Bug 56834 - [ivb DP dual link DVI dongle regression] [drm:intel_dp_i2c_aux_ch] *ERROR* too many retries, giving up
Summary: [ivb DP dual link DVI dongle regression] [drm:intel_dp_i2c_aux_ch] *ERROR* to...
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: x86-64 (AMD64) Linux (All)
: high normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-11-07 08:06 UTC by Luke Hutchison
Modified: 2017-07-24 22:59 UTC (History)
8 users (show)

See Also:
i915 platform: IVB
i915 features: display/DP

Attachments
output of dmesg with dmesg.debug=0x06 (238.30 KB, text/plain)
2012-11-07 08:36 UTC, Luke Hutchison 		no flags Details
Xorg.0.log (24.86 KB, text/x-log)
2012-11-07 08:38 UTC, Luke Hutchison 		no flags Details
nightly build test. (210.40 KB, text/plain)
2013-05-11 18:37 UTC, Jorge Sanchez 		no flags Details
dmesg output with drm.debug=0xe (185.85 KB, text/plain)
2013-12-20 16:20 UTC, Anthony Foiani 		no flags Details
dmesg output with drm.debug=0xe (drm-intel g3ee3d1eea4d6) (182.72 KB, text/plain)
2013-12-23 07:17 UTC, Anthony Foiani 		no flags Details
3.15-rc3 syslog (122.63 KB, text/plain)
2014-05-02 11:30 UTC, Anthony Foiani 		no flags Details
3.15-rc3 syslog with more logging (249.94 KB, text/plain)
2014-05-08 02:32 UTC, Anthony Foiani 		no flags Details
dmesg output with drm.debug=0xe (237.74 KB, text/plain)
2014-05-17 20:59 UTC, freeseek 		no flags Details
dmesg output with drm.debug=0xe (126.17 KB, text/plain)
2014-06-13 19:56 UTC, freeseek 		no flags Details
dmesg output with drm.debug=0xe (236.30 KB, text/plain)
2014-06-13 19:57 UTC, freeseek 		no flags Details
attachment-18664-0.html (1.40 KB, text/html)
2015-10-21 16:53 UTC, Luke Hutchison 		no flags Details
Show Obsolete (1) View All

Description Luke Hutchison 2012-11-07 08:06:38 UTC 
When connecting a machine with an Ivy Bridge HD 4000 graphics chip to a 2560x1440 monitor using an Accell active DisplayPort to Dual-Link DVI adapter, the resolution is not properly detected, and my log is full of the following messages:

[drm:intel_dp_i2c_aux_ch] *ERROR* too many retries, giving up

vesafb is able to detect the resolution correctly, so grub2 and early-stage boot are displayed at full resolution, but then once the Fedora graphical boot starts, the resolution drops to 1024x768. The Xorg Intel driver is also unable to probe EDID properly after that, and the resolution stays at 1024x768 in Xorg.

There was a patch in 2010 to fix I2C/EDID handling with active DisplayPort to DVI converters:

http://lists.freedesktop.org/archives/intel-gfx/2010-December/008927.html
> This fixes broken EDID reading when using an active DisplayPort to
> duallink DVI converter.  If the AUX CH replier chooses to defer the
> transaction, a short read occurs and erroneous data is returned as
> the i2c reply due to a lack of length checking and failure to check
> for AUX ACK.

I checked the source for my kernel (kernel-3.6.6-3.fc18.x86_64), and it already includes this patch.

I can properly read the EDID using monitor-edid, and then manually set up the mode using xrandr.
Comment 1 Luke Hutchison 2012-11-07 08:36:08 UTC 
Created attachment 69654 [details]
output of dmesg with dmesg.debug=0x06
Comment 2 Luke Hutchison 2012-11-07 08:36:57 UTC 
Monitor EDID obtained by running monitor-edid:

Name: QHD DVI
EISA ID: VMO1501
EDID version: 1.3
EDID extension blocks: 0
Screen size: 160.0 cm x 90.0 cm (72.27 inches, aspect ratio 16/9 = 1.78)
Gamma: 2.2
Digital signal
Max video bandwidth: 250 MHz

	HorizSync 30-91
	VertRefresh 56-61

	# Monitor preferred modeline (60.0 Hz vsync, 88.8 kHz hsync, ratio 16/9, 40 dpi)
	ModeLine "2560x1440" 241.5 2560 2600 2632 2720 1440 1443 1448 1481 -hsync +vsync

	# Monitor supported modeline (60.0 Hz vsync, 74.0 kHz hsync, ratio 16/10, 30x33 dpi)
	ModeLine "1920x1200" 154 1920 1968 2000 2080 1200 1202 1208 1235 -hsync +vsync
Comment 3 Luke Hutchison 2012-11-07 08:38:17 UTC 
Created attachment 69655 [details]
Xorg.0.log
Comment 4 Daniel Vetter 2012-11-07 10:44:04 UTC 
Can you please retest on latest drm-intel-nightly branch from

http://cgit.freedesktop.org/~danvet/drm-intel

That contains some reworked EDID reading over dp aux code.
Comment 5 Luke Hutchison 2012-11-12 11:13:37 UTC 
I compiled i915.ko from drm-intel-nightly, and I get the same error message "[drm:intel_dp_i2c_aux_ch] *ERROR* too many retries, giving up", and the same symptom of X starting in 1024x768 mode.
Comment 6 Braden 2012-12-12 19:17:03 UTC 
I am seeing this as well using version 2.20.14 of the Intel driver and kernel 3.6.9 (Fedora 18 prerelease).
Comment 7 Jesse Barnes 2012-12-12 19:23:10 UTC 
Braden, do you see this with the same DP->DVI adaptor?  I guess we just need to get one...
Comment 8 Braden 2012-12-12 19:34:36 UTC 
Yes, I'm using an Accell adapter as well; this one, specifically:

http://www.accellcables.com/products/DisplayPort/DP/dp330_dvid.htm

Like Luke, I'm also using an Ivy Bridge Intel CPU with HD4000 graphics.  monitor-edid says this:

Name: LCD3090WQXi
EISA ID: NEC672a
EDID version: 1.3
EDID extension blocks: 0
Screen size: 64.1 cm x 40.1 cm (29.77 inches, aspect ratio 16/10 = 1.60)
Gamma: 2.2
Digital signal
Max video bandwidth: 270 MHz

	HorizSync 31-99
	VertRefresh 50-85

	# Monitor preferred modeline (59.9 Hz vsync, 98.5 kHz hsync, ratio 16/10, 101 dpi)
	ModeLine "2560x1600" 268 2560 2608 2640 2720 1600 1603 1609 1646 +hsync +vsync
Comment 9 Daniel Vetter 2012-12-12 20:16:05 UTC 
(In reply to comment #8)
> Yes, I'm using an Accell adapter as well; this one, specifically:
> 
> http://www.accellcables.com/products/DisplayPort/DP/dp330_dvid.htm

Another DP adapter for Jani to buy.
Comment 10 Luke Hutchison 2012-12-13 00:42:46 UTC 
That looks like the same adapter I was using, the Accell B087B-007B. It's a widely-recommended adapter, being listed specifically by both AMD and nVidia as recommended for high-res displays and 3D support, so you can probably expect a lot of users to start using it once more machines standardize on DisplayPort.

Unfortunately my adapter got flakier and flakier the longer I used it (it kept blanking out, and there was a lot of static in the blue channel), so I had to send it back, and can't help diagnose this further :-/
Comment 11 Mathias Fußenegger 2013-01-27 21:44:31 UTC 
I am experiencing the same issue. Also Ivy Bridge HD 4000 (i7-3667U).
I am using an Apple Mini Displayport to VGA adapter.

xrandr -q doesn't show any modes. But if I add a mode e.g. xrandr --newmode DP1 1024x768  && xrandr output DP1 --right-of eDP1 --mode 1024x768 it works.

The problem appeared with kernel 3.7.4, everything works fine with 3.6.11.

libdrm version is 2.4.41, but I've also tried the actual version directly from git.
xf86-video-intel is 2.20.19
Comment 12 Daniel Vetter 2013-01-27 22:02:25 UTC 
(In reply to comment #11)
> I am experiencing the same issue. Also Ivy Bridge HD 4000 (i7-3667U).
> I am using an Apple Mini Displayport to VGA adapter.
> 
> xrandr -q doesn't show any modes. But if I add a mode e.g. xrandr --newmode
> DP1 1024x768  && xrandr output DP1 --right-of eDP1 --mode 1024x768 it works.
> 
> The problem appeared with kernel 3.7.4, everything works fine with 3.6.11.
> 
> libdrm version is 2.4.41, but I've also tried the actual version directly
> from git.
> xf86-video-intel is 2.20.19

Different adapter, so might be different bug. Until we know for sure it's simpler to keep a separate report. Mathias, can you please file one with drm.debug=0xe dmesg when the issue happens and the above details.

