DPDK patches and discussions
 help / color / mirror / Atom feed
From: Ivan Malov <ivan.malov@arknetworks.am>
To: "Morten Brørup" <mb@smartsharesystems.com>
Cc: dev@dpdk.org, Thomas Monjalon <thomas@monjalon.net>,
	 Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>,
	 Konstantin Ananyev <konstantin.ananyev@huawei.com>
Subject: Re: [PATCH] ethdev: Reject conflicting TX offloads configuration
Date: Thu, 31 Jul 2025 13:27:58 +0400 (+04)	[thread overview]
Message-ID: <a2542c91-6a37-94c3-98e9-c237dc0c1d5b@arknetworks.am> (raw)
In-Reply-To: <20250731090731.671589-1-mb@smartsharesystems.com>

[-- Attachment #1: Type: text/plain, Size: 6681 bytes --]

Hi Morten,

On Thu, 31 Jul 2025, Morten Brørup wrote:

> When an ethdev port is configured for fast mbuf release, the driver can
> use a TX burst function relying on the fast mbuf release preconditions.
> Thus, also configuring this port or a queue on the port for transmitting
> segmented packets is prohibited.
> Checks for these conflicting configurations have been added to the ethdev
> library, so the drivers don't have to implement them.
>
> Also, the descriptions of the RX and TX offloads have been improved, to
> reflect that they are not only for device capability reporting, but also
> for device and queue configuration purposes.
>
> Signed-off-by: Morten Brørup <mb@smartsharesystems.com>
> ---
> lib/ethdev/rte_ethdev.c | 38 ++++++++++++++++++++++++++++++++++++++
> lib/ethdev/rte_ethdev.h | 20 ++++++++++----------
> 2 files changed, 48 insertions(+), 10 deletions(-)
>
> diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c
> index dd7c00bc94..6570517711 100644
> --- a/lib/ethdev/rte_ethdev.c
> +++ b/lib/ethdev/rte_ethdev.c
> @@ -1531,6 +1531,18 @@ rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q,
> 		goto rollback;
> 	}
>
> +	/* MBUF_FAST_FREE preconditions conflict with MULTI_SEGS support. */
> +	if ((dev_conf->txmode.offloads & RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE) &&
> +	    (dev_conf->txmode.offloads & RTE_ETH_TX_OFFLOAD_MULTI_SEGS)) {
> +		RTE_ETHDEV_LOG_LINE(ERR,
> +			"Ethdev port_id=%d Tx offload %s conflicts with Tx offload %s",

Is it required to spell 'Ethdev ' at start of the line? It was my impression
that 'ethdev' would be printed anyway by [1] (the '# t' part). Am I wrong?

[1] https://github.com/DPDK/dpdk/blob/b222395561638f89562e4ef42e1eabf2d6db43dd/lib/log/rte_log.h#L334

Thank you.

