Thank you for your contribution! Sashiko AI review found 6 potential issue(s) 
to consider:

New issues:
- [High] The return value of `pm_runtime_get_sync()` is ignored in 
`dw_dp_audio_startup()`, violating PM runtime API contracts and risking 
hardware access while suspended.
- [High] Missing error checks on `clk_prepare_enable()` in 
`dw_dp_audio_prepare()` lead to hardware access failures and clock reference 
count underflows.
- [High] SDP register bank slots are permanently leaked because 
`dw_dp_audio_unprepare()` does not free them.
- [Medium] Prematurely updating `dp->audio_interface` before hardware 
validation causes an unbalanced clock disable in the teardown path on error.
- [Medium] `dw_dp_audio_prepare()` explicitly rejects valid intermediate 
channel counts, breaking multi-channel playback.

Pre-existing issues:
- [High] Concurrent accesses to the `dp->sdp_reg_bank` bitmap between DRM and 
ALSA threads lack locking, leading to SDP register corruption.
--

commit 91c310c87d4ddad349da89d89896e7fa9319465d
Author: Sebastian Reichel <[email protected]>

drm/bridge: synopsys: dw-dp: Add audio support

Implement audio support for the Synopsys DesignWare DisplayPort
controller.

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-dp.c 
b/drivers/gpu/drm/bridge/synopsys/dw-dp.c
index 7f4f36c614844..f1946f2c945da 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-dp.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-dp.c

[ ... ]

> +static int dw_dp_audio_startup(struct drm_bridge *bridge,
> +                            struct drm_connector *connector)
> +{
> +     struct dw_dp *dp = bridge_to_dp(bridge);
> +
> +     dev_dbg(dp->dev, "audio startup\n");
> +     pm_runtime_get_sync(dp->dev);

[Severity: High]
Does this code need to check the return value of pm_runtime_get_sync()?

If the power domain or clock fails to resume, it returns a negative error
code. Ignoring this might lead to accessing suspended hardware registers when
ALSA triggers startup.

> +
> +     return 0;
> +}
> +
> +static void dw_dp_audio_unprepare(struct drm_bridge *bridge,
> +                               struct drm_connector *connector)
> +{
> +     struct dw_dp *dp = bridge_to_dp(bridge);
> +
> +     /* Disable all audio streams */
> +     regmap_update_bits(dp->regmap, DW_DP_AUD_CONFIG1, AUDIO_DATA_IN_EN,
> +                        FIELD_PREP(AUDIO_DATA_IN_EN, 0));
> +
> +     if (dp->audio_interface == DW_DP_AUDIO_SPDIF)
> +             clk_disable_unprepare(dp->spdif_clk);
> +     else if (dp->audio_interface == DW_DP_AUDIO_I2S)
> +             clk_disable_unprepare(dp->i2s_clk);
> +
> +     dp->audio_interface = DW_DP_AUDIO_UNUSED;
> +}

[Severity: High]
Does dw_dp_audio_unprepare() leak SDP register bank slots?

dw_dp_audio_prepare() allocates a new SDP slot by calling dw_dp_send_sdp(),
which sets a bit in dp->sdp_reg_bank. However, dw_dp_audio_unprepare() never
clears the allocated bit.

Could this exhaust the register bank slots if userspace starts and stops audio
multiple times?

[ ... ]

> +static int dw_dp_audio_prepare(struct drm_bridge *bridge,
> +                            struct drm_connector *connector,
> +                            struct hdmi_codec_daifmt *daifmt,
> +                            struct hdmi_codec_params *params)
> +{
> +     struct dw_dp *dp = bridge_to_dp(bridge);
> +     u8 audio_data_in_en, supported_audio_interfaces;
> +     u32 cfg1;
> +     int ret;
> +
> +     /*
> +      * prepare might be called multiple times, so release the clocks
> +      * from previous calls to keep the calls in balance.
> +      */
> +     if (dp->audio_interface != DW_DP_AUDIO_UNUSED)
> +             dw_dp_audio_unprepare(bridge, connector);
> +
> +     dp->audio_channels = params->cea.channels;
> +     switch (params->cea.channels) {
> +     case 1:
> +     case 2:
> +             audio_data_in_en = AUDIO_DATA_IN_EN_CHANNEL12;
> +             break;
> +     case 8:
> +             audio_data_in_en = AUDIO_DATA_IN_EN_CHANNEL12 |
> +                                AUDIO_DATA_IN_EN_CHANNEL34 |
> +                                AUDIO_DATA_IN_EN_CHANNEL56 |
> +                                AUDIO_DATA_IN_EN_CHANNEL78;
> +             break;
> +     default:
> +             dev_err(dp->dev, "invalid audio channels %d\n", 
> dp->audio_channels);
> +             return -EINVAL;
> +     }

