Bug 89327

Please attach your xorg log and dmesg output.

(In reply to John from comment #0)
> Since moving to 4.0-rc1 when my screen goes off (energy saving), I lose the
> hdmi audio, standard stuff. 
> When I move the mouse I get the screen back but not the audio anymore.
> Switching it back on with xrandr works.
> 
> Thanks

Can you please attach the 'right' xrandr line, please?
I do not hear anything after screen goes off (energy saving) with drm-next-4.1-wip on r600/NI/Turks, too.

So it seems to be not only with radeonsi.

-Dieter

Created attachment 113830 [details]
dmesg-drm-next-4.1-wip.log.xz

Created attachment 113831 [details]
Xorg.0.log-drm-next-4.1-wip.xz

Please don't compress attachments unless they are really large, it makes them much harder to review.

Can you bisect?

(In reply to Alex Deucher from comment #6)
> Can you bisect?

So I have to go before your -rc1 merge?

(In reply to Dieter Nützel from comment #7)
> (In reply to Alex Deucher from comment #6)
> > Can you bisect?
> 
> So I have to go before your -rc1 merge?

Alex,

my reportet drm-next-4.1-wip error
[drm:radeon_atom_pick_dig_encoder [radeon]] *ERROR* chosen encoder in use 4
is related to this.

If I reenable audio with
xrandr --output HDMI-0 --set audio on
I get next line in dmesg.

2cond time do NOT work but
xrandr --output HDMI-0 --set audio auto
do?!

Is this right?
[    4.722072] snd_hda_intel 0000:01:00.1: Handle VGA-switcheroo audio client

There is no other 'enable audio' in dmesg.

Created attachment 113833 [details]
john's dmesg

Created attachment 113834 [details]
john's xorg.log

(In reply to Dieter Nützel from comment #2)
> (In reply to John from comment #0)
> > Since moving to 4.0-rc1 when my screen goes off (energy saving), I lose the
> > hdmi audio, standard stuff. 
> > When I move the mouse I get the screen back but not the audio anymore.
> > Switching it back on with xrandr works.
> > 
> > Thanks
> 
> Can you please attach the 'right' xrandr line, please?
> I do not hear anything after screen goes off (energy saving) with
> drm-next-4.1-wip on r600/NI/Turks, too.
> 
> So it seems to be not only with radeonsi.
> 
> -Dieter

xrandr --output HDMI-0 --set audio on
works for me

Bisected

6f945693be7eea24b1a8e5ce252a96df98d55a5c is the first bad commit
commit 6f945693be7eea24b1a8e5ce252a96df98d55a5c
Author: Slava Grigorev <slava.grigorev@amd.com>
Date:   Wed Dec 10 14:52:43 2014 -0500

    radeon/audio: applied audio_dpms() and audio_mode_set() calls
    
    Reviewed-by: Christian König <christian.koenig@amd.com>
    Signed-off-by: Slava Grigorev <slava.grigorev@amd.com>
    Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

:040000 040000 33622643fd03f2afac723aa1356fe9c638a8fb3c fac1a0f988f61f513a75af0b0e55cccc4246d500 M drivers

git bisect start
# bad: [308de41ba3bde80d50833718a2e37e49be868f5a] drm/radeon: use NULL rather then 0 in audio detect
git bisect bad 308de41ba3bde80d50833718a2e37e49be868f5a
# good: [49ead33b9f0e0ba0b528951a05986486196848ef] drm/radeon: comment out some currently unused 7xx dpm code
git bisect good 49ead33b9f0e0ba0b528951a05986486196848ef
# good: [64424d6e45aeee311a4231def7e125bcc2de0855] radeon/audio: consolidate update_acr() functions (v2)
git bisect good 64424d6e45aeee311a4231def7e125bcc2de0855
# good: [3be2e7d0e705621c1bb41eeabb63b122d50ecff3] radeon/audio: moved mute programming to a separate function
git bisect good 3be2e7d0e705621c1bb41eeabb63b122d50ecff3
# bad: [6f945693be7eea24b1a8e5ce252a96df98d55a5c] radeon/audio: applied audio_dpms() and audio_mode_set() calls
git bisect bad 6f945693be7eea24b1a8e5ce252a96df98d55a5c
# good: [6e72376dcc663e4b8a00cdd08f61a8623f572ef1] radeon/audio: consolidate audio_mode_set() functions
git bisect good 6e72376dcc663e4b8a00cdd08f61a8623f572ef1
# first bad commit: [6f945693be7eea24b1a8e5ce252a96df98d55a5c] radeon/audio: applied audio_dpms() and audio_mode_set() calls

But I can't revert commit 6f94569 on top of HEAD.

