Potential atomic update failure on pipe A

Bl4ckb0ne · 2016-06-26 03:25:41

Moderator [ewaller] Merged similar threads

Hi.

I know, that title looks like it cames from a Black Mesa game, but its real.
After a fresh install, I have a random problem with my computer. Sometimes the screen goes on and off, and it's really annoying.

Here's what I got from dmesg

[ 7889.938811] [drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* uncleared fifo underrun on pipe A
[ 7889.938827] [drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
[ 7896.880719] traps: lemonbar[4689] general protection ip:7fc1407ff2b3 sp:7fff659ee580 error:0 in libXft.so.2.3.2[7fc1407f3000+15000]
[ 7959.426394] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 7970.468906] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8047.663147] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8267.233739] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8280.769303] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8284.762325] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8285.235742] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8454.429871] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8623.877335] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8657.254873] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8684.549523] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8718.970322] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8749.111421] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8776.146099] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8790.145096] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8832.857030] [drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* uncleared fifo underrun on pipe A
[ 8832.857047] [drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
[ 8836.911602] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8840.840982] traps: lemonbar[23151] general protection ip:7fd0644c32b3 sp:7ffc2fd49430 error:0 in libXft.so.2.3.2[7fd0644b7000+15000]
[ 8861.166543] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8875.238818] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8879.148532] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8885.264745] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8900.653612] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8917.152393] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8920.792123] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8936.920934] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8949.976624] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8993.376754] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[ 8993.376784] [drm:intel_pipe_update_end [i915]] *ERROR* Atomic update failure on pipe A (start=123581 end=123582) time 5 us, min 1789, max 1799, scanline start 1799, end 1800
[ 9178.602103] [drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* uncleared fifo underrun on pipe A
[ 9178.602114] [drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
[ 9247.501418] traps: lemonbar[2041] general protection ip:7f024bb4d2b3 sp:7fff86d0f1e0 error:0 in libXft.so.2.3.2[7f024bb41000+15000]
[ 9277.335767] [drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A

So I'm running bspwm with a xrandr, and the really amazing lemonbuddy on the lemonbar-sm package.

IDK what's going on, but I can't really use my computer because I have to logout/login every few 10 minutes.

Thanks in advance

Last edited by ewaller (2016-07-04 17:53:12)

Lone_Wolf · 2016-06-26 17:57:49

search for "intel_set_cpu_fifo_underrun" and you'll find plenty of references that point to an issue with the intel kernel driver.

Is intel-microcode installed and configured, see https://wiki.archlinux.org/index.php/Microcode ?
Does your system run latest bios/uefi firmware ?

Last edited by Lone_Wolf (2016-06-26 17:58:22)

Bl4ckb0ne · 2016-06-27 03:02:14

I didn't search for "intel_set_cpu_fifo_underrun", thanks.

intel-microcode is installed, and enabled, but I don't think that my bios are up to date. I'll give it a shot.

KenMacD · 2016-06-29 21:42:38

Since upgrading to 4.6.3-1-ARCH my laptop screen will randomly start flickering back and forth to blank.

dmesg at the time shows:

[drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* uncleared fifo underrun on pipe A
[drm:intel_cpu_fifo_underrun_irq_handler [i915]] *ERROR* CPU pipe A FIFO underrun
[drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A
[drm:intel_pipe_update_start [i915]] *ERROR* Potential atomic update failure on pipe A

I have no i915 kernel options on boot, xorg.conf.d doesn't have any options.

/etc/modprobe.d/i915.conf contains:

options i915 drm.debug=14

When this happens I can get it back to normal by disabling that screen, then re-enabling it. One of the scripts that I'm using for it is:

xrandr --output HDMI1 --off --output VIRTUAL1 --off --output eDP1 --off
xrandr --output HDMI1 --off --output VIRTUAL1 --off --output eDP1 --mode 3200x1800 --pos 0x0 --rotate normal

Anyone have any suggestions?

Slithery · 2016-06-30 00:01:57

I've no answer for you but have you seen...
https://bbs.archlinux.org/viewtopic.php?id=214177

KenMacD · 2016-06-30 14:00:14

slithery wrote:

I've no answer for you but have you seen...
https://bbs.archlinux.org/viewtopic.php?id=214177

Thank you. I did but the bios is up to date, and the microcode is installed.

