DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] [RFC] PMD for performance measurement
@ 2014-09-19 12:27 mukawa
  2014-09-19 12:27 ` [dpdk-dev] [RFC] librte_pmd_null: Add null PMD mukawa
  0 siblings, 1 reply; 4+ messages in thread
From: mukawa @ 2014-09-19 12:27 UTC (permalink / raw)
  To: dev; +Cc: nakajima.yoshihiro, masutani.hitoshi

From: Tetsuya Mukawa <mukawa@igel.co.jp>

Hi,

I want to measure throughputs like following cases.

- path connected by RING PMDs
- path connected by librte_vhost and virtio-net PMD
- path connected by MEMNIC PMDs
- .....

Is there anyone want to do same thing?

Anyway, I guess those throughputs may be too high for some devices like ixia.
But it's a bit pain to write or fix applications just for measuring.

I guess a PMD like '/dev/null' and testpmd application will help.
This patch set is RFC of a PMD like '/dev/null'.
Please see the first commit of this patch set.


Here is a my plan to use this PMD

+-------------------------------+
|           testpmd1            |
+-------------+------+----------+
| Target PMD1 |      | null PMD |
+---++--------+      +----------+
    ||
    || Target path
    ||
+---++--------+      +----------+
| Target PMD2 |      | null PMD |
+-------------+------+----------+
|           testpmd2            |
+-------------------------------+

The testpmd1 or testpmd2 will start with "start tx_first". It causes huge
transfers.

The result is not thuroughput of PMD1 or PMD2, but throughput
between PMD1 and PMD2. But I guess it's enough to know rough thuroughput.
Also it's nice for measuing that the same environment can be used.

Any suggestions or comments?

Thanks,
Tetsuya

--
Tetsuya Mukawa (1):
  librte_pmd_null: Add null PMD

 config/common_bsdapp               |   5 +
 config/common_linuxapp             |   5 +
 lib/Makefile                       |   1 +
 lib/librte_pmd_null/Makefile       |  58 +++++
 lib/librte_pmd_null/rte_eth_null.c | 474 +++++++++++++++++++++++++++++++++++++
 5 files changed, 543 insertions(+)
 create mode 100644 lib/librte_pmd_null/Makefile
 create mode 100644 lib/librte_pmd_null/rte_eth_null.c

-- 
1.9.1

^ permalink raw reply	[flat|nested] 4+ messages in thread

* [dpdk-dev] [RFC] librte_pmd_null: Add null PMD
  2014-09-19 12:27 [dpdk-dev] [RFC] PMD for performance measurement mukawa
@ 2014-09-19 12:27 ` mukawa
  2014-09-22  7:46   ` Ivan Boule
  0 siblings, 1 reply; 4+ messages in thread
From: mukawa @ 2014-09-19 12:27 UTC (permalink / raw)
  To: dev; +Cc: nakajima.yoshihiro, masutani.hitoshi

From: Tetsuya Mukawa <mukawa@igel.co.jp>

'null PMD' is a virtual device driver particulary designed to measure
performance of DPDK applications and DPDK PMDs. When an application call rx,
null PMD just allocate mbufs and return those. Also tx, the PMD just free
mbufs.

The PMD has following options.
- size: specify packe size allocated by RX. Default packet size is 64.
- copy: specify 1 or 0 to enable or disable copy while RX and TX.
	Default value is 0(disbaled).
	This option is used for emulating more realistic data transfer.
	Copy size is equal to packet size.

Signed-off-by: Tetsuya Mukawa <mukawa@igel.co.jp>
---
 config/common_bsdapp               |   5 +
 config/common_linuxapp             |   5 +
 lib/Makefile                       |   1 +
 lib/librte_pmd_null/Makefile       |  58 +++++
 lib/librte_pmd_null/rte_eth_null.c | 474 +++++++++++++++++++++++++++++++++++++
 5 files changed, 543 insertions(+)
 create mode 100644 lib/librte_pmd_null/Makefile
 create mode 100644 lib/librte_pmd_null/rte_eth_null.c

diff --git a/config/common_bsdapp b/config/common_bsdapp
index 645949f..a86321f 100644
--- a/config/common_bsdapp
+++ b/config/common_bsdapp
@@ -226,6 +226,11 @@ CONFIG_RTE_LIBRTE_PMD_PCAP=y
 CONFIG_RTE_LIBRTE_PMD_BOND=y
 
 #
+# Compile null PMD
+#
+CONFIG_RTE_LIBRTE_PMD_NULL=y
+
+#
 # Do prefetch of packet data within PMD driver receive function
 #
 CONFIG_RTE_PMD_PACKET_PREFETCH=y
diff --git a/config/common_linuxapp b/config/common_linuxapp
index 5bee910..e3bd8c0 100644
--- a/config/common_linuxapp
+++ b/config/common_linuxapp
@@ -254,6 +254,11 @@ CONFIG_RTE_LIBRTE_PMD_BOND=y
 CONFIG_RTE_LIBRTE_PMD_XENVIRT=n
 
