DPDK patches and discussions
 help / color / mirror / Atom feed
From: oulijun <oulijun@huawei.com>
To: Steve Yang <stevex.yang@intel.com>, <dev@dpdk.org>
Cc: <thomas@monjalon.net>, <ferruh.yigit@intel.com>,
	<andrew.rybchenko@oktetlabs.ru>, <linuxarm@openeuler.org>
Subject: Re: [dpdk-dev] [PATCH v3 01/22] ethdev: fix MTU size exceeds max rx packet length
Date: Fri, 15 Jan 2021 18:44:23 +0800
Message-ID: <c79413fe-06ab-a2eb-96ec-d55d968fd2b2@huawei.com> (raw)
In-Reply-To: <20210114094537.13661-1-stevex.yang@intel.com>

Hi Steve
This is a very good job! But I have some question and suggestions.
Please check it.

在 2021/1/14 17:45, Steve Yang 写道:
> Ethdev is using default Ethernet overhead to decide if provided
> 'max_rx_pkt_len' value is bigger than max (non jumbo) MTU value,
> and limits it to MAX if it is.
> 
> Since the application/driver used Ethernet overhead is different than
> the ethdev one, check result is wrong.
> 
> If the driver is using Ethernet overhead bigger than the default one,
> the provided 'max_rx_pkt_len' is trimmed down, and in the driver when
> correct Ethernet overhead is used to convert back, the resulting MTU is
> less than the intended one, causing some packets to be dropped.
> 
> Like,
> app     -> max_rx_pkt_len = 1500/*mtu*/ + 22/*overhead*/ = 1522
> ethdev  -> 1522 > 1518/*MAX*/; max_rx_pkt_len = 1518
> driver  -> MTU = 1518 - 22 = 1496
> Packets with size 1497-1500 are dropped although intention is to be able
> to send/receive them.
> 
> The fix is to make ethdev use the correct Ethernet overhead for port,
> instead of default one.
> 
> Fixes: 59d0ecdbf0e1 ("ethdev: MTU accessors")
> 
> Signed-off-by: Steve Yang <stevex.yang@intel.com>
> ---
>   lib/librte_ethdev/rte_ethdev.c | 27 ++++++++++++++++++++++++---
>   1 file changed, 24 insertions(+), 3 deletions(-)
> 
> diff --git a/lib/librte_ethdev/rte_ethdev.c b/lib/librte_ethdev/rte_ethdev.c
> index 17ddacc78d..19ca4c4512 100644
> --- a/lib/librte_ethdev/rte_ethdev.c
> +++ b/lib/librte_ethdev/rte_ethdev.c
> @@ -1292,8 +1292,10 @@ rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q,
>   	struct rte_eth_dev *dev;
>   	struct rte_eth_dev_info dev_info;
>   	struct rte_eth_conf orig_conf;
> +	uint16_t overhead_len;
>   	int diag;
>   	int ret;
> +	uint16_t old_mtu;
>   
>   	RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
>   
> @@ -1319,10 +1321,20 @@ rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q,
>   		memcpy(&dev->data->dev_conf, dev_conf,
>   		       sizeof(dev->data->dev_conf));
>   
> +	/* Backup mtu for rollback */
> +	old_mtu = dev->data->mtu;
> +
>   	ret = rte_eth_dev_info_get(port_id, &dev_info);
>   	if (ret != 0)
>   		goto rollback;
>   
> +	/* Get the real Ethernet overhead length */
> +	if (dev_info.max_mtu != UINT16_MAX &&
> +	    dev_info.max_rx_pktlen > dev_info.max_mtu)
> +		overhead_len = dev_info.max_rx_pktlen - dev_info.max_mtu;
> +	else
> +		overhead_len = RTE_ETHER_HDR_LEN + RTE_ETHER_CRC_LEN;
> +
The Ethernet frame header length supported by each NIC may be different, 
which cause the difference of allowed packets size and drop you say.
The diference of Ethernet frame header length have an impact for the 
maximum Ethernet frame length, which is a boundary value
of enabling jumbo frame through the ' max_rx_pkt_len '.
However, we need to do the same thing you do above to get the 
overhead_len every time, which will
cause a lot of duplicate code in the framework and app. For examples, 
parsing and processing for '--max-pkt-len=xxx' parameter,
  and " cmd_config_max_pkt_len_parsed " in testpmd, and here modifying 
here dev_configure API.
It's a little redundant and troublesome.

