouveau 0000:01:00.0: DRM: TMDS table version 2.0 [ 11.937164] nouveau 0000:01:00.0: DRM: DCB version 4.0 [ 11.937165] nouveau 0000:01:00.0: DRM: DCB outp 01: 02811fa6 04420010 [ 11.937167] nouveau 0000:01:00.0: DRM: DCB outp 02: 02011f62 04420010 [ 11.937168] nouveau 0000:01:00.0: DRM: DCB outp 03: 08822fc6 04420010 [ 11.937169] nouveau 0000:01:00.0: DRM: DCB outp 04: 08022f82 04420010 [ 11.937170] nouveau 0000:01:00.0: DRM: DCB conn 01: 00010146 [ 11.937171] nouveau 0000:01:00.0: DRM: DCB conn 02: 01000246 [ 12.126552] nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies [ 12.233419] [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on minor 1 ``` Should I open a new issue for this one?
(In reply to Bruno Pagani from comment #8) >=20 > Should I open a new issue for this one? Yes please, where there appears a new register on differing hardware family. It's an easy operation to merge bugs later on should they be due to the same underlying cause, but almost impossible to separate them out.
Done: https://bugs.freedesktop.org/show_bug.= cgi?id=3D104835
Hi folks, I have the very similar issue about drm, bus read FAULT, etc. However, my problem is more severe than just red lines in dmesg. I have ideapad 700 with HD Graphics 530 as the integrated and GeForce GTX 9= 50M as the discrete card. I sit on gentoo with Linux-4.15.1. The first point is that if I use nvidia drivers than my brightness buttons = do not work (not visible for anything). Summing this up with the total no need= of a powerful graphics I decided to go to the bumblebee just in case I will in= deed need a better gpu in action. That is to fire up the discrete card on demand. My problem is that upon starting some glx-aware program at the moment when bumblebee is trying to start a separate hidden X server with nouveau as the driver my primary X server restarts. After this restart and in attempt to run some glx-aware program again I see= X errors and nothing good. They boil down to the fact that drm cannot ecquire= an access to card0 and card1 because of permission issues. Well, the primary X server now does not restart though. I tried all what bumblebee guys suggest - no way. Also, everything is smooth with the nvidia driver in this config. I think it is those red messages in dmesg are the primary indicator of my troubles.
@Alex You should not use Bumblebee with nouveau. Also your= bug is similar to an old of mine: https://bugs.freedesktop.org/show_bug.c= gi?id=3D98384. So part of your issue is that your system is likely not up-to-date. But the correct solution is to not use nouveau with Bumblebee. And in any c= ase, it has likely nothing to do with this issue.=E2=80=AF;)
@Bruno, thanks for the comment. I will check it out. Bumbl= ebee: you see, I'm fine with intel, I do not need extreme speeds and I have 16GB in RAM ... nvidia - I could manage to install it either globally or in bumblebee but it blocks the brightness control when used permanently - very weird, so only bumblebee would go. Something went wrong with setting nouveau globally and I have given up. Will retry. But it will not suit me if the brightness keys are blocked ... And yes, my system is nearly the tomorrow's version. It is up-to date. At l= east the kernel is the one released few days ago ...
(In reply to Alex from comment #13) > @Bruno, thanks for the comment. I will check= it out. Bumblebee: you see, I'm > fine with intel, I do not need extreme speeds and I have 16GB in RAM .= .. > nvidia - I could manage to install it either globally or in bumblebee = but it > blocks the brightness control when used permanently - very weird, so o= nly > bumblebee would go. With nouveau loaded (assuming things go properly), your GPU should suspend = when not in use.
> Something when wrong with nouveau g= lobaly If you have a quite recent system, maybe you are affected by https://bu= gzilla.kernel.org/show_bug.cgi?id=3D156341.
Still exists in 4.16.0-rc2: nouveau 0000:01:00.0: fb: 2048 MiB GDDR5 nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 022554 [ IBUS ] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 10ac08 [ IBUS ] Machine is HP ZBook 17 G3, the rest (bootlogs, etc) can be found in BUG 100213 report.
still there with 4.13.0-37 ubuntu on HP Spectre, Core i7 7500U= , intel HD Graphics 620, GM108M (GeForce 940MX) : nouveau 0000:01:00.0: NVIDIA GM108 (1180d0a2)=20=20=20=20=20=20=20=20 nouveau 0000:01:00.0: bios: version 82.08.62.00.15=20=20=20 nouveau 0000:01:00.0: fb: 2048 MiB GDDR5=20=20=20=20=20=20=20=20=20=20=20= =20=20 nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 612004 [ IBUS ]= =20=20=20=20=20=20=20 nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 10ac08 [ IBUS ]= =20=20=20=20=20=20=20 nvme0n1: p1 p2 p3=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20= =20=20=20=20=20=20=20=20=20=20=20=20=20=20=20=20 [drm] Memory usable by graphics device =3D 4096M
(In reply to Mikael T from comment #17) > still there with 4.13.0-37 ubuntu on HP Spectre,= Core i7 7500U, intel HD > Graphics 620, GM108M (GeForce 940MX) : >=20 > nouveau 0000:01:00.0: NVIDIA GM108 (1180d0a2) > nouveau 0000:01:00.0: bios: version 82.08.62.00.15 > nouveau 0000:01:00.0: fb: 2048 MiB GDDR5 > nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 612004 [ IBU= S ] > nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 10ac08 [ IBU= S ] Those are differents errors. You should open a new bug report for the first one. The second one is https://bugs.freedesktop.org/show_bug.= cgi?id=3D104835.