If you can bisect in addition where your regression has been introduced, that would be really helpful in tracking down the bug ...
Comment 13 Imre Deak 2013-04-30 06:51:05 UTC 
Since the last report of failures, there have been DP AUX IRQ fixes merged, so could the reporters try the latest drm-intel tree if this is still reproducible?
Comment 14 Luke Hutchison 2013-04-30 07:03:04 UTC 
Unfortunately the problem was so bad I had to return the device, so I can no longer test this. Sorry!
Comment 15 Daniel Vetter 2013-04-30 08:31:25 UTC 
Thanks for reporting back - I guess we'll have to close this now, since the hw is gone.
Comment 16 Braden 2013-04-30 12:08:32 UTC 
I've still got one of these.  I'll test it this evening.
Comment 17 Jorge Sanchez 2013-05-10 12:46:41 UTC 
I own a laptop that has this issue, not using a cable but a por replicator/dock.

I'll try it, should I use the "next" archive from here: http://cgit.freedesktop.org/~danvet/drm-intel?
Comment 18 Daniel Vetter 2013-05-10 12:57:44 UTC 
On Fri, May 10, 2013 at 2:46 PM,  <bugzilla-daemon@freedesktop.org> wrote:
> --- Comment #17 from Jorge Sanchez <jsanchezsilvera@gmail.com> ---
> I own a laptop that has this issue, not using a cable but a por
> replicator/dock.
>
> I'll try it, should I use the "next" archive from here:
> http://cgit.freedesktop.org/~danvet/drm-intel?

drm-intel-nightly is the recommended branch for testing.
--
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
Comment 19 Jorge Sanchez 2013-05-11 18:37:50 UTC 
Created attachment 79174 [details]
nightly build test.

Compiled and fired up the nightly build, still no luck, bug persists. I added an attachement to dmesg. 

Output to tty signaling error is given when the notebook is recently plugged & unplugged to the dock.

I saved that kernel, tell me if you need any info I still can boot on it.

Cheers
Comment 20 Anthony Foiani 2013-05-14 15:30:35 UTC 
I'm having the same issue, using an outboard DisplayPort to Dual-Link DVI-D adapter.

Adapter: Accell B087B-007B
http://www.amazon.com/gp/product/B00856WJH8

Computer is a Dell Optiplex 990 USFF ("ultra small form factor"), CPU is i7-2600S, using CPU graphics (HD2000 I think?).  OS is Fedora 18 x86_64, fully up-to-date.  More system info:
http://paste.fedoraproject.org/12071/45046136/

For whatever it's worth, Windows 7 sees the full resolution without a problem.  I'm pretty sure that grub sees it correctly as well.

If I connect this computer directly to the monitor using a DisplayPort cable, then it works fine.  (I prefer to use the DL DVI-D so I can use my KVM switch.)

It might be worth noting that Luke originally filed a downstream bug:
https://bugzilla.redhat.com/show_bug.cgi?id=842412
Not much more information there, but it can't hurt to have them linked.

As I noted in that bug, and as others mentioned elsewhere in this bug, I can work past the erorr by using xrandr directly:

xrandr --newmode '2560x1440_0.0' 241.50  2560 2608 2640 2720  1440 1443 1448 1481 +hsync -vsync
xrandr --addmode DP1 '2560x1440_0.0'

Is there any info I can provide to help fix this?

Thanks,
Tony
Comment 21 freeseek 2013-05-16 14:48:34 UTC 
I have the same problem with a Lenovo X220 (which uses a Sandy Bridge HD 3000) and a BizLink diplay port adapter. By default, xrandr does not detect the correct resolution of the HP LP3065 monitor. I am using a regular version of Ubuntu Raring (3.8.0-19-generic x86_64). I receive 340 "[drm:intel_dp_i2c_aux_ch] *ERROR* too many retries, giving up" messages at startup (through dmesg). The following command: "sudo get-edid | parse-edid" detects the HP LP3065 monitor resolutions though. My workaround so far has been to run the following commands after booting the laptop:
xrandr --newmode 2560x1600 268.00 2560 2608 2640 2720 1600 1603 1609 1646 -hsync +vsync
xrandr --addmode DP1 2560x1600
xrandr --output LVDS1 --off --output DP1 --mode 2560x1600
Where the above parameters are specific to the parameters received from the "sudo get-edid | parse-edid" command. The same monitor plugged through DVI without the display port adapter worked just fine on a different Linux machine.
Comment 22 Anthony Foiani 2013-08-25 09:23:25 UTC 
Since I'm doing some tests to address a separate bug on a separate machine, I thought I'd update this one.

This problem is still occurring as of 3.10.9-100.fc18.x86_64.

Same symptoms and related data.

Thanks.
Comment 23 Jani Nikula 2013-09-12 10:15:58 UTC 
Please try intel-dp-fixes branch of git://gitorious.org/jani/drm.git. It's intel-drm-nightly plus some DP aux fixes.
Comment 24 Jani Nikula 2013-12-16 14:21:08 UTC 
Please try drm-intel-nightly branch of http://cgit.freedesktop.org/~danvet/drm-intel/log/?h=drm-intel-nightly which has plenty of DP dongle fixes since the last try.
Comment 25 Jani Nikula 2013-12-20 15:25:06 UTC 
(In reply to comment #17)
> I own a laptop that has this issue, not using a cable but a por
> replicator/dock.

(In reply to comment #19)
> Created attachment 79174 [details]
> nightly build test.
> 
> Compiled and fired up the nightly build, still no luck, bug persists. I
> added an attachement to dmesg. 

Jorge, a port replicator or dock is likely a display port branch device, and a different bug. I can't confirm this because your dmesg wasn't with drm.debug=0xe. Please file a new bug if the problem persists.
Comment 26 Jani Nikula 2013-12-20 15:26:22 UTC 
(In reply to comment #22)
> Since I'm doing some tests to address a separate bug on a separate machine,
> I thought I'd update this one.
> 
> This problem is still occurring as of 3.10.9-100.fc18.x86_64.

Anthony, please try a recent kernel or the referenced drm-intel-nightly, and attach dmesg with drm.debug=0xe. Thanks.
Comment 27 Anthony Foiani 2013-12-20 16:18:49 UTC 
(In reply to comment #26)
> Anthony, please try a recent kernel or the referenced drm-intel-nightly, and
> attach dmesg with drm.debug=0xe. Thanks.

I don't have the bandwidth to rebuild right now, but the problem is still occurring as of 3.11.10-200.fc19.x86_64.

Any idea when these changes will hit Fedora 19, or whether there's a test kernel RPM I can use?

I might have time to do a build this weekend, although I don't recall how successful I was the last time I tried a custom build -- the Fedora kernel RPM is getting scarier by the day...

Will attach current dmesg in a moment.
Comment 28 Anthony Foiani 2013-12-20 16:20:27 UTC 
Created attachment 91048 [details]
dmesg output with drm.debug=0xe
Comment 29 Anthony Foiani 2013-12-23 07:16:31 UTC 
Ok, I think I got the latest version built and running.  Still not detecting correct resolution.  Will upload dmesg in a moment.
Comment 30 Anthony Foiani 2013-12-23 07:17:57 UTC 
Created attachment 91143 [details]
dmesg output with drm.debug=0xe (drm-intel g3ee3d1eea4d6)

$ git log -1
commit 3ee3d1eea4d67551be1c10b0acef5fd9540cc052
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Fri Dec 13 17:57:45 2013 +0100

    drm-intel-nightly: 2013y-12m-13d-17h-57m-30s integration manifest
Comment 31 Jani Nikula 2014-04-10 13:57:14 UTC 
There's been considerable improvements in the dp aux and i2c-over-aux code lately. They're on their way to 3.15, but still in a few disjoint repos. I pushed a merge of the relevant bits together into the fdo-56834 branch to my repo at [1]. Please give it a spin and report back.

[1] http://cgit.freedesktop.org/~jani/drm/log/?h=fdo-56834
Comment 32 Jani Nikula 2014-04-28 10:32:54 UTC 
Ping for a retest on 3.15-rc3. We've merged some fixes that have resolved other related bugs.
Comment 33 Anthony Foiani 2014-05-02 11:19:57 UTC 
Well, it's *different* with 3.15-rc3.

Not in a good way, though.

Current fully-up-to-date Fedora 19 boots to some low resolution, but at least it prompts the external DP->DVI2 converter to provide some sort of signal.

3.15-rc3 resulted in no display whatsoever.

Trace output below, going to reboot with drm debugging enabled.

