linux-kernel - Re: [PATCH] media: rkvdec: Fix an error handling path in rkvdec

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <0a8391cb368653b91ea73a51e2c0dee35cceb128.camel@collabora.com>
Date: Mon, 28 Jul 2025 18:50:36 -0400
From: Nicolas Dufresne <nicolas.dufresne@...labora.com>
To: Christophe JAILLET <christophe.jaillet@...adoo.fr>, Detlev Casanova	
 <detlev.casanova@...labora.com>, Mauro Carvalho Chehab
 <mchehab@...nel.org>,  Heiko Stuebner	 <heiko@...ech.de>, Hans Verkuil
 <hverkuil@...nel.org>
Cc: linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org, 
	linux-media@...r.kernel.org, linux-rockchip@...ts.infradead.org, 
	linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH] media: rkvdec: Fix an error handling path in
 rkvdec_probe()

Hi,

Le dimanche 27 juillet 2025 à 12:02 +0200, Christophe JAILLET a écrit :
> If an error occurs after a successful iommu_paging_domain_alloc() call, it
> should be undone by a corresponding iommu_domain_free() call, as already
> done in the remove function.
> 
> Fixes: ff8c5622f9f7 ("media: rkvdec: Restore iommu addresses on errors")
> Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>
> ---
> Compile tested only
> ---
>  drivers/media/platform/rockchip/rkvdec/rkvdec.c | 11 ++++++++---
>  1 file changed, 8 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/media/platform/rockchip/rkvdec/rkvdec.c
> b/drivers/media/platform/rockchip/rkvdec/rkvdec.c
> index d707088ec0dc..eb0d41f85d89 100644
> --- a/drivers/media/platform/rockchip/rkvdec/rkvdec.c
> +++ b/drivers/media/platform/rockchip/rkvdec/rkvdec.c
> @@ -1169,15 +1169,17 @@ static int rkvdec_probe(struct platform_device *pdev)
>  	vb2_dma_contig_set_max_seg_size(&pdev->dev, DMA_BIT_MASK(32));
>  
>  	irq = platform_get_irq(pdev, 0);
> -	if (irq <= 0)
> -		return -ENXIO;
> +	if (irq <= 0) {
> +		ret = -ENXIO;
> +		goto err_free_domain;
> +	}
>  
>  	ret = devm_request_threaded_irq(&pdev->dev, irq, NULL,
>  					rkvdec_irq_handler, IRQF_ONESHOT,
>  					dev_name(&pdev->dev), rkvdec);
>  	if (ret) {
>  		dev_err(&pdev->dev, "Could not request vdec IRQ\n");
> -		return ret;
> +		goto err_free_domain;
>  	}
>  
>  	pm_runtime_set_autosuspend_delay(&pdev->dev, 100);

Have you considered moving the allocation of the domain right above the above
line instead ? The empty domain can't possibly be used unless the probe have
fully completed.

Nicolas

> @@ -1193,6 +1195,9 @@ static int rkvdec_probe(struct platform_device *pdev)
>  err_disable_runtime_pm:
>  	pm_runtime_dont_use_autosuspend(&pdev->dev);
>  	pm_runtime_disable(&pdev->dev);
> +err_free_domain:
> +	if (rkvdec->empty_domain)
> +		iommu_domain_free(rkvdec->empty_domain);
>  	return ret;
>  }
>  

Download attachment "signature.asc" of type "application/pgp-signature" (196 bytes)