 #
+# Compile null PMD
+#
+CONFIG_RTE_LIBRTE_PMD_NULL=y
+
+#
 # Do prefetch of packet data within PMD driver receive function
 #
 CONFIG_RTE_PMD_PACKET_PREFETCH=y
diff --git a/lib/Makefile b/lib/Makefile
index 10c5bb3..61d6ed1 100644
--- a/lib/Makefile
+++ b/lib/Makefile
@@ -50,6 +50,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += librte_pmd_pcap
 DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += librte_pmd_virtio
 DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += librte_pmd_vmxnet3
 DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += librte_pmd_xenvirt
+DIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += librte_pmd_null
 DIRS-$(CONFIG_RTE_LIBRTE_HASH) += librte_hash
 DIRS-$(CONFIG_RTE_LIBRTE_LPM) += librte_lpm
 DIRS-$(CONFIG_RTE_LIBRTE_ACL) += librte_acl
diff --git a/lib/librte_pmd_null/Makefile b/lib/librte_pmd_null/Makefile
new file mode 100644
index 0000000..e017918
--- /dev/null
+++ b/lib/librte_pmd_null/Makefile
@@ -0,0 +1,58 @@
+#   BSD LICENSE
+#
+#   Copyright (C) 2014 Nippon Telegraph and Telephone Corporation.
+#   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 $(RTE_SDK)/mk/rte.vars.mk
+
+#
+# library name
+#
+LIB = librte_pmd_null.a
+
+CFLAGS += -O3
+CFLAGS += $(WERROR_FLAGS)
+
+#
+# all source are stored in SRCS-y
+#
+SRCS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += rte_eth_null.c
+
+#
+# Export include files
+#
+SYMLINK-y-include +=
+
+# this lib depends upon:
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_mbuf
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_ether
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_malloc
+DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_kvargs
+
+include $(RTE_SDK)/mk/rte.lib.mk
diff --git a/lib/librte_pmd_null/rte_eth_null.c b/lib/librte_pmd_null/rte_eth_null.c
new file mode 100644
index 0000000..1a81843
--- /dev/null
+++ b/lib/librte_pmd_null/rte_eth_null.c
@@ -0,0 +1,474 @@
+/*-
+ *   BSD LICENSE
+ *
+ *   Copyright (C) 2014 Nippon Telegraph and Telephone Corporation.
+ *   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 <rte_mbuf.h>
+#include <rte_ethdev.h>
+#include <rte_malloc.h>
+#include <rte_memcpy.h>
+#include <rte_dev.h>
+#include <rte_kvargs.h>
+
+#define ETH_NULL_PACKET_SIZE_ARG	"size"
+#define ETH_NULL_PACKET_COPY_ARG	"copy"
+
+static unsigned default_packet_size = 64;
+static unsigned default_packet_copy;
+
+static const char const *valid_arguments[] = {
+	ETH_NULL_PACKET_SIZE_ARG,
+	ETH_NULL_PACKET_COPY_ARG,
+	NULL
+};
+
+struct pmd_internals;
+
+struct null_queue {
+	struct pmd_internals *internals;
+
+	struct rte_mempool *mb_pool;
+	struct rte_mbuf *dummy_packet;
+
+	rte_atomic64_t rx_pkts;
+	rte_atomic64_t tx_pkts;
+	rte_atomic64_t err_pkts;
+};
+
+struct pmd_internals {
+	unsigned packet_size;
+	unsigned packet_copy;
+	unsigned numa_node;
+
+	unsigned nb_rx_queues;
+	unsigned nb_tx_queues;
+
+	struct null_queue rx_null_queues[1];
+	struct null_queue tx_null_queues[1];
+};
+
+
+static struct ether_addr eth_addr = { .addr_bytes = {0} };
+static const char *drivername = "Null PMD";
+static struct rte_eth_link pmd_link = {
+	.link_speed = 10000,
+	.link_duplex = ETH_LINK_FULL_DUPLEX,
+	.link_status = 0
+};
+
+static uint16_t
+eth_null_rx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
+{
+	int i;
+	struct null_queue *h = q;
+	unsigned packet_size = h->internals->packet_size;
+
+	for (i = 0; i < nb_bufs; i++) {
+		bufs[i] = rte_pktmbuf_alloc(h->mb_pool);
+		if (!bufs[i])
+			break;
+		bufs[i]->data_len = (uint16_t)packet_size;
+		bufs[i]->pkt_len = packet_size;
+		bufs[i]->nb_segs = 1;
+		bufs[i]->next = NULL;
+	}
+
+	rte_atomic64_add(&(h->rx_pkts), i);
+
+	return i;
+}
+
+static uint16_t
+eth_null_copy_rx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
+{
+	int i;
+	struct null_queue *h = q;
+	unsigned packet_size = h->internals->packet_size;
+
+	for (i = 0; i < nb_bufs; i++) {
+		bufs[i] = rte_pktmbuf_alloc(h->mb_pool);
+		if (!bufs[i])
+			break;
+		rte_memcpy(rte_pktmbuf_mtod(bufs[i], void *), h->dummy_packet,
+					packet_size);
+		bufs[i]->data_len = (uint16_t)packet_size;
+		bufs[i]->pkt_len = packet_size;
+		bufs[i]->nb_segs = 1;
+		bufs[i]->next = NULL;
+	}
+
+	rte_atomic64_add(&(h->rx_pkts), i);
+
+	return i;
+}
+
+static uint16_t
+eth_null_tx(void *q, struct rte_mbuf **bufs __rte_unused, uint16_t nb_bufs)
+{
+	int i;
+	struct null_queue *h = q;
+
+	for (i = 0; i < nb_bufs; i++)
+		rte_pktmbuf_free(bufs[i]);
+