May  2 04:12:17 opti990 kernel: [    2.108060] ------------[ cut here ]------------
May  2 04:12:17 opti990 kernel: [    2.108092] WARNING: CPU: 1 PID: 227 at drivers/gpu/drm/i915/intel_display.c:9766 intel_modeset_check_state+0x61f/0x7d0 [i915]()
May  2 04:12:17 opti990 kernel: [    2.108093] encoder's hw state doesn't match sw tracking (expected 1, found 0)
May  2 04:12:17 opti990 kernel: [    2.108100] Modules linked in: i915(+) i2c_algo_bit drm_kms_helper drm i2c_core video
May  2 04:12:17 opti990 kernel: [    2.108103] CPU: 1 PID: 227 Comm: systemd-udevd Not tainted 3.15.0-0.rc3.git3.1.fc19.x86_64 #1
May  2 04:12:17 opti990 kernel: [    2.108104] Hardware name: Dell Inc. OptiPlex 990/0PGKWF, BIOS A07 09/10/2011
May  2 04:12:17 opti990 kernel: [    2.108108]  0000000000000000 00000000127708fd ffff880036e933c8 ffffffff817d5f5a
May  2 04:12:17 opti990 kernel: [    2.108112]  ffff880036e93410 ffff880036e93400 ffffffff8109592d 0000000000000001
May  2 04:12:17 opti990 kernel: [    2.108115]  ffff880036e93490 ffff88021fdd67e8 ffff88021cb41000 ffff88021fdd6000
May  2 04:12:17 opti990 kernel: [    2.108116] Call Trace:
May  2 04:12:17 opti990 kernel: [    2.108122]  [<ffffffff817d5f5a>] dump_stack+0x4d/0x66
May  2 04:12:17 opti990 kernel: [    2.108126]  [<ffffffff8109592d>] warn_slowpath_common+0x7d/0xa0
May  2 04:12:17 opti990 kernel: [    2.108128]  [<ffffffff810959ac>] warn_slowpath_fmt+0x5c/0x80
May  2 04:12:17 opti990 kernel: [    2.108148]  [<ffffffffa00eaaff>] intel_modeset_check_state+0x61f/0x7d0 [i915]
May  2 04:12:17 opti990 kernel: [    2.108164]  [<ffffffffa00ead45>] intel_set_mode+0x25/0x30 [i915]
May  2 04:12:17 opti990 kernel: [    2.108179]  [<ffffffffa00ebc0d>] intel_crtc_set_config+0x8ed/0xdd0 [i915]
May  2 04:12:17 opti990 kernel: [    2.108183]  [<ffffffff810f734d>] ? trace_hardirqs_on+0xd/0x10
May  2 04:12:17 opti990 kernel: [    2.108195]  [<ffffffffa0033ec7>] drm_mode_set_config_internal+0x67/0x110 [drm]
May  2 04:12:17 opti990 kernel: [    2.108200]  [<ffffffffa00877d3>] drm_fb_helper_restore_fbdev_mode+0xb3/0xe0 [drm_kms_helper]
May  2 04:12:17 opti990 kernel: [    2.108205]  [<ffffffffa00893f1>] drm_fb_helper_set_par+0x31/0x80 [drm_kms_helper]
May  2 04:12:17 opti990 kernel: [    2.108209]  [<ffffffff81430254>] fbcon_init+0x514/0x5a0
May  2 04:12:17 opti990 kernel: [    2.108213]  [<ffffffff814c6bf8>] visual_init+0xb8/0x120
May  2 04:12:17 opti990 kernel: [    2.108216]  [<ffffffff814c9363>] do_bind_con_driver+0x163/0x330
May  2 04:12:17 opti990 kernel: [    2.108218]  [<ffffffff814c9b14>] do_take_over_console+0x114/0x1c0
May  2 04:12:17 opti990 kernel: [    2.108221]  [<ffffffff8142b8d3>] do_fbcon_takeover+0x63/0xd0
May  2 04:12:17 opti990 kernel: [    2.108224]  [<ffffffff81430d35>] fbcon_event_notify+0x6b5/0x800
May  2 04:12:17 opti990 kernel: [    2.108227]  [<ffffffff817e4156>] notifier_call_chain+0x66/0x150
May  2 04:12:17 opti990 kernel: [    2.108230]  [<ffffffff810c87e3>] __blocking_notifier_call_chain+0x53/0x70
May  2 04:12:17 opti990 kernel: [    2.108234]  [<ffffffff810c8816>] blocking_notifier_call_chain+0x16/0x20
May  2 04:12:17 opti990 kernel: [    2.108236]  [<ffffffff8143784b>] fb_notifier_call_chain+0x1b/0x20
May  2 04:12:17 opti990 kernel: [    2.108239]  [<ffffffff814399bc>] register_framebuffer+0x1fc/0x350
May  2 04:12:17 opti990 kernel: [    2.108243]  [<ffffffffa008977e>] drm_fb_helper_initial_config+0x33e/0x510 [drm_kms_helper]
May  2 04:12:17 opti990 kernel: [    2.108246]  [<ffffffff810f7145>] ? mark_held_locks+0x75/0xa0
May  2 04:12:17 opti990 kernel: [    2.108249]  [<ffffffff817dee76>] ? _raw_spin_unlock_irqrestore+0x36/0x70
May  2 04:12:17 opti990 kernel: [    2.108252]  [<ffffffff810f7275>] ? trace_hardirqs_on_caller+0x105/0x1d0
May  2 04:12:17 opti990 kernel: [    2.108271]  [<ffffffffa00f6f6f>] intel_fbdev_initial_config+0x1f/0x30 [i915]
May  2 04:12:17 opti990 kernel: [    2.108292]  [<ffffffffa0118269>] i915_driver_load+0xf39/0xfc0 [i915]
May  2 04:12:17 opti990 kernel: [    2.108295]  [<ffffffff810f7145>] ? mark_held_locks+0x75/0xa0
May  2 04:12:17 opti990 kernel: [    2.108298]  [<ffffffff817dc78d>] ? __mutex_unlock_slowpath+0xbd/0x190
May  2 04:12:17 opti990 kernel: [    2.108301]  [<ffffffff810f7275>] ? trace_hardirqs_on_caller+0x105/0x1d0
May  2 04:12:17 opti990 kernel: [    2.108304]  [<ffffffff810f7145>] ? mark_held_locks+0x75/0xa0
May  2 04:12:17 opti990 kernel: [    2.108307]  [<ffffffff817dee76>] ? _raw_spin_unlock_irqrestore+0x36/0x70
May  2 04:12:17 opti990 kernel: [    2.108310]  [<ffffffff810f7275>] ? trace_hardirqs_on_caller+0x105/0x1d0
May  2 04:12:17 opti990 kernel: [    2.108313]  [<ffffffff810f734d>] ? trace_hardirqs_on+0xd/0x10
May  2 04:12:17 opti990 kernel: [    2.108322]  [<ffffffffa002e08d>] drm_dev_register+0xad/0x100 [drm]
May  2 04:12:17 opti990 kernel: [    2.108330]  [<ffffffffa0030df9>] drm_get_pci_dev+0x89/0x200 [drm]
May  2 04:12:17 opti990 kernel: [    2.108347]  [<ffffffffa00925fb>] i915_pci_probe+0x3b/0x60 [i915]
May  2 04:12:17 opti990 kernel: [    2.108350]  [<ffffffff8140ad35>] local_pci_probe+0x45/0xa0
May  2 04:12:17 opti990 kernel: [    2.108352]  [<ffffffff8140bfe5>] ? pci_match_device+0xc5/0xd0
May  2 04:12:17 opti990 kernel: [    2.108355]  [<ffffffff8140c129>] pci_device_probe+0xf9/0x150
May  2 04:12:17 opti990 kernel: [    2.108359]  [<ffffffff814f7665>] driver_probe_device+0x125/0x3a0
May  2 04:12:17 opti990 kernel: [    2.108362]  [<ffffffff814f79b3>] __driver_attach+0x93/0xa0
May  2 04:12:17 opti990 kernel: [    2.108364]  [<ffffffff814f7920>] ? __device_attach+0x40/0x40
May  2 04:12:17 opti990 kernel: [    2.108367]  [<ffffffff814f53a3>] bus_for_each_dev+0x73/0xc0
May  2 04:12:17 opti990 kernel: [    2.108370]  [<ffffffff814f6fee>] driver_attach+0x1e/0x20
May  2 04:12:17 opti990 kernel: [    2.108372]  [<ffffffff814f6b98>] bus_add_driver+0x188/0x260
May  2 04:12:17 opti990 kernel: [    2.108375]  [<ffffffffa0165000>] ? 0xffffffffa0164fff
May  2 04:12:17 opti990 kernel: [    2.108377]  [<ffffffff814f8054>] driver_register+0x64/0xf0
May  2 04:12:17 opti990 kernel: [    2.108380]  [<ffffffffa0165000>] ? 0xffffffffa0164fff
May  2 04:12:17 opti990 kernel: [    2.108382]  [<ffffffff8140a6c0>] __pci_register_driver+0x60/0x70
May  2 04:12:17 opti990 kernel: [    2.108390]  [<ffffffffa003107a>] drm_pci_init+0x10a/0x140 [drm]
May  2 04:12:17 opti990 kernel: [    2.108393]  [<ffffffffa0165000>] ? 0xffffffffa0164fff
May  2 04:12:17 opti990 kernel: [    2.108407]  [<ffffffffa016506a>] i915_init+0x6a/0x6c [i915]
May  2 04:12:17 opti990 kernel: [    2.108410]  [<ffffffff8100216a>] do_one_initcall+0xfa/0x1b0
May  2 04:12:17 opti990 kernel: [    2.108414]  [<ffffffff81062e53>] ? set_memory_nx+0x43/0x50
May  2 04:12:17 opti990 kernel: [    2.108418]  [<ffffffff811302ae>] load_module+0x1e1e/0x2700
May  2 04:12:17 opti990 kernel: [    2.108421]  [<ffffffff8112b840>] ? store_uevent+0x70/0x70
May  2 04:12:17 opti990 kernel: [    2.108424]  [<ffffffff81237610>] ? kernel_read+0x50/0x80
May  2 04:12:17 opti990 kernel: [    2.108428]  [<ffffffff81130d66>] SyS_finit_module+0xa6/0xd0
May  2 04:12:17 opti990 kernel: [    2.108432]  [<ffffffff817e9a69>] system_call_fastpath+0x16/0x1b
May  2 04:12:17 opti990 kernel: [    2.108434] ---[ end trace 50dbbbf7920940d9 ]---
Comment 34 Anthony Foiani 2014-05-02 11:30:05 UTC 
Created attachment 98334 [details]
3.15-rc3 syslog
Comment 35 Jani Nikula 2014-05-07 12:22:31 UTC 
(In reply to comment #34)
> Created attachment 98334 [details]
> 3.15-rc3 syslog

