Re: [dpdk-dev] [PATCH v2] distributor_app: new sample app

["Ananyev, Konstantin" <konstantin.ananyev@intel.com>]

> -----Original Message-----
> From: Pattan, Reshma
> Sent: Tuesday, September 30, 2014 9:03 AM
> To: Ananyev, Konstantin; De Lara Guarch, Pablo; dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH v2] distributor_app: new sample app
> 
> Hi Konstantin,
> 
> Any comments on below Pablos comment? If so please provide.
> 
> Thanks,
> Reshma
> 

Hi Reshma,

No really.
I would just change printf to what I suggested.
For your app, I believe that is more than enough.
But if you'd like to introduce some sort of rate limiting for logging -
sure go ahead would be interesting to see that patch.
Konstantin

> 
> 
> -----Original Message-----
> From: De Lara Guarch, Pablo
> Sent: Monday, September 29, 2014 2:35 PM
> To: Ananyev, Konstantin; Pattan, Reshma; dev@dpdk.org
> Subject: RE: [dpdk-dev] [PATCH v2] distributor_app: new sample app
> 
> 
> 
> > -----Original Message-----
> > From: Ananyev, Konstantin
> > Sent: Monday, September 29, 2014 2:07 PM
> > To: Pattan, Reshma; De Lara Guarch, Pablo; dev@dpdk.org
> > Subject: RE: [dpdk-dev] [PATCH v2] distributor_app: new sample app
> >
> >
> >
> > > -----Original Message-----
> > > From: Pattan, Reshma
> > > Sent: Monday, September 29, 2014 1:40 PM
> > > To: Ananyev, Konstantin; De Lara Guarch, Pablo; dev@dpdk.org
> > > Subject: RE: [dpdk-dev] [PATCH v2] distributor_app: new sample app
> > >
> > >
> > >
> > > -----Original Message-----
> > > From: Ananyev, Konstantin
> > > Sent: Friday, September 26, 2014 4:52 PM
> > > To: De Lara Guarch, Pablo; Pattan, Reshma; dev@dpdk.org
> > > Subject: RE: [dpdk-dev] [PATCH v2] distributor_app: new sample app
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of De Lara
> > > > Guarch, Pablo
> > > > Sent: Friday, September 26, 2014 4:12 PM
> > > > To: Pattan, Reshma; dev@dpdk.org
> > > > Subject: Re: [dpdk-dev] [PATCH v2] distributor_app: new sample app
> > > >
> > > > Hi,
> > > >
> > > > > -----Original Message-----
> > > > > From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of reshmapa
> > > > > Sent: Wednesday, September 24, 2014 3:17 PM
> > > > > To: dev@dpdk.org
> > > > > Subject: [dpdk-dev] [PATCH v2] distributor_app: new sample app
> > > > >
> > > > > From: Reshma Pattan <reshma.pattan@intel.com>
> > > > >
> > > > > A new sample app that shows the usage of the distributor library.
> > > > > This app works as follows:
> > > > >
> > > > > * An RX thread runs which pulls packets from each ethernet port in turn
> > > > >   and passes those packets to worker using a distributor component.
> > > > > * The workers take the packets in turn, and determine the output port
> > > > >   for those packets using basic l2forwarding doing an xor on the source
> > > > >   port id.
> > > > > * The RX thread takes the returned packets from the workers and
> > enqueue
> > > > >   those packets into an rte_ring structure.
> > > > > * A TX thread pulls the packets off the rte_ring structure and then
> > > > >   sends each packet out the output port specified previously by
> > > > > the worker
> > > > > * Command-line option support provided only for portmask.
