Re: [PATCH net-next] net: enetc: fix array underflow in error handling code

[Vladimir Oltean <vladimir.oltean@nxp.com>]

On Fri, Apr 09, 2021 at 03:24:28PM +0300, Dan Carpenter wrote:
> This loop will try to unmap enetc_unmap_tx_buff[-1] and crash.
> 
> Fixes: 9d2b68cc108d ("net: enetc: add support for XDP_REDIRECT")
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> ---
>  drivers/net/ethernet/freescale/enetc/enetc.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/freescale/enetc/enetc.c b/drivers/net/ethernet/freescale/enetc/enetc.c
> index 57049ae97201..d86395775ed0 100644
> --- a/drivers/net/ethernet/freescale/enetc/enetc.c
> +++ b/drivers/net/ethernet/freescale/enetc/enetc.c
> @@ -895,7 +895,7 @@ static int enetc_xdp_frame_to_xdp_tx_swbd(struct enetc_bdr *tx_ring,
>  		dma = dma_map_single(tx_ring->dev, data, len, DMA_TO_DEVICE);
>  		if (unlikely(dma_mapping_error(tx_ring->dev, dma))) {
>  			/* Undo the DMA mapping for all fragments */
> -			while (n-- >= 0)
> +			while (--n >= 0)
>  				enetc_unmap_tx_buff(tx_ring, &xdp_tx_arr[n]);
>  
>  			netdev_err(tx_ring->ndev, "DMA map error\n");
> -- 
> 2.30.2
> 

I have no idea what went wrong. I distinctly remember "testing" this by
writing up a small program:

#include <stdio.h>

int main(void)
{
	int n = 5;

	while (n-- >= 0)
		printf("%d\n", n);

	return 0;
}

I find myself doing that most of the time when there is a need for an
unrolling loop. I do see that this goes down to -1, I wonder how I
missed it.

Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>