The log doesn't have high enough loglevel to have the debug output. Please increase the loglevel and provide the dmesg.
Comment 36 Anthony Foiani 2014-05-08 02:32:06 UTC 
Created attachment 98657 [details]
3.15-rc3 syslog with more logging

This one came up to the "VESA" resolution, but when I tried to manually force the native resolution, it didn't work.

Force script is described up-bug, I think, but in case it's not, it's just:
$ cat fixup-display.sh 
xrandr --newmode '2560x1440_0.0' 241.50  2560 2608 2640 2720  1440 1443 1448 1481 +hsync -vsync
xrandr --addmode DP1 '2560x1440_0.0'

(Timings were extracted by connecting directly via displayport, in which case the box comes up to native resolution w/o a problem.)
Comment 37 Anthony Foiani 2014-05-08 03:25:36 UTC 
Let me clarify that "didn't work".  It booted, gave me a login screen, and let me log in (all at some low resolution).

When I attempted to force the native (2560x1440) resolution, my screen went into sleep mode (indicating that the computer failed to successfully set the resolution).  I had to log in from a remote box and restart that way.

Thanks for looking into this!
Comment 38 Daniel Vetter 2014-05-15 21:30:36 UTC 
Increasing to regression since we seem to get worse at this :(
Comment 39 freeseek 2014-05-17 19:49:49 UTC 
I have the same problem with a Lenovo X220 and a BizLink diplay port adapter, as I upgraded to Ubuntu 14.04. I am currently using 3.15.0-031500rc5-generic. As per comment #36 I also need to add the correct external monitor display with xrandr. I then use the following command to switch from the laptop screen to the monitor screen: "xrandr --output LVDS1 --off --output VGA1 --off --output HDMI1 --off --output DP1 --mode 2560x1600".

Every other time this will fail, and it looks like the laptop tried to switch to DP1 unsuccessfully and then reverted to LVDS1. Occasionally, this will also crash the machine requiring a hard reset. Once the laptop finally succeeds in using DP1, running a program like xrandr or unity-control-center will immediately trigger something and revert the display to LVDS1 (100% reproducibility), possibly crashing the system. Each time the laptop reverts to LVDS1 it triggers a message like the one in Comment #33 in dmesg.

I will also add that "get-edid" completely fails and shows no useful output. It says something like "Looks like no busses have an EDID. Sorry!" and "The EDID data should not be trusted as the VBE call failed". The same Bizlink adapter works flawlessly on a Mac.
Comment 40 freeseek 2014-05-17 20:59:30 UTC 
Created attachment 99247 [details]
dmesg output with drm.debug=0xe
Comment 41 freeseek 2014-05-17 21:04:29 UTC 
Part of the issue seems to me that four incorrect resolutions are detected (1024x768 800x600 848x480 640x480) and the system keeps trying on its own to revert to 1024x768 anytime something like xrandr is run, and which is not supported by the monitor (which only supports 2560x1800 and 1280x800). This behavior is new to Ubuntu 14.04 somehow, but I attempted to re-install previous kernels, all the way up to 3.5.0-49-generic, and the problem persists.
Comment 42 Anthony Foiani 2014-05-18 01:30:57 UTC 
I don't know much about it, but I don't think those resolutions are being "detected", as such -- I think those are just the fallback (old VESA) resolutions.

My monitor (Dell U2713HM, I think) will gracefully display those lower resolutions scaled up to the native 2560x1440.

If your monitor supports native DisplayPort, does it work with your graphics output?  That's where I got my mode line settings (although I would prefer to have it work with DL-DVI, since there aren't yet any reasonable KVMs for DP.

Either way, good luck!
Comment 43 Anthony Foiani 2014-05-18 01:32:04 UTC 
BTW, I have a second DP->DL-DVI-D adaptor (bought for my MacBook Pro).  If I can find the appropriate full-size to mini DP matcher, would it help if I sent one of the developers my existing adaptor, so they can iterate more quickly?
Comment 44 Anthony Foiani 2014-05-31 04:22:07 UTC 
Ok, that's ... peculiar.

I got the appropriate adaptor for my other adapter (a mini-displayport one), and it works fine on my Fedora 19 box.  This adapter is:

Kanex iAdapt C30 MDPC30 Mini DisplayPort to DVI Adapter + USB for Apple Cinema Display 30-Inch
http://www.amazon.com/Kanex-MDPC30-DisplayPort-Adapter-Display/dp/B0045I85GY

It worked with my macbook as well (although it would sometimes fail to come out of sleep properly -- haven't had this one connected here long enough to tell... but the one I'm replacing would sometimes come back in an unstable state as well, with glitchy pixels everywhere.)

It's not as reliable with Windows 7 on the same machine, though; the previous adapter was rock solid under Windows, while this one has had issues (fine vertical white lines once, sleep mode the second time -- but power cycling the adapter solved the latter).

If one of the devs would like the original adapter (that does not auto-play with linux at all, and seems to have gotten worse recently), I can try to ship it to them.
Comment 45 Jani Nikula 2014-06-05 14:30:52 UTC 
(In reply to comment #36)
> Created attachment 98657 [details]
> 3.15-rc3 syslog with more logging
> 
> This one came up to the "VESA" resolution, but when I tried to manually
> force the native resolution, it didn't work.
> 
> Force script is described up-bug, I think, but in case it's not, it's just:
> $ cat fixup-display.sh 
> xrandr --newmode '2560x1440_0.0' 241.50  2560 2608 2640 2720  1440 1443 1448
> 1481 +hsync -vsync
> xrandr --addmode DP1 '2560x1440_0.0'
> 
> (Timings were extracted by connecting directly via displayport, in which
> case the box comes up to native resolution w/o a problem.)