> > > > >
> > > > > Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
> > > > > Signed-off-by: Reshma Pattan <reshma.pattan@intel.com>
> > > > > ---
> > > > >  examples/Makefile                 |   1 +
> > > > >  examples/distributor_app/Makefile |  57 ++++
> > > > >  examples/distributor_app/main.c   | 585
> > > > > ++++++++++++++++++++++++++++++++++++++
> > > > >  examples/distributor_app/main.h   |  46 +++
> > > > >  4 files changed, 689 insertions(+)  create mode 100644
> > > > > examples/distributor_app/Makefile  create mode
> > > > > 100644 examples/distributor_app/main.c  create mode 100644
> > > > > examples/distributor_app/main.h
> > > > >
> > > > > diff --git a/examples/Makefile b/examples/Makefile index
> > > > > 6245f83..2ba82b0 100644
> > > > > --- a/examples/Makefile
> > > > > +++ b/examples/Makefile
> > > > > @@ -66,5 +66,6 @@ DIRS-y += vhost
> > > > >  DIRS-$(CONFIG_RTE_LIBRTE_XEN_DOM0) += vhost_xen  DIRS-y +=
> > vmdq
> > > > > DIRS-y += vmdq_dcb
> > > > > +DIRS-$(CONFIG_RTE_LIBRTE_DISTRIBUTOR) += distributor_app
> > > > >
> > > > >  include $(RTE_SDK)/mk/rte.extsubdir.mk diff --git
> > > > > a/examples/distributor_app/Makefile
> > > > > b/examples/distributor_app/Makefile
> > > > > new file mode 100644
> > > > > index 0000000..394785d
> > > > > --- /dev/null
> > > > > +++ b/examples/distributor_app/Makefile
> > > > > @@ -0,0 +1,57 @@
> > > > > +#   BSD LICENSE
> > > > > +#
> > > > > +#   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> > > > > +#   All rights reserved.
> > > > > +#
> > > > > +#   Redistribution and use in source and binary forms, with or without
> > > > > +#   modification, are permitted provided that the following conditions
> > > > > +#   are met:
> > > > > +#
> > > > > +#     * Redistributions of source code must retain the above copyright
> > > > > +#       notice, this list of conditions and the following disclaimer.
> > > > > +#     * Redistributions in binary form must reproduce the above
> > copyright
> > > > > +#       notice, this list of conditions and the following disclaimer in
> > > > > +#       the documentation and/or other materials provided with the
> > > > > +#       distribution.
> > > > > +#     * Neither the name of Intel Corporation nor the names of its
> > > > > +#       contributors may be used to endorse or promote products
> > derived
> > > > > +#       from this software without specific prior written permission.
> > > > > +#
> > > > > +#   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
> > > > > CONTRIBUTORS
> > > > > +#   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
> > BUT
> > > > > NOT
> > > > > +#   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
> > > > > FITNESS FOR
> > > > > +#   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> > > > > COPYRIGHT
> > > > > +#   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> > > > > INCIDENTAL,
> > > > > +#   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
> > BUT
> > > > > NOT
> > > > > +#   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
> > LOSS
> > > > > OF USE,
> > > > > +#   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
> > CAUSED
> > > > > AND ON ANY
> > > > > +#   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
> > > > > TORT
> > > > > +#   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
> > OUT OF
> > > > > THE USE
> > > > > +#   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
> > > > > DAMAGE.
> > > > > +
> > > > > +ifeq ($(RTE_SDK),)
> > > > > +$(error "Please define RTE_SDK environment variable") endif
> > > > > +
> > > > > +# Default target, can be overriden by command line or
> > > > > +environment RTE_TARGET ?= x86_64-default-linuxapp-gcc
> > > >
> > > > This target is not present anymore. Change it to
> > > > x86_64-native-linuxapp-
> > gcc.
> > > >
> > > > > +
> > > > > +include $(RTE_SDK)/mk/rte.vars.mk
> > > > > +
> > > > > +# binary name
> > > > > +APP = distributor_app
> > > > > +
> > > > > +# all source are stored in SRCS-y SRCS-y := main.c
> > > > > +
> > > > > +CFLAGS += $(WERROR_FLAGS)
> > > > > +
> > > > > +# workaround for a gcc bug with noreturn attribute #
> > > > > +http://gcc.gnu.org/bugzilla/show_bug.cgi?id=12603
> > > > > +ifeq ($(CONFIG_RTE_TOOLCHAIN_GCC),y) CFLAGS_main.o +=
> > > > > +-Wno-return-type endif
> > > > > +
> > > > > +EXTRA_CFLAGS += -O3 -Wfatal-errors
> > > > > +
> > > > > +include $(RTE_SDK)/mk/rte.extapp.mk
> > > > > diff --git a/examples/distributor_app/main.c
> > > > > b/examples/distributor_app/main.c new file mode 100644 index
> > > > > 0000000..628810a
> > > > > --- /dev/null
> > > > > +++ b/examples/distributor_app/main.c
> > > > > @@ -0,0 +1,585 @@
> > > > > +/*-
> > > > > + *   BSD LICENSE
> > > > > + *
> > > > > + *   Copyright(c) 2010-2014 Intel Corporation. All rights reserved.
