Bug 92371

Summary: [hsw dp mst] external displays unusable on HP Pro x2 612 G1: drm:intel_dp_start_link_train [i915]] *ERROR* too many full retries, give up
Product: Drivers Reporter: Henning Rohde (Rohde.Henning)
Component: Video(DRI - Intel)Assignee: intel-gfx-bugs (intel-gfx-bugs)
Status: RESOLVED MOVED    
Severity: normal CC: intel-gfx-bugs, lachlan.00, leho, m.mcnutt, Rohde.Henning
Priority: P3    
Hardware: x86-64   
OS: Linux   
URL: https://bugs.freedesktop.org/show_bug.cgi?id=90963
Kernel Version: 3.19.0-0.rc6.git2.1 (also on 3.17.7-300) Subsystem:
Regression: No Bisected commit-id:
Attachments: # journalctl -kmb | grep -i "dmi:\|drm\|graphic\|intel\|i915\|video" | grep -iv "bluetooth\|[...]
full dmesg as display is connected after X11 is up and is activated with xrandr
log from booting 4.3.0-0.rc7.git0.1.fc24.x86_64
log from booting 4.3.0-0.rc7.git2.1.fc24 with drm.debug=14
log from booting 4.4.0-0.rc0.git4.2.fc24.x86_64 with drm.debug=14
screenshot booting 4.4.0-0.rc0.git4.2.fc24.x86_64 with drm.debug=14

Description Henning Rohde 2015-01-30 22:49:00 UTC 
Created attachment 165311 [details]
# journalctl -kmb |  grep -i "dmi:\|drm\|graphic\|intel\|i915\|video" | grep -iv "bluetooth\|[...]

Hello,

after connecting any display, either via DP or VGA, the tablet slows down on booting while mass-issuing following error:
drm:intel_dp_start_link_train [i915]] *ERROR* too many full retries, give up

dmesg on drm.debug=0xe and i915.mmio_debug=1 is attached.

Same error occurs if display is connected after X11 is up and is activated with xrandr - the screen on the external display stays off:
$ xrandr --current
Screen 0: minimum 8 x 8, current 1920 x 1200, maximum 32767 x 32767
eDP1 connected primary 1920x1080+0+0 (normal left inverted right x axis y axis) 276mm x 156mm
   1920x1080     60.03*+
   1400x1050     59.98  
   1280x1024     60.02  
   1280x960      60.00  
   1024x768      60.00  
   800x600       60.32    56.25  
   640x480       59.94  
DP1 disconnected (normal left inverted right x axis y axis)
DP1-1 connected 1920x1200+0+0 (normal left inverted right x axis y axis) 546mm x 352mm
   1920x1200     59.95*+
   1024x768      60.00  
   800x600       60.32    56.25  
   848x480       60.00  
   640x480       59.94  
DP1-2 disconnected (normal left inverted right x axis y axis)
DP1-3 disconnected (normal left inverted right x axis y axis)
DP2 disconnected (normal left inverted right x axis y axis)
HDMI1 disconnected (normal left inverted right x axis y axis)
HDMI2 disconnected (normal left inverted right x axis y axis)
VIRTUAL1 disconnected (normal left inverted right x axis y axis)

