Bug 103518 - [KBL] GPU HANG: ecode 9:0:0x83ffef77, in webbrowser-app [5914], reason: No progress on rcs0, action: reset
Summary: [KBL] GPU HANG: ecode 9:0:0x83ffef77, in webbrowser-app [5914], reason: No pr...
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2017-10-30 19:38 UTC by Bernd Nottelmann
Modified: 2018-01-05 16:55 UTC (History)
1 user (show)

See Also:
i915 platform: KBL
i915 features: GPU hang


Attachments
/sys/class/drm/card0/error (21.37 KB, application/x-bzip)
2017-10-30 19:38 UTC, Bernd Nottelmann
no flags Details

Description Bernd Nottelmann 2017-10-30 19:38:47 UTC
Created attachment 135168 [details]
/sys/class/drm/card0/error

Starting webbrowser-app from ubuntu 16.04 HWE causes this GPU hang (excrpt from dmesg):

[ 6953.959227] asynchronous wait on fence i915:cinnamon[1678]/1:88b6 timed out
[ 6957.038174] [drm] GPU HANG: ecode 9:0:0x83ffef77, in webbrowser-app [5914], reason: No progress on rcs0, action: reset
[ 6957.038175] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 6957.038175] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 6957.038176] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 6957.038176] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[ 6957.038176] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 6957.038181] i915 0000:00:02.0: Resetting rcs0 after gpu hang
[ 6972.839195] i915 0000:00:02.0: Resetting rcs0 after gpu hang
[ 6983.911084] asynchronous wait on fence i915:Xorg[957]/1:266a timed out
[ 6988.839078] i915 0000:00:02.0: Resetting rcs0 after gpu hang
[ 7004.839031] i915 0000:00:02.0: Resetting rcs0 after gpu hang
[ 7020.838912] i915 0000:00:02.0: Resetting rcs0 after gpu hang
[ 7020.907468] kauditd_printk_skb: 15 callbacks suppressed

