DPDK patches and discussions
 help / color / mirror / Atom feed
From: Ferruh Yigit <ferruh.yigit@intel.com>
To: Vadim Podovinnikov <podovinnikov@protei.ru>, linville@tuxdriver.com
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH v4] add drop statistic for af_packet
Date: Mon, 2 Dec 2019 09:40:02 +0000	[thread overview]
Message-ID: <7de9f57d-c7c4-f40c-b93b-7372a990f3d2@intel.com> (raw)
In-Reply-To: <20191129145925.26628-1-podovinnikov@protei.ru>

On 11/29/2019 2:59 PM, Vadim Podovinnikov wrote:
> Signed-off-by: Vadim Podovinnikov <podovinnikov@protei.ru>
> ---
>  drivers/net/af_packet/rte_eth_af_packet.c | 33 +++++++++++++++++------
>  1 file changed, 25 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/net/af_packet/rte_eth_af_packet.c b/drivers/net/af_packet/rte_eth_af_packet.c
> index eee0fbce2..2aa7c0fcc 100644
> --- a/drivers/net/af_packet/rte_eth_af_packet.c
> +++ b/drivers/net/af_packet/rte_eth_af_packet.c
> @@ -52,6 +52,7 @@ struct pkt_rx_queue {
>  
>  	volatile unsigned long rx_pkts;
>  	volatile unsigned long rx_bytes;
> +	volatile unsigned long rx_drop;
>  };
>  
>  struct pkt_tx_queue {
> @@ -322,6 +323,25 @@ eth_dev_info(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
>  	return 0;
>  }
>  
> +static void
> +fill_eth_drop_stats(struct rte_eth_dev *dev)
> +{
> +	unsigned int i, imax;
> +	struct pmd_internals *internal = dev->data->dev_private;
> +	socklen_t sock_len = sizeof(struct tpacket_stats);
> +	struct tpacket_stats st;
> +
> +	imax = (internal->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS ?
> +			internal->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS);
> +	for (i = 0; i < imax; i++) {
> +		memset(&st, 0, sock_len);
> +		int rc = getsockopt(internal->rx_queue[i].sockfd, SOL_PACKET,
> +					PACKET_STATISTICS, &st, &sock_len);
> +		if (rc == 0)
> +			internal->rx_queue[i].rx_drop += st.tp_drops;
> +	}
> +}
> +
>  static int
>  eth_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *igb_stats)
>  {
> @@ -329,22 +349,18 @@ eth_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *igb_stats)
>  	unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
>  	unsigned long rx_bytes_total = 0, tx_bytes_total = 0, rx_drop = 0;
>  	const struct pmd_internals *internal = dev->data->dev_private;
> -	socklen_t sock_len = sizeof(struct tpacket_stats);
> -	struct tpacket_stats st;
> +
> +	fill_eth_drop_stats(dev);

There is already a loop per queue, instead of maxing a function for all queues
and duplicate 'imax' calculation, what about making a function that gets the
queue_id as parameter and called in the blow loop?

>  
>  	imax = (internal->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS ?
>  	        internal->nb_queues : RTE_ETHDEV_QUEUE_STAT_CNTRS);
>  	for (i = 0; i < imax; i++) {
>  		igb_stats->q_ipackets[i] = internal->rx_queue[i].rx_pkts;
>  		igb_stats->q_ibytes[i] = internal->rx_queue[i].rx_bytes;
> +		igb_stats->q_errors[i] = internal->rx_queue[i].rx_drop;

'q_errors' is for the packets that are received but erroneous, what is the
af_packet 'tp_drops' stats for? Is it for wrong packets or the packets dropped
because receive rate was too high?

>  		rx_total += igb_stats->q_ipackets[i];
>  		rx_bytes_total += igb_stats->q_ibytes[i];
> -
> -		memset(&st, 0, sock_len);
> -		int rc = getsockopt(internal->rx_queue[i].sockfd, SOL_PACKET,
> -					PACKET_STATISTICS, &st, &sock_len);
> -		if (rc == 0)
> -			rx_drop += st.tp_drops;

The patch looks like on top of your previous version, please make new versions
always on top of latest head, so not incremental patches but from scratch each time.

> +		rx_drop += igb_stats->q_errors[i];

What are you doing with 'rx_drop'? Not using it at all?

>  	}
>  
>  	imax = (internal->nb_queues < RTE_ETHDEV_QUEUE_STAT_CNTRS ?
> @@ -375,6 +391,7 @@ eth_stats_reset(struct rte_eth_dev *dev)
>  	for (i = 0; i < internal->nb_queues; i++) {
>  		internal->rx_queue[i].rx_pkts = 0;
>  		internal->rx_queue[i].rx_bytes = 0;
> +		internal->rx_queue[i].rx_drop = 0;
>  	}
>  
>  	for (i = 0; i < internal->nb_queues; i++) {
> 


      reply	other threads:[~2019-12-02  9:40 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20191125180325.2E2284C9D@dpdk.org>
2019-11-26 14:32 ` [dpdk-dev] [PATCH v2] " Vadim
2019-11-26 14:45   ` Ferruh Yigit
2019-11-26 15:47     ` [dpdk-dev] [PATCH v3] " Vadim Podovinnikov
2019-11-28 15:03       ` Ferruh Yigit
2019-11-29 14:59         ` [dpdk-dev] [PATCH v4] " Vadim Podovinnikov
2019-12-02  9:40           ` Ferruh Yigit [this message]

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=7de9f57d-c7c4-f40c-b93b-7372a990f3d2@intel.com \
    --to=ferruh.yigit@intel.com \
    --cc=dev@dpdk.org \
    --cc=linville@tuxdriver.com \
    --cc=podovinnikov@protei.ru \
    /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).