$ xrandr --properties
Screen 0: minimum 8 x 8, current 1920 x 1080, maximum 32767 x 32767
eDP1 connected primary 1920x1080+0+0 (normal left inverted right x axis y axis) 276mm x 156mm
	EDID: 
		00ffffffffffff0030e4470400000000
		00170104951c1078024bb59b5b548e25
		1e505400000001010101010101010101
		010101010101143780c270381f403c2e
		3500149c1000001b0000000000000000
		00000000000000000000000000000000
		00000000000000000000000000000002
		000c33ff0f3c961c0f2496000000006d
	BACKLIGHT: 562 
		range: (0, 937)
	Backlight: 562 
		range: (0, 937)
	scaling mode: Full aspect 
		supported: None, Full, Center, Full aspect
	Broadcast RGB: Automatic 
		supported: Automatic, Full, Limited 16:235
	audio: auto 
		supported: force-dvi, off, auto, on
   1920x1080     60.03*+
   1400x1050     59.98  
   1280x1024     60.02  
   1280x960      60.00  
   1024x768      60.00  
   800x600       60.32    56.25  
   640x480       59.94  
DP1 disconnected (normal left inverted right x axis y axis)
	Broadcast RGB: Automatic 
		supported: Automatic, Full, Limited 16:235
	audio: auto 
		supported: force-dvi, off, auto, on
DP1-1 connected (normal left inverted right x axis y axis)
	EDID: 
		00ffffffffffff0022f0692801010101
		05150104a53623782efc81a4554d9d25
		125054210800814081809500a940b300
		d1c001010101283c80a070b023403020
		360022602100001a000000fd003b3d18
		5011000a202020202020000000fc0048
		50205a523234770a20202020000000ff
		00434e54313035313138480a20200041
	Broadcast RGB: Automatic 
		supported: Automatic, Full, Limited 16:235
	audio: auto 
		supported: force-dvi, off, auto, on
   1920x1200     59.95 +
   1920x1080     59.99  
   1600x1200     60.00  
   1680x1050     59.95  
   1280x1024     60.02  
   1440x900      59.89  
   1280x960      60.00  
   1024x768      60.00  
   800x600       60.32  
   640x480       60.00  
DP1-2 disconnected (normal left inverted right x axis y axis)
[...]
Comment 1 Henning Rohde 2015-01-30 22:51:19 UTC 
Created attachment 165321 [details]
full dmesg as display is connected after X11 is up and is activated with xrandr
Comment 2 Matt 2015-08-07 03:43:34 UTC 
This precise behaviour is still present on this hardware on Fedora 22 Kernel Version 4.1.3-201.fc22.x86_64

However loading an older Fedora 20 3.16.7 kernel and the external displayport is able to function correctly
Comment 3 Jani Nikula 2015-10-20 08:18:17 UTC 
Please try the latest kernel.
Comment 4 Matt 2015-10-26 23:06:01 UTC 
Trying again with Kernel 4.2.3-200.fc22.x86_64

I get an image but it shakes horizontally and is ghosted with another version of the image
Comment 5 Matt 2015-10-26 23:07:54 UTC 
Further to that it no longer throws the drm:intel_dp_train error
Comment 6 Jani Nikula 2015-10-27 08:34:17 UTC 
(In reply to Matt from comment #4)
> Trying again with Kernel 4.2.3-200.fc22.x86_64

And with latest, I meant latest, v4.3-rc7.
Comment 7 Henning Rohde 2015-10-27 22:09:36 UTC 
Having tried 4.3.0-0.rc7.git0.1.fc24.x86_64 there are no errors anymore until X starts up. On activating the external display the internal one shows blue and white stripes and the computer hung up - had to power down hard.
Comment 8 Henning Rohde 2015-10-27 22:10:30 UTC 
Created attachment 191331 [details]
log from booting 4.3.0-0.rc7.git0.1.fc24.x86_64
Comment 9 Jani Nikula 2015-10-28 12:31:26 UTC 
(In reply to Henning Rohde from comment #7)
> Having tried 4.3.0-0.rc7.git0.1.fc24.x86_64 there are no errors anymore
> until X starts up.

How should I interpret that? Are there errors after X starts up?

> On activating the external display the internal one shows
> blue and white stripes and the computer hung up - had to power down hard.

Can you ssh into the machine?

(In reply to Henning Rohde from comment #8)
> Created attachment 191331 [details]
> log from booting 4.3.0-0.rc7.git0.1.fc24.x86_64

Please add drm.debug=14 module parameter and attach the same.
Comment 10 Henning Rohde 2015-11-02 06:36:38 UTC 
Created attachment 191821 [details]
log from booting 4.3.0-0.rc7.git2.1.fc24 with drm.debug=14

Hi!

Booted with "drm.debug=14", again been able to do login graphically until X11 got up.
Computer hangs completely, no login via ssh possible.