> > > > > + *   All rights reserved.
> > > > > + *
> > > > > + *   Redistribution and use in source and binary forms, with or without
> > > > > + *   modification, are permitted provided that the following conditions
> > > > > + *   are met:
> > > > > + *
> > > > > + *     * Redistributions of source code must retain the above copyright
> > > > > + *       notice, this list of conditions and the following disclaimer.
> > > > > + *     * Redistributions in binary form must reproduce the above
> > copyright
> > > > > + *       notice, this list of conditions and the following disclaimer in
> > > > > + *       the documentation and/or other materials provided with the
> > > > > + *       distribution.
> > > > > + *     * Neither the name of Intel Corporation nor the names of its
> > > > > + *       contributors may be used to endorse or promote products
> > derived
> > > > > + *       from this software without specific prior written permission.
> > > > > + *
> > > > > + *   THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
> > > > > CONTRIBUTORS
> > > > > + *   "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING,
> > BUT
> > > > > NOT
> > > > > + *   LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
> > > > > FITNESS FOR
> > > > > + *   A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> > > > > COPYRIGHT
> > > > > + *   OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> > > > > INCIDENTAL,
> > > > > + *   SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
> > (INCLUDING, BUT
> > > > > NOT
> > > > > + *   LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
> > LOSS
> > > > > OF USE,
> > > > > + *   DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
> > CAUSED
> > > > > AND ON ANY
> > > > > + *   THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
> > OR
> > > > > TORT
> > > > > + *   (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
> > OUT OF
> > > > > THE USE
> > > > > + *   OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH
> > > > > DAMAGE.
> > > > > + */
> > > > > +
> > > > > +#include <stdint.h>
> > > > > +#include <inttypes.h>
> > > > > +#include <unistd.h>
> > > > > +#include <signal.h>
> > > > > +#include <getopt.h>
> > > > > +
> > > > > +#include <rte_eal.h>
> > > > > +#include <rte_ethdev.h>
> > > > > +#include <rte_cycles.h>
> > > > > +#include <rte_malloc.h>
> > > > > +#include <rte_debug.h>
> > > > > +#include <rte_distributor.h>
> > > > > +
> > > > > +#include "main.h"
> > > > > +
> > > > > +#define RX_RING_SIZE 256
> > > > > +#define RX_FREE_THRESH 32
> > > > > +#define RX_PTHRESH 8
> > > > > +#define RX_HTHRESH 8
> > > > > +#define RX_WTHRESH 0
> > > > > +
> > > > > +#define TX_RING_SIZE 512
> > > > > +#define TX_FREE_THRESH 32
> > > > > +#define TX_PTHRESH 32
> > > > > +#define TX_HTHRESH 0
> > > > > +#define TX_WTHRESH 0
> > > > > +#define TX_RSBIT_THRESH 32
> > > > > +#define TX_Q_FLAGS (ETH_TXQ_FLAGS_NOMULTSEGS |
> > > > > ETH_TXQ_FLAGS_NOVLANOFFL |\
> > > > > +	ETH_TXQ_FLAGS_NOXSUMSCTP | ETH_TXQ_FLAGS_NOXSUMUDP |
> > > > > \
> > > > > +	ETH_TXQ_FLAGS_NOXSUMTCP)
> > > > > +
> > > > > +#define NUM_MBUFS ((64*1024)-1) #define MBUF_SIZE (2048 +
