From: Ajit Khaparde <ajit.khaparde@broadcom.com>
To: Ferruh Yigit <ferruh.yigit@intel.com>
Cc: dev@dpdk.org, Shahaf Shuler <shahafs@mellanox.com>
Subject: Re: [dpdk-dev] [PATCH] net/bnxt: switch to the new offload API
Date: Wed, 21 Mar 2018 12:27:19 -0700 [thread overview]
Message-ID: <CACZ4nhuGWMSvp7hfPnzG2PNOOOcd1nusSyj8euc4pVLPWZMsKA@mail.gmail.com> (raw)
In-Reply-To: <ebbdcf65-02a8-22b1-0612-2a5d4b31b6d6@intel.com>
Hi Shahaf,
Well, the application could.
While we can't ignore the advantages of enabling the offloads.
I felt if we should really mandate applications to request all the
supported offloads?
Thanks
Ajit
On Wed, Mar 21, 2018 at 12:08 PM, Ferruh Yigit <ferruh.yigit@intel.com>
wrote:
> On 3/6/2018 7:40 PM, Ajit Khaparde wrote:
> > Update bnxt PMD to new ethdev offloads API.
> > Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
> > ---
> > drivers/net/bnxt/bnxt_ethdev.c | 59 +++++++++++++++++++++++++++++-
> ------------
> > 1 file changed, 41 insertions(+), 18 deletions(-)
> >
> > diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_
> ethdev.c
> > index 21c46f833..cca4ef40c 100644
> > --- a/drivers/net/bnxt/bnxt_ethdev.c
> > +++ b/drivers/net/bnxt/bnxt_ethdev.c
> > @@ -146,6 +146,27 @@ static const struct rte_pci_id bnxt_pci_id_map[] = {
> > ETH_RSS_NONFRAG_IPV6_TCP | \
> > ETH_RSS_NONFRAG_IPV6_UDP)
> >
> > +#define BNXT_DEV_TX_OFFLOAD_SUPPORT (DEV_TX_OFFLOAD_VLAN_INSERT | \
> > + DEV_TX_OFFLOAD_IPV4_CKSUM | \
> > + DEV_TX_OFFLOAD_TCP_CKSUM | \
> > + DEV_TX_OFFLOAD_UDP_CKSUM | \
> > + DEV_TX_OFFLOAD_TCP_TSO | \
> > + DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM | \
> > + DEV_TX_OFFLOAD_VXLAN_TNL_TSO | \
> > + DEV_TX_OFFLOAD_GRE_TNL_TSO | \
> > + DEV_TX_OFFLOAD_IPIP_TNL_TSO | \
> > + DEV_TX_OFFLOAD_GENEVE_TNL_TSO | \
> > + DEV_TX_OFFLOAD_MULTI_SEGS)
> > +
> > +#define BNXT_DEV_RX_OFFLOAD_SUPPORT (DEV_RX_OFFLOAD_VLAN_FILTER | \
> > + DEV_RX_OFFLOAD_VLAN_STRIP | \
> > + DEV_RX_OFFLOAD_IPV4_CKSUM | \
> > + DEV_RX_OFFLOAD_UDP_CKSUM | \
> > + DEV_RX_OFFLOAD_TCP_CKSUM | \
> > + DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM | \
> > + DEV_RX_OFFLOAD_JUMBO_FRAME | \
> > + DEV_RX_OFFLOAD_CRC_STRIP)
> > +
> > static int bnxt_vlan_offload_set_op(struct rte_eth_dev *dev, int mask);
> > static void bnxt_print_link_info(struct rte_eth_dev *eth_dev);
> >
> > @@ -430,21 +451,14 @@ static void bnxt_dev_info_get_op(struct
> rte_eth_dev *eth_dev,
> > dev_info->min_rx_bufsize = 1;
> > dev_info->max_rx_pktlen = BNXT_MAX_MTU + ETHER_HDR_LEN +
> ETHER_CRC_LEN
> > + VLAN_TAG_SIZE;
> > - dev_info->rx_offload_capa = DEV_RX_OFFLOAD_VLAN_STRIP |
> > - DEV_RX_OFFLOAD_IPV4_CKSUM |
> > - DEV_RX_OFFLOAD_UDP_CKSUM |
> > - DEV_RX_OFFLOAD_TCP_CKSUM |
> > - DEV_RX_OFFLOAD_OUTER_IPV4_CKSUM;
> > - dev_info->tx_offload_capa = DEV_TX_OFFLOAD_VLAN_INSERT |
> > - DEV_TX_OFFLOAD_IPV4_CKSUM |
> > - DEV_TX_OFFLOAD_TCP_CKSUM |
> > - DEV_TX_OFFLOAD_UDP_CKSUM |
> > - DEV_TX_OFFLOAD_TCP_TSO |
> > - DEV_TX_OFFLOAD_OUTER_IPV4_CKSUM |
> > - DEV_TX_OFFLOAD_VXLAN_TNL_TSO |
> > - DEV_TX_OFFLOAD_GRE_TNL_TSO |
> > - DEV_TX_OFFLOAD_IPIP_TNL_TSO |
> > - DEV_TX_OFFLOAD_GENEVE_TNL_TSO;
> > +
> > + dev_info->rx_queue_offload_capa = BNXT_DEV_RX_OFFLOAD_SUPPORT;
>
> Hi Ajit,
>
> Is bnxt supports queue level offloads? If not PMD should return 0 as
> rx_queue_offload_capa and tx_queue_offload_capa.
