[SDW] timeouts with RT711 (non-SDCA) registers - jack init disable issue?

[plbossart]

Intel internal daily tests 1118 shows the following errors

kernel: [10520.802203] sof-audio-pci 0000:00:1f.3: ipc tx succeeded: 0x40020000: GLB_PM_MSG: CTX_RESTORE
kernel: [10520.802228] sof-audio-pci 0000:00:1f.3: ipc tx: 0x40010000: GLB_PM_MSG: CTX_SAVE
kernel: [10520.802364] sof-audio-pci 0000:00:1f.3: ipc tx succeeded: 0x40010000: GLB_PM_MSG: CTX_SAVE
kernel: [10520.802482] sof-audio-pci 0000:00:1f.3: FW Poll Status: reg=0x20140000 successful
kernel: [10521.358781] intel-sdw intel-sdw.0: IO transfer timed out, cmd 3 device 1 addr 3501 len 1
kernel: [10521.358796] soundwire sdw-master-0: trf on Slave 1 failed:-110 write addr 3501 count 0
kernel: [10521.862761] intel-sdw intel-sdw.0: IO transfer timed out, cmd 3 device 1 addr 3819 len 1
kernel: [10521.862777] soundwire sdw-master-0: trf on Slave 1 failed:-110 write addr 3819 count 0
kernel: [10522.366733] intel-sdw intel-sdw.0: IO transfer timed out, cmd 3 device 1 addr 3821 len 1
kernel: [10522.366747] soundwire sdw-master-0: trf on Slave 1 failed:-110 write addr 3821 count 0
kernel: [10522.870708] intel-sdw intel-sdw.0: IO transfer timed out, cmd 3 device 1 addr 3855 len 1
kernel: [10522.870723] soundwire sdw-master-0: trf on Slave 1 failed:-110 write addr 3855 count 0
kernel: [10522.870734] rt711 sdw:0:25d:711:0: in rt711_jack_init disable
kernel: [10523.374683] intel-sdw intel-sdw.0: IO transfer timed out, cmd 3 device 1 addr 3501 len 1
kernel: [10523.374698] soundwire sdw-master-0: trf on Slave 1 failed:-110 write addr 3501 count 0
kernel: [10523.375638] sof-audio-pci 0000:00:1f.3: FW Poll Status: reg=0x9010f0f successful
kernel: [10523.375816] sof-audio-pci 0000:00:1f.3: FW Poll Status: reg=0xf0f successful
kernel: [10523.375832] sof-audio-pci 0000:00:1f.3: DSP core(s) enabled? 0 : core_mask 1

Unclear if it's related to the calibration issues? @bardliao @shumingfan FYI

[shumingfan]

@plbossart Do you have the whole log?
Is it the disabled jack detection case?

[plbossart]

Here's the log. that's for a kmod load/unload test on TGL RVP with the topic/sof-dev branch
cmd: TPLG=sof-tgl-rt711-rt1308-4ch.tplg ~/sof-test/test-case/check-kmod-load-unload.sh -l 50

dmesg.txt

It doesn't contain the dynamic debug traces for rt711 so it's of limited use.
@bardliao @fredoh9 what does it take to add rt711 information to the sof-dyndbg.conf?

[shumingfan]

@plbossart I checked the dmesg.txt, but I'm not sure the root cause for now.
I suspect that the timing of the DSP core going to suspend mode and the jack init disable is too close.
Is it possible that the DSP core suspended/disabled already?

[shumingfan]

I feel this issue is the same as #2627.

[plbossart]

@shumingfan this happens with the regular RT711, not the SDCA version. I updated the description to make it clearer.

[bardliao]

This is fixed by #2640