+	rte_atomic64_add(&(h->tx_pkts), i);
+
+	return i;
+}
+
+static uint16_t
+eth_null_copy_tx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
+{
+	int i;
+	struct null_queue *h = q;
+	unsigned packet_size = h->internals->packet_size;
+
+	for (i = 0; i < nb_bufs; i++) {
+		rte_memcpy(h->dummy_packet, rte_pktmbuf_mtod(bufs[i], void *),
+					packet_size);
+		rte_pktmbuf_free(bufs[i]);
+	}
+
+	rte_atomic64_add(&(h->tx_pkts), i);
+
+	return i;
+}
+
+static int
+eth_dev_configure(struct rte_eth_dev *dev __rte_unused) { return 0; }
+
+static int
+eth_dev_start(struct rte_eth_dev *dev)
+{
+	dev->data->dev_link.link_status = 1;
+	return 0;
+}
+
+static void
+eth_dev_stop(struct rte_eth_dev *dev)
+{
+	dev->data->dev_link.link_status = 0;
+}
+
+static int
+eth_rx_queue_setup(struct rte_eth_dev *dev, uint16_t rx_queue_id,
+		uint16_t nb_rx_desc __rte_unused,
+		unsigned int socket_id __rte_unused,
+		const struct rte_eth_rxconf *rx_conf __rte_unused,
+		struct rte_mempool *mb_pool __rte_unused)
+{
+	struct rte_mbuf *dummy_packet;
+	struct pmd_internals *internals = dev->data->dev_private;
+	unsigned packet_size = internals->packet_size;
+
+	if (rx_queue_id != 0)
+			return -ENODEV;
+
+	internals->rx_null_queues[rx_queue_id].mb_pool = mb_pool;
+	dev->data->rx_queues[rx_queue_id] =
+		&internals->rx_null_queues[rx_queue_id];
+	dummy_packet = rte_zmalloc_socket(NULL,
+			packet_size, 0, internals->numa_node);
+	if (dummy_packet == NULL)
+		return -ENOMEM;
+
+	internals->rx_null_queues[rx_queue_id].internals = internals;
+	internals->rx_null_queues[rx_queue_id].dummy_packet = dummy_packet;
+
+	return 0;
+}
+
+static int
+eth_tx_queue_setup(struct rte_eth_dev *dev, uint16_t tx_queue_id,
+		uint16_t nb_tx_desc __rte_unused,
+		unsigned int socket_id __rte_unused,
+		const struct rte_eth_txconf *tx_conf __rte_unused)
+{
+	struct rte_mbuf *dummy_packet;
+	struct pmd_internals *internals = dev->data->dev_private;
+	unsigned packet_size = internals->packet_size;
+
+	if (tx_queue_id != 0)
+			return -ENODEV;
+
+	dev->data->tx_queues[tx_queue_id] =
+		&internals->tx_null_queues[tx_queue_id];
+	dummy_packet = rte_zmalloc_socket(NULL,
+			packet_size, 0, internals->numa_node);
+	if (dummy_packet == NULL)
+		return -ENOMEM;
+
+	internals->tx_null_queues[tx_queue_id].internals = internals;
+	internals->tx_null_queues[tx_queue_id].dummy_packet = dummy_packet;
+
+	return 0;
+}
+
+
+static void
+eth_dev_info(struct rte_eth_dev *dev,
+		struct rte_eth_dev_info *dev_info)
+{
+	struct pmd_internals *internals = dev->data->dev_private;
+
+	dev_info->driver_name = drivername;
+	dev_info->max_mac_addrs = 1;
+	dev_info->max_rx_pktlen = (uint32_t)-1;
+	dev_info->max_rx_queues = (uint16_t)internals->nb_rx_queues;
+	dev_info->max_tx_queues = (uint16_t)internals->nb_tx_queues;
+	dev_info->min_rx_bufsize = 0;
+	dev_info->pci_dev = NULL;
+}
+
+static void
+eth_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *igb_stats)
+{
+	unsigned i;
+	unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
+	const struct pmd_internals *internal = dev->data->dev_private;
+
+	memset(igb_stats, 0, sizeof(*igb_stats));
+	for (i = 0; i < RTE_ETHDEV_QUEUE_STAT_CNTRS &&
+			i < internal->nb_rx_queues; i++) {
+		igb_stats->q_ipackets[i] =
+			internal->rx_null_queues[i].rx_pkts.cnt;
+		rx_total += igb_stats->q_ipackets[i];
+	}
+
+	for (i = 0; i < RTE_ETHDEV_QUEUE_STAT_CNTRS &&
+			i < internal->nb_tx_queues; i++) {
+		igb_stats->q_opackets[i] =
+			internal->tx_null_queues[i].tx_pkts.cnt;
+		igb_stats->q_errors[i] =
+			internal->tx_null_queues[i].err_pkts.cnt;
+		tx_total += igb_stats->q_opackets[i];
+		tx_err_total += igb_stats->q_errors[i];
+	}
+
+	igb_stats->ipackets = rx_total;
+	igb_stats->opackets = tx_total;
+	igb_stats->oerrors = tx_err_total;
+}
+
+static void
+eth_stats_reset(struct rte_eth_dev *dev)
+{
+	unsigned i;
+	struct pmd_internals *internal = dev->data->dev_private;
+
+	for (i = 0; i < internal->nb_rx_queues; i++)
+		internal->rx_null_queues[i].rx_pkts.cnt = 0;
+	for (i = 0; i < internal->nb_tx_queues; i++) {
+		internal->tx_null_queues[i].tx_pkts.cnt = 0;
+		internal->tx_null_queues[i].err_pkts.cnt = 0;
+	}
+}
+
+static void
+eth_queue_release(void *q __rte_unused) { ; }
+static int
+eth_link_update(struct rte_eth_dev *dev __rte_unused,
+		int wait_to_complete __rte_unused) { return 0; }
+
+static struct eth_dev_ops ops = {
+		.dev_start = eth_dev_start,
+		.dev_stop = eth_dev_stop,
+		.dev_configure = eth_dev_configure,