Here, the aux channel seems to stop working completely. It's odd.
Comment 46 Jani Nikula 2014-06-05 14:32:30 UTC 
(In reply to comment #40)
> Created attachment 99247 [details]
> dmesg output with drm.debug=0xe

And same here. freeseeker, which kernel were you running? Care to post full dmesg from early boot with drm.debug=0xe, running a recent kernel?
Comment 47 freeseek 2014-06-13 19:56:50 UTC 
Created attachment 101009 [details]
dmesg output with drm.debug=0xe

/var/log/dmesg
Comment 48 freeseek 2014-06-13 19:57:14 UTC 
Created attachment 101011 [details]
dmesg output with drm.debug=0xe

dmesg
Comment 49 freeseek 2014-06-13 20:01:14 UTC 
I don't know how to post the full dmesg from early boot, as with the drm.debug=0xe option I get so many of the following messages:
1) [drm:drm_dp_i2c_do_msg] native defer[drm:drm_dp_i2c_do_msg] native defer
2) [drm:drm_dp_i2c_do_msg] native defer
3) [drm:drm_dp_i2c_do_msg] too many retries, giving up
That it fills the buffer. However, the /var/log/dmesg file seems to contain the first 128Kb of dmesg, so that should contain the information you require. I just updated to and I am running linux 3.15.0-031500-generic.
Comment 50 freeseek 2014-06-13 20:53:20 UTC 
Also, I am unable now to generate an attachment like 99247 with the "Call Trace", as right now, using kernel 3.15.0-031500-generic, when I run xrandr or unity-control-center the system always and reproducibly crashes and I am unable to generate any log. Do you have any suggestions for a DVI->DP adapter that works with Sandybridge 3000 HD? If so, I would be happy to buy one and send my BizLink adapter (http://www.amazon.com/dp/B003XYBA72/) to you if needed. My monitor is an HP LP3065 (http://www.amazon.com/dp/B000KB6E3G/) but I cannot send you that. ;-)

I have also consistently noticed that after the system crashes, if I reboot (with Ctrl+Alt+SysRq+B) rather than turning the laptop off and on, the system is more unstable, as if some weird state is preserved across reboots. It does not make sense to me, but maybe it does make sense to you.
Comment 51 freeseek 2014-07-03 15:57:26 UTC 
For some weird reasons, by deleting all configuration files in my home directory, the behavior reverted to what it was with Ubuntu 13.10. I cannot reproduce the crash anymore. I still need to add the correct resolution with xrandr as it is not detected correctly. Somehow there is a lurking bug in i915, but this must have been triggered by some weird settings I had in Unity. This is my best guess.
Comment 52 freeseek 2014-07-11 14:53:42 UTC 
Never mind, while the situation has improved, I still get errors like this one:
[   28.067101] [drm:intel_dp_aux_native_write] *ERROR* too many retries, giving up
[   28.067105] [drm:intel_dp_start_link_train] *ERROR* failed to enable link training
[   28.106453] [drm:intel_dp_complete_link_train] *ERROR* failed to train DP, aborting
[   28.106458] ------------[ cut here ]------------
[   28.106489] WARNING: CPU: 0 PID: 1449 at /build/buildd/linux-3.13.0/drivers/gpu/drm/i915/intel_dp.c:2747 intel_dp_link_down+0x1d2/0x210 [i915]()
[   28.106490] Modules linked in: pci_stub vboxpci(OF) vboxnetadp(OF) vboxnetflt(OF) vboxdrv(OF) rfcomm bnep binfmt_misc asix usbnet mii intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel uvcvideo videobuf2_vmall
oc videobuf2_memops videobuf2_core aesni_intel videodev aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd joydev serio_raw btusb bluetooth arc4 snd_hda_codec_hdmi iwldvm mac80211 snd_hda_codec_conexant thinkpad_acpi nvram snd_seq_midi snd_hda_intel snd_hda_codec snd_seq_midi_ev
ent snd_hwdep iwlwifi cfg80211 snd_rawmidi snd_pcm lpc_ich snd_page_alloc mei_me mei snd_seq snd_seq_device snd_timer snd soundcore mac_hid parport_pc ppdev lp parport raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 multipath linear
 hid_generic usbhid hid i915 psmouse ahci libahci sdhci_pci sdhci i2c_algo_bit e1000e drm_kms_helper ptp drm pps_core wmi video
[   28.106539] CPU: 0 PID: 1449 Comm: Xorg Tainted: GF          O 3.13.0-30-generic #55-Ubuntu
[   28.106540] Hardware name: LENOVO 42872WU/42872WU, BIOS 8DET69WW (1.39 ) 07/18/2013
[   28.106542]  0000000000000009 ffff8800c1df5a30 ffffffff8171a324 0000000000000000
[   28.106545]  ffff8800c1df5a68 ffffffff810676bd ffff8800362e40c8 ffff880212685800
[   28.106548]  ffff88021246f000 0000000080180004 ffff88020cc94000 ffff8800c1df5a78
[   28.106550] Call Trace:
[   28.106556]  [<ffffffff8171a324>] dump_stack+0x45/0x56
[   28.106560]  [<ffffffff810676bd>] warn_slowpath_common+0x7d/0xa0
[   28.106562]  [<ffffffff8106779a>] warn_slowpath_null+0x1a/0x20
[   28.106577]  [<ffffffffa0165142>] intel_dp_link_down+0x1d2/0x210 [i915]
[   28.106591]  [<ffffffffa0169b4f>] intel_dp_complete_link_train+0x11f/0x350 [i915]
[   28.106603]  [<ffffffffa0169dec>] intel_enable_dp+0x6c/0xa0 [i915]
[   28.106614]  [<ffffffffa0169e32>] g4x_enable_dp+0x12/0x30 [i915]
[   28.106627]  [<ffffffffa015373b>] ironlake_crtc_enable+0x34b/0x520 [i915]
[   28.106640]  [<ffffffffa0155769>] __intel_set_mode+0x859/0x9d0 [i915]
[   28.106653]  [<ffffffffa0158426>] intel_set_mode+0x16/0x30 [i915]
[   28.106666]  [<ffffffffa0158cdb>] intel_crtc_set_config+0x7ab/0x9a0 [i915]
[   28.106680]  [<ffffffffa0029ebd>] drm_mode_set_config_internal+0x5d/0xe0 [drm]
[   28.106693]  [<ffffffffa002cda7>] drm_mode_setcrtc+0xf7/0x5e0 [drm]
[   28.106703]  [<ffffffffa001dc22>] drm_ioctl+0x502/0x630 [drm]
[   28.106709]  [<ffffffff811cfa20>] do_vfs_ioctl+0x2e0/0x4c0
[   28.106713]  [<ffffffff8109dd94>] ? vtime_account_user+0x54/0x60
[   28.106716]  [<ffffffff811cfc81>] SyS_ioctl+0x81/0xa0
[   28.106720]  [<ffffffff8172aeff>] tracesys+0xe1/0xe6
[   28.106722] ---[ end trace b9ba88310aede0da ]---
[   28.136947] [drm:cpt_verify_modeset] *ERROR* mode set failed: pipe A stuck
[   28.197004] ------------[ cut here ]------------
[   28.197027] WARNING: CPU: 0 PID: 1449 at /build/buildd/linux-3.13.0/drivers/gpu/drm/i915/intel_display.c:9226 intel_modeset_check_state+0x62d/0x780 [i915]()
[   28.197028] encoder's hw state doesn't match sw tracking (expected 1, found 0)
[   28.197029] Modules linked in: pci_stub vboxpci(OF) vboxnetadp(OF) vboxnetflt(OF) vboxdrv(OF) rfcomm bnep binfmt_misc asix usbnet mii intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel uvcvideo videobuf2_vmall
oc videobuf2_memops videobuf2_core aesni_intel videodev aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd joydev serio_raw btusb bluetooth arc4 snd_hda_codec_hdmi iwldvm mac80211 snd_hda_codec_conexant thinkpad_acpi nvram snd_seq_midi snd_hda_intel snd_hda_codec snd_seq_midi_ev
ent snd_hwdep iwlwifi cfg80211 snd_rawmidi snd_pcm lpc_ich snd_page_alloc mei_me mei snd_seq snd_seq_device snd_timer snd soundcore mac_hid parport_pc ppdev lp parport raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 multipath linear
 hid_generic usbhid hid i915 psmouse ahci libahci sdhci_pci sdhci i2c_algo_bit e1000e drm_kms_helper ptp drm pps_core wmi video
[   28.197065] CPU: 0 PID: 1449 Comm: Xorg Tainted: GF       W  O 3.13.0-30-generic #55-Ubuntu
[   28.197066] Hardware name: LENOVO 42872WU/42872WU, BIOS 8DET69WW (1.39 ) 07/18/2013
[   28.197067]  0000000000000009 ffff8800c1df5b68 ffffffff8171a324 ffff8800c1df5bb0
[   28.197070]  ffff8800c1df5ba0 ffffffff810676bd ffff8800c1df5c20 0000000000000001
[   28.197072]  ffff880212685800 ffff880212685af8 ffff8800362e4000 ffff8800c1df5c00
[   28.197074] Call Trace:
[   28.197079]  [<ffffffff8171a324>] dump_stack+0x45/0x56
[   28.197081]  [<ffffffff810676bd>] warn_slowpath_common+0x7d/0xa0
[   28.197083]  [<ffffffff8106772c>] warn_slowpath_fmt+0x4c/0x50
[   28.197094]  [<ffffffffa015824d>] intel_modeset_check_state+0x62d/0x780 [i915]
[   28.197104]  [<ffffffffa0158435>] intel_set_mode+0x25/0x30 [i915]
[   28.197113]  [<ffffffffa0158cdb>] intel_crtc_set_config+0x7ab/0x9a0 [i915]
[   28.197124]  [<ffffffffa0029ebd>] drm_mode_set_config_internal+0x5d/0xe0 [drm]
[   28.197133]  [<ffffffffa002cda7>] drm_mode_setcrtc+0xf7/0x5e0 [drm]
[   28.197140]  [<ffffffffa001dc22>] drm_ioctl+0x502/0x630 [drm]
[   28.197145]  [<ffffffff811cfa20>] do_vfs_ioctl+0x2e0/0x4c0
[   28.197147]  [<ffffffff8109dd94>] ? vtime_account_user+0x54/0x60
[   28.197150]  [<ffffffff811cfc81>] SyS_ioctl+0x81/0xa0
[   28.197152]  [<ffffffff8172aeff>] tracesys+0xe1/0xe6
[   28.197154] ---[ end trace b9ba88310aede0db ]---
[   28.197167] [drm:intel_pipe_config_compare] *ERROR* mismatch in has_dp_encoder (expected 1, found 0)
[   28.197168] ------------[ cut here ]------------
[   28.197177] WARNING: CPU: 0 PID: 1449 at /build/buildd/linux-3.13.0/drivers/gpu/drm/i915/intel_display.c:9299 check_crtc_state+0x255/0x340 [i915]()
[   28.197178] pipe state doesn't match!
[   28.197179] Modules linked in: pci_stub vboxpci(OF) vboxnetadp(OF) vboxnetflt(OF) vboxdrv(OF) rfcomm bnep binfmt_misc asix usbnet mii intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel uvcvideo videobuf2_vmall
oc videobuf2_memops videobuf2_core aesni_intel videodev aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd joydev serio_raw btusb bluetooth arc4 snd_hda_codec_hdmi iwldvm mac80211 snd_hda_codec_conexant thinkpad_acpi nvram snd_seq_midi snd_hda_intel snd_hda_codec snd_seq_midi_event snd_hwdep iwlwifi cfg80211 snd_rawmidi snd_pcm lpc_ich snd_page_alloc mei_me mei snd_seq snd_seq_device snd_timer snd soundcore mac_hid parport_pc ppdev lp parport raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 multipath linear hid_generic usbhid hid i915 psmouse ahci libahci sdhci_pci sdhci i2c_algo_bit e1000e drm_kms_helper ptp drm pps_core wmi video
[   28.197210] CPU: 0 PID: 1449 Comm: Xorg Tainted: GF       W  O 3.13.0-30-generic #55-Ubuntu
[   28.197211] Hardware name: LENOVO 42872WU/42872WU, BIOS 8DET69WW (1.39 ) 07/18/2013
[   28.197212]  0000000000000009 ffff8800c1df58a8 ffffffff8171a324 ffff8800c1df58f0
[   28.197214]  ffff8800c1df58e0 ffffffff810676bd ffff88021246f6e0 ffff8800c1df5970
[   28.197216]  ffff880212685800 ffff88021246f000 ffff880212685b28 ffff8800c1df5940
[   28.197218] Call Trace:
[   28.197220]  [<ffffffff8171a324>] dump_stack+0x45/0x56
[   28.197222]  [<ffffffff810676bd>] warn_slowpath_common+0x7d/0xa0
[   28.197223]  [<ffffffff8106772c>] warn_slowpath_fmt+0x4c/0x50
[   28.197233]  [<ffffffffa0148fcb>] ? intel_pipe_config_compare+0x5fb/0xac0 [i915]
[   28.197242]  [<ffffffffa014a7d5>] check_crtc_state+0x255/0x340 [i915]
[   28.197252]  [<ffffffffa0157edb>] intel_modeset_check_state+0x2bb/0x780 [i915]
[   28.197261]  [<ffffffffa0158435>] intel_set_mode+0x25/0x30 [i915]
[   28.197269]  [<ffffffffa0158cdb>] intel_crtc_set_config+0x7ab/0x9a0 [i915]
[   28.197278]  [<ffffffffa0029ebd>] drm_mode_set_config_internal+0x5d/0xe0 [drm]
[   28.197287]  [<ffffffffa002cda7>] drm_mode_setcrtc+0xf7/0x5e0 [drm]
[   28.197294]  [<ffffffffa001dc22>] drm_ioctl+0x502/0x630 [drm]
[   28.197298]  [<ffffffff811cfa20>] do_vfs_ioctl+0x2e0/0x4c0
[   28.197300]  [<ffffffff8109dd94>] ? vtime_account_user+0x54/0x60
[   28.197302]  [<ffffffff811cfc81>] SyS_ioctl+0x81/0xa0
[   28.197304]  [<ffffffff8172aeff>] tracesys+0xe1/0xe6
[   28.197305] ---[ end trace b9ba88310aede0dc ]---
Could the problem be related to this: http://www.spinics.net/lists/intel-gfx/msg46896.html
Comment 53 Olivier Berger 2014-09-03 13:19:56 UTC 
We have experienced similar issues on a Dell Latitude E7240, and as the article at http://www.dell.com/support/troubleshooting/an/en/andhs1/KCS/KcsArticles/ArticleIframeView?docid=636946&doclang=EN suggests to apply a firmware upgrade for the "Synaptics VMM2320 MST HUB", we installed it (from Windows), and it seems our problems were solved.

