Bug 700782 – Kernel >=5.3.11 hangs intel graphics card

• Product: Gentoo Linux
• Status: RESOLVED FIXED

[Description cazzantonio 2019-11-20 15:22:55 UTC]

Updating from gentoo-sources-5.3.10 to 5.3.11, sometimes the desktop hangs for few seconds becoming unresponsive and I get this error in dmesg:

[   36.202250] i915 0000:00:02.0: Resetting rcs0 for hang on rcs0

A way to reproduce this 100% of times is opening at least 2-3 times a document with libreoffice.

I tried to get the vanilla sources and go with git bisect (following this: https://wiki.gentoo.org/wiki/Kernel_git-bisect) and I got the incriminated commit:

77fc9100fc5768ca01ca2dd2cc5a515a4723a58a is the first bad commit
commit 77fc9100fc5768ca01ca2dd2cc5a515a4723a58a
Author: Jon Bloomfield <[e-mail]>
Date:   Thu Sep 27 10:23:17 2018 -0700

    drm/i915/cmdparser: Use explicit goto for error paths
   
    commit 0546a29cd884fb8184731c79ab008927ca8859d0 upstream.
   
    In the next patch we will be adding a second valid
    termination condition which will require a small
    amount of refactoring to share logic with the BB_END
    case.
   
    Refactor all error conditions to jump to a dedicated
    exit path, with 'break' reserved only for a successful
    parse.
   
    Cc: Tony Luck <[e-mail]>
    Cc: Dave Airlie <[e-mail]>
    Cc: Takashi Iwai <[e-mail]>
    Cc: Tyler Hicks <[e-mail]>
    Signed-off-by: Jon Bloomfield <[e-mail]>
    Reviewed-by: Chris Wilson <[e-mail]>
    Signed-off-by: Greg Kroah-Hartman <[e-mail]>

 drivers/gpu/drm/i915/i915_cmd_parser.c | 25 +++++++++++++------------
 1 file changed, 13 insertions(+), 12 deletions(-)

Please patch this or notify upstream in order to avoid this annoying problem. The hangs tend to happen quite frequently.

My system:
heavensdoor ~ # inxi -Fzm
System:    Host: heavensdoor Kernel: 5.3.8-gentoo x86_64 bits: 64 Desktop: Xfce 4.14.1 Distro: Gentoo Base System release 2.6
Machine:   Type: Laptop System: Notebook product: P9XXRC v: N/A serial: N/A
           Mobo: Notebook model: P9XXRC serial: N/A UEFI: INSYDE v: 1.07.04 date: 05/03/2019
Battery:   ID-1: BAT0 charge: 59.2 Wh condition: 59.2/56.2 Wh (105%)
Memory:    RAM: total: 31.08 GiB used: 803.1 MiB (2.5%)
           Array-1: capacity: 32 GiB slots: 2 EC: None
           Device-1: ChannelA-DIMM0 size: 16 GiB speed: 3000 MT/s
           Device-2: ChannelB-DIMM0 size: 16 GiB speed: 3000 MT/s
CPU:       Topology: 6-Core model: Intel Core i7-9750H bits: 64 type: MT MCP L2 cache: 12.0 MiB
           Speed: 800 MHz min/max: 800/4500 MHz Core speeds (MHz): 1: 800 2: 800 3: 801 4: 800 5: 801 6: 800 7: 800 8: 800
           9: 800 10: 800 11: 801 12: 801
Graphics:  Device-1: Intel UHD Graphics 630 driver: i915 v: kernel
           Device-2: NVIDIA TU116M [GeForce GTX 1660 Mobile] driver: N/A
           Display: server: X.Org 1.20.5 driver: intel unloaded: modesetting,vesa resolution: 1920x1080~144Hz
           OpenGL: renderer: Mesa DRI Intel UHD Graphics 630 (Coffeelake 3x8 GT2) v: 4.5 Mesa 19.1.8
Audio:     Device-1: Intel Cannon Lake PCH cAVS driver: snd_hda_intel
           Device-2: NVIDIA driver: snd_hda_intel
           Sound Server: ALSA v: k5.3.8-gentoo
Network:   Device-1: Qualcomm Atheros AR9462 Wireless Network Adapter driver: ath9k
           IF: wlp8s0 state: up mac: <filter>
           Device-2: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet driver: r8169
           IF: enp9s0 state: down mac: <filter>
           IF-ID-1: sit0 state: down mac: <filter>
           IF-ID-2: tunl0 state: down mac: <filter>
Drives:    Local Storage: total: 1.38 TiB used: 173.88 GiB (12.3%)
           ID-1: /dev/nvme0n1 vendor: Samsung model: SSD 970 PRO 512GB size: 476.94 GiB
           ID-2: /dev/sda vendor: Samsung model: SSD 860 EVO 1TB size: 931.51 GiB
Partition: ID-1: / size: 100.00 GiB used: 7.93 GiB (7.9%) fs: btrfs dev: /dev/dm-0
           ID-2: /home size: 276.45 GiB used: 4.57 GiB (1.7%) fs: btrfs dev: /dev/dm-1
Sensors:   System Temperatures: cpu: 58.0 C mobo: N/A
           Fan Speeds (RPM): N/A
Info:      Processes: 296 Uptime: 44m Shell: bash inxi: 3.0.36