[Severity: Medium]
Will this explicit rejection break valid multi-channel playback?

The driver sets bridge->hdmi_audio_max_i2s_playback_channels = 8 indicating
hardware support, but this switch statement returns -EINVAL for intermediate
channel counts like 4 or 6. Is it intended to reject valid multi-channel
audio setups?

> +
> +     switch (daifmt->fmt) {
> +     case HDMI_SPDIF:
> +             dp->audio_interface = DW_DP_AUDIO_SPDIF;
> +             break;
> +     case HDMI_I2S:
> +             /*
> +              * It is recommended to use SPDIF instead of I2S, since I2S 
> mode requires
> +              * manually inserting PCUV control bits from userspace and this 
> is done
> +              * automatically in hardware for SPDIF mode.
> +              */
> +             dp->audio_interface = DW_DP_AUDIO_I2S;
> +             break;
> +     default:
> +             dev_err(dp->dev, "invalid DAI format %d\n", daifmt->fmt);
> +             return -EINVAL;
> +     }
> +
> +     regmap_read(dp->regmap, DW_DP_CONFIG_REG1, &cfg1);
> +     supported_audio_interfaces = FIELD_GET(AUDIO_SELECT, cfg1);
> +
> +     if (supported_audio_interfaces != DW_DP_AUDIO_I2S_AND_SPDIF &&
> +         supported_audio_interfaces != dp->audio_interface) {
> +             dev_err(dp->dev, "unsupported DAI %d\n", daifmt->fmt);
> +             return -EINVAL;
> +     }

[Severity: Medium]
Could this leave dp->audio_interface in a corrupted state on failure?

If an unsupported DAI format is provided, dp->audio_interface is updated prior
to the validation check. If this function then returns -EINVAL, the ALSA core
later calls dw_dp_audio_shutdown() -> dw_dp_audio_unprepare(), which acts on
the corrupted state and attempts to disable a clock that was never enabled.

> +
> +     clk_prepare_enable(dp->spdif_clk);
> +     clk_prepare_enable(dp->i2s_clk);

[Severity: High]
Does this code need to check the return values of clk_prepare_enable()?

If clk_prepare_enable() fails, the driver continues to program hardware
registers, which could lead to hardware lockups. Furthermore, the teardown
path unconditionally disables the failed clock, causing reference count
underflows.

[ ... ]

> +     ret = dw_dp_audio_infoframe_send(dp);
> +     if (ret < 0)
> +             dev_err(dp->dev, "failed to send audio infoframe\n");

[Severity: High]
This isn't a bug introduced by this patch, but adding audio support introduces
a concurrent access path that makes a pre-existing race possible.

dw_dp_audio_infoframe_send() calls dw_dp_send_sdp(), which allocates a slot
using non-atomic test and set:

        nr = find_first_zero_bit(dp->sdp_reg_bank, SDP_REG_BANK_SIZE);
        if (nr < SDP_REG_BANK_SIZE)
                set_bit(nr, dp->sdp_reg_bank);

Can ALSA audio preparation running concurrently with DRM display updates lead
to slot allocation collisions and SDP hardware register corruption?

-- 
Sashiko AI review ยท 
https://sashiko.dev/#/patchset/20260612-synopsys-dw-dp-improvements-v3-0-dc61e6352...@collabora.com?part=10

Reply via email to