I also found some people found 4.6.3 fixed the issue for them, but no such luck here.

KenMacD · 2016-07-04 14:28:56

I'm not sure if you're still seeing this. I am and posted at:

https://bbs.archlinux.org/viewtopic.php?id=214317

If you are, try installing `intel-gpu-tools` then run `intel_gpu_top`. It seems to have made it stable for me, but only while it's running.

KenMacD · 2016-07-04 14:30:19

Interestingly if I run

intel_gpu_top

from the intel-gpu-tools package it will stop the laptop screen from flashing and make it stable, but only while it's running.

ewaller · 2016-07-04 17:53:31

Merged similar threads

skiguy0123 · 2016-07-19 15:56:17

I'm having a similar problem. After anywhere from 1 min to several hours of use, the display will turn off or flicker unless I move the mouse constantly. The work around usggested by KenMacD works for now, but it's not a particularly satisfying long term solution. Anyone have any thoughts on how to get to the bottom of this?
I wonder if it's similar to this.

skiguy0123 · 2016-07-24 08:46:29

Update: based on advice from this post, I added /etc/modprobe.d/i915.conf which contains:

options i915 enable_psr=0

This seems to have fixed the problem

Last edited by skiguy0123 (2016-07-24 08:47:14)

R00KIE · 2016-07-24 14:06:51

On my machine I have a strong suspicion that a similar problem is related to deeper package sleep states, although more than one problem might be involved.

jadelord · 2016-08-02 07:21:29

KenMacD wrote:

When this happens I can get it back to normal by disabling that screen, then re-enabling it. One of the scripts that I'm using for it is:
xrandr --output HDMI1 --off --output VIRTUAL1 --off --output eDP1 --off
xrandr --output HDMI1 --off --output VIRTUAL1 --off --output eDP1 --mode 3200x1800 --pos 0x0 --rotate normal
Anyone have any suggestions?

I noticed the error appears only while running an xrandr script to use multiple monitors. The errors could be seen by running:

sudo journalctl -p err..alert -b

However, for me, setting a modprobe option did not do anything. And some users have commented the same here:
https://bugs.freedesktop.org/show_bug.cgi?id=91883

I managed to stop the error messages appearing after setting kernel parameter during boot, i915.enable_psr=0. (as mentioned in https://wiki.archlinux.org/index.php/De … owersaving)

brix · 2016-08-17 15:45:57

jadelord wrote:

[...] for me, setting a modprobe option did not do anything. And some users have commented the same here:
https://bugs.freedesktop.org/show_bug.cgi?id=91883
I managed to stop the error messages appearing after setting kernel parameter during boot, i915.enable_psr=0. (as mentioned in https://wiki.archlinux.org/index.php/De … owersaving)

Thanks very much for this! Works for me, too, as the modprobe option didn't.

[ i5-4200U - Intel Haswell-ULT Integrated Graphics Controller - linux-4.7-1 (though presumably it would have solved the identical issue with Linux 4.6 as well) ]

KenMacD · 2016-09-01 15:28:46

Just to update, this still occurs with 4.7.2-1

tom0000 · 2016-09-02 03:45:48

Same thing has been happening to me for months: screen randomly starts flashing and is only temporarily stabilized by mouse movement. I've noticed changing the refresh rate on the display temporarily resolves the issue (maybe for a few hours), but I like your script to turn off/on the panel. Maybe tie that to a key combination as a quick workaround?

after searching a while back I concluded it was a hardware issue. My computer is the lenovo yoga 2 pro, but some people in lenovo's forum claim the same problem is happening to some dell model also that uses the same lcd panel. In those threads there are an assortment of OSes and versions of those OSes, which is why I assume HW issue.. There were dozens of people complaining, and all those under warranty claimed lenovo was replacing parts- so they believed it to be a hardware issue as well.

brix · 2016-09-02 04:15:38

tom0000 wrote:

Same thing has been happening to me for months: screen randomly starts flashing and is only temporarily stabilized by mouse movement. I've noticed changing the refresh rate on the display temporarily resolves the issue (maybe for a few hours), but I like your script to turn off/on the panel. Maybe tie that to a key combination as a quick workaround?
after searching a while back I concluded it was a hardware issue. My computer is the lenovo yoga 2 pro, but some people in lenovo's forum claim the same problem is happening to some dell model also that uses the same lcd panel. In those threads there are an assortment of OSes and versions of those OSes, which is why I assume HW issue.. There were dozens of people complaining, and all those under warranty claimed lenovo was replacing parts- so they believed it to be a hardware issue as well.