Maybe, it is necessary for driver to directly report the supported 
maximum Ethernet frame length by rte_dev_info_get API.
As following:
struct rte_eth_dev_info {
	xxxx
	/**
	 * The maximum Ethernet frame length supported by each
	 * driver varies with the Ethernet header length.
	 */
	uint16_t eth_max_len;
	xxxx
}

int
rte_eth_dev_info_get(uint16_t port_id, struct rte_eth_dev_info *dev_info)
{
     xxx
	dev_info->min_mtu = RTE_ETHER_MIN_MTU;
	dev_info->max_mtu = UINT16_MAX;
	dev_info->eth_max_len = RTE_ETHER_MAX_LEN;
     xxx
}

And then:
xxx_devv_infos_get(struct rte_eth_dev *eth_dev, struct rte_eth_dev_info 
*info)
{
	xxx
	info->eth_max_len = xxx_ETHER_MAX _LEN;
	xxx
}
What do you think?
>   	/* If number of queues specified by application for both Rx and Tx is
>   	 * zero, use driver preferred values. This cannot be done individually
>   	 * as it is valid for either Tx or Rx (but not both) to be zero.
> @@ -1410,11 +1422,18 @@ rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q,
>   			goto rollback;
>   		}
>   	} else {
> -		if (dev_conf->rxmode.max_rx_pkt_len < RTE_ETHER_MIN_LEN ||
> -			dev_conf->rxmode.max_rx_pkt_len > RTE_ETHER_MAX_LEN)
> +		uint16_t pktlen = dev_conf->rxmode.max_rx_pkt_len;
> +		if (pktlen < RTE_ETHER_MIN_MTU + overhead_len ||
> +		    pktlen > RTE_ETHER_MTU + overhead_len)
>   			/* Use default value */
>   			dev->data->dev_conf.rxmode.max_rx_pkt_len =
> -							RTE_ETHER_MAX_LEN;
> +						RTE_ETHER_MTU + overhead_len;
> +	}
> +
> +	/* Scale the MTU size to adapt max_rx_pkt_len */
> +	if (dev_conf->rxmode.offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) {
> +		dev->data->mtu = dev->data->dev_conf.rxmode.max_rx_pkt_len -
> +				overhead_len;
>   	}
Now that we update mtu here when jumbo frame enabled. It is necessary to 
check validity of max_rx_pkt_len.
As following:
if (dev_conf->rxmode.offloads & DEV_RX_OFFLOAD_JUMBO_FRAME) {
	if (dev_conf->rxmode.max_rx_pkt_len > dev_info.max_rx_pktlen ||
		dev_conf->rxmode.max_rx_pkt_len <=  RTE_ETHER_MTU + overhead_len) {
			xxxx
		}
} else {
		xxx
}

If it does not thing above, which will cause an unreasonable case.
Like,
dev_conf->rxmode.offloads & DEV_RX_OFFLOAD_JUMBO_FRAME == 1
dev_conf->rxmode.max_rx_pkt_len = 1500
overhead_len = 26
dev->data->mtu = dev_conf->rxmode.max_rx_pkt_len - overhead_len = 1500 - 
26 = 1474

In fact, DEV_RX_OFFLOAD_JUMBO_FRAME is set to rxmode.offloads when mtu > 
1500.
>   
>   	/*
> @@ -1549,6 +1568,8 @@ rte_eth_dev_configure(uint16_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q,
>   	eth_dev_tx_queue_config(dev, 0);
>   rollback:
>   	memcpy(&dev->data->dev_conf, &orig_conf, sizeof(dev->data->dev_conf));
> +	if (old_mtu != dev->data->mtu)
> +		dev->data->mtu = old_mtu;
>   
>   	rte_ethdev_trace_configure(port_id, nb_rx_q, nb_tx_q, dev_conf, ret);
>   	return ret;
> 

  parent reply	other threads:[~2021-01-15 10:44 UTC|newest]