+		.dev_infos_get = eth_dev_info,
+		.rx_queue_setup = eth_rx_queue_setup,
+		.tx_queue_setup = eth_tx_queue_setup,
+		.rx_queue_release = eth_queue_release,
+		.tx_queue_release = eth_queue_release,
+		.link_update = eth_link_update,
+		.stats_get = eth_stats_get,
+		.stats_reset = eth_stats_reset,
+};
+
+static int
+eth_dev_null_create(const char *name __rte_unused,
+		const unsigned numa_node,
+		unsigned packet_size,
+		unsigned packet_copy)
+{
+	const unsigned nb_rx_queues = 1;
+	const unsigned nb_tx_queues = 1;
+	struct rte_eth_dev_data *data = NULL;
+	struct rte_pci_device *pci_dev = NULL;
+	struct pmd_internals *internals = NULL;
+	struct rte_eth_dev *eth_dev = NULL;
+
+	RTE_LOG(INFO, PMD, "Creating null ethdev on numa socket %u\n",
+			numa_node);
+
+	/* now do all data allocation - for eth_dev structure, dummy pci driver
+	 * and internal (private) data
+	 */
+	data = rte_zmalloc_socket(name, sizeof(*data), 0, numa_node);
+	if (data == NULL)
+		goto error;
+
+	pci_dev = rte_zmalloc_socket(name, sizeof(*pci_dev), 0, numa_node);
+	if (pci_dev == NULL)
+		goto error;
+
+	internals = rte_zmalloc_socket(name, sizeof(*internals), 0, numa_node);
+	if (internals == NULL)
+		goto error;
+
+	/* reserve an ethdev entry */
+	eth_dev = rte_eth_dev_allocate(name);
+	if (eth_dev == NULL)
+		goto error;
+
+	/* now put it all together
+	 * - store queue data in internals,
+	 * - store numa_node info in pci_driver
+	 * - point eth_dev_data to internals and pci_driver
+	 * - and point eth_dev structure to new eth_dev_data structure
+	 */
+	/* NOTE: we'll replace the data element, of originally allocated eth_dev
+	 * so the nulls are local per-process */
+
+	internals->nb_rx_queues = nb_rx_queues;
+	internals->nb_tx_queues = nb_tx_queues;
+	internals->packet_size = packet_size;
+	internals->packet_copy = packet_copy;
+	internals->numa_node = numa_node;
+
+	pci_dev->numa_node = numa_node;
+
+	data->dev_private = internals;
+	data->port_id = eth_dev->data->port_id;
+	data->nb_rx_queues = (uint16_t)nb_rx_queues;
+	data->nb_tx_queues = (uint16_t)nb_tx_queues;
+	data->dev_link = pmd_link;
+	data->mac_addrs = &eth_addr;
+
+	eth_dev->data = data;
+	eth_dev->dev_ops = &ops;
+	eth_dev->pci_dev = pci_dev;
+
+	/* finally assign rx and tx ops */
+	if (packet_copy) {
+		eth_dev->rx_pkt_burst = eth_null_copy_rx;
+		eth_dev->tx_pkt_burst = eth_null_copy_tx;
+	} else {
+		eth_dev->rx_pkt_burst = eth_null_rx;
+		eth_dev->tx_pkt_burst = eth_null_tx;
+	}
+
+	return 0;
+
+error:
+	if (data)
+		rte_free(data);
+	if (pci_dev)
+		rte_free(pci_dev);
+	if (internals)
+		rte_free(internals);
+	return -1;
+}
+
+static inline int
+get_packet_size_arg(const char *key __rte_unused,
+		const char *value, void *extra_args)
+{
+	const char *a = value;
+	unsigned *packet_size = extra_args;
+
+	*packet_size = (unsigned)strtoul(a, NULL, 0);
+	if (*packet_size == UINT_MAX)
+		return -1;
+
+	return 0;
+}
+
+static inline int
+get_packet_copy_arg(const char *key __rte_unused,
+		const char *value, void *extra_args)
+{
+	const char *a = value;
+	unsigned *packet_copy = extra_args;
+
+	*packet_copy = (unsigned)strtoul(a, NULL, 0);
+	if (*packet_copy == UINT_MAX)
+		return -1;
+
+	return 0;
+}
+
+static int
+rte_pmd_null_devinit(const char *name, const char *params)
+{
+	unsigned numa_node;
+	unsigned packet_size = default_packet_size;
+	unsigned packet_copy = default_packet_copy;
+	struct rte_kvargs *kvlist;
+	int ret;
+
+	RTE_LOG(INFO, PMD, "Initializing pmd_null for %s\n", name);
+
+	numa_node = rte_socket_id();
+
+	kvlist = rte_kvargs_parse(params, valid_arguments);
+	if (kvlist == NULL)
+		return -1;
+
+	if (rte_kvargs_count(kvlist, ETH_NULL_PACKET_SIZE_ARG) == 1) {
+
+		ret = rte_kvargs_process(kvlist, ETH_NULL_PACKET_SIZE_ARG,
+				&get_packet_size_arg, &packet_size);
+		if (ret < 0)
+			return -1;
+	}
+
+	if (rte_kvargs_count(kvlist, ETH_NULL_PACKET_COPY_ARG) == 1) {
+
+		ret = rte_kvargs_process(kvlist, ETH_NULL_PACKET_COPY_ARG,
+				&get_packet_copy_arg, &packet_copy);
+		if (ret < 0)
+			return -1;
+	}
+
+	RTE_LOG(INFO, PMD, "Configure pmd_null: packet size is %d, "
+			"packet copy is %s\n", packet_size,
+			packet_copy ? "enabled" : "disabled");
+
+	return eth_dev_null_create(name, numa_node, packet_size, packet_copy);
+}
+
+static struct rte_driver pmd_null_drv = {
+	.name = "eth_null",
+	.type = PMD_VDEV,
+	.init = rte_pmd_null_devinit,
+};
+
+PMD_REGISTER_DRIVER(pmd_null_drv);
-- 
1.9.1

