2019-09-28 00:28:53

by Yizhuo Zhai

Subject: [PATCH] iio: adc: imx25-gcq: Variable could be uninitialized if regmap_read() fails

In function mx25_gcq_irq(), local variable "stats" could
be uninitialized if function regmap_read() returns -EINVAL.
However, this value is used in if statement, which is
potentially unsafe. The same case applied to the variable
"data" in function mx25_gcq_get_raw_value() in the same file.

Signed-off-by: Yizhuo <[email protected]>
---
drivers/iio/adc/fsl-imx25-gcq.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/iio/adc/fsl-imx25-gcq.c b/drivers/iio/adc/fsl-imx25-gcq.c
index df19ecae52f7..dbf3e8e85aba 100644
--- a/drivers/iio/adc/fsl-imx25-gcq.c
+++ b/drivers/iio/adc/fsl-imx25-gcq.c
@@ -74,7 +74,10 @@ static irqreturn_t mx25_gcq_irq(int irq, void *data)
struct mx25_gcq_priv *priv = data;
u32 stats;

- regmap_read(priv->regs, MX25_ADCQ_SR, &stats);
+ int ret = regmap_read(priv->regs, MX25_ADCQ_SR, &stats);
+
+ if (ret)
+ return ret;

if (stats & MX25_ADCQ_SR_EOQ) {
regmap_update_bits(priv->regs, MX25_ADCQ_MR,
@@ -121,7 +124,10 @@ static int mx25_gcq_get_raw_value(struct device *dev,
return -ETIMEDOUT;
}

- regmap_read(priv->regs, MX25_ADCQ_FIFO, &data);
+ int ret = regmap_read(priv->regs, MX25_ADCQ_FIFO, &data);
+
+ if (ret)
+ return ret;

*val = MX25_ADCQ_FIFO_DATA(data);

--
2.17.1

2019-09-30 07:45:17

by Marco Felsch

Subject: Re: [PATCH] iio: adc: imx25-gcq: Variable could be uninitialized if regmap_read() fails

Hi Yizhuo,

thanks for your patch.

On 19-09-27 17:28, Yizhuo wrote:
> In function mx25_gcq_irq(), local variable "stats" could
> be uninitialized if function regmap_read() returns -EINVAL.
> However, this value is used in if statement, which is
> potentially unsafe. The same case applied to the variable
> "data" in function mx25_gcq_get_raw_value() in the same file.

IMHO the commit header should be something like: "iio: adc: imx25-gcq:
fix uninitialized variable usage"...

and please add a fixes tag.

>
> Signed-off-by: Yizhuo <[email protected]>
> ---
> drivers/iio/adc/fsl-imx25-gcq.c | 10 ++++++++--
> 1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/iio/adc/fsl-imx25-gcq.c b/drivers/iio/adc/fsl-imx25-gcq.c
> index df19ecae52f7..dbf3e8e85aba 100644
> --- a/drivers/iio/adc/fsl-imx25-gcq.c
> +++ b/drivers/iio/adc/fsl-imx25-gcq.c
> @@ -74,7 +74,10 @@ static irqreturn_t mx25_gcq_irq(int irq, void *data)
> struct mx25_gcq_priv *priv = data;
> u32 stats;
>
> - regmap_read(priv->regs, MX25_ADCQ_SR, &stats);
> + int ret = regmap_read(priv->regs, MX25_ADCQ_SR, &stats);

Please don't do this. First declare the variable and then use it.

Regards,
Marco

> +
> + if (ret)
> + return ret;
>
> if (stats & MX25_ADCQ_SR_EOQ) {
> regmap_update_bits(priv->regs, MX25_ADCQ_MR,
> @@ -121,7 +124,10 @@ static int mx25_gcq_get_raw_value(struct device *dev,
> return -ETIMEDOUT;
> }
>
> - regmap_read(priv->regs, MX25_ADCQ_FIFO, &data);
> + int ret = regmap_read(priv->regs, MX25_ADCQ_FIFO, &data);
> +
> + if (ret)
> + return ret;
>
> *val = MX25_ADCQ_FIFO_DATA(data);
>
> --
> 2.17.1
>
>
>

--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |

2019-10-01 08:55:51

by Jonathan Cameron

Subject: Re: [PATCH] iio: adc: imx25-gcq: Variable could be uninitialized if regmap_read() fails

On Mon, 30 Sep 2019 09:44:12 +0200
Marco Felsch <[email protected]> wrote:

> Hi Yizhuo,
>
> thanks for your patch.
>
> On 19-09-27 17:28, Yizhuo wrote:
> > In function mx25_gcq_irq(), local variable "stats" could
> > be uninitialized if function regmap_read() returns -EINVAL.
> > However, this value is used in if statement, which is
> > potentially unsafe. The same case applied to the variable
> > "data" in function mx25_gcq_get_raw_value() in the same file.
>
> IMHO the commit header should be something like: "iio: adc: imx25-gcq:
> fix uninitialized variable usage"...
>
> and please add a fixes tag.
As with the others, before adding a fixes tag, please verify there
is an actual path to trigger this.

In this case it's an mmio regmap with no clock. For those, I'm not sure
if there is a failure path.

Still a worthwhile hardening / cleanup patch, but shouldn't be called
a fix or marked with a fixes tag because we don't want people to think
it is necessary to backport it.

Thanks,

Jonathan


>
> >
> > Signed-off-by: Yizhuo <[email protected]>
> > ---
> > drivers/iio/adc/fsl-imx25-gcq.c | 10 ++++++++--
> > 1 file changed, 8 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/iio/adc/fsl-imx25-gcq.c b/drivers/iio/adc/fsl-imx25-gcq.c
> > index df19ecae52f7..dbf3e8e85aba 100644
> > --- a/drivers/iio/adc/fsl-imx25-gcq.c
> > +++ b/drivers/iio/adc/fsl-imx25-gcq.c
> > @@ -74,7 +74,10 @@ static irqreturn_t mx25_gcq_irq(int irq, void *data)
> > struct mx25_gcq_priv *priv = data;
> > u32 stats;
> >
> > - regmap_read(priv->regs, MX25_ADCQ_SR, &stats);
> > + int ret = regmap_read(priv->regs, MX25_ADCQ_SR, &stats);
>
> Please don't do this. First declare the variable and then use it.
>
> Regards,
> Marco
>
> > +
> > + if (ret)
> > + return ret;
> >
> > if (stats & MX25_ADCQ_SR_EOQ) {
> > regmap_update_bits(priv->regs, MX25_ADCQ_MR,
> > @@ -121,7 +124,10 @@ static int mx25_gcq_get_raw_value(struct device *dev,
> > return -ETIMEDOUT;
> > }
> >
> > - regmap_read(priv->regs, MX25_ADCQ_FIFO, &data);
> > + int ret = regmap_read(priv->regs, MX25_ADCQ_FIFO, &data);
> > +
> > + if (ret)
> > + return ret;
> >
> > *val = MX25_ADCQ_FIFO_DATA(data);
> >
> > --
> > 2.17.1
> >
> >
> >
>