>
> > + dev_info->rx_offload_capa = BNXT_DEV_RX_OFFLOAD_SUPPORT;
> > + if (bp->flags & BNXT_FLAG_PTP_SUPPORTED)
> > + dev_info->rx_offload_capa |= DEV_RX_OFFLOAD_TIMESTAMP;
> > + dev_info->tx_queue_offload_capa = BNXT_DEV_TX_OFFLOAD_SUPPORT;
> > + dev_info->tx_offload_capa = BNXT_DEV_TX_OFFLOAD_SUPPORT;
> > + dev_info->flow_type_rss_offloads = BNXT_ETH_RSS_SUPPORT;
> >
> > /* *INDENT-OFF* */
> > dev_info->default_rxconf = (struct rte_eth_rxconf) {
> > @@ -454,7 +468,8 @@ static void bnxt_dev_info_get_op(struct rte_eth_dev
> *eth_dev,
> > .wthresh = 0,
> > },
> > .rx_free_thresh = 32,
> > - .rx_drop_en = 0,
> > + /* If no descriptors available, pkts are dropped by
> default */
> > + .rx_drop_en = 1,
> > };
> >
> > dev_info->default_txconf = (struct rte_eth_txconf) {
> > @@ -465,8 +480,6 @@ static void bnxt_dev_info_get_op(struct rte_eth_dev
> *eth_dev,
> > },
> > .tx_free_thresh = 32,
> > .tx_rs_thresh = 32,
> > - .txq_flags = ETH_TXQ_FLAGS_NOMULTSEGS |
> > - ETH_TXQ_FLAGS_NOOFFLOADS,
> > };
> > eth_dev->data->dev_conf.intr_conf.lsc = 1;
> >
> > @@ -510,6 +523,16 @@ static void bnxt_dev_info_get_op(struct rte_eth_dev
> *eth_dev,
> > static int bnxt_dev_configure_op(struct rte_eth_dev *eth_dev)
> > {
> > struct bnxt *bp = (struct bnxt *)eth_dev->data->dev_private;
> > + uint64_t tx_offloads = eth_dev->data->dev_conf.txmode.offloads;
> > + uint64_t rx_offloads = eth_dev->data->dev_conf.rxmode.offloads;
> > +
> > + if (tx_offloads != BNXT_DEV_TX_OFFLOAD_SUPPORT)
> > + PMD_DRV_LOG(ERR, "Tx offloads requested 0x%lx supported
> 0x%x\n",
> > + tx_offloads, BNXT_DEV_TX_OFFLOAD_SUPPORT);
>
> Isn't is possible that application request only a subset of the supported
> offloads?
>
> And should API return an error for this case?
>
> > +
> > + if (rx_offloads != BNXT_DEV_RX_OFFLOAD_SUPPORT)
> > + PMD_DRV_LOG(ERR, "Rx offloads requested 0x%lx supported
> 0x%x\n",
> > + rx_offloads, BNXT_DEV_RX_OFFLOAD_SUPPORT);
> >
> > bp->rx_queues = (void *)eth_dev->data->rx_queues;
> > bp->tx_queues = (void *)eth_dev->data->tx_queues;
> >
>
>
next prev parent reply other threads:[~2018-03-21 19:28 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-03-06 19:40 Ajit Khaparde
2018-03-21 19:08 ` Ferruh Yigit
2018-03-21 19:27 ` Ajit Khaparde [this message]
2018-03-22 6:22 ` Shahaf Shuler
2018-03-26 22:47 ` [dpdk-dev] [PATCH v2] " Ajit Khaparde
2018-03-30 10:38 ` Ferruh Yigit
2018-04-02 22:25 ` [dpdk-dev] [PATCH v3 1/2] " Ajit Khaparde
2018-04-02 22:25 ` [dpdk-dev] [PATCH v3 2/2] net/bnxt: fix rx_drop_en setting Ajit Khaparde
2018-04-05 14:36 ` [dpdk-dev] [PATCH v3 1/2] net/bnxt: switch to the new offload API Ferruh Yigit
2018-04-06 0:22 ` [dpdk-dev] [PATCH v4 1/2] net/bnxt: fix rx_drop_en setting Ajit Khaparde
2018-04-06 0:22 ` [dpdk-dev] [PATCH v4 2/2] net/bnxt: switch to the new offload API Ajit Khaparde
2018-04-06 15:11 ` Ferruh Yigit
2018-04-07 17:40 ` [dpdk-dev] [PATCH v5 1/2] net/bnxt: fix rx_drop_en setting Ajit Khaparde
2018-04-07 17:40 ` [dpdk-dev] [PATCH v5 2/2] net/bnxt: switch to the new offload API Ajit Khaparde
2018-04-09 16:53 ` [dpdk-dev] [dpdk-stable] [PATCH v5 1/2] net/bnxt: fix rx_drop_en setting Ferruh Yigit
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CACZ4nhuGWMSvp7hfPnzG2PNOOOcd1nusSyj8euc4pVLPWZMsKA@mail.gmail.com \
--to=ajit.khaparde@broadcom.com \
--cc=dev@dpdk.org \
--cc=ferruh.yigit@intel.com \
--cc=shahafs@mellanox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).