Created attachment 143620 [details] dmesg output on booted Arch with 5.0.0-rc7-drm-intel-next-git-b4bf44d2dcbd and enabled debug flag == Machine details == Host: Aspire E5-574 V1.14 CPU: Intel i5-6200U (4) @ 2.700GHz GPU: Intel Skylake GT2 [HD Graphics 520] Memory: 15904MiB (2 * 8GB) Resolution: 1920x1080 == Issue description == This problem was already solved for me once. When I bought this laptop I immediately changed its hardware configuration. As far as I remember, that problem was introduced when I extended RAM memory from 1x4GB to 2x8GB. Any Linux OS with any kernel version (that supported Skylake graphics at all) had that issue. System starts normally and image rendering is totally fine, but sometimes I can see that some part of the screen is black for a small fraction of a second. Sometimes I can see some little scratches and distortions too. Back then I commented on existing, related bug https://bugs.freedesktop.org/show_bug.cgi?id=94605#c87 I posted some dmesg outputs and then Paulo Zanoni posted kernel patches I tested and first patch was enough for me to not experiencing this issue anymore. For now I modified my hardware configuration again by replacing harddrive (from 256G SSD to 512GB SSD). Unfortunately now the problem is back. I already checked Windows and it works just fine without any flickering nor any distortions at all. I don't know why it happens, maybe new harddrive consumes more power or something. Anyway there was also something that Paulo Zanoni suggested - sometimes first row of pixels is broken while playing video. Back when we exchanged comments I wasn't able to reproduce this issue, but on regular basis I started to notice green line on top of the screen showin ocasionally when playing fullscreen video. == Steps to reproduce == Simply run any graphical environment with i915 and sooner or later I can experience some broken frames. It feels like it happens the more often, the more GPU resources are used - e.g. playing YouTube video and running Virtual Machine with other OS makes it happening pretty often. I doesn't seem like X.org related problem, because I can easily reproduce it on the Wayland session. == Logs == When this happens, this appears in dmesg output: [ 2609.752028] [drm:intel_hpd_irq_handler [i915]] HPD interrupt storm detected on PIN 6 [ 2609.752539] [drm] HPD interrupt storm detected on connector DP-1: switching from hotplug detection to polling The `dmesg.txt` attachment has full dmesg output. I added kernel parameters drm.debug=0xe and log_buf_len=4M so maybe those logs would be any useful. == Environments == I tried with few different kernel versions and distros and it's the same everywhere: Arch Linux: Linux 5.0 Arch Linux: Linux 5.0-rc7 (linux-drm-intel-next) KDE Neon (Ubuntu?): Linux ~4.16 Debian Stable: Linux 4.8 == Summary == Since this is not any blocking sort of bug, it is really annoying and breaks the overall experience. I would be very grateful for help / fixes. From my side, if I only have some time, I can make any tests, compile kernels, provide logs and so on.
There's clearly something fishy going on with your machine on account of those interrupt storms. You have no external displays connected right? The flickers could be underruns (not seeing those in the logs though), or they could be some kind of eDP signal integrity issues. First thing I would suggest is a BIOS update. Can you try that?
Actually there's no any external screen connected, but sometimes I use external screens and works like a harm both with VGA and HDMI port. Here's my xrandr --query output Screen 0: minimum 320 x 200, current 1920 x 1080, maximum 8192 x 8192 eDP-1 connected primary 1920x1080+0+0 (normal left inverted right x axis y axis) 344mm x 193mm 1920x1080 60.05 + 60.01* 59.97 59.96 59.93 1680x1050 59.95 59.88 1400x1050 59.98 1600x900 59.99 59.94 59.95 59.82 1280x1024 60.02 1400x900 59.96 59.88 1280x960 60.00 1440x810 60.00 59.97 1368x768 59.88 59.85 1280x800 59.99 59.97 59.81 59.91 1280x720 60.00 59.99 59.86 59.74 1024x768 60.04 60.00 960x720 60.00 928x696 60.05 896x672 60.01 1024x576 59.95 59.96 59.90 59.82 960x600 59.93 60.00 960x540 59.96 59.99 59.63 59.82 800x600 60.00 60.32 56.25 840x525 60.01 59.88 864x486 59.92 59.57 700x525 59.98 800x450 59.95 59.82 640x512 60.02 700x450 59.96 59.88 640x480 60.00 59.94 720x405 59.51 58.99 684x384 59.88 59.85 640x400 59.88 59.98 640x360 59.86 59.83 59.84 59.32 512x384 60.00 512x288 60.00 59.92 480x270 59.63 59.82 400x300 60.32 56.34 432x243 59.92 59.57 320x240 60.05 360x202 59.51 59.13 320x180 59.84 59.32 HDMI-1 disconnected (normal left inverted right x axis y axis) DP-1 disconnected (normal left inverted right x axis y axis) I think I made BIOS upgrade once, but it was like 2 years ago. I'll check if theres some update available and let you know.
