[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20170424071624.GA30553@pali>
Date: Mon, 24 Apr 2017 09:16:24 +0200
From: Pali Rohár <pali.rohar@...il.com>
To: Pan Bian <bianpan2016@....com>
Cc: Sebastian Reichel <sre@...nel.org>, linux-pm@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/1] power: Fix unchecked return value of devm_kzalloc
On Monday 24 April 2017 08:34:08 Pan Bian wrote:
> Function devm_kzalloc() will return a NULL pointer. However, in function
> isp1704_charger_probe(), the return value of devm_kzalloc() is directly
> used without validation. This may result in a bad memory access bug.
>
> Signed-off-by: Pan Bian <bianpan2016@....com>
> ---
> drivers/power/supply/isp1704_charger.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/power/supply/isp1704_charger.c b/drivers/power/supply/isp1704_charger.c
> index 4cd6899..2f02463 100644
> --- a/drivers/power/supply/isp1704_charger.c
> +++ b/drivers/power/supply/isp1704_charger.c
> @@ -418,6 +418,8 @@ static int isp1704_charger_probe(struct platform_device *pdev)
>
> pdata = devm_kzalloc(&pdev->dev,
> sizeof(struct isp1704_charger_data), GFP_KERNEL);
> + if (pdata)
> + goto fail0;
> pdata->enable_gpio = gpio;
>
> dev_info(&pdev->dev, "init gpio %d\n", pdata->enable_gpio);
fail0 returns ret value value which is initialized to -ENODEV. It could
be fragile if somebody change default initialization ret value to
something else (wort to 0). In your case before goto fail0 it is needed
to explicitly set ret to -ENOMEM so prevent future problems.
--
Pali Rohár
pali.rohar@...il.com
Powered by blists - more mailing lists