During some restarts I got this in dmesg:
[    4.159787] [drm:btc_dpm_set_power_state [radeon]] *ERROR* rv770_restrict_performance_levels_before_switch failed

And my reported
r600 - NI/Turks: flickering with 'geom-outlining-150'
seems to be related!?

It was first gone under drm-next-4.1-wip but arise with going bak to drm-next-3.20-wip (3.19-rc5). I had to power down my system to get it back after 'git bisect reset' and got the above messages during some power cycles.

(In reply to Dieter Nützel from comment #12)
> Bisected
> 
> 6f945693be7eea24b1a8e5ce252a96df98d55a5c is the first bad commit
> commit 6f945693be7eea24b1a8e5ce252a96df98d55a5c
> Author: Slava Grigorev <slava.grigorev@amd.com>
> Date:   Wed Dec 10 14:52:43 2014 -0500
> 
>     radeon/audio: applied audio_dpms() and audio_mode_set() calls
>     
>     Reviewed-by: Christian König <christian.koenig@amd.com>
>     Signed-off-by: Slava Grigorev <slava.grigorev@amd.com>
>     Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
> 
> :040000 040000 33622643fd03f2afac723aa1356fe9c638a8fb3c
> fac1a0f988f61f513a75af0b0e55cccc4246d500 M drivers
> 
> git bisect start
> # bad: [308de41ba3bde80d50833718a2e37e49be868f5a] drm/radeon: use NULL
> rather then 0 in audio detect
> git bisect bad 308de41ba3bde80d50833718a2e37e49be868f5a
> # good: [49ead33b9f0e0ba0b528951a05986486196848ef] drm/radeon: comment out
> some currently unused 7xx dpm code
> git bisect good 49ead33b9f0e0ba0b528951a05986486196848ef
> # good: [64424d6e45aeee311a4231def7e125bcc2de0855] radeon/audio: consolidate
> update_acr() functions (v2)
> git bisect good 64424d6e45aeee311a4231def7e125bcc2de0855
> # good: [3be2e7d0e705621c1bb41eeabb63b122d50ecff3] radeon/audio: moved mute
> programming to a separate function
> git bisect good 3be2e7d0e705621c1bb41eeabb63b122d50ecff3
> # bad: [6f945693be7eea24b1a8e5ce252a96df98d55a5c] radeon/audio: applied
> audio_dpms() and audio_mode_set() calls
> git bisect bad 6f945693be7eea24b1a8e5ce252a96df98d55a5c
> # good: [6e72376dcc663e4b8a00cdd08f61a8623f572ef1] radeon/audio: consolidate
> audio_mode_set() functions
> git bisect good 6e72376dcc663e4b8a00cdd08f61a8623f572ef1
> # first bad commit: [6f945693be7eea24b1a8e5ce252a96df98d55a5c] radeon/audio:
> applied audio_dpms() and audio_mode_set() calls

Based on the commit's name I'd say it looks like a good match.
Thank you for doing the bisecting instead of me :)

Created attachment 113872 [details] [review]
patch 1/5

Please try the attached patch set.

Created attachment 113873 [details] [review]
patch 2/5

Created attachment 113874 [details] [review]
patch 3/5

Created attachment 113875 [details] [review]
patch 4/5

Created attachment 113876 [details] [review]
patch 5/5

Thank you Alex,

it is under works.
But could you offer a git tree?
Can't find them in drm-next-4.1-wip.
I couldn't apply patches to my old former drm-next-3.19 and now moved (during system change) to drm-next-4.1-wip tree any longer?! - Any hints?

Maybe I have to work with a new checked out one.
Or could it be, that I only have forgotten to add me to the git 'system'?
Have to investigate.

-Dieter

These plus some additional fixes:
http://cgit.freedesktop.org/~agd5f/linux/log/?h=audio-fixes

Patches 1-5 are
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
on
drm-next-4.1-wip

Got all applied fine with 'git am ...'
I think that I couldn't apply 'extern' patches (e.g. from Dave) on the tree.

Thanks a lot Alex!

Cheers,
  Dieter

I have patched rc2 and so far after more than a day I have not needed to re-enable the audio, so I'd guess the patches do work fine.

Thanks for the quick fix!

It seems rc3 includes these already, so that's all good.