On Sat, Mar 29, 2025 at 08:14:33PM +0100, Martin Blumenstingl wrote:
> Commit 2fef64eec23a ("ASoC: hdmi-codec: Add a prepare hook") added a
> prepare implementation. Back then the new callback was only integrated
> with hdmi_codec_i2s_dai_ops (which is used by the I2S input code-path).
> It was not added to hdmi_codec_spdif_dai_ops (which is used by the SPDIF
> input code-path).
>
> With commit baf616647fe6 ("drm/connector: implement generic HDMI audio
> helpers") the DRM subsystem has gained a helper framework which can be
> used by HDMI controller drivers. HDMI controller drivers are often
> tightly coupled with the hdmi-codec because of the so-called HDMI audio
> infoframe (which is often managed by the display controller).
>
> To allow the new DRM HDMI audio framework to work with the hdmi-codec
> driver for SPDIF inputs we also need to hook up the prepare callback to
> hdmi_codec_spdif_dai_ops. Just hooking into the hw_params callback would
> not be enough as hw_params (is called too early and) doesn't have access
> to the HDMI audio infoframe contents.
>
> Suggested-by: Dmitry Baryshkov <dmitry.barysh...@linaro.org>
> Signed-off-by: Martin Blumenstingl <martin.blumensti...@googlemail.com>
> ---
> Changes since v1 at [0]:
> - re-sending since there was no feedback
>
>
> The following three upstream drivers can use the hdmi-codec with SPDIF
> inputs:
> - drivers/gpu/drm/bridge/adv7511/adv7511_audio.c
> - drivers/gpu/drm/rockchip/cdn-dp-core.c
> - drivers/gpu/drm/tegra/hdmi.c
>
> It would be great if any of the maintainers of these platforms (Cc'ed)
> could confirm that this patch doesn't break anything.
>
> Mark, Dmitry: who of you would take this patch (Mark via the ASoC
> tree or Dmitry via drm-misc)?
There is no reason for it to go through drm-misc, we picked up only
several API changes in order to ease merging of the patchset. This patch
should go normally via Mark's tree.
--
With best wishes
Dmitry
