[PATCH v3, 04/15] media: mtk-vcodec: Use component framework to manage each hardware information

[Tzung-Bi Shih]

On Tue, Jul 27, 2021 at 06:10:40PM +0800, Yunfei Dong wrote:
> +static struct component_match *mtk_vcodec_match_add(
> +     struct mtk_vcodec_dev *vdec_dev)
> + {
Remove the extra space before {.

> +     struct platform_device *pdev = vdec_dev->plat_dev;
> +     struct component_match *match = NULL;
> +     int i;
> +
> +     for (i = 0; i < ARRAY_SIZE(mtk_vdec_drv_ids); i++) {
> +             struct device_node *comp_node;
> +             enum mtk_vdec_hw_id comp_idx;
> +             const struct of_device_id *of_id;
Failed to see benifits to declare here in the case.  To be neat, move
the variable declaration to the beginning of function.

> +
> +             comp_node = of_find_compatible_node(NULL, NULL,
> +                     mtk_vdec_drv_ids[i].compatible);
> +             if (!comp_node)
> +                     continue;
If moving of_id assignment to the beginning of for-loop, use of_id->compatible.

> +
> +             if (!of_device_is_available(comp_node)) {
> +                     of_node_put(comp_node);
> +                     dev_err(&pdev->dev, "Fail to get MMSYS node\n");
> +                     continue;
> +             }
> +
> +             of_id = &mtk_vdec_drv_ids[i];
> +             if (!of_id) {
> +                     dev_err(&pdev->dev, "Failed to get match node\n");
> +                     return ERR_PTR(-EINVAL);
> +             }
Move the block to the beginning of for-loop.  Looking to other blocks
around, does it make more sense to use `continue` instead of `return`?

> +
> +             comp_idx = (enum mtk_vdec_hw_id)of_id->data;
> +             mtk_v4l2_debug(4, "Get component:hw_id(%d),vdec_dev(0x%p),comp_node(0x%p)\n",
> +                     comp_idx, vdec_dev, comp_node);
> +             vdec_dev->component_node[comp_idx] = comp_node;
> +
> +             component_match_add_release(&pdev->dev, &match, mtk_vdec_release_of,
> +                     mtk_vdec_compare_of, comp_node);
> +             of_node_put(comp_node);
I thought it shouldn't call of_node_put(...) which is already
delegated to component framework (mtk_vdec_release_of).

> +     }
> +
> +   return match;
Fix the indent.

> +static int mtk_vcodec_init_dec_params(struct mtk_vcodec_dev *dev)
> +{
> +     struct platform_device *pdev = dev->plat_dev;
> +     int ret;
> +
> +     ret = mtk_vcodec_get_reg_bases(dev);
> +     if (ret)
> +             return ret;
> +
> +     if (!dev->is_comp_supported) {
> +             dev->dec_irq = platform_get_irq(pdev, 0);
> +             if (dev->dec_irq < 0) {
> +                     dev_err(&pdev->dev, "failed to get irq number");
> +                     return dev->dec_irq;
> +             }
> +
> +             irq_set_status_flags(dev->dec_irq, IRQ_NOAUTOEN);
> +             ret = devm_request_irq(&pdev->dev, dev->dec_irq,
> +                             mtk_vcodec_dec_irq_handler, 0, pdev->name, dev);
> +             if (ret) {
> +                     dev_err(&pdev->dev, "failed to install dev->dec_irq %d (%d)",
> +                             dev->dec_irq, ret);
> +                     return ret;
> +             }
> +
> +             ret = mtk_vcodec_init_dec_pm(pdev, &dev->pm);
> +             if (ret < 0) {
> +                     dev_err(&pdev->dev, "failed to get mt vcodec clock source");
> +                     return ret;
> +             }
> +     }
> +
> +     return ret;
To be concise, return 0.

> +     comp_node =
> +             of_find_compatible_node(NULL, NULL, "mediatek,mtk-vcodec-core");
> +     if (!comp_node)
> +             dev->is_comp_supported = false;
> +     else
> +             dev->is_comp_supported = true;
> +     of_node_put(comp_node);
Looking up "mediatek,mtk-vcodec-core" to determine if it uses
component framework sounds like a sub-optimal method.

> @@ -311,7 +429,6 @@ static int mtk_vcodec_probe(struct platform_device *pdev)
>               MTK_VCODEC_DEC_NAME);
>       video_set_drvdata(vfd_dec, dev);
>       dev->vfd_dec = vfd_dec;
> -     platform_set_drvdata(pdev, dev);
>
>       dev->m2m_dev_dec = v4l2_m2m_init(&mtk_vdec_m2m_ops);
>       if (IS_ERR((__force void *)dev->m2m_dev_dec)) {
> @@ -362,8 +479,17 @@ static int mtk_vcodec_probe(struct platform_device *pdev)
>       mtk_v4l2_debug(0, "decoder registered as /dev/video%d",
>               vfd_dec->num);
>
> -     return 0;
> +     if (dev->is_comp_supported) {
> +             ret = mtk_vcodec_init_master(dev);
> +             if (ret < 0)
> +                     goto err_component_match;
> +     } else {
> +             platform_set_drvdata(pdev, dev);
> +     }
Has asked the same question in [1].  Why it removes the
platform_set_drvdata() above?  mtk_vcodec_init_master() also calls
platform_set_drvdata().

[1]: https://patchwork.linuxtv.org/project/linux-media/patch/20210707062157.21176-4-yunfei.dong@mediatek.com/

> +     return 0;
> +err_component_match:
> +     video_unregister_device(vfd_dec);
Why video_unregister_device()?  It is already called [2].

[2]: https://elixir.bootlin.com/linux/v5.14-rc3/source/drivers/media/platform/mtk-vcodec/mtk_vcodec_dec_drv.c#L344

> @@ -379,9 +505,8 @@ static int mtk_vcodec_probe(struct platform_device *pdev)
>  err_dec_alloc:
>       v4l2_device_unregister(&dev->v4l2_dev);
>  err_res:
> -     mtk_vcodec_release_dec_pm(&dev->pm);
Shouldn't remove it.  mtk_vcodec_init_dec_params() also needs to undo
in the path.

Refactoring them to mtk_vcodec_init_dec_params() makes the error
handling more complicated.

> diff --git a/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h b/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h
> index 973b0b3649c6..d6bb723db106 100644
> --- a/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h
> +++ b/drivers/media/platform/mtk-vcodec/mtk_vcodec_drv.h
> @@ -17,6 +17,11 @@
>  #include <media/videobuf2-core.h>
>  #include "mtk_vcodec_util.h"
>
> +#define VDEC_HW_ACTIVE       0x10
> +#define VDEC_IRQ_CFG 0x11
> +#define VDEC_IRQ_CLR 0x10
> +#define VDEC_IRQ_CFG_REG     0xa4
If moving to mtk_vcodec_dec_hw.h or mtk_vcodec_dec_hw.c makes more sense?