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [dpdk-dev] [RFC] librte_pmd_null: Add null PMD
  2014-09-19 12:27 ` [dpdk-dev] [RFC] librte_pmd_null: Add null PMD mukawa
@ 2014-09-22  7:46   ` Ivan Boule
  2014-09-22  8:07     ` Tetsuya Mukawa
  0 siblings, 1 reply; 4+ messages in thread
From: Ivan Boule @ 2014-09-22  7:46 UTC (permalink / raw)
  To: mukawa, dev; +Cc: nakajima.yoshihiro, masutani.hitoshi

On 09/19/2014 02:27 PM, mukawa@igel.co.jp wrote:
> From: Tetsuya Mukawa <mukawa@igel.co.jp>
>
> 'null PMD' is a virtual device driver particulary designed to measure
> performance of DPDK applications and DPDK PMDs. When an application call rx,
> null PMD just allocate mbufs and return those. Also tx, the PMD just free
> mbufs.
>
> The PMD has following options.
> - size: specify packe size allocated by RX. Default packet size is 64.
> - copy: specify 1 or 0 to enable or disable copy while RX and TX.
> 	Default value is 0(disbaled).
> 	This option is used for emulating more realistic data transfer.
> 	Copy size is equal to packet size.
Hi Tetsuya,

This is a "nice to have" PMD.