I too have a Lenovo Yoga 2 Pro, was experiencing the same (or very similar) problems and beginning to wonder if a hardware fault was to blame.

I'm very curious to know if you've tried booting with the addition of boot parameter

i915.enable_psr=0

It solved the problem for me (see upthread).

xedchou · 2016-09-02 18:54:37

KenMacD wrote:

Just to update, this still occurs with 4.7.2-1

Yes, I just updated to 4.7.2, the problem is still there. But I found 4.5.7 didn't have the problem.

bit101 · 2016-09-27 12:18:45

So glad I found this post. I also have a Yoga 2 Pro and have been experiencing this exact issue. Another quick workaround I've found is that closing the lid and opening it, forcing a suspend / resume clears things up - at least for another hour or two.

But I just applied the i915.enable_psr=0 fix. I won't really know the results for a few hours though. Crossing my fingers though. This was my first full Arch install and it's working great outside this one annoying issue.

Awebb · 2016-09-27 13:55:34

Also: https://bugs.freedesktop.org/show_bug.cgi?id=91883

bit101 · 2016-09-27 14:34:55

Thanks. It's been about 3.5 hours so far, and all is solid. Hoping for the best.

bit101 · 2016-09-27 18:11:01

Well, it went past 5 hours with no glitches. I rebooted in the process of installing some other modules. So the count is restarted. But I'm going to say that i915.enable_psr=0 has handled things for me. Thanks so much!

tom0000 · 2016-09-29 22:40:47

brix wrote:

I'm very curious to know if you've tried booting with the addition of boot parameter
i915.enable_psr=0
It solved the problem for me (see upthread).

I hadn't bothered because all the windows users I read about were turning that param off with mixed results.
But I did finally try it and seems to have solved it for me at least... been weeks now without the screen flicker issue or the atomic update failure message.

Arch Linux

#1 2016-06-26 03:25:41

Potential atomic update failure on pipe A

#2 2016-06-26 17:57:49

Re: Potential atomic update failure on pipe A

#3 2016-06-27 03:02:14

Re: Potential atomic update failure on pipe A

#4 2016-06-29 21:42:38

Re: Potential atomic update failure on pipe A

#5 2016-06-30 00:01:57

Re: Potential atomic update failure on pipe A

#6 2016-06-30 14:00:14

Re: Potential atomic update failure on pipe A

#7 2016-07-04 14:28:56

Re: Potential atomic update failure on pipe A

#8 2016-07-04 14:30:19

Re: Potential atomic update failure on pipe A

#9 2016-07-04 17:53:31

Re: Potential atomic update failure on pipe A

#10 2016-07-19 15:56:17

Re: Potential atomic update failure on pipe A

#11 2016-07-24 08:46:29

Re: Potential atomic update failure on pipe A

#12 2016-07-24 14:06:51

Re: Potential atomic update failure on pipe A

#13 2016-08-02 07:21:29

Re: Potential atomic update failure on pipe A

#14 2016-08-17 15:45:57

Re: Potential atomic update failure on pipe A

#15 2016-09-01 15:28:46

Re: Potential atomic update failure on pipe A

#16 2016-09-02 03:45:48

Re: Potential atomic update failure on pipe A

#17 2016-09-02 04:15:38

Re: Potential atomic update failure on pipe A

#18 2016-09-02 18:54:37

Re: Potential atomic update failure on pipe A

#19 2016-09-27 12:18:45

Re: Potential atomic update failure on pipe A

#20 2016-09-27 13:55:34

Re: Potential atomic update failure on pipe A

#21 2016-09-27 14:34:55

Re: Potential atomic update failure on pipe A

#22 2016-09-27 18:11:01

Re: Potential atomic update failure on pipe A

#23 2016-09-29 22:40:47

Re: Potential atomic update failure on pipe A

Board footer