Nouveau - Sep 2019 - [Bug 111724] New: NVE6 (GK106) memory re-clocking breaks GpuTest plot3d benchmark

https://bugs.freedesktop.org/show_bug.cgi?id=111724

            Bug ID: 111724
           Summary: NVE6 (GK106) memory re-clocking breaks GpuTest plot3d
                    benchmark
           Product: xorg
           Version: unspecified
          Hardware: x86-64 (AMD64)
                OS: Linux (All)
            Status: NEW
          Severity: normal
          Priority: not set
         Component: Driver/nouveau
          Assignee: nouveau at lists.freedesktop.org
          Reporter: mmenzyns at redhat.com
        QA Contact: xorg-team at lists.x.org

I have stepped upon a problem with NVE6 (GK106) in GpuTest
https://www.geeks3d.com/gputest/ plot3d benchmark that occurs only in plot3d
and nowhere else. There are visible glitches and when left for a longer time
Nouveau seems to crash.

The GPU has 4 profiles: 
07: core 324 MHz memory 648 MHz
0a: core 324-862 MHz memory 1620 MHz
0d: core 549-1228 MHz memory 6008 MHz
0f: core 549-1228 MHz memory 6008 MHz

The problem occurs when switching re-clocking profile directly from 648 MHz to
6008 MHz skipping the 0xA 1620 MHz profile. If gone through 0xA profile
everything works fine.

If the memory re-clocking is disabled, it works fine. If there is 0xF profile
set directly (breaking the benchmark) with memory re-clocking enabled, then the
nouveau gets unloaded, and nouveau gets loaded back with memory re-clocking
disabled, when changing re-clocking profiles it still glitches. Which implies
something that breaks this is only touched when the memory re-clocking is
enabled.

I have gone through all nouveau pmu scripts traces, checked every difference
(of the scripts) with Nvidia driver and nothing seemed to affect this problem
that has different values than Nvidia. Actual code which was changing the
values for 0xf profile to be same as Nvidia is here:
https://github.com/mmenzyns/nouveau/tree/linux-5.2_gk106_memory_issues. The
scripts for the highest-profile should be almost identical between Nvidia and
Nouveau.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/60bf375f/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #1 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145395
  --> https://bugs.freedesktop.org/attachment.cgi?id=145395&action=edit
mmiotrace log when changing directly from 0x7 to 0xf

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/d3d047fd/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #2 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145396
  --> https://bugs.freedesktop.org/attachment.cgi?id=145396&action=edit
pmu log from dmesg when changing directly from 0x7 to 0xf

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/7a6ea93f/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #3 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145397
  --> https://bugs.freedesktop.org/attachment.cgi?id=145397&action=edit
pmu log from dmesg with noveau code modified so the values are same as with
nvidia driver

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/b4c89812/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #4 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145398
  --> https://bugs.freedesktop.org/attachment.cgi?id=145398&action=edit
mmiotrace log when changing from 0x7 to 0xa and then to 0xf, benchmark works in
this case

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/544bb5c3/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #5 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145399
  --> https://bugs.freedesktop.org/attachment.cgi?id=145399&action=edit
pmu log from dmesg when changin from 0x7 to 0xa and then to 0xf, benchmark
working

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/1bb63f27/attachment-0001.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #6 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145400
  --> https://bugs.freedesktop.org/attachment.cgi?id=145400&action=edit
mmiotrace log from nvidia driver

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/4c7769a9/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #7 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145401
  --> https://bugs.freedesktop.org/attachment.cgi?id=145401&action=edit
strap_peek

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/9d5d1ca0/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #8 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145402
  --> https://bugs.freedesktop.org/attachment.cgi?id=145402&action=edit
vbios.rom

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/48a70990/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #9 from Ilia Mirkin <imirkin at alum.mit.edu> ---
Mark - try using blob ctxsw firmware. Perhaps ours misses something. There are
some GK106's which just die immediately with out firmware... (See
VideoAcceleration wiki page for how to extract firmware from blob drivers.)

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/075c9630/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #10 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145403
  --> https://bugs.freedesktop.org/attachment.cgi?id=145403&action=edit
part of dmesg when running plot3d fullscreen

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/a3d915df/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #11 from Mark Menzynski <mmenzyns at redhat.com> ---
Created attachment 145404
  --> https://bugs.freedesktop.org/attachment.cgi?id=145404&action=edit
screenshot from the glitchy benchmark

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/e2a6de35/attachment.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

--- Comment #12 from Mark Menzynski <mmenzyns at redhat.com> ---
(In reply to Ilia Mirkin from comment #9)
> Mark - try using blob ctxsw firmware. Perhaps ours misses something. There
> are some GK106's which just die immediately with out firmware... (See
> VideoAcceleration wiki page for how to extract firmware from blob drivers.)
Doesn't work. Same problem.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20190917/86a7d7e2/attachment-0001.html>

https://bugs.freedesktop.org/show_bug.cgi?id=111724

Martin Peres <martin.peres at free.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |MOVED
             Status|NEW                         |RESOLVED

--- Comment #13 from Martin Peres <martin.peres at free.fr> ---
-- 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/xorg/driver/xf86-video-nouveau/issues/503.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://lists.freedesktop.org/archives/nouveau/attachments/20191204/7e16c00e/attachment.html>