linux-kernel - Re: [PATCH] drivers: loongson: fix memory leak in loongson2_guts

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: <20622cce-7bd0-fd2a-16b1-064f572d52dc@xen0n.name>
Date:   Mon, 6 Mar 2023 10:27:16 +0800
From:   WANG Xuerui <kernel@...0n.name>
To:     Dongliang Mu <dzm91@...t.edu.cn>, Yinbo Zhu <zhuyinbo@...ngson.cn>
Cc:     loongarch@...ts.linux.dev, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] drivers: loongson: fix memory leak in
 loongson2_guts_probe/remove

Hi,

On 2023/3/6 09:22, Dongliang Mu wrote:
> drivers/soc/loongson/loongson2_guts.c:150 loongson2_guts_probe()
> warn: 'guts->regs' from ioremap() not released on lines: 131,135,139,143.
> 
> Fix this by invoking iounmap() in the error handling code and the remove
> function.
> 
> Note that, this patch is not tested due to the loongson architecture.

I can't test either, AFAIK no one outside of Loongson have working 
LoongArch Loongson-2 hardware running new world system. Yinbo will have 
to take care of this.

> 
> Signed-off-by: Dongliang Mu <dzm91@...t.edu.cn> > ---
>   drivers/soc/loongson/loongson2_guts.c | 18 +++++++++++++-----
>   1 file changed, 13 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/soc/loongson/loongson2_guts.c b/drivers/soc/loongson/loongson2_guts.c
> index bace4bc8e03b..ba330adb555c 100644
> --- a/drivers/soc/loongson/loongson2_guts.c
> +++ b/drivers/soc/loongson/loongson2_guts.c
> @@ -98,6 +98,7 @@ static int loongson2_guts_probe(struct platform_device *pdev)
>   	const struct loongson2_soc_die_attr *soc_die;
>   	const char *machine;
>   	u32 svr;
> +	int rc = -ENOMEM;
>   
>   	/* Initialize guts */
>   	guts = devm_kzalloc(dev, sizeof(*guts), GFP_KERNEL);
> @@ -128,19 +129,21 @@ static int loongson2_guts_probe(struct platform_device *pdev)
>   		soc_dev_attr.family = devm_kasprintf(dev, GFP_KERNEL, "Loongson");
>   	}
>   	if (!soc_dev_attr.family)
> -		return -ENOMEM;
> +		goto iounmap;
>   	soc_dev_attr.soc_id = devm_kasprintf(dev, GFP_KERNEL,
>   					     "svr:0x%08x", svr);
>   	if (!soc_dev_attr.soc_id)
> -		return -ENOMEM;
> +		goto iounmap;
>   	soc_dev_attr.revision = devm_kasprintf(dev, GFP_KERNEL, "%d.%d",
>   					       (svr >>  4) & 0xf, svr & 0xf);
>   	if (!soc_dev_attr.revision)
> -		return -ENOMEM;
> +		goto iounmap;

Instead of so many goto's, it's probably better to just change the 
ioremap to devm_ioremap and let the driver core take care of the rest. 
Much more maintainable this way IMO.

-- 
WANG "xen0n" Xuerui

Linux/LoongArch mailing list: https://lore.kernel.org/loongarch/