On Saturday 30 January 2016 19:47:36 Vasily Averin wrote:
> Patch makes safe an access to 'oiph' and 'oudph' variables
> if they was not initilized.
>
> Signed-off-by: Vasily Averin <v...@virtuozzo.com>
> ---
> drivers/net/ethernet/intel/i40e/i40e_txrx.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
> b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
> index 47bd8b3..779f77e 100644
> --- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
> +++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
> @@ -2363,7 +2363,8 @@ static void i40e_tx_enable_csum(struct sk_buff *skb,
> u32 *tx_flags,
> }
> if ((tx_ring->flags & I40E_TXR_FLAGS_OUTER_UDP_CSUM) &&
> (l4_tunnel == I40E_TXD_CTX_UDP_TUNNELING) &&
> - (*cd_tunneling & I40E_TXD_CTX_QW0_EXT_IP_MASK)) {
> + (*cd_tunneling & I40E_TXD_CTX_QW0_EXT_IP_MASK) &&
> + oudph && oiph) {
> oudph->check = ~csum_tcpudp_magic(oiph->saddr,
> oiph->daddr,
> (skb->len -
> skb_transport_offset(skb)),
If we can actually get here with oudph==NULL or oiph==NULL, we should back
my patch instead and fix it another way.
In the version we have in net-next, that is not possible.
Arnd