Attached: the crash dump /sys/class/drm/card0/error
Comment 1 Elizabeth 2017-10-31 21:55:48 UTC
Looks like bug 101991, from error state:
...
render command stream:
  START: 0x03c56000
  HEAD:  0x00000464 [0x00000428]
    head = 0x00000464, wraps = 0
  TAIL:  0x000004a8 [0x00000480, 0x000004a8]
  CTL:   0x00003001
    len=16384, enabled
  MODE:  0x00000000
  HWS:   0xfffe8000
  ACTHD: 0x00000008 ef18ff00
    at ring: 0x00000008
  IPEIR: 0x00000000
  IPEHR: 0x7c281088
  INSTDONE: 0xffd7ffff
    busy: CS
    busy: GAM
...
  seqno: 0x0000af1d
  last_seqno: 0x0000af1f
  waiting: yes
  ring->head: 0x00000000
  ring->tail: 0x000004a8
  hangcheck stall: yes
  hangcheck action: active head
  hangcheck action timestamp: 4296628496, 917600 ms ago
  engine reset count: 0
  ELSP[0]:  pid 5914, ban score 0, seqno        6:0000af1e, emitted 919292ms ago, head 00000428, tail 000004a8
  ELSP[1]:  pid 1678, ban score 0, seqno        5:0000af1f, emitted 919272ms ago, head 00001820, tail 00001898
  Active context: webbrowser-app[5914] user_handle 0 hw_id 6, ban score 0 guilty 0 active 0
...
batch (rcs0 (submitted by webbrowser-app [5914], ctx 0 [6], score 0)) at 0x00000000_00040000
0x00040000:      0x7d990000: 3D UNKNOWN: 3d_965 opcode = 0x7d99
0x00040004:      0x00000000: MI_NOOP
0x00040008:      0x7d9a0000: 3D UNKNOWN: 3d_965 opcode = 0x7d9a
0x0004000c:      0x00000000: MI_NOOP
0x00040010:      0x7d980000: 3D UNKNOWN: 3d_965 opcode = 0x7d98
0x00040014:      0x00000000: MI_NOOP
0x00040018:      0x7d890002: 3D UNKNOWN: 3d_965 opcode = 0x7d89
0x0004001c:      0x89800000: UNKNOWN
0x00040020:      0x00000000: MI_NOOP
0x00040024:      0x00000000: MI_NOOP
0x00040028:      0x7c281088: 3D UNKNOWN: 3d_965 opcode = 0x7c28
0x0004002c:      0x7c291099: 3D UNKNOWN: 3d_965 opcode = 0x7c29
0x00040030:      0x7c2a10aa: 3D UNKNOWN: 3d_965 opcode = 0x7c2a
0x00040034:      0x7c2b10bb: 3D UNKNOWN: 3d_965 opcode = 0x7c2b
0x00040038:      0x7d8c0000: 3D UNKNOWN: 3d_965 opcode = 0x7d8c
Bad length (2) in MI_STORE_DATA_IMM, [3, 4]
0x0004003c:      0x10000000: MI_STORE_DATA_IMM
0x00040040:      0x7d8c0000:    dword 1
0x00040044:      0x30000000: UNKNOWN
0x00040048:      0x7d8c0000: 3D UNKNOWN: 3d_965 opcode = 0x7d8c
0x0004004c:      0x50000000: COLOR_BLT
Bad count in COLOR_BLT
0x00040050:      0x7d8c0000:    dword 1
0x00040054:      0x70000000: 3D UNKNOWN: 3d_965 opcode = 0x7000
0x00040058:      0x7d8b0000: 3D UNKNOWN: 3d_965 opcode = 0x7d8b
0x0004005c:      0xa0000000: UNKNOWN
0x00040060:      0x7d8d0001: 3D UNKNOWN: 3d_965 opcode = 0x7d8d
0x00040064:      0x00000008: MI_NOOP
0x00040068:      0x3f800000: UNKNOWN
0x0004006c:      0x7d010000: 3D UNKNOWN: 3d_965 opcode = 0x7d01
0x00040070:      0x80808080: UNKNOWN
0x00040074:      0x6700a176: 3D UNKNOWN: 3d_965 opcode = 0x6700
0x00040078:      0x682008a1: 3D UNKNOWN: 3d_965 opcode = 0x6820
0x0004007c:      0x6f402100: 3D UNKNOWN: 3d_965 opcode = 0x6f40
0x00040080:      0x62120aa9: 3D UNKNOWN: 3d_965 opcode = 0x6212
0x00040084:      0x76b3ffff: 3D UNKNOWN: 3d_965 opcode = 0x76b3
0x00040088:      0x6c818a01: 3D UNKNOWN: 3d_965 opcode = 0x6c81
0x0004008c:      0x6ba008a1: 3D UNKNOWN: 3d_965 opcode = 0x6ba0
0x00040090:      0x69802100: 3D UNKNOWN: 3d_965 opcode = 0x6980
0x00040094:      0x63a00aaa: 3D UNKNOWN: 3d_965 opcode = 0x63a0
0x00040098:      0x6423070e: 3D UNKNOWN: 3d_965 opcode = 0x6423
0x0004009c:      0x66014142: 3D UNKNOWN: 3d_965 opcode = 0x6601
0x000400a0:      0x75000000: 3D UNKNOWN: 3d_965 opcode = 0x7500
0x000400a4:      0x7d880000: 3D UNKNOWN: 3d_965 opcode = 0x7d88
0x000400a8:      0x00000000: MI_NOOP
0x000400ac:      0x65000042: 3D UNKNOWN: 3d_965 opcode = 0x6500
0x000400b0:      0x6a000000: 3D UNKNOWN: 3d_965 opcode = 0x6a00
0x000400b4:      0x7d020000: 3D UNKNOWN: 3d_965 opcode = 0x7d02
0x000400b8:      0x00000000: MI_NOOP
0x000400bc:      0x7d8e0001: 3D UNKNOWN: 3d_965 opcode = 0x7d8e
0x000400c0:      0x03400400: MI UNKNOWN
0x000400c4:      0x00000000: MI_NOOP
0x000400c8:      0x7d850000: 3D UNKNOWN: 3d_965 opcode = 0x7d85
0x000400cc:      0x00880300: MI UNKNOWN
0x000400d0:      0x7d810001: 3D UNKNOWN: 3d_965 opcode = 0x7d81
0x000400d4:      0x00000000: MI_NOOP
0x000400d8:      0x000f0063: MI_NOOP
0x000400dc:      0x7c800002: 3D UNKNOWN: 3d_965 opcode = 0x7c80
0x000400e0:      0x7d800003: 3D UNKNOWN: 3d_965 opcode = 0x7d80
0x000400e4:      0x00000000: MI_NOOP
0x000400e8:      0x00000000: MI_NOOP
0x000400ec:      0x000f00ff: MI_NOOP
0x000400f0:      0x00000000: MI_NOOP
0x000400f4:      0x00000000: MI_NOOP
0x000400f8:      0x7d830000: 3D UNKNOWN: 3d_965 opcode = 0x7d83
0x000400fc:      0x00000000: MI_NOOP
0x00040100:      0x6d021181: 3D UNKNOWN: 3d_965 opcode = 0x6d02
0x00040104:      0x6d060101: 3D UNKNOWN: 3d_965 opcode = 0x6d06
0x00040108:      0x6e008045: 3D UNKNOWN: 3d_965 opcode = 0x6e00
0x0004010c:      0x6e048045: 3D UNKNOWN: 3d_965 opcode = 0x6e04
0x00040110:      0x7f0c000f: 3D UNKNOWN: 3d_965 opcode = 0x7f0c
0x00040114:      0x00000000: MI_NOOP
0x00040118:      0x00000000: MI_NOOP
0x0004011c:      0x3f000000: UNKNOWN
0x00040120:      0x00ff00ff: MI UNKNOWN
0x00040124:      0x43800000: 2D UNKNOWN
0x00040128:      0x00000000: MI_NOOP
0x0004012c:      0x3f000000: UNKNOWN
0x00040130:      0x00000000: MI_NOOP
0x00040134:      0x43800000: 2D UNKNOWN
0x00040138:      0x41800000: 2D UNKNOWN
0x0004013c:      0x3f000000: UNKNOWN
0x00040140:      0x00000000: MI_NOOP
0x00040144:      0x00000000: MI_NOOP
0x00040148:      0x41800000: 2D UNKNOWN
0x0004014c:      0x3f000000: UNKNOWN
0x00040150:      0x00000000: MI_NOOP
0x00040154:      0x02000000: MI_FLUSH
0x00040158:      0x02000000: MI_FLUSH
0x0004015c:      0x05000000: MI_BATCH_BUFFER_END
Comment 2 Chris Wilson 2017-11-23 21:25:29 UTC
The test is disabled by default, and perchance maybe the w/a was relevant.
Comment 3 Chris Wilson 2017-11-23 21:27:26 UTC
(In reply to Chris Wilson from comment #2)
> The test is disabled by default, and perchance maybe the w/a was relevant.

Oh dear, which bug was I trying to reply to?

commit 4f0aa1fa3e3849caee450ee5d14fcc289cf16703
Author: Anusha Srivatsa <anusha.srivatsa@intel.com>
Date:   Thu Nov 9 10:51:43 2017 -0800

    drm/i915/dmc: DMC 1.04 for Kabylake

and a new dmc-1.04 required.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.