Huawei MateBook D15 (BOHB-WAX9-PCB-B2): ES8336 speakers require SPEAKERS_EN_GPIO1 quirk

Huawei MateBook D15 (BOHB-WAX9-PCB-B2): ES8336 speakers require SPEAKERS_EN_GPIO1 quirk

[Thilo Schuetz, Berlin]

Dear team, 

I would like to report what appears to be an incorrect quirk assignment
for a Huawei MateBook D15 with an ES8336 codec. 

Please note: The machine here is my mother's notebook and I encountered
the error while switching her from Windows to Debian. Research and the
fix were done with support of ChatGPT, so please treat this accordingly,
even though I've tried to be careful and complete myself. (I'm only a
user, after all.) 

System: 

 	* Laptop: Huawei MateBook D15
 	* DMI board name: BOHB-WAX9-PCB-B2
 	* Distribution: Debian 13
 	* Kernel: 6.12.94-amd64
 	* Audio driver: snd_soc_sof_es8336

Problem:
With the default driver configuration, the internal speakers produce no
sound. Bluetooth audio works normally. 

The driver reports:

sof-essx8336: quirk mask 0x1a0

From the current upstream source, this DMI entry selects:

SOF_ES8336_HEADPHONE_GPIO |
SOC_ES8336_HEADSET_MIC1

Investigation: 

 	* The DSDT contains an ESSX8336 device with two GPIO outputs.
 	* No HWSP0001 ACPI device is present.

	* An I²C scan shows no devices at addresses 0x58 or 0x5b, so the
HWSP0001-based amplifier workaround is not applicable to this hardware.
(https://patchew.org/linux/20260614161028.10405-1-tgaraev653%40gmail.com)
	* PipeWire, SOF firmware and the ES8336 codec were detected and loaded
correctly.

Workaround:
Adding the following modprobe option:

options snd_soc_sof_es8336 quirk=0x1b0

adds the SPEAKERS_EN_GPIO1 quirk. 

After reboot, dmesg reports:

sof-essx8336: Overriding quirk 0x1a0 => 0x1b0
sof-essx8336: quirk mask 0x1b0
sof-essx8336: quirk SSP0
sof-essx8336: quirk DMIC enabled
sof-essx8336: Speakers GPIO1 quirk enabled
sof-essx8336: quirk headphone GPIO enabled
sof-essx8336: quirk headset at mic1 port enabled

With this override, the internal speakers work immediately after boot
and normal audio playback (Firefox/YouTube) functions correctly. 

Limitations:
Unfortunately I did not have wired headphones available, so I could not
verify headphone jack behavior after enabling the additional quirk. 

Because this board already has an upstream DMI entry, I am not
suggesting a specific patch. However, the results above suggest that at
least some BOHB-WAX9-PCB-B2 systems require the SPEAKERS_EN_GPIO1 quirk
for the internal speakers to function. 

If additional logs or tests would be helpful, I would be happy to
provide them. This might take a little, as I will need to visit my mom
for that then. 

Thank you for your work on the driver. 

Kind regards,
Thilo

-- 
Dipl.Ing. Thilo Schuetz
Georg-Wilhelm-Strasse 20
10711 Berlin
GERMANY

thilo.schuetz@berlin.de
+49 178 3060200  (Mobile Germany)
+63 929 7155840  (Mobile Philippines)