If booted later with display attached via uvcvideo i run into the following deadlock:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: ======================================================
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: [ INFO: possible circular locking dependency detected ]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 4.3.0-0.rc7.git2.1.fc24.x86_64 #1 Not tainted
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: -------------------------------------------------------
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: Xorg/1465 is trying to acquire lock:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  (&dmabuf->lock){+.+.+.}, at: [<ffffffff815a53bf>] dma_buf_vunmap+0x3f/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                            but task is already holding lock:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  (&dev->struct_mutex){+.+.+.}, at: [<ffffffffa007ba9e>] drm_gem_object_handle_unreference_unlocked+0x13e/0x170 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                            which lock already depends on the new lock.
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                            the existing dependency chain (in reverse order) is:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                            -> #1 (&dev->struct_mutex){+.+.+.}:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff811078ae>] lock_acquire+0xce/0x1c0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff81860a26>] mutex_lock_interruptible_nested+0x86/0x470
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa0184a19>] i915_mutex_lock_interruptible+0x29/0x50 [i915]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa0178c18>] i915_gem_dmabuf_vmap+0x38/0x200 [i915]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff815a52e5>] dma_buf_vmap+0x55/0xf0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa07cdd73>] udl_gem_vmap+0x23/0xa0 [udl]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa07cd1b2>] udl_handle_damage+0x322/0x380 [udl]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa07cb873>] udl_crtc_mode_set+0x783/0x7f0 [udl]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa012f86c>] drm_crtc_helper_set_mode+0x29c/0x4d0 [drm_kms_helper]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa0130985>] drm_crtc_helper_set_config+0xa45/0xbd0 [drm_kms_helper]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa00876d5>] drm_mode_set_config_internal+0x65/0x110 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa008bd62>] drm_mode_setcrtc+0x3e2/0x500 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa007c8a8>] drm_ioctl+0x138/0x520 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff812825ae>] do_vfs_ioctl+0x2ee/0x550
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff81282889>] SyS_ioctl+0x79/0x90
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff81864c72>] entry_SYSCALL_64_fastpath+0x12/0x76
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                            -> #0 (&dmabuf->lock){+.+.+.}:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff81106caf>] __lock_acquire+0x191f/0x1ba0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff811078ae>] lock_acquire+0xce/0x1c0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff81860e96>] mutex_lock_nested+0x86/0x400
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff815a53bf>] dma_buf_vunmap+0x3f/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa07cde18>] udl_gem_vunmap+0x28/0x50 [udl]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa07cde5c>] udl_gem_free_object+0x1c/0x60 [udl]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa007b170>] drm_gem_object_free+0x30/0x50 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa007bab8>] drm_gem_object_handle_unreference_unlocked+0x158/0x170 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa007bc56>] drm_gem_object_release_handle+0x56/0x70 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff8141d058>] idr_for_each+0xc8/0x130
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa007c310>] drm_gem_release+0x20/0x30 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffffa007b04e>] drm_release+0x3de/0x4d0 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff8126f315>] __fput+0xf5/0x200
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff8126f45e>] ____fput+0xe/0x10
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff810cdcd3>] task_work_run+0x73/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff810ad714>] do_exit+0x354/0xc30
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff810ae080>] do_group_exit+0x50/0xc0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff810ae104>] SyS_exit_group+0x14/0x20
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        [<ffffffff81864c72>] entry_SYSCALL_64_fastpath+0x12/0x76
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                            other info that might help us debug this:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  Possible unsafe locking scenario:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        CPU0                    CPU1
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:        ----                    ----
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:   lock(&dev->struct_mutex);
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:                                lock(&dmabuf->lock);
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:                                lock(&dev->struct_mutex);
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:   lock(&dmabuf->lock);
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                             *** DEADLOCK ***
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 2 locks held by Xorg/1465:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  #0:  (drm_global_mutex){+.+.+.}, at: [<ffffffffa007acaa>] drm_release+0x3a/0x4d0 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  #1:  (&dev->struct_mutex){+.+.+.}, at: [<ffffffffa007ba9e>] drm_gem_object_handle_unreference_unlocked+0x13e/0x170 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: 
                                            stack backtrace:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: CPU: 1 PID: 1465 Comm: Xorg Not tainted 4.3.0-0.rc7.git2.1.fc24.x86_64 #1
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: Hardware name: Hewlett-Packard HP Pro x2 612 G1 Tablet/221B, BIOS M83 Ver. 01.31 07/21/2015
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  0000000000000000 00000000b4a9c6af ffff88009248ba20 ffffffff8141bef9
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  ffffffff82c39250 ffff88009248ba60 ffffffff81103973 ffff88009248bac0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  ffff8802362fbf20 0000000000000002 0000000000000002 ffff8802362fb240
Nov 02 07:17:38 Fedora22.AmaliX.net kernel: Call Trace:
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff8141bef9>] dump_stack+0x4b/0x72
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81103973>] print_circular_bug+0x1e3/0x250
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81106caf>] __lock_acquire+0x191f/0x1ba0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81123a7d>] ? debug_lockdep_rcu_enabled+0x1d/0x20
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81123a7d>] ? debug_lockdep_rcu_enabled+0x1d/0x20
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff811078ae>] lock_acquire+0xce/0x1c0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff815a53bf>] ? dma_buf_vunmap+0x3f/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81860e96>] mutex_lock_nested+0x86/0x400
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff815a53bf>] ? dma_buf_vunmap+0x3f/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81861099>] ? mutex_lock_nested+0x289/0x400
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff815a53bf>] ? dma_buf_vunmap+0x3f/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81104fa9>] ? trace_hardirqs_on_caller+0x129/0x1b0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff815a53bf>] dma_buf_vunmap+0x3f/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa07cde18>] udl_gem_vunmap+0x28/0x50 [udl]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa07cde5c>] udl_gem_free_object+0x1c/0x60 [udl]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa007b170>] drm_gem_object_free+0x30/0x50 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa007bab8>] drm_gem_object_handle_unreference_unlocked+0x158/0x170 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa0096625>] ? drm_vma_node_revoke+0x55/0x80 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa007bc56>] drm_gem_object_release_handle+0x56/0x70 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff8141d058>] idr_for_each+0xc8/0x130
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa007bc00>] ? drm_gem_dumb_destroy+0x20/0x20 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa007c310>] drm_gem_release+0x20/0x30 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffffa007b04e>] drm_release+0x3de/0x4d0 [drm]
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff8126f315>] __fput+0xf5/0x200
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff8126f45e>] ____fput+0xe/0x10
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff810cdcd3>] task_work_run+0x73/0xa0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff810ad714>] do_exit+0x354/0xc30
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81104fa9>] ? trace_hardirqs_on_caller+0x129/0x1b0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff810ae080>] do_group_exit+0x50/0xc0
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff810ae104>] SyS_exit_group+0x14/0x20
Nov 02 07:17:38 Fedora22.AmaliX.net kernel:  [<ffffffff81864c72>] entry_SYSCALL_64_fastpath+0x12/0x76

