[SRU][S/U][PATCH 0/1] Support TI TAC5xx2 audio codec on Intel NVL platform
Timo Aaltonen
timo.aaltonen at canonical.com
Wed Sep 23 14:06:16 UTC 2026
Timo Aaltonen kirjoitti 23.9.2026 klo 17.04:
> Chris Chiu via kernel-team kirjoitti 18.9.2026 klo 13.25:
>> BugLink: https://bugs.launchpad.net/bugs/2167689
>>
>> [Impact]
>> Systems using Intel SoundWire codecs from the tac5xx2 family on Intel
>> NVL may fail to wake from suspend. The codec was not treated as a
>> wake-capable device. Symptom: system does not resume with normal wake
>> inputs and dmesg shows no SoundWire wake/resume messages.
>>
>> [Fix]
>> Add tac5xx2 family to the intel_auxdevice wake_capable_list so the
>> codec is recognized as wake-capable. Upstream commit: 1d4d3198ccce. No
>> released v* tag contains this commit (appears in the upstream tree).
>>
>> [Test Plan]
>> Suspend and wake 10 times:
>> $ sudo rtcwake -m mem -s 20
>> Use the usual wake input (power button, lid, keyboard).
>> Without patch: system fails to resume or needs manual power. With
>> patch: system resumes normally and dmesg shows SoundWire wake/resume
>> messages.
>>
>> [Where problems could occur]
>> If a non-wake-capable device is added, machines may see immediate
>> wakeups or fail to enter deep suspend. This would show as repeated
>> spurious wakes or inability to reach low-power suspend while running
>> the intel_auxdevice SoundWire driver.
>>
>> [Other Info]
>> Subsystem: drivers/soundwire (intel_auxdevice). Target: mainline and
>> backports for Ubuntu kernels as needed.
>>
>> Niranjan H Y (1):
>> soundwire: intel_auxdevice: add tac5xx2 family to wake_capable_list
>>
>> drivers/soundwire/intel_auxdevice.c | 2 ++
>> 1 file changed, 2 insertions(+)
>>
>
> Shouldn't this be SAUCE? Where is it from?
nevermind, it's in linux-next
More information about the kernel-team
mailing list