Re: [dpdk-dev] [PATCH] net/bnx2x: convert to new Rx offloads API

["Mody, Rasesh" <Rasesh.Mody@cavium.com>]

> From: Ferruh Yigit [mailto:ferruh.yigit@intel.com]
> Sent: Thursday, April 05, 2018 6:40 AM
> 
> On 4/1/2018 7:03 AM, Rasesh Mody wrote:
> > From: Harish Patil <harish.patil@cavium.com>
> >
> > Ethdev RX offloads API has changed since:
> > commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API")
> >
> > This patch makes use of DEV_RX_OFFLOAD_JUMBO_FRAME offload flag to
> > advertise jumbo frame support.
> >
> > Signed-off-by: Harish Patil <harish.patil@cavium.com>
> > ---
> >  drivers/net/bnx2x/bnx2x_ethdev.c |    5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/bnx2x/bnx2x_ethdev.c
> > b/drivers/net/bnx2x/bnx2x_ethdev.c
> > index 483d5a1..fada4bd 100644
> > --- a/drivers/net/bnx2x/bnx2x_ethdev.c
> > +++ b/drivers/net/bnx2x/bnx2x_ethdev.c
> > @@ -140,11 +140,13 @@ struct rte_bnx2x_xstats_name_off {
> > bnx2x_dev_configure(struct rte_eth_dev *dev)  {
> >  	struct bnx2x_softc *sc = dev->data->dev_private;
> > +	struct rte_eth_rxmode *rxmode = &dev->data->dev_conf.rxmode;
> > +
> >  	int mp_ncpus = sysconf(_SC_NPROCESSORS_CONF);
> >
> >  	PMD_INIT_FUNC_TRACE();
> >
> > -	if (dev->data->dev_conf.rxmode.jumbo_frame)
> > +	if (rxmode & DEV_RX_OFFLOAD_JUMBO_FRAME)
> 
> I guess you would mean "rxmode->offloads".
> 
> Isn't this caught by automated build?

You are right, will send a v2 with proper offload check.

I do not think this was caught by automated build.

> 
> >  		sc->mtu = dev->data->dev_conf.rxmode.max_rx_pkt_len;
> >
> >  	if (dev->data->nb_tx_queues > dev->data->nb_rx_queues) { @@ -
> 454,6
> > +456,7 @@ struct rte_bnx2x_xstats_name_off {
> >  	dev_info->max_rx_pktlen  = BNX2X_MAX_RX_PKT_LEN;
> >  	dev_info->max_mac_addrs  = BNX2X_MAX_MAC_ADDRS;
> >  	dev_info->speed_capa = ETH_LINK_SPEED_10G |
> ETH_LINK_SPEED_20G;
> > +	dev_info->rx_offload_capa = DEV_RX_OFFLOAD_JUMBO_FRAME;
> >  }
> >
> >  static int
> >