Thanks for your works!
Comment 11 Henning Rohde 2015-11-09 18:54:22 UTC 
Good evening!

> How should I interpret that? Are there errors after X starts up?
Sorry, don't know - it crashed as soon as the kde-session get's loaded.

Today i did further testing with 4.4.0-0.rc0.git4.2.fc24.x86_64.
If the display is connected to the docking-station, the computer still crashes.
http://store.hp.com/us/en/pdp/business-solutions/hp-2013-ultraslim-docking-station

But if the display is directly attached to the keyboard-unit, kde gets up albeit with a distorted graphics.

Logs and picture are attached - thanks for your works in advance!
Comment 12 Henning Rohde 2015-11-09 19:01:08 UTC 
Created attachment 192571 [details]
log from booting 4.4.0-0.rc0.git4.2.fc24.x86_64 with drm.debug=14

display is attached to keyboard-unit, not docking-station as before
Comment 13 Henning Rohde 2015-11-09 19:03:56 UTC 
Created attachment 192581 [details]
screenshot booting 4.4.0-0.rc0.git4.2.fc24.x86_64 with drm.debug=14
Comment 14 lachlan.00 2015-12-03 02:22:37 UTC 
Is there anything i can add to help on this bug?

I have the same PC and docking station connected to dual displayport connections.

This error has been happening for me from 3.18 to 4.x

Right now on Debian's 4.3-1~exp1 kernel i just get a hard crash on boot
Comment 15 Jani Nikula 2016-01-13 13:07:48 UTC 
Sorry, no updates. However I think we have a dupe at the freedesktop.org bugzilla where we're consolidating all the bug tracking. Please follow-up there: https://bugs.freedesktop.org/show_bug.cgi?id=90963

Also, please file any new bug reports at the same place,
https://bugs.freedesktop.org/enter_bug.cgi?product=DRI&component=DRM/Intel