I think I'm also seeing this with Fedora 29 Beta, I was also s= eeing errors with updated Fedora 28. [ 1.614055] nouveau: detected PR support, will not use DSM [ 1.614243] nouveau 0000:01:00.0: NVIDIA GM108 (118070a2) [ 1.629639] nouveau 0000:01:00.0: bios: version 82.08.3b.00.4b [ 1.714600] nouveau 0000:01:00.0: fb: 2048 MiB DDR3 [ 1.714622] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 60= 13d4 [ IBUS ] [ 1.729900] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 10= ac08 [ IBUS ] [ 2.372521] nouveau 0000:01:00.0: DRM: VRAM: 2048 MiB [ 2.372522] nouveau 0000:01:00.0: DRM: GART: 1048576 MiB [ 2.372526] nouveau 0000:01:00.0: DRM: Pointer to TMDS table invalid [ 2.372543] nouveau 0000:01:00.0: DRM: DCB version 4.0 [ 2.373356] nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies [ 2.373363] [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on minor 1 [ 8.044183] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 61= 9444 [ IBUS ] [ 9.126208] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 61= 9444 [ IBUS ] [ 93.901280] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 60= 13d4 [ IBUS ] [ 101.777394] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 60= 13d4 [ IBUS ] [ 112.733004] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 60= 13d4 [ IBUS ] I'm also on a laptop (Asus Zenbook UX303U) with hybrid graphics (Skylake GT2 [HD Graphics 520] (rev 07) / GM108M [GeForce 940M] (rev a2)).
(In reply to boldingd from comment #19) > I think I'm also seeing this with Fedora 29 Beta= , I was also seeing errors > with updated Fedora 28. No, yours are different issues. > [ 1.729900] nouveau 0000:01:00.0: bus: MMIO r= ead of 00000000 FAULT at > 10ac08 [ IBUS ] This is https://bugs.freedesktop.org/show_bug.= cgi?id=3D104835 > [ 8.044183] nouveau 0000:01:00.0: bus: MMIO r= ead of 00000000 FAULT at > 619444 [ IBUS ] > [ 9.126208] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT = at > 619444 [ IBUS ] I haven=E2=80=99t found any reference for this one. You might open a new bu= greport for it. > [ 1.714622] nouveau 0000:01:00.0: bus: MMIO r= ead of 00000000 FAULT at > 6013d4 [ IBUS ] > [ 93.901280] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT = at > 6013d4 [ IBUS ] > [ 101.777394] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT = at > 6013d4 [ IBUS ] > [ 112.733004] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT = at > 6013d4 [ IBUS ] This one also appears (and is briefly discussed) in https://bugs.freedesktop.org/show_bug.c= gi?id=3D98386
From my dmesg. [ 12.189523] ax88179_178a 2-2:1.0 enp0s20f0u2: ax88179 - Link status is: 0 [ 15.645518] ax88179_178a 2-2:1.0 enp0s20f0u2: ax88179 - Link status is: 1 [ 29.886049] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT at 60= 13d4 [ IBUS ] [ 57.810423] logitech-hidpp-device 0003:046D:4013.0009: HID++ 2.0 device connected. [ 60.146429] logitech-hidpp-device 0003:046D:4002.000A: HID++ 2.0 device connected. I am also getting this error kernel 4.19.5-1-default openSUSE Tumbleweed
(In reply to Cameron from comment #21) > From my dmesg. >=20 > [ 12.189523] ax88179_178a 2-2:1.0 enp0s20f0u2: ax88179 - Link status= is: 0 > [ 15.645518] ax88179_178a 2-2:1.0 enp0s20f0u2: ax88179 - Link status= is: 1 > [ 29.886049] nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAULT = at > 6013d4 [ IBUS ] > [ 57.810423] logitech-hidpp-device 0003:046D:4013.0009: HID++ 2.0 de= vice > connected. > [ 60.146429] logitech-hidpp-device 0003:046D:4002.000A: HID++ 2.0 de= vice > connected. >=20 >=20 > I am also getting this error >=20 > kernel 4.19.5-1-default > openSUSE Tumbleweed No, as I said just above your post, this would be https://bugs.freedesktop.org/show_bug.c= gi?id=3D98386 The hex value after =E2=80=9Cat=E2=80=9D does matter.=E2=80=AF;)