Hope this helps.
Comment 54 Jani Nikula 2014-09-04 06:35:42 UTC 
(In reply to comment #53)
> We have experienced similar issues on a Dell Latitude E7240, and as the
> article at
> http://www.dell.com/support/troubleshooting/an/en/andhs1/KCS/KcsArticles/
> ArticleIframeView?docid=636946&doclang=EN suggests to apply a firmware
> upgrade for the "Synaptics VMM2320 MST HUB", we installed it (from Windows),
> and it seems our problems were solved.
> 
> Hope this helps.

I doubt you'll be able to install firmware upgrades to dongles...
Comment 55 freeseek 2014-10-28 13:07:31 UTC 
The situation is still not improved with Ubuntu 14.10. However, with kernel 3.16.0-23-generic I get a somewhat different output in dmesg (this is when I switch off the laptop screen to have the output only on the external monitor):

[   47.631564] [drm:intel_dp_start_link_train] *ERROR* failed to enable link training
[   47.638442] [drm:intel_dp_complete_link_train] *ERROR* failed to start channel equalization
[   47.670056] [drm:cpt_verify_modeset] *ERROR* mode set failed: pipe B stuck
[   47.670115] ------------[ cut here ]------------
[   47.670146] WARNING: CPU: 0 PID: 1724 at /build/buildd/linux-3.16.0/drivers/gpu/drm/i915/intel_display.c:10120 intel_modeset_check_state+0x559/0x7c0 [i915]()
[   47.670147] encoder's hw state doesn't match sw tracking (expected 1, found 0)
[   47.670148] Modules linked in: ipt_MASQUERADE iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack ip_tables x_tables pci_stub vboxpci(OE) vboxnetadp(OE) vboxnetflt(OE) vboxdrv(OE) rfcomm bnep bluetooth 6lowpan_iphc binfmt_misc asix usbnet mii snd_hda_codec_hdmi snd_hda_codec_conexant snd_hda_codec_generic uvcvideo videobuf2_vmalloc videobuf2_memops videobuf2_core v4l2_common videodev media intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel arc4 aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd iwldvm snd_hda_intel snd_hda_controller joydev snd_hda_codec mac80211 serio_raw snd_hwdep snd_pcm iwlwifi snd_seq_midi cfg80211 snd_seq_midi_event lpc_ich snd_rawmidi snd_seq snd_seq_device snd_timer
[   47.670166]  thinkpad_acpi nvram snd shpchp soundcore mei_me mei mac_hid parport_pc ppdev lp parport raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 multipath linear hid_generic usbhid hid i915 ahci psmouse e1000e libahci i2c_algo_bit drm_kms_helper sdhci_pci drm ptp sdhci pps_core wmi video
[   47.670180] CPU: 0 PID: 1724 Comm: Xorg Tainted: G        W  OE 3.16.0-23-generic #31-Ubuntu
[   47.670181] Hardware name: LENOVO 42872WU/42872WU, BIOS 8DET69WW (1.39 ) 07/18/2013
[   47.670182]  0000000000000009 ffff880206243b98 ffffffff8177fcbc ffff880206243be0
[   47.670184]  ffff880206243bd0 ffffffff8106fd8d ffff88021263b800 ffff880206243c50
[   47.670186]  0000000000000001 ffff88021263bb10 ffff88020eda1800 ffff880206243c30
[   47.670187] Call Trace:
[   47.670193]  [<ffffffff8177fcbc>] dump_stack+0x45/0x56
[   47.670196]  [<ffffffff8106fd8d>] warn_slowpath_common+0x7d/0xa0
[   47.670198]  [<ffffffff8106fdfc>] warn_slowpath_fmt+0x4c/0x50
[   47.670211]  [<ffffffffc049d809>] intel_modeset_check_state+0x559/0x7c0 [i915]
[   47.670221]  [<ffffffffc049db05>] intel_set_mode+0x25/0x30 [i915]
[   47.670231]  [<ffffffffc049ebba>] intel_crtc_set_config+0xa4a/0xda0 [i915]
[   47.670241]  [<ffffffffc035d6d1>] drm_mode_set_config_internal+0x61/0xe0 [drm]
[   47.670250]  [<ffffffffc0361413>] drm_mode_setcrtc+0x283/0x580 [drm]
[   47.670256]  [<ffffffffc0351a4f>] drm_ioctl+0x1df/0x680 [drm]
[   47.670260]  [<ffffffff811f4bc8>] do_vfs_ioctl+0x2c8/0x4a0
[   47.670262]  [<ffffffff811e31a1>] ? __sb_end_write+0x31/0x60
[   47.670264]  [<ffffffff811e0d12>] ? vfs_write+0x1b2/0x1f0
[   47.670266]  [<ffffffff811f4e21>] SyS_ioctl+0x81/0xa0
[   47.670268]  [<ffffffff81787ced>] system_call_fastpath+0x1a/0x1f
[   47.670269] ---[ end trace 898071e4e63bb9c2 ]---
[   47.670285] [drm:intel_pipe_config_compare] *ERROR* mismatch in has_dp_encoder (expected 1, found 0)
[   47.670286] ------------[ cut here ]------------
[   47.670296] WARNING: CPU: 0 PID: 1724 at /build/buildd/linux-3.16.0/drivers/gpu/drm/i915/intel_display.c:10192 check_crtc_state+0x2f5/0x3f0 [i915]()
[   47.670297] pipe state doesn't match!
[   47.670297] Modules linked in: ipt_MASQUERADE iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack ip_tables x_tables pci_stub vboxpci(OE) vboxnetadp(OE) vboxnetflt(OE) vboxdrv(OE) rfcomm bnep bluetooth 6lowpan_iphc binfmt_misc asix usbnet mii snd_hda_codec_hdmi snd_hda_codec_conexant snd_hda_codec_generic uvcvideo videobuf2_vmalloc videobuf2_memops videobuf2_core v4l2_common videodev media intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel arc4 aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd iwldvm snd_hda_intel snd_hda_controller joydev snd_hda_codec mac80211 serio_raw snd_hwdep snd_pcm iwlwifi snd_seq_midi cfg80211 snd_seq_midi_event lpc_ich snd_rawmidi snd_seq snd_seq_device snd_timer
[   47.670314]  thinkpad_acpi nvram snd shpchp soundcore mei_me mei mac_hid parport_pc ppdev lp parport raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 multipath linear hid_generic usbhid hid i915 ahci psmouse e1000e libahci i2c_algo_bit drm_kms_helper sdhci_pci drm ptp sdhci pps_core wmi video
[   47.670324] CPU: 0 PID: 1724 Comm: Xorg Tainted: G        W  OE 3.16.0-23-generic #31-Ubuntu
[   47.670325] Hardware name: LENOVO 42872WU/42872WU, BIOS 8DET69WW (1.39 ) 07/18/2013
[   47.670326]  0000000000000009 ffff8802062438d0 ffffffff8177fcbc ffff880206243918
[   47.670327]  ffff880206243908 ffffffff8106fd8d ffff8802062439a0 ffff88021263b800
[   47.670329]  ffff88021263bb40 ffff880212579000 ffff8802125796f8 ffff880206243968
[   47.670330] Call Trace:
[   47.670332]  [<ffffffff8177fcbc>] dump_stack+0x45/0x56
[   47.670334]  [<ffffffff8106fd8d>] warn_slowpath_common+0x7d/0xa0
[   47.670336]  [<ffffffff8106fdfc>] warn_slowpath_fmt+0x4c/0x50
[   47.670346]  [<ffffffffc048ce5b>] ? intel_pipe_config_compare+0x64b/0xba0 [i915]
[   47.670356]  [<ffffffffc048ea35>] check_crtc_state+0x2f5/0x3f0 [i915]
[   47.670359]  [<ffffffff813a5776>] ? sprintf+0x46/0x50
[   47.670369]  [<ffffffffc049d5c9>] intel_modeset_check_state+0x319/0x7c0 [i915]
[   47.670379]  [<ffffffffc049db05>] intel_set_mode+0x25/0x30 [i915]
[   47.670388]  [<ffffffffc049ebba>] intel_crtc_set_config+0xa4a/0xda0 [i915]
[   47.670397]  [<ffffffffc035d6d1>] drm_mode_set_config_internal+0x61/0xe0 [drm]
[   47.670405]  [<ffffffffc0361413>] drm_mode_setcrtc+0x283/0x580 [drm]
[   47.670411]  [<ffffffffc0351a4f>] drm_ioctl+0x1df/0x680 [drm]
[   47.670414]  [<ffffffff811f4bc8>] do_vfs_ioctl+0x2c8/0x4a0
[   47.670415]  [<ffffffff811e31a1>] ? __sb_end_write+0x31/0x60
[   47.670417]  [<ffffffff811e0d12>] ? vfs_write+0x1b2/0x1f0
[   47.670418]  [<ffffffff811f4e21>] SyS_ioctl+0x81/0xa0
[   47.670420]  [<ffffffff81787ced>] system_call_fastpath+0x1a/0x1f
[   47.670421] ---[ end trace 898071e4e63bb9c3 ]---
[   47.677066] ------------[ cut here ]------------
[   47.677155] WARNING: CPU: 0 PID: 1724 at /build/buildd/linux-3.16.0/drivers/gpu/drm/i915/intel_dp.c:3125 intel_dp_link_down+0x1d1/0x230 [i915]()
