Re: [BUG] Dell Inspiron 3501 (Tiger Lake) - No analog audio output, only HDMI detected

[Cezary Rojewski <cezary.rojewski@intel.com>]

On 2026-03-20 2:31 AM, Enrique Orozco wrote:
> Dear sound/ALSA maintainers,
> 
> I am reporting an issue with audio on a Dell Inspiron 3501 laptop. The
> system detects the Intel HDA controller but only HDMI outputs appear;
> the internal speakers and headphone jack are not detected at all.
> 
> System Information:
> - Laptop: Dell Inspiron 3501
> - Audio Controller: Intel Corporation Tiger Lake-LP Smart Sound Technology
>    Audio Controller [8086:a0c8] (rev 20)
> - Subsystem: Dell Device [1028:0a25]
> - Kernel: 6.18.12-1-MANJARO (also tested with 6.18.12)
> - Distribution: Manjaro (Arch-based)
> 
> Problem Description:
> After boot, only HDMI audio outputs are available. The internal speakers
> and headphone jack do not appear in ALSA or PulseAudio/PipeWire. The
> kernel detects the audio hardware but fails to initialize the analog
> codec.
> 
> Observations:
> 1. `aplay -l` shows only HDMI devices (devices 3,7,8,9)
> 2. No analog output device (device 0) appears
> 3. The codec detected is only "Intel Tigerlake HDMI" - no analog codec
> 4. The following error appears in dmesg:
>     `snd_soc_avs 0000:00:1f.3: core_mask 1 power off failed: -110`
> 5. Attempted various snd_hda_intel model parameters (dell-headset-multi,
>     dell-headset-dock, generic) with no success
> 6. Forcing legacy driver with `snd-intel-dspcfg dsp_driver=1` also
>     didn't help
> 
> Reproduction Steps:
> 1. Boot any recent Linux kernel (tested with 5.15.x and 6.18.x)
> 2. Check audio outputs with `aplay -l`
> 3. Observe only HDMI outputs are present
> 4. No sound from internal speakers or headphone jack
> 
> Expected Behavior:
> The system should detect and provide both analog output (internal
> speakers/headphone) and HDMI output.
> 
> Additional Information:
> - ALSA diagnostics:
> http://alsa-project.org/db/?f=8eb23206a6d8aa76d24136f224b1ad93dab53651


Hi Enrique,

Sorry to hear about the problems.  Your diagnostics seem to be devoid of 
dmesg, it would be vital to provide it.  You could also open a bugzilla 
and attach the logs there.

Now, this report seems a bit all over the place, mentioning all the 
sound drivers at once.  Let's start with context - their purpose:

- snd_hda_intel driver default for all general, non-DSP cases
- for TGL, snd_sof_xxx driver should probe by default for all DSP cases
- snd_soc_avs is utilized on designs older than TGL for all DSP cases

Your TGL of device ID=a0c8 points to PCI_DEVICE_ID_INTEL_HDA_TGL_LP 
(include/linux/pci_ids.h).  This means snd_sof_xxx should have probed 
first.  Frankly, there should be no attempts for snd_soc_avs unless a 
user forced the enumeration.  Such enumeration is destined to fail - not 
because of the driver, but because there is no firmware binary paired 
with it available on linux-firmware..

..unless one copies a binary from a Windows device.

Last but not least:
	`snd_soc_avs 0000:00:1f.3: core_mask 1 power off failed: -110`

While without the dmesg I cannot say for certain, failing such "basic" 
operation as power-off MASTER_CORE (mask 1) usually means that the 
laptop, despite being -LP, has AudioDSP functionality disabled in BIOS. 
Forced enumeration won't workaround that either.


Kind regards,
Czarek