> > > > > +sizeof(struct rte_mbuf) +
> > > > > RTE_PKTMBUF_HEADROOM)
> > > > > +#define MBUF_CACHE_SIZE 250
> > > > > +#define BURST_SIZE 32
> > > > > +#define RTE_RING_SZ 1024
> > > > > +
> > > > > +/* mask of enabled ports */
> > > > > +static uint32_t enabled_port_mask = 0;
> > > > > +
> > > > > +static volatile struct app_stats {
> > > > > +	struct {
> > > > > +		uint64_t rx_pkts;
> > > > > +		uint64_t returned_pkts;
> > > > > +		uint64_t enqueued_pkts;
> > > > > +	} rx __rte_cache_aligned;
> > > > > +
> > > > > +	struct {
> > > > > +		uint64_t dequeue_pkts;
> > > > > +		uint64_t tx_pkts;
> > > > > +	} tx __rte_cache_aligned;
> > > > > +} app_stats;
> > > > > +
> > > > > +static const struct rte_eth_conf port_conf_default = {
> > > > > +	.rxmode = {
> > > > > +		.mq_mode = ETH_MQ_RX_RSS,
> > > > > +		.max_rx_pkt_len = ETHER_MAX_LEN,
> > > > > +		.split_hdr_size = 0,
> > > > > +		.header_split   = 0, /**< Header Split disabled */
> > > > > +		.hw_ip_checksum = 0, /**< IP checksum offload enabled */
> > > > > +		.hw_vlan_filter = 0, /**< VLAN filtering disabled */
> > > > > +		.jumbo_frame    = 0, /**< Jumbo Frame Support disabled */
> > > > > +		.hw_strip_crc   = 0, /**< CRC stripped by hardware */
> > > > > +	},
> > > > > +	.txmode = {
> > > > > +		.mq_mode = ETH_MQ_TX_NONE,
> > > > > +	},
> > > > > +	.lpbk_mode = 0,
> > > > > +	.rx_adv_conf = {
> > > > > +			.rss_conf = {
> > > > > +				.rss_hf = ETH_RSS_IPV4 | ETH_RSS_IPV6 |
> > > > > +					ETH_RSS_IPV4_TCP |
> > > > > ETH_RSS_IPV4_UDP |
> > > > > +					ETH_RSS_IPV6_TCP |
> > > > > ETH_RSS_IPV6_UDP,
> > > > > +			}
> > > > > +	},
> > > > > +};
> > > > > +
> > > > > +static const struct rte_eth_rxconf rx_conf_default = {
> > > > > +	.rx_thresh = {
> > > > > +		.pthresh = RX_PTHRESH,
> > > > > +		.hthresh = RX_HTHRESH,
> > > > > +		.wthresh = RX_WTHRESH,
> > > > > +	},
> > > > > +	.rx_free_thresh = RX_FREE_THRESH,
> > > > > +	.rx_drop_en = 0,
> > > > > +};
> > > > > +
> > > > > +static const struct rte_eth_txconf tx_conf_default = {
> > > > > +	.tx_thresh = {
> > > > > +		.pthresh = TX_PTHRESH,
> > > > > +		.hthresh = TX_HTHRESH,
> > > > > +		.wthresh = TX_WTHRESH,
> > > > > +	},
> > > > > +	.tx_free_thresh = TX_FREE_THRESH,
> > > > > +	.tx_rs_thresh = TX_RSBIT_THRESH,
> > > > > +	.txq_flags = TX_Q_FLAGS
> > > > > +
> > > > > +};
> > > > > +
> > > > > +struct output_buffer {
> > > > > +	unsigned count;
> > > > > +	struct rte_mbuf *mbufs[BURST_SIZE]; };
> > > > > +
> > > > > +/*
> > > > > + * Initialises a given port using global settings and with the
> > > > > +rx buffers
> > > > > + * coming from the mbuf_pool passed as parameter  */ static
> > > > > +inline int port_init(uint8_t port, struct rte_mempool *mbuf_pool) {
> > > > > +	struct rte_eth_conf port_conf = port_conf_default;
> > > > > +	const uint16_t rxRings = 1, txRings = rte_lcore_count() - 1;
> > > > > +	int retval;
> > > > > +	uint16_t q;
> > > > > +
> > > > > +	if (port >= rte_eth_dev_count())
> > > > > +		return -1;
> > > > > +
> > > > > +	retval = rte_eth_dev_configure(port, rxRings, txRings, &port_conf);
> > > > > +	if (retval != 0)
