William, please attach kernel log.

Created attachment 126112 [details]
kernel log of FIFO underrun error.

*** Bug 98503 has been marked as a duplicate of this bug. ***

Reporter, ss this still valid bug?

Created attachment 128476 [details]
dmesg with drm.debug=0x1e

(In reply to Jani Saarinen from comment #4)
> Reporter, ss this still valid bug?

Still valid with drm-intel-nightly commit "f6a248e drm-tip: 2016y-12m-10d-21h-47m-23s UTC integration manifest".

I have attached a log excerpt that hopefully contains everything relevant.
If a full log or more information is needed, please let me know.

I'm sorry about this delay until getting back to you. M.G, William - lot's of code has been submitted to this area causing these messages to be cleaned (seen on https://intel-gfx-ci.01.org/CI/ at least). Can you please retest with the latest kernel (preferable from drm-tip) and mark status RESOLVED if problem does not exist anymore, REOPENED (with the latest dmesg logs) if problem still persist.

M.G, is issue still valid?

(In reply to Jani Saarinen from comment #7)
> M.G, is issue still valid?

Sorry for not responding earlier, I have been very busy. The issue still persists with kernel 4.11. I didn't had time yet to test drm-tip. I will do this as soon as possible, but it might take some days. Additionally it is difficult to reproduce the issue because it only seems to occur after a longer uptime.

OK, thanks reporting back. We will wait for your feedback.

Any feedabck now, and sorry for the push. Just trying to keep only relevant bugs open.

(In reply to Jani Saarinen from comment #10)
> Any feedabck now, and sorry for the push. Just trying to keep only relevant
> bugs open.

I am still testing drm-tip (commit 028a0b3762b318a72ab04675355db97c854c230b from 2017-05-22). So far I did not see any underrun. However this could be due to my changed setup. Because I moved I currently don't have any external monitor connected. Please allow me another 14 days to test (then with an external monitor).

If you prefer, you can also close the bug and I will reopen it if I encounter the problem again.

Created attachment 132020 [details]
drm-tip dmesg with drm.debug=0xe

The bug is still present in drm-tip commit 028a0b3762b318a72ab04675355db97c854c230b (from 2017-05-25 - not 2017-05-22, sorry). Please see attached dmesg log. Please also note the kernel warning including the trace regarding a backlight issue. Or is this a subsequent error due to the underrun?

Here is what I did when the bug was triggered:
- I did a Suspend to RAM while no external monitor was connected
- I connected an external monitor via HDMI (plus keyboard and mouse via USB)
- I opened my laptop to resume from suspend. Output on internal LCD screen was OK and I executed a script in order to switch output from internal LCD to external screen. The script basically does:

 xrandr --output eDP1 --off
 xrandr --output HDMI1 --auto --left-of eDP1 --dpi 96x96

I got only a black screen on the external monitor (backlight on). To solve the problem I executed another script using a hotkey, which does:

 xrandr --output HDMI1 --off
 xrandr --output eDP1 --auto --primary --dpi 96x96

Then I executed my first script again and output on external monitor was OK.

First of all. Sorry about spam.
This is mass update for our bugs. 

Sorry if you feel this annoying but with this trying to understand if bug still valid or not.
If bug investigation still in progress, please ignore this and I apologize!

If you think this is not anymore valid, please comment to the bug that can be closed.
If you haven't tested with our latest pre-upstream tree(drm-tip), can you do that also to see if issue is valid there still and if you cannot see issue there, please comment to the bug.

The bug seems to be fixed in newer kernel versions. I did not see anymore underruns with kernel 4.14.x in the last two months. According to the logs last week one underrun happened with kernel 4.15.8. However, I did not notice any problems. I still see messages like

[drm:intel_pipe_update_end [i915]] *ERROR* Atomic update failure on pipe B (start=37800 end=37801) time 2437 us, min 1192, max 1199, scanline start 1130, end 75

in the logs, but they don't seem to be critical and there are already other bug reports regarding this issue. So I mark this bug as RESOLVED. If problems should arise again I will reopen the bug.

Thank you, closing.