Re: [PATCH 0/2] ALSA: sc6000: fix board configuration state and resume handling

From: Takashi Iwai

Date: Fri Apr 10 2026 - 02:37:49 EST


On Fri, 10 Apr 2026 05:54:31 +0200,
Cássio Gabriel wrote:
>
> The SC-6000 driver has to program board-level DSP routing before the
> WSS codec can operate correctly. Today that setup has two related
> problems.
>
> First, the probe path may auto-select IRQ and DMA resources, but the
> SC-6000 soft configuration is still derived from the raw module
> parameter arrays. When irq=auto or dma=auto is used, the WSS codec can
> be created with the selected live resources while the board itself is
> programmed with unresolved values.
>
> Second, the driver still lacks suspend/resume support. The WSS layer can
> save and restore codec registers, but the SC-6000-specific DSP routing
> and MSS initialization are done only during probe and are not replayed
> after resume.
>
> This series fixes the probe-time resource mismatch first and then reuses
> that cached board state to restore the hardware on resume. Keeping the
> board programming in a shared helper also keeps the old/new DSP split in
> one place so probe and resume cannot drift apart.
>
> - Patch 1 stores the computed SC-6000 board state in card-private data and
> builds it from the live resources selected at probe time.
> - Patch 2 wires ISA suspend/resume callbacks and restores the cached
> board setup before the WSS codec resume path runs.
>
> Signed-off-by: Cássio Gabriel <cassiogabrielcontato@xxxxxxxxx>
> ---
> Cássio Gabriel (2):
> ALSA: sc6000: Keep the programmed board state in card-private data
> ALSA: sc6000: Restore board setup across suspend

Applied both patches now. Thanks.


Takashi