> +			port_id,
> +			rte_eth_dev_tx_offload_name(RTE_ETH_TX_OFFLOAD_MULTI_SEGS),
> +			rte_eth_dev_tx_offload_name(RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE));
> +		ret = -EINVAL;
> +		goto rollback;
> +	}
> +
> 	dev->data->dev_conf.rx_adv_conf.rss_conf.rss_hf =
> 		rte_eth_rss_hf_refine(dev_conf->rx_adv_conf.rss_conf.rss_hf);
>
> @@ -2709,6 +2721,32 @@ rte_eth_tx_queue_setup(uint16_t port_id, uint16_t tx_queue_id,
> 		return -EINVAL;
> 	}
>
> +	/*
> +	 * If the driver uses a Tx function with MBUF_FAST_FREE preconditions,
> +	 * per-queue MULTI_SEGS support is not possible.
> +	 */
> +	if ((dev->data->dev_conf.txmode.offloads & RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE) &&
> +			(local_conf.offloads & RTE_ETH_TX_OFFLOAD_MULTI_SEGS)) {
> +		RTE_ETHDEV_LOG_LINE(ERR,
> +			"Ethdev port_id=%d tx_queue_id=%d, Tx offload %s conflicts with per-port Tx offload %s",
> +			port_id, tx_queue_id,
> +			rte_eth_dev_tx_offload_name(RTE_ETH_TX_OFFLOAD_MULTI_SEGS),
> +			rte_eth_dev_tx_offload_name(RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE));
> +		return -EINVAL;
> +	}
> +	/*
> +	 * If the driver uses a Tx function with MULTI_SEGS support,
> +	 * runtime support for per-queue MBUF_FAST_FREE optimization depends on the driver.
> +	 */
> +	if ((dev->data->dev_conf.txmode.offloads & RTE_ETH_TX_OFFLOAD_MULTI_SEGS) &&
> +			(local_conf.offloads & RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE))
> +		RTE_ETHDEV_LOG_LINE(DEBUG,
> +			"Ethdev port_id=%d tx_queue_id=%d, Tx offload %s potential conflict with per-port Tx offload %s, "
> +			"runtime support depends on the driver",
> +			port_id, tx_queue_id,
> +			rte_eth_dev_tx_offload_name(RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE),
> +			rte_eth_dev_tx_offload_name(RTE_ETH_TX_OFFLOAD_MULTI_SEGS));
> +
> 	rte_ethdev_trace_txq_setup(port_id, tx_queue_id, nb_tx_desc, tx_conf);
> 	return eth_err(port_id, dev->dev_ops->tx_queue_setup(dev,
> 		       tx_queue_id, nb_tx_desc, socket_id, &local_conf));
> diff --git a/lib/ethdev/rte_ethdev.h b/lib/ethdev/rte_ethdev.h
> index f9fb6ae549..760a1d2675 100644
> --- a/lib/ethdev/rte_ethdev.h
> +++ b/lib/ethdev/rte_ethdev.h
> @@ -1550,7 +1550,7 @@ struct rte_eth_conf {
> };
>
> /**
> - * Rx offload capabilities of a device.
> + * Rx offload capabilities/configuration of a device or queue.
>  */
> #define RTE_ETH_RX_OFFLOAD_VLAN_STRIP       RTE_BIT64(0)
> #define RTE_ETH_RX_OFFLOAD_IPV4_CKSUM       RTE_BIT64(1)
> @@ -1585,12 +1585,12 @@ struct rte_eth_conf {
> 			     RTE_ETH_RX_OFFLOAD_QINQ_STRIP)
>
> /*
> - * If new Rx offload capabilities are defined, they also must be
> + * If new Rx offloads are defined, they also must be
>  * mentioned in rte_rx_offload_names in rte_ethdev.c file.
>  */
>
> /**
> - * Tx offload capabilities of a device.
> + * Tx offload capabilities/configuration of a device or queue.
>  */
> #define RTE_ETH_TX_OFFLOAD_VLAN_INSERT      RTE_BIT64(0)
> #define RTE_ETH_TX_OFFLOAD_IPV4_CKSUM       RTE_BIT64(1)
> @@ -1611,10 +1611,10 @@ struct rte_eth_conf {
>  * Tx queue without SW lock.
>  */
> #define RTE_ETH_TX_OFFLOAD_MT_LOCKFREE      RTE_BIT64(14)
> -/** Device supports multi segment send. */
> +/** Multi segment send. */
> #define RTE_ETH_TX_OFFLOAD_MULTI_SEGS       RTE_BIT64(15)
> /**
> - * Device supports optimization for fast release of mbufs.
> + * Optimization for fast release of mbufs.
>  * When set application must guarantee that per-queue all mbufs come from the same mempool,
>  * are direct, have refcnt=1, next=NULL and nb_segs=1, as done by rte_pktmbuf_prefree_seg().
>  *
> @@ -1623,27 +1623,27 @@ struct rte_eth_conf {
> #define RTE_ETH_TX_OFFLOAD_MBUF_FAST_FREE   RTE_BIT64(16)
> #define RTE_ETH_TX_OFFLOAD_SECURITY         RTE_BIT64(17)
> /**
> - * Device supports generic UDP tunneled packet TSO.
> + * Generic UDP tunneled packet TSO.
>  * Application must set RTE_MBUF_F_TX_TUNNEL_UDP and other mbuf fields required
>  * for tunnel TSO.
>  */
> #define RTE_ETH_TX_OFFLOAD_UDP_TNL_TSO      RTE_BIT64(18)
> /**
> - * Device supports generic IP tunneled packet TSO.
> + * Generic IP tunneled packet TSO.
>  * Application must set RTE_MBUF_F_TX_TUNNEL_IP and other mbuf fields required
>  * for tunnel TSO.
>  */
> #define RTE_ETH_TX_OFFLOAD_IP_TNL_TSO       RTE_BIT64(19)
> -/** Device supports outer UDP checksum */
> +/** Outer UDP checksum. Used for tunneling packet. */
> #define RTE_ETH_TX_OFFLOAD_OUTER_UDP_CKSUM  RTE_BIT64(20)
> /**
> - * Device sends on time read from RTE_MBUF_DYNFIELD_TIMESTAMP_NAME
> + * Send on time read from RTE_MBUF_DYNFIELD_TIMESTAMP_NAME
>  * if RTE_MBUF_DYNFLAG_TX_TIMESTAMP_NAME is set in ol_flags.
>  * The mbuf field and flag are registered when the offload is configured.
>  */
> #define RTE_ETH_TX_OFFLOAD_SEND_ON_TIMESTAMP RTE_BIT64(21)
> /*
> - * If new Tx offload capabilities are defined, they also must be
> + * If new Tx offloads are defined, they also must be
>  * mentioned in rte_tx_offload_names in rte_ethdev.c file.
>  */
>
> -- 
> 2.43.0
>
>

  parent reply	other threads:[~2025-07-31  9:28 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-07-31  9:07 Morten Brørup
2025-07-31  9:24 ` Bruce Richardson
2025-07-31 10:07   ` Andrew Rybchenko
2025-07-31  9:27 ` Ivan Malov [this message]
2025-07-31  9:34   ` Morten Brørup
2025-07-31 11:32 ` Konstantin Ananyev
2025-07-31 12:56 ` [PATCH v2 1/3] testpmd: Do not enable mbuf fast release TX offload by default Morten Brørup
2025-07-31 12:56   ` [PATCH v2 2/3] ethdev: Improve descriptions of RX and TX offloads Morten Brørup
2025-07-31 12:56   ` [PATCH v2 3/3] ethdev: Reject conflicting TX offloads configuration Morten Brørup
2025-08-01  1:09     ` zhoumin

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=a2542c91-6a37-94c3-98e9-c237dc0c1d5b@arknetworks.am \
    --to=ivan.malov@arknetworks.am \
    --cc=andrew.rybchenko@oktetlabs.ru \
    --cc=dev@dpdk.org \
    --cc=konstantin.ananyev@huawei.com \
    --cc=mb@smartsharesystems.com \
    --cc=thomas@monjalon.net \
    /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).