[Guado/wolf/lumpy/Gandof] Audio input node is always switching to 3.5 after Suspend/Resume. |
||||||
Issue description
Chrome Version: 10176.5.0 / 64.0.3282.11
OS: ChromeOS
Device: Guado
What steps will reproduce the problem?
(1) Sign in to the device.
(2) Plug 3.5 mm headset.
(3) Plug USB headset.
(4) Suspend the device.
powerd_dbus_suspend
(5) Press any key to resume the device.
(6) check the audio nodes form cras_test_client.
What is the expected result?
USB audio input and output nodes should be active.
What happens instead?
USB audio output and 3.5 audio input nodes are active.
Please use labels and text to provide additional information.
chronos@localhost / $ cras_test_client --dump_server_info
System Volume (0-100): 100
Capture Gain (-17.25 - 30.00): 18.77dB
User muted: Not muted
Output Devices:
ID Name
14 Microsoft LifeChat LX-3000: USB Audio:3,0
9 HDA Intel PCH: ALC283 Analog:1,0
8 HDA Intel HDMI: HDMI 2:0,8
7 HDA Intel HDMI: HDMI 1:0,7
6 HDA Intel HDMI: HDMI 0:0,3
Output Nodes:
Stable Id ID Vol Plugged L/R swapped Time Hotword Type Name
(a50fedcf) 14:0 75 yes no 1512602705 USB *Microsoft LifeChat LX-3000: USB Audio:3,0: Speaker
(d5da8553) 9:0 100 yes no 1512602440 HEADPHONE Headphone Jack
(5280b26d) 8:0 100 no no 0 HDMI HDMI/DP,pcm=8 Jack
(78297d51) 7:0 75 yes no 1512602846 HDMI ASUS VE258
(f2070728) 6:0 100 no no 0 HDMI HDMI/DP,pcm=3 Jack
(2fb77ef4) 6:1 100 no no 0 HDMI HDMI
Input Devices:
ID Name
15 Microsoft LifeChat LX-3000: USB Audio:3,0
10 HDA Intel PCH: ALC283 Analog:1,0
5 Post DSP Loopback
4 Post Mix Pre DSP Loopback
Input Nodes:
Stable Id ID Gain Plugged L/R swapped Time Hotword Type Name
(0054823c) 15:0 0 yes no 1512602705 USB Microsoft LifeChat LX-3000: USB Audio:3,0: Mic
(ca711b33) 10:0 0 yes no 1512602846 MIC *Mic Jack
(8c74f766) 5:0 0 yes no 0 POST_DSP_LOOPBACK Post DSP Loopback
(8a04af91) 4:0 0 yes no 0 POST_MIX_LOOPBACK Post Mix Pre DSP Loopback
Attached clients:
ID pid uid
17 2737 1000
18 2737 1000
19 2737 1000
21 9471 1000
Num active streams: 3
Last audio active time: 4648, 470673317
,
Dec 6 2017
,
Dec 6 2017
,
Dec 7 2017
Is this a regression?
,
Dec 7 2017
The doc for node selection logic: https://docs.google.com/document/d/1BTTl_9yL6c9DQVLvA-Btuwms07xK6jHh33w4XHIlWAw/edit?ts=599cb3f1 Note that if 3.5mm headphone is hot-plugged, it will be selected as active node. I am not sure whether on resume, CRAS will report a NodesChanged event. I think it depends on driver. For example, on bob, across suspend/resume, there will be no plug/unplug event of 3.5mm headphone. Maybe on Guado, the 3.5mm headphone will be detected as unplugged and then plugged across suspend resume. If that is true, 3.5mm headphone will be selected. Hi sontis, could you please check if it is a regression, and if it behave consistently ? Thanks!
,
Dec 15 2017
Able to reproduce this issue consistently on Guado, Wolf and Lumpy devices.
Both audio input and output nodes should switch to 3.5mm headset, If suspend/resume initiates plug/unplug event. But actually only the audio input node is changing to 3.5mm hadset.
Repo steps:
1. sign in to the device.
2. Plug 3.5mm headset.
Observation:3.5 is selected as audio input and output
3. Plug USB headset.
Observation:USB is selected as audio input and output
4. Suspend/resume the device.
Observation: USB is selected as audio out and 3.5 is selected as audio input.
,
Dec 15 2017
There was additional work done by jennyz@, to get the 3.5mm microphone to comply to the behavior described in #5. Jenny, could https://chromium-review.googlesource.com/c/chromium/src/+/762141 be affecting the hotplug behavior in way to introduce this as a regression
,
Jan 23 2018
Facing same issue with M65 10323.1.0_gandof
,
Mar 14 2018
Facing same issue on M66_10452.11.0_Zako
,
Mar 14 2018
#7: Yes, it is likely caused by the change to prioritizing 3.5mm headset input, combined with the timing of audio nodes re-discovery sequence after resuming, which could make 3.5mm input show up like just hot plugged. I will need to debug to check the event sequence.
,
Mar 21 2018
Hi Jenny, let me know if you need any help with this!
,
Sep 10
Still able to reproduce this issue on M70 build 11021.11.0_lulu
,
Dec 6
Still able to reproduce this issue on M72 11316.6.0_wolf build. |
||||||
►
Sign in to add a comment |
||||||
Comment 1 by sontis@chromium.org
, Dec 6 2017