[Bug 107600] New: Regression: stuttering H.264 video w/ mythtv on old laptop w/ 915GM igp

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=107600

            Bug ID: 107600
           Summary: Regression: stuttering H.264 video w/ mythtv on old
                    laptop w/ 915GM igp
           Product: DRI
           Version: DRI git
          Hardware: x86 (IA32)
                OS: Linux (All)
            Status: NEW
          Severity: normal
          Priority: medium
         Component: DRM/Intel
          Assignee: intel-gfx-bugs at lists.freedesktop.org
          Reporter: ubuntu at hbuus.com
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
                CC: intel-gfx-bugs at lists.freedesktop.org

Created attachment 141158
  --> https://bugs.freedesktop.org/attachment.cgi?id=141158&action=edit
Dmesg from 0.000

I use mythtv (mythtv.org) to record and playback TV and for playing a video
library I have ripped from content I own using handbrake (handbrake.fr). My
current production environment consists of a server (backend in mythtv terms)
which handles recording and storage for the video library and clients (mythtv
frontends) that playback the content delivered by the server. All are running
mythtv 0.27 on ubuntu 14.04, based on kernel 3.13.

One frontend, an old 32 bit Dell laptop from 2005, was working well, even
playing 1080i mpeg2 content, until I tried to upgrade to ubuntu 18.04 based on
kernel 4.15 and mythtv 29. Specifically, there is a slight but noticeable and
distracting stutter in panning scenes when playing some H.264 standard
definition content. I cannot reproduce the problem using vlc or mplayer.

FWIW, I always upgrade by doing clean installs in separate partitions so it is
easy for me to switch back to a production environment and forward to evaluate
a new environment. I found that when I installed and ran a newer kernel on the
working production environment, I could reproduce the problem. After installing
and testing and git bisecting lots of kernels. I found what I think is the
commit that introduced the regression, in the 3.19 kernel, namely:

commit 6a2c4232ece145d8b5a8f95f767bd6d0d2d2f2bb
Author: Chris Wilson <chris at chris-wilson.co.uk>
Date:   Tue Nov 4 04:51:40 2014 -0800
drm/i915: Make the physical object coherent with GTT

And the last good commit:

commit 132f3f1767dbabfb01f3c9bd63098c65d91eeac9
Author: Imre Deak <imre.deak at intel.com>
Date:   Mon Nov 10 15:34:33 2014 +0200
drm/i915: WARN if we receive any gen9 rps interrupts

I have since compiled the latest kernel as of 8/12/2018, commit
62dae6e374af7d09a7b7a6f9e7979351d03ed19b, using a clone of drm-tip and have
confirmed the problem still happens with that kernel. 

I realize this is an old laptop, but it's working well, except for this issue.
My production environment won't be getting security updates after April, 2019,
so I really would like to upgrade before then. I welcome thoughts on a
workaround, if a kernel patch is not in the cards.

* How often does the steps listed above trigger the issue?
Always

* The following information about your system:
-- system architecture: ("uname -m"): i686
-- kernel version: ("uname -r"): 4.18.0-rc8-62dae6e374af
-- Linux distribution: Ubuntu 18.04.1 LTS
-- Machine or mother board model: Dell Inspiron B130 (Owner's Manual Copyright
2005)
-- Display connector: Built in 15.4" 1280x800, 262,144 colors, 60Hz

-- A full dmesg with debug information and/or a GPU crash dump: dmesg[12].txt

* Other attachments if relevant:
-- screenshot or photo (a picture is worth a thousand words): not applicable
-- output of "xrandr --verbose" for display mode issue: xrandr.txt
-- intel_reg_dumper output: intel_reg_dump.txt
-- VBIOS dump: vbios.dump
-- for GPU hang, get the last batch buffer: not applicable
-- for suspend/resume problems, refer to the guide: not applicable

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20180816/f950e7a3/attachment.html>