Ok, thanks to my ext-hdd windows I could quickly upgrade BIOS downloaded from acer.com and I upgraded Changelog: (not full info available) * Improve touchpad performance * Update CPU Microcode for Intel SGX security issue Unfortunately issue remains unchanged. Oh, I forgot - this kernel message: [ 2609.752539] [drm] HPD interrupt storm detected on connector DP-1: switching from hotplug detection to polling It was always here even when there wasn't any (visible?) flickering. And yes - it was always strange that it complains about DP-1 connector which is not even used.
Ok, so today I've done some simple tests and now I think those DP-1 connector messages are not related to eDP-1 internal laptop screen flickering. First of all I tried to disable this VGA port by doing echo off > /sys/class/drm/card0-DP-1/status No change either with those messages and screen flickering, but when I connected ext display to it, it wasn't working until I wrote 'detect' to the same file. External screen connected to VGA output is normally detected, I can set it up using Plasma Desktop Settings, it works just fine. When it's connected there are no more logs about it in dmesg output. My internal screen still flickers, so I don't think this is related.
It looks like running older kernel with i915.enable_rc6=0 does the trick. I tried Debian Stable on kernel v4.9 with XFCE and it flickers with rc6 enabled while it stops when I disable it. Unfortunately newer kernels does not respect this parameter anymore. I managed to compile Linux 4.9 on my Arch, but for some reason it cannot load Qt5 libraries, so my Desktop can't start. I was able to run Gnome and I was starring at glxgears (~4500fps) for 15 minutes straight. No flickering... So highly possible this is RC6 related :/
Have you verified the issue with drmtip? (https://cgit.freedesktop.org/drm-tip) Can you attach the dmesg from boot with drmtip kernel?
Reporter, do you still have the issue with latest drmtip? Can you please confirm?
I haven't been using this laptop for a while, but I haven't sell it yet. I replaced the harddrive so I need to install some distro and I will check this out.
(In reply to Dominik Klementowski from comment #8) > I haven't been using this laptop for a while, but I haven't sell it yet. I > replaced the harddrive so I need to install some distro and I will check > this out. Any updates here? Dropping the priority to Medium until it's confirmed that issue persists with drmtip.
I can only reproduce this issue if I extend this laptop RAM to 16GB (2x8GB DIMM). I installed some Ubuntu-based distro and I used this: https://kernel.ubuntu.com/~kernel-ppa/mainline/drm-tip/2019-09-22/ fresh build of kernel. I confirm, those ocasional flickers and glitches still happen.
(In reply to Dominik Klementowski from comment #10) > I can only reproduce this issue if I extend this laptop RAM to 16GB (2x8GB > DIMM). I installed some Ubuntu-based distro and I used this: > https://kernel.ubuntu.com/~kernel-ppa/mainline/drm-tip/2019-09-22/ fresh > build of kernel. I confirm, those ocasional flickers and glitches still > happen. Issue is not reproducible when the machine has got 8GB RAM? Can you please attach the latest dmesg (from drmtip) from boot with kernel parameters drm.debug=0x1e log_buf_len=4M?
Created attachment 145536 [details] dmesg dump after tearing So I booted up with additional kernel parameters. I run some WebGL benchmark to stress the GPU a little, then I was just watching a YouTube video. The log file I attached should contain some logs that were spit while those artifacts happened. Last glitches were visible like 10 seconds before I dumped those logs. I'll try to reproduce the problem with 8GB of RAM to make me completely sure if it's not happening with less amount of memory, but that was my impression.
Created attachment 145597 [details] Dmesg dump on SteamOS Ok, now I tried to run various different things on that laptop. I run SteamOS and now it seems like the behavior is dependent on what is being rendered. The Steam full screen menu causes a lot of flickering - even few blinks per second. It's the same with both default kernel (which is 4.19) and manually compiled current drm-tip revision.
(In reply to Dominik Klementowski from comment #13) > Created attachment 145597 [details] > Dmesg dump on SteamOS > I couldn't find any issues from the logs, is the dmesg from drmtip kernel?
Yes, I cloned and compiled https://github.com/freedesktop/drm-tip
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/intel/issues/242.
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.