Re: [PATCH] ALSA: hda/realtek: Fix speaker mute LED on Lenovo ThinkBook 14 G8+ IPH

[Takashi Iwai <tiwai@suse.de>]

On Thu, 30 Jul 2026 09:14:53 +0200,
Padhia Luo wrote:
> 
> On the ThinkBook 14 G8+ IPH (SSID 17aa:393e, ALC287) the F1 speaker mute
> LED never lights up, while the F4 mic mute LED works.
> 
> Both LEDs are platform LEDs registered by lenovo-wmi-hotkey-utilities and
> default to the audio-mute / audio-micmute triggers, so the speaker LED only
> follows a control carrying SNDRV_CTL_ELEM_ACCESS_SPK_LED. No control on
> this machine has that flag set, so snd_ctl_led never attaches anything and
> /sys/class/sound/ctl-led/speaker/card0/list stays empty. The mic LED is
> unaffected because MIC_LED is set from the SOF topology on the DMIC
> control, which does not go through the codec fixups at all.
> 
> The pin configuration of this machine matches the ThinkPad pin quirk that
> selects ALC285_FIXUP_THINKPAD_HEADSET_JACK - pin_config_match() masks out
> the sequence/association nibbles, so 0x14=0x90170120 still matches the
> 0x90170110 in the table. That fixup chains into ALC269_FIXUP_THINKPAD_ACPI,
> but hda_fixup_thinkpad_acpi() returns early because is_thinkpad() is false:
> a ThinkBook exposes neither LEN0068/LEN0268 nor IBM0068. Therefore
> snd_hda_gen_add_mute_led_cdev() is never called and spec->vmaster_mute_led
> stays 0.
> 
> The vendor fallback SND_PCI_QUIRK_VENDOR(0x17aa, "Lenovo XPAD",
> ALC269_FIXUP_LENOVO_XPAD_ACPI) would have handled this correctly - the
> machine does expose LHK2019 and VPC2004, so is_ideapad() is true - but it
> never runs: the pin quirk has already set codec->fixup_id, and
> snd_hda_pick_fixup() returns immediately in that case.
> 
> Add an SSID quirk selecting a fixup that keeps everything the machine
> currently gets (headset jack handling plus the X1 Gen7 DAC routing) and
> additionally runs the ideapad ACPI setup. It chains into
> ALC287_FIXUP_LENOVO_YOGA_PRO7, which already combines
> alc285_fixup_thinkpad_x1_gen7 with ALC269_FIXUP_LENOVO_XPAD_ACPI, so the
> resulting chain differs from the current one only by the added ideapad step
> and cannot regress the analog output or the headset jack.
> 
> Tested on 7.1.5 on the affected machine: the speaker LED group is now
> populated at probe time without any userspace help, and the F1 LED
> follows the mute state.  Compared against a boot with the previous fixup
> selection, the mixer control list (names and numids) and the registered
> jack input devices are identical.
> 
> Note that the underlying mismatch is not specific to this SSID. Any Lenovo
> non-ThinkPad whose pins collide with a ThinkPad pin quirk loses its mute
> LED the same way. Letting hda_fixup_thinkpad_acpi() fall back to the
> ideapad check would cover the whole class at once, but that touches a
> helper shared with every ThinkPad, so this patch only fixes the machine
> that was actually tested.
> 
> Signed-off-by: Padhia Luo <lcj20010426@gmail.com>

Applied to for-next branch.  Thanks.


Takashi