http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c1 Bernhard Wiedemann changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |oholecek@suse.com, | |sbrabec@suse.com, | |sreeves@suse.com --- Comment #1 from Bernhard Wiedemann --- might be related to pulseaudio -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c3 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |darren@freemaninstruments.c | |om Flags| |needinfo?(darren@freemanins | |truments.com) --- Comment #3 from Takashi Iwai --- As a start, please run alsa-info.sh (with --no-upload option) as root before and after suspend, and upload the outputs to Bugzilla. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c5 --- Comment #5 from Darren Freeman --- Created attachment 664294 --> http://bugzilla.opensuse.org/attachment.cgi?id=664294&action=edit "alsa-info.sh --no-upload" before suspend -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c7 --- Comment #7 from Darren Freeman --- Created attachment 664297 --> http://bugzilla.opensuse.org/attachment.cgi?id=664297&action=edit "pactl list sinks" before suspend -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c9 Darren Freeman changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(darren@freemanins | |truments.com) | --- Comment #9 from Darren Freeman --- Performed alsa-info.sh --no-upload, performed pactl list sinks >file1.txt . Then suspended, resumed. Sound was still coming out of the earphone jack. Suspended, resumed. Now sound was coming out of the internal speakers, which is not the correct behaviour. Performed alsa-info.sh --no-upload, performed pactl list sinks >file2.txt . Then I removed and reinserted the audio jack, the problem went away. Hopefully you can tell what's going wrong. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c11 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(darren@freemanins | |truments.com) --- Comment #11 from Takashi Iwai --- Right, something went wrong wrt the jack detection after resume. Could you install hda-verb and run like below before and after suspend as root? hda-verb /dev/snd/hwC1D0 0x25 GET_PIN_SENSE 0 It should show the value with bit 31 on when the headphone jack is detected. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c13 --- Comment #13 from Darren Freeman --- (All as root..) echo 1 > /sys/kernel/debug/tracing/events/hda/enable hda-verb /dev/snd/hwC1D0 0x25 GET_PIN_SENSE 0 nid = 0x25, verb = 0xf09, param = 0x0 value = 0x80000000 suspend/resume.. bug is triggered on the first attempt.. hda-verb /dev/snd/hwC1D0 0x25 GET_PIN_SENSE 0 nid = 0x25, verb = 0xf09, param = 0x0 value = 0x80000000 cat /sys/kernel/debug/tracing/trace > trace.log echo 0 > /sys/kernel/debug/tracing/events/hda/enable -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c15 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(darren@freemanins | |truments.com) --- Comment #15 from Takashi Iwai --- I prepared a test KMP in OBS home:tiwai:bnc964621 repo. After installing it, you should have /sys/module/snd_hda_codec_via/parameters/delay1 and /sys/module/snd_hda_codec_via/parameters/delay2 files. These are delays to be put at resume. As default, both are 10 ms. At first, try as is, whether it works. It has an extra delay after init, so this might help already. If this doesn't work, try to increase the values. You can change it like: echo -n 100 > /sys/module/snd_hda_codec_via/parameters/delay1 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c17 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(darren@freemanins | |truments.com) --- Comment #17 from Takashi Iwai --- What git commit ID has FACTORY kernel? Check "rpm -qi kernel-default-4.4.0-3" and you see it in Git revision. Also, double-check whether it happens depending on the kernel version. That is, try to uninstall KMP again, reinstall the kernel you saw the issue, and retest. Meanwhile, I enabled the build of KMP for Kernel:stable. If the problem is reproduced reliably with Kernel:stable, try KMP again. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c19 --- Comment #19 from Takashi Iwai --- (In reply to Darren Freeman from comment #18)

The package name is kernel-default-4.4.0-3.1.x86_64

It is from Tumbleweed OSS.

And what shows rpm -qi with it? -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c21 --- Comment #21 from Takashi Iwai --- (In reply to Darren Freeman from comment #20)

GIT Revision: 9f68b909e79a8520cee723b728fec80e33eaef6b

So, this is identical with 4.4.0-8.g9f68b90-default you reported (see the commit ID). That means that it must be something else than the kernel that changed the behavior. Please double-check in which situation the problem happens reliably. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c23 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(darren@freemanins | |truments.com) --- Comment #23 from Takashi Iwai --- (In reply to Darren Freeman from comment #22)

This bug is definitely not fixed. I just rebooted for the first time in 11 days, to the latest kernel in OBS Kernel:stable. And after a suspend/resume, sound came out of the internal speakers.

uname output: 4.4.1-5.g6398c2d-default package name: kernel-default-4.4.1-5.1.g6398c2d.x86_64

GIT commit: 6398c2df356e9052b52ba35e636955cf7a7154d9

I'm guessing the KMP fix isn't built yet for this kernel. Due to a bug in YaST that I'm about to submit, I can't get it to update that repo, so maybe it has been built but I can't see it yet.

Anyway I think that proves that the latest kernel still has this issue, please make sure the KMP is built for the above version, and I'll try to confirm that this is definitely the fix.

The KMP rebuild may take time, but it's done full automatically, so just be patient. It's possible, though, that the old KMP still works for the new kernel as long as the kernel ABI is compatible. Check /sys/module/snd_hda_codec_via/parameters/delay1 and delay2 files are present. If yes, try to adjust them as was suggested. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=964621 http://bugzilla.opensuse.org/show_bug.cgi?id=964621#c25 Takashi Iwai changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(darren@freemanins | |truments.com) --- Comment #25 from Takashi Iwai --- The KMP isn't provided for a solution but to measure what values make the things working. So, wait for a while, then install the KMP when it matches with your running kernel. Then test with various delay values. Once when we figure out the necessary delay values, they can be put into the upstream driver code and openSUSE Kernel as well. Without that information, there can be any fix. So please give it a test. -- You are receiving this mail because: You are on the CC list for the bug.