Thread overview: 115+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-12-09  3:16 [dpdk-dev] [PATCH v1 00/12] fix rx packets dropped issue Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 01/12] net/dpaa2: fix the jumbo frame flag condition for mtu set Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 02/12] net/e1000: " Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 03/12] net/hns3: " Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 04/12] net/i40e: fix the jumbo frame flag condition Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 05/12] net/iavf: " Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 06/12] net/ice: " Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 07/12] net/ipn3ke: fix the jumbo frame flag condition for mtu set Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 08/12] net/octeontx: " Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 09/12] net/octeontx2: fix the jumbo frame flag condition for mtu Steve Yang
2020-12-21  7:16   ` [dpdk-dev] [EXT] " Sunil Kumar Kori
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 10/12] net/qede: fix the jumbo frame flag condition for mtu set Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 11/12] net/sfc: " Steve Yang
2020-12-09  3:16 ` [dpdk-dev] [PATCH v1 12/12] net/thunderx: " Steve Yang
2020-12-11  4:31 ` [dpdk-dev] [PATCH v1 00/12] fix rx packets dropped issue Guo, Jia
2020-12-14 17:44 ` Ferruh Yigit
2020-12-17  9:22 ` [dpdk-dev] [PATCH v2 00/22] " Steve Yang
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 01/22] ethdev: fix MTU size exceeds max rx packet length Steve Yang
2020-12-28 14:51     ` Andrew Rybchenko
2021-01-13 10:32       ` Ferruh Yigit
2020-12-30 10:19     ` oulijun
     [not found]       ` <DM6PR11MB43627F10DDAFC9801816FF5BF9D00@DM6PR11MB4362.namprd11.prod.outlook.com>
2021-01-13 10:25         ` Ferruh Yigit
2021-01-13 11:04         ` Ferruh Yigit
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 02/22] app/testpmd: fix max rx packet length for VLAN packets Steve Yang
2021-01-13 11:26     ` Ferruh Yigit
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 03/22] net/dpaa: fix the jumbo frame flag condition for mtu set Steve Yang
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 04/22] net/dpaa2: " Steve Yang
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 05/22] net/e1000: " Steve Yang
2020-12-18  2:42     ` Guo, Jia
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 06/22] net/hns3: " Steve Yang
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 07/22] net/i40e: fix the jumbo frame flag condition Steve Yang
2020-12-18  2:44     ` Guo, Jia
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 08/22] net/iavf: " Steve Yang
2020-12-17  9:22   ` [dpdk-dev] [PATCH v2 09/22] net/ice: " Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 10/22] net/ipn3ke: fix the jumbo frame flag condition for mtu set Steve Yang
2020-12-19  0:54     ` Xu, Rosen
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 11/22] net/octeontx: " Steve Yang
2020-12-21 15:04     ` [dpdk-dev] [EXT] " Harman Kalra
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 12/22] net/octeontx2: fix the jumbo frame flag condition for mtu Steve Yang
2020-12-18 10:15     ` Nithin Dabilpuram
2020-12-21  7:19     ` [dpdk-dev] [EXT] " Sunil Kumar Kori
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 13/22] net/qede: fix the jumbo frame flag condition for mtu set Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 14/22] net/sfc: " Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 15/22] net/thunderx: " Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 16/22] net/ixgbe: fix the jumbo frame flag condition Steve Yang
2020-12-18  2:43     ` Guo, Jia
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 17/22] net/cxgbe: " Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 18/22] net/axgbe: fix the jumbo frame flag condition for mtu set Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 19/22] net/enetc: " Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 20/22] net/hinic: " Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 21/22] net/nfp: " Steve Yang
2020-12-17  9:23   ` [dpdk-dev] [PATCH v2 22/22] net/liquidio: " Steve Yang
2021-01-13 11:32   ` [dpdk-dev] [PATCH v2 00/22] fix rx packets dropped issue Ferruh Yigit
2021-01-14  9:45   ` [dpdk-dev] [PATCH v3 " Steve Yang
2021-01-14  9:45   ` [dpdk-dev] [PATCH v3 01/22] ethdev: fix MTU size exceeds max rx packet length Steve Yang
2021-01-14 16:36     ` Ferruh Yigit
2021-01-14 17:13       ` Ferruh Yigit
2021-01-14 17:29         ` Andrew Boyer
2021-01-14 20:44           ` Ferruh Yigit
2021-01-15 10:44     ` oulijun [this message]
2021-01-18 10:42       ` Ferruh Yigit
2021-01-19  8:46         ` oulijun
2021-01-23  9:05           ` oulijun