Just a minor comment.
I guess it should be NTT instead of Intel in the following comment
"#     * Neither the name of Intel Corporation nor the names of its".

Regards,
Ivan

>
> Signed-off-by: Tetsuya Mukawa <mukawa@igel.co.jp>
> ---
>   config/common_bsdapp               |   5 +
>   config/common_linuxapp             |   5 +
>   lib/Makefile                       |   1 +
>   lib/librte_pmd_null/Makefile       |  58 +++++
>   lib/librte_pmd_null/rte_eth_null.c | 474 +++++++++++++++++++++++++++++++++++++
>   5 files changed, 543 insertions(+)
>   create mode 100644 lib/librte_pmd_null/Makefile
>   create mode 100644 lib/librte_pmd_null/rte_eth_null.c
>
> diff --git a/config/common_bsdapp b/config/common_bsdapp
> index 645949f..a86321f 100644
> --- a/config/common_bsdapp
> +++ b/config/common_bsdapp
> @@ -226,6 +226,11 @@ CONFIG_RTE_LIBRTE_PMD_PCAP=y
>   CONFIG_RTE_LIBRTE_PMD_BOND=y
>
>   #
> +# Compile null PMD
> +#
> +CONFIG_RTE_LIBRTE_PMD_NULL=y
> +
> +#
>   # Do prefetch of packet data within PMD driver receive function
>   #
>   CONFIG_RTE_PMD_PACKET_PREFETCH=y
> diff --git a/config/common_linuxapp b/config/common_linuxapp
> index 5bee910..e3bd8c0 100644
> --- a/config/common_linuxapp
> +++ b/config/common_linuxapp
> @@ -254,6 +254,11 @@ CONFIG_RTE_LIBRTE_PMD_BOND=y
>   CONFIG_RTE_LIBRTE_PMD_XENVIRT=n
>
>   #
> +# Compile null PMD
> +#
> +CONFIG_RTE_LIBRTE_PMD_NULL=y
> +
> +#
>   # Do prefetch of packet data within PMD driver receive function
>   #
>   CONFIG_RTE_PMD_PACKET_PREFETCH=y
> diff --git a/lib/Makefile b/lib/Makefile
> index 10c5bb3..61d6ed1 100644
> --- a/lib/Makefile
> +++ b/lib/Makefile
> @@ -50,6 +50,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += librte_pmd_pcap
>   DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += librte_pmd_virtio
>   DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += librte_pmd_vmxnet3
>   DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += librte_pmd_xenvirt
> +DIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += librte_pmd_null
>   DIRS-$(CONFIG_RTE_LIBRTE_HASH) += librte_hash
>   DIRS-$(CONFIG_RTE_LIBRTE_LPM) += librte_lpm
>   DIRS-$(CONFIG_RTE_LIBRTE_ACL) += librte_acl
> diff --git a/lib/librte_pmd_null/Makefile b/lib/librte_pmd_null/Makefile
> new file mode 100644
> index 0000000..e017918
> --- /dev/null
> +++ b/lib/librte_pmd_null/Makefile
> @@ -0,0 +1,58 @@
> +#   BSD LICENSE
> +#
> +#   Copyright (C) 2014 Nippon Telegraph and Telephone Corporation.
> +#   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 $(RTE_SDK)/mk/rte.vars.mk
> +
> +#
> +# library name
> +#
> +LIB = librte_pmd_null.a
> +
> +CFLAGS += -O3
> +CFLAGS += $(WERROR_FLAGS)
> +
> +#
> +# all source are stored in SRCS-y
> +#
> +SRCS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += rte_eth_null.c
> +
> +#
> +# Export include files
> +#
> +SYMLINK-y-include +=
> +
> +# this lib depends upon:
> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_mbuf
> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_ether
> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_malloc
> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_kvargs
> +
> +include $(RTE_SDK)/mk/rte.lib.mk
> diff --git a/lib/librte_pmd_null/rte_eth_null.c b/lib/librte_pmd_null/rte_eth_null.c
> new file mode 100644
> index 0000000..1a81843
> --- /dev/null
> +++ b/lib/librte_pmd_null/rte_eth_null.c
> @@ -0,0 +1,474 @@
> +/*-
> + *   BSD LICENSE
> + *
> + *   Copyright (C) 2014 Nippon Telegraph and Telephone Corporation.
> + *   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 <rte_mbuf.h>
