Re: [PATCH v3 11/33] drm/mediatek: mtk_hdmi: Unregister audio platform device on failure

[AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>]

Il 23/12/24 06:22, CK Hu (胡俊光) ha scritto:
> Hi, Angelo:
> 
> On Tue, 2024-12-17 at 16:43 +0100, AngeloGioacchino Del Regno wrote:
>> External email : Please do not click links or open attachments until you have verified the sender or the content.
>>
>>
>> The probe function of this driver may fail after registering the
>> audio platform device: in that case, the state is not getting
>> cleaned up, leaving this device registered.
>>
>> Adding up to the mix, should the probe function of this driver
>> return a probe deferral for N times, we're registering up to N
>> audio platform devices and, again, never freeing them up.
>>
>> To fix this, add a pointer to the audio platform device in the
>> mtk_hdmi structure, and add a devm action to unregister it upon
>> driver removal or probe failure.
>>
>> Fixes: 8f83f26891e1 ("drm/mediatek: Add HDMI support")
>> Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
>> ---
>>   drivers/gpu/drm/mediatek/mtk_hdmi.c | 27 +++++++++++++++++++++------
>>   1 file changed, 21 insertions(+), 6 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/mediatek/mtk_hdmi.c b/drivers/gpu/drm/mediatek/mtk_hdmi.c
>> index 59a8a81fa0ec..9131c2fdfba8 100644
>> --- a/drivers/gpu/drm/mediatek/mtk_hdmi.c
>> +++ b/drivers/gpu/drm/mediatek/mtk_hdmi.c
>> @@ -172,6 +172,7 @@ struct mtk_hdmi {
>>          unsigned int sys_offset;
>>          void __iomem *regs;
>>          enum hdmi_colorspace csp;
>> +       struct platform_device *audio_pdev;
>>          struct hdmi_audio_param aud_param;
>>          bool audio_enable;
>>          bool powered;
>> @@ -1662,6 +1663,11 @@ static const struct hdmi_codec_ops mtk_hdmi_audio_codec_ops = {
>>          .no_capture_mute = 1,
>>   };
>>
>> +static void mtk_hdmi_unregister_audio_driver(void *data)
>> +{
>> +       platform_device_unregister(data);
>> +}
>> +
>>   static int mtk_hdmi_register_audio_driver(struct device *dev)
>>   {
>>          struct mtk_hdmi *hdmi = dev_get_drvdata(dev);
>> @@ -1671,13 +1677,22 @@ static int mtk_hdmi_register_audio_driver(struct device *dev)
>>                  .i2s = 1,
>>                  .data = hdmi,
>>          };
>> -       struct platform_device *pdev;
>> +       int ret;
>>
>> -       pdev = platform_device_register_data(dev, HDMI_CODEC_DRV_NAME,
>> -                                            PLATFORM_DEVID_AUTO, &codec_data,
>> -                                            sizeof(codec_data));
>> -       if (IS_ERR(pdev))
>> -               return PTR_ERR(pdev);
>> +       hdmi->audio_pdev = platform_device_register_data(dev,
>> +                                                        HDMI_CODEC_DRV_NAME,
>> +                                                        PLATFORM_DEVID_AUTO,
>> +                                                        &codec_data,
>> +                                                        sizeof(codec_data));
>> +       if (IS_ERR(hdmi->audio_pdev))
>> +               return PTR_ERR(hdmi->audio_pdev);
>> +
>> +       ret = devm_add_action_or_reset(dev, mtk_hdmi_unregister_audio_driver,
>> +                                      hdmi->audio_pdev);
>> +       if (ret) {
>> +               platform_device_unregister(hdmi->audio_pdev);
> 
> When devm_add_action_or_reset() fail, it would call mtk_hdmi_unregister_audio_driver() inside it.
> So it's not necessary to call platform_device_unregister() here.
> 

Oh, right. Yeah, that was quite an overlook from me.

I'll fix that in v4.

Thanks,
Angelo

> Regards,
> CK
> 
>> +               return ret;
>> +       }
>>
>>          DRM_INFO("%s driver bound to HDMI\n", HDMI_CODEC_DRV_NAME);
>>          return 0;
>> --
>> 2.47.0
>>
>