Re: [PATCH 1/2] spi: s3c64xx: fix potential segmentation fault

[Krzysztof Kozlowski]

On Tue, Feb 07, 2017 at 05:10:24PM +0900, Andi Shyti wrote:
> The return value of dmaengine_prep_slave_sg is not checked, but
> eventually it can fail and in that case return 'NULL' causing a
> segmentation fault.

NULL pointer dereference instead of segmentation fault.

> 
> Check dmaengine_prep_slave_sg return value and exit in case of
> failure. For doing this all the 'void' functions involved has
> been turned to 'int'.
> 
> This patch fixes '1397997 Dereference null return value' from
> scan.coverity.com
> 
> Signed-off-by: Andi Shyti <andi.shyti@xxxxxxxxxxx>
> ---
>  drivers/spi/spi-s3c64xx.c | 23 ++++++++++++++++++-----
>  1 file changed, 18 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/spi/spi-s3c64xx.c b/drivers/spi/spi-s3c64xx.c
> index b392cca8fa4f..f6ea9ae047ec 100644
> --- a/drivers/spi/spi-s3c64xx.c
> +++ b/drivers/spi/spi-s3c64xx.c
> @@ -276,7 +276,7 @@ static void s3c64xx_spi_dmacb(void *data)
>  	spin_unlock_irqrestore(&sdd->lock, flags);
>  }
>  
> -static void prepare_dma(struct s3c64xx_spi_dma_data *dma,
> +static int prepare_dma(struct s3c64xx_spi_dma_data *dma,
>  			struct sg_table *sgt)
>  {
>  	struct s3c64xx_spi_driver_data *sdd;
> @@ -305,12 +305,16 @@ static void prepare_dma(struct s3c64xx_spi_dma_data *dma,
>  
>  	desc = dmaengine_prep_slave_sg(dma->ch, sgt->sgl, sgt->nents,
>  				       dma->direction, DMA_PREP_INTERRUPT);
> +	if (!desc)
> +		return -EINVAL;
>  
>  	desc->callback = s3c64xx_spi_dmacb;
>  	desc->callback_param = dma;
>  
>  	dmaengine_submit(desc);
>  	dma_async_issue_pending(dma->ch);
> +
> +	return 0;
>  }
>  
>  static void s3c64xx_spi_set_cs(struct spi_device *spi, bool enable)
> @@ -360,12 +364,13 @@ static bool s3c64xx_spi_can_dma(struct spi_master *master,
>  	return xfer->len > (FIFO_LVL_MASK(sdd) >> 1) + 1;
>  }
>  
> -static void enable_datapath(struct s3c64xx_spi_driver_data *sdd,
> +static int enable_datapath(struct s3c64xx_spi_driver_data *sdd,
>  				struct spi_device *spi,
>  				struct spi_transfer *xfer, int dma_mode)
>  {
>  	void __iomem *regs = sdd->regs;
>  	u32 modecfg, chcfg;
> +	int ret;
>  
>  	modecfg = readl(regs + S3C64XX_SPI_MODE_CFG);
>  	modecfg &= ~(S3C64XX_SPI_MODE_TXDMA_ON | S3C64XX_SPI_MODE_RXDMA_ON);
> @@ -391,7 +396,9 @@ static void enable_datapath(struct s3c64xx_spi_driver_data *sdd,
>  		chcfg |= S3C64XX_SPI_CH_TXCH_ON;
>  		if (dma_mode) {
>  			modecfg |= S3C64XX_SPI_MODE_TXDMA_ON;
> -			prepare_dma(&sdd->tx_dma, &xfer->tx_sg);
> +			ret = prepare_dma(&sdd->tx_dma, &xfer->tx_sg);
> +			if (ret)
> +				return ret;
>  		} else {
>  			switch (sdd->cur_bpw) {
>  			case 32:
> @@ -423,12 +430,16 @@ static void enable_datapath(struct s3c64xx_spi_driver_data *sdd,
>  			writel(((xfer->len * 8 / sdd->cur_bpw) & 0xffff)
>  					| S3C64XX_SPI_PACKET_CNT_EN,
>  					regs + S3C64XX_SPI_PACKET_CNT);
> -			prepare_dma(&sdd->rx_dma, &xfer->rx_sg);
> +			ret = prepare_dma(&sdd->rx_dma, &xfer->rx_sg);
> +			if (ret)

Don't you have to terminate the transfer started from tx prepapre_dma()?

> +				return ret;
>  		}
>  	}
>  
>  	writel(modecfg, regs + S3C64XX_SPI_MODE_CFG);
>  	writel(chcfg, regs + S3C64XX_SPI_CH_CFG);
> +
> +	return 0;
>  }
>  
>  static u32 s3c64xx_spi_wait_for_timeout(struct s3c64xx_spi_driver_data *sdd,
> @@ -677,7 +688,9 @@ static int s3c64xx_spi_transfer_one(struct spi_master *master,
>  	sdd->state &= ~RXBUSY;
>  	sdd->state &= ~TXBUSY;
>  
> -	enable_datapath(sdd, spi, xfer, use_dma);
> +	status = enable_datapath(sdd, spi, xfer, use_dma);
> +	if (status)
> +		return status;

This looks really untested...

You have a spin lock here so you cannot just bail out. Instead, the
error path should clean it up by terminating transfers and unlocking.

Best regards,
Krzysztof