> +#include <rte_ethdev.h>
> +#include <rte_malloc.h>
> +#include <rte_memcpy.h>
> +#include <rte_dev.h>
> +#include <rte_kvargs.h>
> +
> +#define ETH_NULL_PACKET_SIZE_ARG	"size"
> +#define ETH_NULL_PACKET_COPY_ARG	"copy"
> +
> +static unsigned default_packet_size = 64;
> +static unsigned default_packet_copy;
> +
> +static const char const *valid_arguments[] = {
> +	ETH_NULL_PACKET_SIZE_ARG,
> +	ETH_NULL_PACKET_COPY_ARG,
> +	NULL
> +};
> +
> +struct pmd_internals;
> +
> +struct null_queue {
> +	struct pmd_internals *internals;
> +
> +	struct rte_mempool *mb_pool;
> +	struct rte_mbuf *dummy_packet;
> +
> +	rte_atomic64_t rx_pkts;
> +	rte_atomic64_t tx_pkts;
> +	rte_atomic64_t err_pkts;
> +};
> +
> +struct pmd_internals {
> +	unsigned packet_size;
> +	unsigned packet_copy;
> +	unsigned numa_node;
> +
> +	unsigned nb_rx_queues;
> +	unsigned nb_tx_queues;
> +
> +	struct null_queue rx_null_queues[1];
> +	struct null_queue tx_null_queues[1];
> +};
> +
> +
> +static struct ether_addr eth_addr = { .addr_bytes = {0} };
> +static const char *drivername = "Null PMD";
> +static struct rte_eth_link pmd_link = {
> +	.link_speed = 10000,
> +	.link_duplex = ETH_LINK_FULL_DUPLEX,
> +	.link_status = 0
> +};
> +
> +static uint16_t
> +eth_null_rx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
> +{
> +	int i;
> +	struct null_queue *h = q;
> +	unsigned packet_size = h->internals->packet_size;
> +
> +	for (i = 0; i < nb_bufs; i++) {
> +		bufs[i] = rte_pktmbuf_alloc(h->mb_pool);
> +		if (!bufs[i])
> +			break;
> +		bufs[i]->data_len = (uint16_t)packet_size;
> +		bufs[i]->pkt_len = packet_size;
> +		bufs[i]->nb_segs = 1;
> +		bufs[i]->next = NULL;
> +	}
> +
> +	rte_atomic64_add(&(h->rx_pkts), i);
> +
> +	return i;
> +}
> +
> +static uint16_t
> +eth_null_copy_rx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
> +{
> +	int i;
> +	struct null_queue *h = q;
> +	unsigned packet_size = h->internals->packet_size;
> +
> +	for (i = 0; i < nb_bufs; i++) {
> +		bufs[i] = rte_pktmbuf_alloc(h->mb_pool);
> +		if (!bufs[i])
> +			break;
> +		rte_memcpy(rte_pktmbuf_mtod(bufs[i], void *), h->dummy_packet,
> +					packet_size);
> +		bufs[i]->data_len = (uint16_t)packet_size;
> +		bufs[i]->pkt_len = packet_size;
> +		bufs[i]->nb_segs = 1;
> +		bufs[i]->next = NULL;
> +	}
> +
> +	rte_atomic64_add(&(h->rx_pkts), i);
> +
> +	return i;
> +}
> +
> +static uint16_t
> +eth_null_tx(void *q, struct rte_mbuf **bufs __rte_unused, uint16_t nb_bufs)
> +{
> +	int i;
> +	struct null_queue *h = q;
> +
> +	for (i = 0; i < nb_bufs; i++)
> +		rte_pktmbuf_free(bufs[i]);
> +
> +	rte_atomic64_add(&(h->tx_pkts), i);
> +
> +	return i;
> +}
> +
> +static uint16_t
> +eth_null_copy_tx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
> +{
> +	int i;
> +	struct null_queue *h = q;
> +	unsigned packet_size = h->internals->packet_size;
> +
> +	for (i = 0; i < nb_bufs; i++) {
> +		rte_memcpy(h->dummy_packet, rte_pktmbuf_mtod(bufs[i], void *),
> +					packet_size);
> +		rte_pktmbuf_free(bufs[i]);
> +	}
> +
> +	rte_atomic64_add(&(h->tx_pkts), i);
> +
> +	return i;
> +}
> +
> +static int
> +eth_dev_configure(struct rte_eth_dev *dev __rte_unused) { return 0; }
> +
> +static int
> +eth_dev_start(struct rte_eth_dev *dev)
> +{
> +	dev->data->dev_link.link_status = 1;
> +	return 0;
> +}
> +
> +static void
> +eth_dev_stop(struct rte_eth_dev *dev)
> +{
> +	dev->data->dev_link.link_status = 0;
> +}
> +
> +static int
> +eth_rx_queue_setup(struct rte_eth_dev *dev, uint16_t rx_queue_id,
> +		uint16_t nb_rx_desc __rte_unused,