[   47.677160] Modules linked in: ipt_MASQUERADE iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack ip_tables x_tables pci_stub vboxpci(OE) vboxnetadp(OE) vboxnetflt(OE) vboxdrv(OE) rfcomm bnep bluetooth 6lowpan_iphc binfmt_misc asix usbnet mii snd_hda_codec_hdmi snd_hda_codec_conexant snd_hda_codec_generic uvcvideo videobuf2_vmalloc videobuf2_memops videobuf2_core v4l2_common videodev media intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel arc4 aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd iwldvm snd_hda_intel snd_hda_controller joydev snd_hda_codec mac80211 serio_raw snd_hwdep snd_pcm iwlwifi snd_seq_midi cfg80211 snd_seq_midi_event lpc_ich snd_rawmidi snd_seq snd_seq_device snd_timer
[   47.677201]  thinkpad_acpi nvram snd shpchp soundcore mei_me mei mac_hid parport_pc ppdev lp parport raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 multipath linear hid_generic usbhid hid i915 ahci psmouse e1000e libahci i2c_algo_bit drm_kms_helper sdhci_pci drm ptp sdhci pps_core wmi video
[   47.677234] CPU: 0 PID: 1724 Comm: Xorg Tainted: G        W  OE 3.16.0-23-generic #31-Ubuntu
[   47.677236] Hardware name: LENOVO 42872WU/42872WU, BIOS 8DET69WW (1.39 ) 07/18/2013
[   47.677238]  0000000000000009 ffff880206243ad8 ffffffff8177fcbc 0000000000000000
[   47.677240]  ffff880206243b10 ffffffff8106fd8d ffff88020e440000 0000000000180304
[   47.677243]  ffff88021263b800 0000000000000001 ffff88020eda18d8 ffff880206243b20
[   47.677246] Call Trace:
[   47.677260]  [<ffffffff8177fcbc>] dump_stack+0x45/0x56
[   47.677269]  [<ffffffff8106fd8d>] warn_slowpath_common+0x7d/0xa0
[   47.677273]  [<ffffffff8106fe6a>] warn_slowpath_null+0x1a/0x20
[   47.677320]  [<ffffffffc04b12d1>] intel_dp_link_down+0x1d1/0x230 [i915]
[   47.677355]  [<ffffffffc04b60c8>] intel_disable_dp+0x68/0x70 [i915]
[   47.677375]  [<ffffffffc049819a>] ironlake_crtc_disable+0x15a/0x980 [i915]
[   47.677380]  [<ffffffff811c2f49>] ? kmem_cache_alloc_trace+0x39/0x1e0
[   47.677396]  [<ffffffffc049a158>] ? __intel_set_mode+0x48/0xab0 [i915]
[   47.677411]  [<ffffffffc049a397>] __intel_set_mode+0x287/0xab0 [i915]
[   47.677427]  [<ffffffffc049daf6>] intel_set_mode+0x16/0x30 [i915]
[   47.677443]  [<ffffffffc049ebba>] intel_crtc_set_config+0xa4a/0xda0 [i915]
[   47.677460]  [<ffffffffc035d6d1>] drm_mode_set_config_internal+0x61/0xe0 [drm]
[   47.677473]  [<ffffffffc0361413>] drm_mode_setcrtc+0x283/0x580 [drm]
[   47.677483]  [<ffffffffc0351a4f>] drm_ioctl+0x1df/0x680 [drm]
[   47.677489]  [<ffffffff811f4bc8>] do_vfs_ioctl+0x2c8/0x4a0
[   47.677493]  [<ffffffff81665ff5>] ? __sys_recvmsg+0x65/0x80
[   47.677496]  [<ffffffff811f4e21>] SyS_ioctl+0x81/0xa0
[   47.677500]  [<ffffffff81787ced>] system_call_fastpath+0x1a/0x1f
[   47.677502] ---[ end trace 898071e4e63bb9c4 ]---
[   47.785947] ------------[ cut here ]------------
[   47.786032] WARNING: CPU: 2 PID: 1724 at /build/buildd/linux-3.16.0/drivers/gpu/drm/i915/intel_display.c:953 intel_wait_for_pipe_off+0x1d6/0x1e0 [i915]()
[   47.786036] pipe_off wait timed out
[   47.786038] Modules linked in: ipt_MASQUERADE iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 nf_nat nf_conntrack ip_tables x_tables pci_stub vboxpci(OE) vboxnetadp(OE) vboxnetflt(OE) vboxdrv(OE) rfcomm bnep bluetooth 6lowpan_iphc binfmt_misc asix usbnet mii snd_hda_codec_hdmi snd_hda_codec_conexant snd_hda_codec_generic uvcvideo videobuf2_vmalloc videobuf2_memops videobuf2_core v4l2_common videodev media intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel arc4 aesni_intel aes_x86_64 lrw gf128mul glue_helper ablk_helper cryptd iwldvm snd_hda_intel snd_hda_controller joydev snd_hda_codec mac80211 serio_raw snd_hwdep snd_pcm iwlwifi snd_seq_midi cfg80211 snd_seq_midi_event lpc_ich snd_rawmidi snd_seq snd_seq_device snd_timer
[   47.786074]  thinkpad_acpi nvram snd shpchp soundcore mei_me mei mac_hid parport_pc ppdev lp parport raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 multipath linear hid_generic usbhid hid i915 ahci psmouse e1000e libahci i2c_algo_bit drm_kms_helper sdhci_pci drm ptp sdhci pps_core wmi video
[   47.786088] CPU: 2 PID: 1724 Comm: Xorg Tainted: G        W  OE 3.16.0-23-generic #31-Ubuntu
[   47.786089] Hardware name: LENOVO 42872WU/42872WU, BIOS 8DET69WW (1.39 ) 07/18/2013
[   47.786090]  0000000000000009 ffff880206243aa0 ffffffff8177fcbc ffff880206243ae8
[   47.786092]  ffff880206243ad8 ffffffff8106fd8d ffff88020e440000 0000000000071008
[   47.786094]  00000000ffff09b1 ffff88021263bb38 ffff88021263bb40 ffff880206243b38
[   47.786095] Call Trace:
[   47.786101]  [<ffffffff8177fcbc>] dump_stack+0x45/0x56
[   47.786105]  [<ffffffff8106fd8d>] warn_slowpath_common+0x7d/0xa0
[   47.786107]  [<ffffffff8106fdfc>] warn_slowpath_fmt+0x4c/0x50
[   47.786118]  [<ffffffffc0492796>] intel_wait_for_pipe_off+0x1d6/0x1e0 [i915]
[   47.786128]  [<ffffffffc049284b>] intel_disable_pipe+0xab/0xc0 [i915]
[   47.786139]  [<ffffffffc04980d8>] ironlake_crtc_disable+0x98/0x980 [i915]
[   47.786141]  [<ffffffff811c2f49>] ? kmem_cache_alloc_trace+0x39/0x1e0
[   47.786151]  [<ffffffffc049a158>] ? __intel_set_mode+0x48/0xab0 [i915]
[   47.786161]  [<ffffffffc049a397>] __intel_set_mode+0x287/0xab0 [i915]
[   47.786171]  [<ffffffffc049daf6>] intel_set_mode+0x16/0x30 [i915]
[   47.786182]  [<ffffffffc049ebba>] intel_crtc_set_config+0xa4a/0xda0 [i915]
[   47.786192]  [<ffffffffc035d6d1>] drm_mode_set_config_internal+0x61/0xe0 [drm]
[   47.786200]  [<ffffffffc0361413>] drm_mode_setcrtc+0x283/0x580 [drm]
[   47.786207]  [<ffffffffc0351a4f>] drm_ioctl+0x1df/0x680 [drm]
[   47.786210]  [<ffffffff811f4bc8>] do_vfs_ioctl+0x2c8/0x4a0
[   47.786212]  [<ffffffff81665ff5>] ? __sys_recvmsg+0x65/0x80
[   47.786214]  [<ffffffff811f4e21>] SyS_ioctl+0x81/0xa0
[   47.786216]  [<ffffffff81787ced>] system_call_fastpath+0x1a/0x1f
[   47.786217] ---[ end trace 898071e4e63bb9c5 ]---
[   47.849930] [drm:ironlake_disable_pch_transcoder] *ERROR* failed to disable transcoder B
Comment 56 freeseek 2015-01-06 23:16:01 UTC 
With kernel 3.18.1-031801-generic the system does not crash anymore. However, it is still the case that using the BizLink diplay port adapter the monitor resolutions are not acquired correctly (while the command "sudo get-edid | parse-edid" provides the correct resolutions) and it is still the case that when plugging the monitor, often but not always the laptop fails to turn the external screen on (previously this often triggered a freeze). When the external screen fails to turn on, usually turning it off and then back on fixed the problem. This is still not ideal but it is a great improvement over kernel 3.16. I still get the following messages in the system log:
[drm:intel_dp_start_link_train] *ERROR* failed to enable link training
[drm:intel_dp_complete_link_train] *ERROR* failed to start channel equalization
Comment 58 Jesse Barnes 2015-03-09 21:56:01 UTC 
Ping, hopefully those patches helped?
Comment 59 Jani Nikula 2015-04-02 16:52:32 UTC 
[Funny, I seem to have linked to the same patch twice.]