> > > > > +		return retval;
> > > > > +
> > > > > +	for (q = 0; q < rxRings; q++) {
> > > > > +		retval = rte_eth_rx_queue_setup(port, q, RX_RING_SIZE,
> > > > > +
> > > > > 	rte_eth_dev_socket_id(port),
> > > > > +						&rx_conf_default,
> > > > > mbuf_pool);
> > > > > +		if (retval < 0)
> > > > > +			return retval;
> > > > > +	}
> > > > > +
> > > > > +	for (q = 0; q < txRings; q++) {
> > > > > +		retval = rte_eth_tx_queue_setup(port, q, TX_RING_SIZE,
> > > > > +
> > > > > 	rte_eth_dev_socket_id(port),
> > > > > +						&tx_conf_default);
> > > > > +		if (retval < 0)
> > > > > +			return retval;
> > > > > +	}
> > > > > +
> > > > > +	retval  = rte_eth_dev_start(port);
> > > > > +	if (retval < 0)
> > > > > +		return retval;
> > > > > +
> > > > > +	struct rte_eth_link link;
> > > > > +	rte_eth_link_get_nowait(port, &link);
> > > > > +	if (!link.link_status) {
> > > > > +		sleep(1);
> > > > > +		rte_eth_link_get_nowait(port, &link);
> > > > > +	}
> > > > > +
> > > > > +	if (!link.link_status) {
> > > > > +		printf("Link down on port %"PRIu8"\n", port);
> > > > > +		return 0;
> > > > > +	}
> > > > > +
> > > > > +	struct ether_addr addr;
> > > > > +	rte_eth_macaddr_get(port, &addr);
> > > > > +	printf("Port %u MAC: %02"PRIx8" %02"PRIx8" %02"PRIx8
> > > > > +			" %02"PRIx8" %02"PRIx8" %02"PRIx8"\n",
> > > > > +			(unsigned)port,
> > > > > +			addr.addr_bytes[0], addr.addr_bytes[1],
> > > > > +			addr.addr_bytes[2], addr.addr_bytes[3],
> > > > > +			addr.addr_bytes[4], addr.addr_bytes[5]);
> > > > > +
> > > > > +	rte_eth_promiscuous_enable(port);
> > > > > +
> > > > > +	return 0;
> > > > > +}
> > > > > +
> > > > > +struct lcore_params {
> > > > > +	unsigned worker_id;
> > > > > +	struct rte_distributor *d;
> > > > > +	struct rte_ring *r;
> > > > > +};
> > > > > +
> > > > > +static __attribute__((noreturn)) void lcore_rx(struct
> > > > > +lcore_params
> > > > > +*p) {
> > > > > +	struct rte_distributor *d = p->d;
> > > > > +	struct rte_ring *r = p->r;
> > > > > +	const uint8_t nb_ports = rte_eth_dev_count();
> > > > > +	const int socket_id = rte_socket_id();
> > > > > +	uint8_t port;
> > > > > +
> > > > > +	for (port = 0; port < nb_ports; port++) {
> > > > > +		/* skip ports that are not enabled */
> > > > > +		if ((enabled_port_mask & (1 << port)) == 0)
> > > > > +			continue;
> > > > > +
> > > > > +		if (rte_eth_dev_socket_id(port) > 0 &&
> > > > > +				rte_eth_dev_socket_id(port) != socket_id)
> > > > > +			printf("WARNING, port %u is on remote NUMA node
> > > > > to "
> > > > > +					"RX thread.\n\tPerformance will not
> > > > > "
> > > > > +					"be optimal.\n", port);
> > > > > +	}
> > > > > +
> > > > > +	printf("\nCore %u doing packet RX.\n", rte_lcore_id());
> > > > > +	port = 0;
> > > > > +	for (;;) {
> > > > > +		/* skip ports that are not enabled */
> > > > > +		if ((enabled_port_mask & (1 << port)) == 0) {
> > > > > +			if (++port == nb_ports)
> > > > > +				port = 0;
> > > > > +			continue;
> > > > > +		}
> > > > > +		struct rte_mbuf *bufs[BURST_SIZE*2];
> > > > > +		const uint16_t nb_rx = rte_eth_rx_burst(port, 0, bufs,
> > > > > +				BURST_SIZE);
> > > > > +		app_stats.rx.rx_pkts += nb_rx;