openSUSE Tumbleweed Kernel 4.19.5-1-default Getting these messages Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: fb: 4096 MiB GDDR5 Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: bus: MMIO read of 00000= 000 FAULT at 022554 [ IBUS ] Dec 05 07:35:28 flisk kernel: fbcon: Taking over console Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: bus: MMIO read of 00000= 000 FAULT at 10ac08 [ IBUS ] Dec 05 07:35:28 flisk kernel: usb 1-4: New USB device found, idVendor=3D808= 7, idProduct=3D0a2b, bcdDevice=3D 0.10 Dec 05 07:35:28 flisk kernel: usb 1-4: New USB device strings: Mfr=3D0, Product=3D0, SerialNumber=3D0 Dec 05 07:35:28 flisk kernel: usb 3-1: New USB device found, idVendor=3D042= 4, idProduct=3D2137, bcdDevice=3D60.80 Dec 05 07:35:28 flisk kernel: usb 3-1: New USB device strings: Mfr=3D1, Product=3D2, SerialNumber=3D0 Dec 05 07:35:28 flisk kernel: usb 3-1: Product: USB2137B Dec 05 07:35:28 flisk kernel: usb 3-1: Manufacturer: SMSC Dec 05 07:35:28 flisk kernel: hub 3-1:1.0: USB hub found Dec 05 07:35:28 flisk kernel: hub 3-1:1.0: 7 ports detected Dec 05 07:35:28 flisk kernel: vga_switcheroo: enabled Dec 05 07:35:28 flisk kernel: [TTM] Zone kernel: Available graphics memory: 16391158 kiB Dec 05 07:35:28 flisk kernel: [TTM] Zone dma32: Available graphics memory: 2097152 kiB Dec 05 07:35:28 flisk kernel: [TTM] Initializing pool allocator Dec 05 07:35:28 flisk kernel: [TTM] Initializing DMA pool allocator Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: DRM: VRAM: 4096 MiB Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: DRM: GART: 1048576 MiB Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: DRM: Pointer to TMDS ta= ble invalid Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: DRM: DCB version 4.0 Dec 05 07:35:28 flisk kernel: nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies Dec 05 07:35:28 flisk kernel: [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on minor 1
I'm getting one of these too on Fedora 28 KDE Plasma spin. Acer Aspire VN7-592G/Aspire VN7-592G, BIOS V1.13 03/21/2018 Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz NVIDIA Corporation GM107M [GeForce GTX 960M] (rev a2) $ cat /proc/version Linux version 4.19.6-200.fc28.x86_64 (mockbuil= d@bkernel04.phx2.fedoraproject.org) (gcc version 8.2.1 20181105 (Red Hat 8.2.1-5) (GCC)) #1 SMP Sun Dec 2 18:03:32 UTC 2018 $ journalctl -b --no-hostname -o short-monotonic | grep nouveau [ 2.515464] kernel: nouveau: detected PR support, will not use DSM [ 2.515480] kernel: nouveau 0000:01:00.0: enabling device (0006 -> 00= 07) [ 2.515620] kernel: nouveau 0000:01:00.0: NVIDIA GM107 (1171b0a2) [ 2.537544] kernel: nouveau 0000:01:00.0: bios: version 82.07.80.00.2a [ 2.586306] kernel: nouveau 0000:01:00.0: fb: 2048 MiB GDDR5 [ 2.586357] kernel: nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAU= LT at 022554 [ IBUS ] [ 2.599343] kernel: nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAU= LT at 10ac08 [ IBUS ] [ 3.260508] kernel: nouveau 0000:01:00.0: DRM: VRAM: 2048 MiB [ 3.260509] kernel: nouveau 0000:01:00.0: DRM: GART: 1048576 MiB [ 3.260511] kernel: nouveau 0000:01:00.0: DRM: Pointer to TMDS table inv= alid [ 3.260512] kernel: nouveau 0000:01:00.0: DRM: DCB version 4.0 [ 3.261005] kernel: nouveau 0000:01:00.0: DRM: MM: using COPY for buffer copies [ 3.261011] kernel: [drm] Initialized nouveau 1.3.1 20120801 for 0000:01:00.0 on minor 1 [ 7.846986] kernel: nouveau 0000:01:00.0: bus: MMIO read of 00000000 FAU= LT at 619444 [ IBUS ]
(In reply to Bruno Pagani from comment #4) > @gene, Johnny B. Goode: From what I know, th= e =E2=80=9Cat=E2=80=9D value matters, because > they have meanings. Though from one generation to another this could be > different values for the same register. >=20 > Also, @Johnny, in your case it=E2=80=99s a write instead of a read= , so I would > definitively open another issue. In the end I opened 2 new issues with MMIO FAULT #108980 #108982 Bug, in different variations, started in kernel 4.10 still exist. There is a thousands searchings for this bug in google.