Re: [PATCH] [PATCH 5.10] iommu/mediatek: remove redundant comparison

[Greg Kroah-Hartman <gregkh@linuxfoundation.org>]

On Thu, Aug 24, 2023 at 04:09:54PM +0300, Alexandra Diupina wrote:
> iommu_device_register always returns 0 in 4.11-5.12, so

What do you mean by "4.11-5.12"?

> we need to remove redundant comparison with 0

Why?

> 
> Found by Linux Verification Center (linuxtesting.org) with SVACE.
> 
> Fixes: b16c0170b53c ("iommu/mediatek: Make use of iommu_device_register interface")

How is this a fix if it isn't actually changing the code at all?

> Signed-off-by: Alexandra Diupina <adiupina@astralinux.ru>
> ---
>  drivers/iommu/mtk_iommu.c | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/drivers/iommu/mtk_iommu.c b/drivers/iommu/mtk_iommu.c
> index 051815c9d2bb..208c47218b75 100644
> --- a/drivers/iommu/mtk_iommu.c
> +++ b/drivers/iommu/mtk_iommu.c
> @@ -748,9 +748,7 @@ static int mtk_iommu_probe(struct platform_device *pdev)
>  	iommu_device_set_ops(&data->iommu, &mtk_iommu_ops);
>  	iommu_device_set_fwnode(&data->iommu, &pdev->dev.of_node->fwnode);
>  
> -	ret = iommu_device_register(&data->iommu);
> -	if (ret)
> -		return ret;
> +	iommu_device_register(&data->iommu);

This is obviously incorrect, please fix your "tool" that is telling you
to make such a broken change.

Unless your tool wants to create buggy code?  If so, it's succeeding
very well.

greg k-h