Reproducible: Always

[Comment 1 cazzantonio 2019-11-21 19:22:32 UTC]

problem still present in 5.3.12

[Comment 2 Chris 2019-11-22 19:36:31 UTC]

Experiencing this exact problem without LibreOffice launch or running. Have not been able to isolate to specific software, but does seem to occur with Firefox and Evolution running (not only during launch).

[Comment 3 Mike Pagano 2019-11-22 23:53:10 UTC]

Can you attach the full dmesg please ?

[Comment 4 cazzantonio 2019-11-23 18:57:25 UTC]

I don't think it depends on the software used. My example with libreoffice was just a quick test. I think it's a call to a drm hardware acceleration feature that hangs the card.
It's already nailed down to that kernel commit.

[Comment 5 cazzantonio 2019-11-23 18:58:57 UTC]

This is my report on linux bugzilla:
https://bugzilla.kernel.org/show_bug.cgi?id=205545

[Comment 6 cazzantonio 2019-11-26 06:50:00 UTC]

Created attachment 597544 [details]
dmesg output

This is the dmesg output of the error happening with that kernel commit

[Comment 7 cazzantonio 2019-11-26 17:32:15 UTC]

Problem seems solved in 5.4.0 (still present in 5.3.13 though)

[Comment 8 Marek Bartosiewicz 2019-11-27 09:10:02 UTC]

5.4.0 is still bad for my Skylake laptop. Keeps its Intel GPU 100% powered on and no cstates:

                    |             GPU     |
                    |                     |
                    | Powered On100,0%    |
                    | RC6         0,0%    |
                    | RC6p        0,0%    |
                    | RC6pp       0,0%    |

[Comment 9 cazzantonio 2019-11-27 21:02:56 UTC]

I just discovered that, after some use, the notebook freezes completely under 5.4.0. No logs, just a total freeze followed my a manual hard reset

[Comment 10 Marek Bartosiewicz 2019-11-29 06:08:09 UTC]

https://bugs.freedesktop.org/show_bug.cgi?id=112315

[Comment 11 Mike Pagano 2019-12-21 23:48:05 UTC]

https://gitlab.freedesktop.org/drm/intel/issues/614

[Comment 12 Mike Pagano 2020-01-31 23:31:23 UTC]

Can you test gentoo-sources-5.5.0 ? The fix is reportedly in this kernel version

[Comment 13 cazzantonio 2020-02-02 13:16:26 UTC]

I'm testing 5.5.0 since a couple of days and didn't experience any hang.
Still I don't know how 5.4.x behaves and that is going to be the longterm kernel for the next year...

[Comment 14 Mike Pagano 2020-04-04 23:17:16 UTC]

Closing for now as it seems you moved to 5.5.X.
If you want to go back to 5.4.X and you still have issues, please re-open