2021-01-25 12:22           ` Ferruh Yigit
2021-01-18  7:04     ` [dpdk-dev] [PATCH v4 00/22] fix rx packets dropped issue Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 01/22] ethdev: fix MTU size exceeds max rx packet length Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 02/22] app/testpmd: fix max rx packet length for VLAN packets Steve Yang
2021-01-21 15:27         ` Lance Richardson
2021-01-22 14:26           ` Lance Richardson
2021-01-25 12:14           ` Ferruh Yigit
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 03/22] net/dpaa: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 04/22] net/dpaa2: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 05/22] net/e1000: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 06/22] net/hns3: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 07/22] net/i40e: fix the jumbo frame flag condition Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 08/22] net/iavf: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 09/22] net/ice: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 10/22] net/ipn3ke: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 11/22] net/octeontx: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 12/22] net/octeontx2: fix the jumbo frame flag condition for mtu Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 13/22] net/qede: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 14/22] net/sfc: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 15/22] net/thunderx: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 16/22] net/ixgbe: fix the jumbo frame flag condition Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 17/22] net/cxgbe: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 18/22] net/axgbe: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 19/22] net/enetc: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 20/22] net/hinic: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 21/22] net/nfp: " Steve Yang
2021-01-18  7:04       ` [dpdk-dev] [PATCH v4 22/22] net/liquidio: " Steve Yang
2021-01-18 11:54       ` [dpdk-dev] [PATCH v4 00/22] fix rx packets dropped issue Ferruh Yigit
2021-01-14  9:45   ` [dpdk-dev] [PATCH v3 02/22] app/testpmd: fix max rx packet length for VLAN packets Steve Yang
2021-01-14  9:45   ` [dpdk-dev] [PATCH v3 03/22] net/dpaa: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-14 10:26     ` Hemant Agrawal
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 04/22] net/dpaa2: " Steve Yang
2021-01-14 10:27     ` Hemant Agrawal
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 05/22] net/e1000: " Steve Yang
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 06/22] net/hns3: " Steve Yang
2021-01-18  1:30     ` oulijun
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 07/22] net/i40e: fix the jumbo frame flag condition Steve Yang
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 08/22] net/iavf: " Steve Yang
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 09/22] net/ice: " Steve Yang
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 10/22] net/ipn3ke: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 11/22] net/octeontx: " Steve Yang
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 12/22] net/octeontx2: fix the jumbo frame flag condition for mtu Steve Yang
2021-01-14  9:46   ` [dpdk-dev] [PATCH v3 13/22] net/qede: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 14/22] net/sfc: " Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 15/22] net/thunderx: " Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 16/22] net/ixgbe: fix the jumbo frame flag condition Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 17/22] net/cxgbe: " Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 18/22] net/axgbe: fix the jumbo frame flag condition for mtu set Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 19/22] net/enetc: " Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 20/22] net/hinic: " Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 21/22] net/nfp: " Steve Yang
2021-01-14  9:47   ` [dpdk-dev] [PATCH v3 22/22] net/liquidio: " Steve Yang

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=c79413fe-06ab-a2eb-96ec-d55d968fd2b2@huawei.com \
    --to=oulijun@huawei.com \
    --cc=andrew.rybchenko@oktetlabs.ru \
    --cc=dev@dpdk.org \
    --cc=ferruh.yigit@intel.com \
    --cc=linuxarm@openeuler.org \
    --cc=stevex.yang@intel.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

DPDK patches and discussions

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://inbox.dpdk.org/dev/0 dev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 dev dev/ https://inbox.dpdk.org/dev \
		dev@dpdk.org
	public-inbox-index dev

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.dev


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git