Skip to content

ASoC: rt722-sdca: Enable speaker playback on Acer Swift Pro SFP14-55 - #6052

Open
shripad-jyothinath wants to merge 1 commit into
thesofproject:topic/sof-devfrom
shripad-jyothinath:rt722-acer-sfp14-55-speaker-fix
Open

shripad-jyothinath wants to merge 1 commit into
thesofproject:topic/sof-devfrom
shripad-jyothinath:rt722-acer-sfp14-55-speaker-fix

Conversation

@shripad-jyothinath

Copy link
Copy Markdown

This patch enables the internal stereo speakers on the Acer Swift Pro SFP14-55 (Meteor Lake-P with Realtek RT722 SDCA SoundWire codec).

Problem

On the Acer Swift Pro SFP14-55, audio output through the internal speakers produces no sound at all under Linux, even though headphones and DMIC work fine.

Root Cause

While debugging on the actual laptop hardware and comparing ACPI SSDT tables against the SoundWire register dumps, I found a few key issues preventing the Class-D amplifier from producing sound:

  1. ACPI Entity mismatch: The ACPI SSDT tables declare the Class-D SmartAmp power domain under Entity 0x41 (Device (E041)), but RT722_SDCA_ENT_PDE23 in rt722-sdca.h was hardcoded to 0x23.
  2. PDE float control: Hardware register 0x61:0x29 (RT722_AMP_PDE_FLOAT_CTL) defaults to 0x2323, which leaves the amplifier power domain disconnected from Entity 0x41. Setting this to 0x4141 connects it to the power domain declared in ACPI.
  3. Bridge control bit: Register 0x20:0x07 (RT722_MISC_CTRL1) bit 15 (0x8000) was unasserted by default. Without setting bit 15, digital audio stream data from the SoundWire bus is gated and does not reach the Class-D stage.
  4. Class-D tuning: The ACPI table specifies tuning registers for the amp (0x20:0x38 = 0x890a, 0x20:0x37 = 0xfc15, 0x20:0x40 = 0x8800).
  5. Power sequencing and EAPD: The amplifier PLL and LDOs in 0x5F:0x0B need to be enabled (0x9800) along with raising EAPD (0x20:0x10 = 0x0360) during SND_SOC_DAPM_POST_PMU. An 80ms settling delay avoids DC pop. On stream stop (PRE_PMD), EAPD is lowered back to 0x0260 and the amp power is cut.

Changes

  • Added a DMI quirk match for Acer Swift Pro SFP14-55 (RT722_QUIRK_ACER_SFP14_55).
  • Configured PDE float control, bridge routing, and ACPI tuning in rt722_sdca_init().
  • Implemented amplifier power and EAPD sequence in DAPM widget event rt722_sdca_spk_event().
  • Corrected RT722_SDCA_ENT_PDE23 define to 0x41.

Verification

  • Verified on Acer Swift Pro SFP14-55 running 7.4.0-rc1-sof-test+.
  • Stream playback tested with ALSA and PipeWire/WirePlumber desktop audio (pw-play).
  • Both acoustic loopback recordings (via onboard DMIC) and listening tests confirm clean, pop-free stereo sound from the internal speakers.

The internal speakers on the Acer Swift Pro SFP14-55 (Meteor Lake-P
with Realtek RT722 SDCA SoundWire codec) produce no sound on Linux.
Investigation revealed several factors preventing audio output:

1. In the ACPI SSDT table, the Class-D SmartAmp power domain is
declared under Entity 0x41 (Device E041), but the upstream rt722
header hardcodes RT722_SDCA_ENT_PDE23 as 0x23. Update the definition
to 0x41.

2. Hardware register 0x61:0x29 (RT722_AMP_PDE_FLOAT_CTL) controls which
entity the Class-D power domain is mapped to. The default is 0x2323.
On Acer SFP14-55, set this to 0x4141 to map it to Entity 0x41.

3. Hardware register 0x20:0x07 (RT722_MISC_CTRL1) bit 15 (0x8000)
must be set to route digital stream audio to the Class-D stage. When
this bit is clear, audio from the bus is blocked at the bridge.

4. Apply the ACPI AF04 FIT Class-D tuning values (0x20:0x38 = 0x890a,
0x20:0x37 = 0xfc15, 0x20:0x40 = 0x8800).

5. On stream start (POST_PMU), power up the amplifier PLL and LDOs
(0x5F:0x0B = 0x9800) and raise EAPD (0x20:0x10 = 0x0360) with an 80ms
settling delay to avoid pops. On stream stop (PRE_PMD), lower EAPD to
0x0260 and cut amplifier power back to 0x0000.

Add a DMI quirk table matching the Acer Swift Pro SFP14-55 to apply
these settings cleanly without affecting other platforms.

Signed-off-by: Shripad J <shripad-jyothinath@users.noreply.github.com>
Copilot AI balanced review requested due to automatic review settings November 22, 2026 07:19

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing.

@sofci

sofci commented Nov 22, 2026

Copy link
Copy Markdown
Collaborator

Can one of the admins verify this patch?

reply test this please to run this test once

@shripad-jyothinath
shripad-jyothinath requested a balanced review from Copilot November 22, 2026 10:24

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.

An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants