Re: [PATCH] ALSA: hda/realtek: Add quirk for IPASON SmartBook S1

[Takashi Iwai <tiwai@suse.de>]

On Wed, 16 Sep 2026 05:06:00 +0200,
周建宇 wrote:
> 
> From 09aa2aa6f566f10eb303dfd4731e1776d0a1c96e Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?=E5=91=A8=E5=BB=BA=E5=AE=87?= <997536907@qq.com>
> Date: Wed, 16 Sep 2026 10:30:17 +0800
> Subject: [PATCH] ALSA: hda/realtek: Add quirk for IPASON SmartBook S1
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
> 
> The internal speaker of the IPASON SmartBook S1 (an Emdoor ODM laptop)
> stays silent under Linux, while the Bluetooth audio and the headset
> output work fine.  The same machine plays sound under Windows with the
> vendor driver, so the hardware and the amplifier are known good.
> 
> The BIOS reports the pin default of the actual internal speaker pin
> (NID 0x1b) as 0x411111f0, i.e. the default association 0xf, which means
> "unconnected".  The pin parser therefore drops that node already at the
> pin-configuration time, hence no mixer control is ever created for it
> and the pin is never driven, no matter how the output pin 0x14 is set
> up.  Enabling 0x1b makes the speaker work immediately.
> 
> The existing ALC256_FIXUP_HONOR_MRB_XXX_M1020_AUDIO quirk makes the
> speaker work, too, but it additionally rewrites the pins 0x14, 0x19 and
> 0x1a with Honor-specific values, which creates a bogus "Dock Mic" device
> that does not exist on this hardware.  Add a minimal fixup that only
> enables 0x1b as the internal speaker.  Touching 0x14 is not needed:
> with the pin configuration overridden at runtime, both keeping its BIOS
> value (0x90170110) and disabling it (0x411111f0) leave the speaker
> working, so 0x1b alone decides the output.
> 
>   Codec:       Realtek ALC256 (0x10ec0256), SSID 0x2782:0x0206
>   BIOS pincfg: 0x14 0x90170110  0x18 0x411111f0  0x19 0x02a19040
>                0x1a 0x411111f0  0x1b 0x411111f0  0x1d 0x4067b945
>                0x1e 0x411111f0  0x21 0x02214020
> 
> Tested on 7.2.0-1-MANJARO and 6.18.45-1-MANJARO, with both the SOF and
> the legacy HDA driver paths.
> 
> Signed-off-by: 周建宇 <997536907@qq.com>

The code change looks mostly OK, but could you try to use western
alphabets for your sign-off and from tags?  That makes easier to
handle for many other people.

Also, you don't have to add always the model name string; nowadays a
quick can be applied via the ID alias (xxx:yyy form).
So drop the addition to alc269_fixup_models[] unless you believe this
must be added inevitably.


thanks,

Takashi