> > > > > +
> > > > > +		rte_distributor_process(d, bufs, nb_rx);
> > > > > +		const uint16_t nb_ret = rte_distributor_returned_pkts(d,
> > > > > +				bufs, BURST_SIZE*2);
> > > > > +		app_stats.rx.returned_pkts += nb_ret;
> > > > > +		if (unlikely(nb_ret == 0))
> > > > > +			continue;
> > > > > +
> > > > > +		uint16_t sent = rte_ring_enqueue_burst(r, (void *)bufs,
> > > > > nb_ret);
> > > > > +		app_stats.rx.enqueued_pkts += sent;
> > > > > +		if (unlikely(sent < nb_ret)) {
> > > > > +			printf("Packet loss due to full ring\n");
> > > >
> > > > I would remove this printf. If core is too busy to process
> > > > packets, this may
> > make performance worse.
> > > > Statistics should reflect this packet loss.
> > >
> > > >Probably use RTE_LOG(DEBUG, ...).
> > > >That's way you can switch it on/off at compile time.
> > >
> > > By default  LOG_LEVEL is 8 in config/common_linuxapp, hence I see
> > RTE_LOG(DEBUG,...) always works.
> > > So until unless change the LOG-LEVEL to lesser value,
> > > RTE_LOG(DEBUG,..)
> > will be logged ..
> > > Please let me know if I can go ahead and remove the printf as of now
> > > until
> > the LOG_LEVEL is fixed. (or) I have to add some logic to log
> > > RTE_LOG(DEBUG,..)  only once.
> >
> > Sorry, I probably wasn't clear enough.
> > What I am saying -  create something like that inside your app code:
> >
> > #ifdef XXX_DEBUG
> > #define XXX_LOG(level, fmt, args...) \
> >         RTE_LOG(level, PMD, "%s(): " fmt "\n", __func__, ## args)
> > #else #define XXX_LOG(level, fmt, args...) do { } while(0) #endif
> >
> > And replace your pritnf()s with it.
> 
> I think we should make some changes in some logs in DPDK, and change the default log level to be INFO or less, so we could use the
> DEBUG level for things like this, but anyway, that would be another patch that I am thinking to submit some other day.
> Anyway, for this one, this option could work, but still that message may be shown hundreds of thousands of times per second,
> considering that CPU is overloaded. So, should we include something like a flag here, so message is shown just once?
> I know that this will only be used when DEBUG is enabled, but still that could flood the screen, as it is in an infinite loop.
> 
> >
> > >
> > > >
> > > > > +			while (sent < nb_ret)
> > > > > +				rte_pktmbuf_free(bufs[sent++]);
> > > > > +		}
> > > > > +		if (++port == nb_ports)
> > > > > +			port = 0;
> > > > > +	}
> > > > > +}
> > > > > +
> > > > > +static inline void
> > > > > +flush_one_port(struct output_buffer *outbuf, uint8_t outp) {
> > > > > +	unsigned nb_tx = rte_eth_tx_burst(outp, 0, outbuf->mbufs,
> > > > > +			outbuf->count);
> > > > > +	app_stats.tx.tx_pkts += nb_tx;
> > > > > +
> > > > > +	if (unlikely(nb_tx < outbuf->count)) {
> > > > > +		printf("Packet loss with tx_burst\n");
> > > >
> > > > Same here. This is fine for debugging, but statistics are updated
> > > > as well,
> > so we do not need this printf.
> > > >
> > > > > +		do {
> > > > > +			rte_pktmbuf_free(outbuf->mbufs[nb_tx]);
> > > > > +		} while (++nb_tx < outbuf->count);
> > > > > +	}
> > > > > +	outbuf->count = 0;
> > > > > +}
> > > > > +