linux-kernel - Re: [PATCH phy-next] phy: apple: apple: Check the actual ioremap return value

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <20260210204822.jr22el4rlyosin2q@skbuf>
Date: Tue, 10 Feb 2026 22:48:22 +0200
From: Vladimir Oltean <olteanv@...il.com>
To: Janne Grunau <j@...nau.net>
Cc: Sven Peter <sven@...nel.org>, Neal Gompa <neal@...pa.dev>,
	Vinod Koul <vkoul@...nel.org>,
	Neil Armstrong <neil.armstrong@...aro.org>,
	Philipp Zabel <p.zabel@...gutronix.de>, asahi@...ts.linux.dev,
	linux-phy@...ts.infradead.org, linux-kernel@...r.kernel.org,
	Dan Carpenter <dan.carpenter@...aro.org>
Subject: Re: [PATCH phy-next] phy: apple: apple: Check the actual ioremap
 return value

Hello Janne,

On Sat, Feb 07, 2026 at 05:40:34PM +0100, Janne Grunau wrote:
> The address where the devm_ioremap_resource() return value is stored is
> always a valid pointer. Check the actual return value instead as that
> that might be an error value.
> 
> Reported-by: Dan Carpenter <dan.carpenter@...aro.org>
> Closes: https://lore.kernel.org/asahi/aYXvX1bYOXtYCgfC@stanley.mountain/
> Fixes: 8e98ca1e74db ("phy: apple: Add Apple Type-C PHY")
> Signed-off-by: Janne Grunau <j@...nau.net>
> ---
>  drivers/phy/apple/atc.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/phy/apple/atc.c b/drivers/phy/apple/atc.c
> index dc867f368b68748ea953e594ad998d7f965d8d1d..c144e273a555a741b49adfccbe046df83d193e03 100644
> --- a/drivers/phy/apple/atc.c
> +++ b/drivers/phy/apple/atc.c
> @@ -2206,7 +2206,7 @@ static int atcphy_map_resources(struct platform_device *pdev, struct apple_atcph
>  	for (int i = 0; i < ARRAY_SIZE(resources); i++) {
>  		res = platform_get_resource_byname(pdev, IORESOURCE_MEM, resources[i].name);
>  		*resources[i].addr = devm_ioremap_resource(&pdev->dev, res);
> -		if (IS_ERR(resources[i].addr))
> +		if (IS_ERR(*resources[i].addr))
>  			return dev_err_probe(atcphy->dev, PTR_ERR(resources[i].addr),
>  					     "Unable to map %s regs", resources[i].name);

This does not seem correct - every call site that tests a pointer for
IS_ERR() also decodes the error using PTR_ERR(). Whereas you are here
creating the pattern "if (IS_ERR(*a)) return PTR_ERR(a)".

Proven practice seems to tell us that insisting to save the
devm_ioremap_resource() return code directly in *resources[i].addr
is too complex here.

Would you consider creating a temporary "void __iomem *addr" local
variable, and set "*resources[i].addr = addr" only once it's validated?
This would make the code obviously correct upon review.

>  
> 
> ---
> base-commit: dbeea86fecef7cf2b93aded4525d74f6277376ef
> change-id: 20260207-phy-apple-resource-err-ptr-5923d1130465
> 
> Best regards,
> -- 
> Janne Grunau <j@...nau.net>
> 
>