A later version of the patch, plus a number of fixes to drm dp aux code, plus a fix to i915 dp aux code, have been merged to drm-intel-nightly. Please re-test using that. I'm fairly optimistic this is fixed.
Comment 60 Elio 2015-08-12 16:45:41 UTC 
Did someone try those patches successfully?
Comment 61 Luke Hutchison 2015-08-20 02:42:02 UTC 
On kernel-4.1.4-100.fc21.x86_64 at least, the problem still persists.
Comment 62 Ville Syrjala 2015-08-27 14:43:01 UTC 
Please test
git://github.com/vsyrjala/linux.git i2c_aux_short_write_4

If there are still i2c-over-aux failures with that you can try setting drm.dp_aux_i2c_speed_khz to a lover value (5 or 1 are sensible candidates, default is 10).
Comment 63 Jani Nikula 2015-10-07 13:38:34 UTC 
Please try current drm-intel-nightly branch of http://cgit.freedesktop.org/drm-intel - there are again several fixes in the i2c-over-aux area.
Comment 64 Jani Nikula 2015-10-21 12:17:01 UTC 
Ping for re-testing.
Comment 65 Luke Hutchison 2015-10-21 16:53:08 UTC 
Created attachment 119044 [details]
attachment-18664-0.html

I spent a couple of hours downloading and compiling after the previous ping
for retest, then I remembered that I don't have this hardware anymore :o)
I hope somebody else can test.
On Oct 21, 2015 5:17 AM, <bugzilla-daemon@freedesktop.org> wrote:

> *Comment # 64 <https://bugs.freedesktop.org/show_bug.cgi?id=56834#c64> on
> bug 56834 <https://bugs.freedesktop.org/show_bug.cgi?id=56834> from Jani
> Nikula <jani.nikula@intel.com> *
>
> Ping for re-testing.
>
> ------------------------------
> You are receiving this mail because:
>
>    - You reported the bug.
>
>
Comment 66 Jani Nikula 2015-10-22 12:00:32 UTC 
(In reply to Luke Hutchison from comment #65)
> I spent a couple of hours downloading and compiling after the previous ping
> for retest, then I remembered that I don't have this hardware anymore :o)
> I hope somebody else can test.

Thanks for the follow-up. We've fixed plenty of similar/related issues in kernels newer than the last one tried here, so I'll be an optimist for a change and claim we've fixed this. That said, please do reopen if the problem persists with latest upstream kernels. Thanks for the report.

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.