ChanServ changed the topic of #linux-msm to:
junari has joined #linux-msm
marvin24_ has joined #linux-msm
marvin24 has quit [Ping timeout: 480 seconds]
f_ has quit [Remote host closed the connection]
f_ has joined #linux-msm
f_ has quit [Remote host closed the connection]
f_ has joined #linux-msm
junari has quit [Ping timeout: 480 seconds]
<aka_[m]> So i have managed to get audio output on headphone jack on sm6115 with wcd9370 but there is an issue
<aka_[m]> once i toggle output to something else switching back doesn't get audio up anymore
<krzk> aklimov_: maybe above is related to your problems somehow with wcd937x? or would help you figure out proper mixer settings?
<aka_[m]> krzk: at some point i managed to kick soundwire in balls and it panicked
<aka_[m]> something about freeing
<aka_[m]> there is also this which happens after i switch outputs... (full message at <https://matrix.org/_matrix/media/v3/download/matrix.org/EsMreGJpxkvbXkganNvoHKEB>)
<aka_[m]> im using sc8280xp card driver
<aka_[m]> krzk: lol i noticed something now
<aka_[m]> when switching outputs during playback it fails
<aka_[m]> and after switching back there is no audio
<aka_[m]> however killing playback stream then 4s later 'wcd937x_codec codec: clk already disabled' appears and audio is working
<aka_[m]> tho this only shows on headphone playback
<aka_[m]> my device seems to wire AUX from WCD into AW87359 and EAR into AW87519
<aka_[m]> HPHL is left for wcd to manage
<aka_[m]> all of this stuff lacks so much of documentation
<aka_[m]> there are like multiple blocks rx/tx/va/wsa but i don't even can figure out what can be supported by each
<aka_[m]> any idea how i can toggle widget for speaker amp?
<aka_[m]> from what i see it shows well that its wired yet output is OFF
<aka_[m]> inputs also are "OFF"
<aka_[m]> wcd937x_codec_enable_ear_pa() differs a bit between DS and ML
<aka_[m]> seems thats just mask
<aka_[m]> so i cannot get ear to work and im not 100% sure its amp or wcd alone
<aka_[m]> i have configured lpass-rx/wcd with these settings(adjusted to mainline)
<krzk> aka_[m]: it seems your hardware setup is almost the same as aklimov_ so you both share some goals or issues
<aka_[m]> Not sure about Konrad, he had sm6115 too but he was probably stuck on awinic i2s codec part
<krzk> aka_[m]: how is your speaker/amplifier connected to WCD937x? Over AUX or EAR?
<krzk> and which bus? I2C?
<aka_[m]> WCD9370 expose HPHR/EAR/AUX and i have two awinic i2c amps connected to EAR and AUX
<krzk> ah? one to EAR and one to AUX? interesting
<aka_[m]> yea
<aka_[m]> AW87359/AW87519
<aka_[m]> 519 one on schematics have HW_EN so i didn't bother with setting it up for 359
<krzk> so answering your questions... lemme grab a drink
<aka_[m]> need pdf with schematics?
<aka_[m]> its online on some site
<aka_[m]> there is chance my dumb driver is off but its so dumb it shouldnt
<krzk> so first, the downstream mixers might not match 100% upstream code. Some mixer controls are different, so you cannot trust them blindly. Also downstream usually had poor audio-routing, so this needs investigating in your DTS.
<krzk> I suggest enabling all soundwire interrupts on wcd937x and (if you have) also on I2C speakers. You might start seeing then interrupts for some errors or "AUX PDM WD" - watchdog interrupt.
<krzk> aka_[m]: your earlier message about clocks suggest some driver issues - imbalance in clock wcd937x handling (DAPM, suspend/resume)
<krzk> if you have wcd937x headphone output (if I read correctly), that's already nice progress and gives some certainty that wcd937x and RX macro are configured.
<krzk> aka_[m]: Now if AW speakers are quiet, I assume you configured them over I2C and they do respond over I2C so they got out of shutdown/reset state, then: mixers, audio-routing or speaker I2C configuration.
<aka_[m]> krzk: i added them to audio-routings and when i switch to earpiece output i see it executes SND_SOC_DAPM_OUT_DRV_E action
<krzk> we can talk again tomorrow during working time, so aklimov_ will join. His setup is exactly the same, except AW->WSA8815
<aka_[m]> this awinic crap just takes "firmware" before enabling
<aka_[m]> which is your avg reg sequence
<aka_[m]> it disables, set fields and enable
f_ is now known as g_
g_ is now known as f_
jhovold has quit [Ping timeout: 480 seconds]