> +		unsigned int socket_id __rte_unused,
> +		const struct rte_eth_rxconf *rx_conf __rte_unused,
> +		struct rte_mempool *mb_pool __rte_unused)
> +{
> +	struct rte_mbuf *dummy_packet;
> +	struct pmd_internals *internals = dev->data->dev_private;
> +	unsigned packet_size = internals->packet_size;
> +
> +	if (rx_queue_id != 0)
> +			return -ENODEV;
> +
> +	internals->rx_null_queues[rx_queue_id].mb_pool = mb_pool;
> +	dev->data->rx_queues[rx_queue_id] =
> +		&internals->rx_null_queues[rx_queue_id];
> +	dummy_packet = rte_zmalloc_socket(NULL,
> +			packet_size, 0, internals->numa_node);
> +	if (dummy_packet == NULL)
> +		return -ENOMEM;
> +
> +	internals->rx_null_queues[rx_queue_id].internals = internals;
> +	internals->rx_null_queues[rx_queue_id].dummy_packet = dummy_packet;
> +
> +	return 0;
> +}
> +
> +static int
> +eth_tx_queue_setup(struct rte_eth_dev *dev, uint16_t tx_queue_id,
> +		uint16_t nb_tx_desc __rte_unused,
> +		unsigned int socket_id __rte_unused,
> +		const struct rte_eth_txconf *tx_conf __rte_unused)
> +{
> +	struct rte_mbuf *dummy_packet;
> +	struct pmd_internals *internals = dev->data->dev_private;
> +	unsigned packet_size = internals->packet_size;
> +
> +	if (tx_queue_id != 0)
> +			return -ENODEV;
> +
> +	dev->data->tx_queues[tx_queue_id] =
> +		&internals->tx_null_queues[tx_queue_id];
> +	dummy_packet = rte_zmalloc_socket(NULL,
> +			packet_size, 0, internals->numa_node);
> +	if (dummy_packet == NULL)
> +		return -ENOMEM;
> +
> +	internals->tx_null_queues[tx_queue_id].internals = internals;
> +	internals->tx_null_queues[tx_queue_id].dummy_packet = dummy_packet;
> +
> +	return 0;
> +}
> +
> +
> +static void
> +eth_dev_info(struct rte_eth_dev *dev,
> +		struct rte_eth_dev_info *dev_info)
> +{
> +	struct pmd_internals *internals = dev->data->dev_private;
> +
> +	dev_info->driver_name = drivername;
> +	dev_info->max_mac_addrs = 1;
> +	dev_info->max_rx_pktlen = (uint32_t)-1;
> +	dev_info->max_rx_queues = (uint16_t)internals->nb_rx_queues;
> +	dev_info->max_tx_queues = (uint16_t)internals->nb_tx_queues;
> +	dev_info->min_rx_bufsize = 0;
> +	dev_info->pci_dev = NULL;
> +}
> +
> +static void
> +eth_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *igb_stats)
> +{
> +	unsigned i;
> +	unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
> +	const struct pmd_internals *internal = dev->data->dev_private;
> +
> +	memset(igb_stats, 0, sizeof(*igb_stats));
> +	for (i = 0; i < RTE_ETHDEV_QUEUE_STAT_CNTRS &&
> +			i < internal->nb_rx_queues; i++) {
> +		igb_stats->q_ipackets[i] =
> +			internal->rx_null_queues[i].rx_pkts.cnt;
> +		rx_total += igb_stats->q_ipackets[i];
> +	}
> +
> +	for (i = 0; i < RTE_ETHDEV_QUEUE_STAT_CNTRS &&
> +			i < internal->nb_tx_queues; i++) {
> +		igb_stats->q_opackets[i] =
> +			internal->tx_null_queues[i].tx_pkts.cnt;
> +		igb_stats->q_errors[i] =
> +			internal->tx_null_queues[i].err_pkts.cnt;
> +		tx_total += igb_stats->q_opackets[i];
> +		tx_err_total += igb_stats->q_errors[i];
> +	}
> +
> +	igb_stats->ipackets = rx_total;
> +	igb_stats->opackets = tx_total;
> +	igb_stats->oerrors = tx_err_total;
> +}
> +
> +static void
> +eth_stats_reset(struct rte_eth_dev *dev)
> +{
> +	unsigned i;
> +	struct pmd_internals *internal = dev->data->dev_private;
> +
> +	for (i = 0; i < internal->nb_rx_queues; i++)
> +		internal->rx_null_queues[i].rx_pkts.cnt = 0;
> +	for (i = 0; i < internal->nb_tx_queues; i++) {
> +		internal->tx_null_queues[i].tx_pkts.cnt = 0;
> +		internal->tx_null_queues[i].err_pkts.cnt = 0;
> +	}
> +}
> +
> +static void
> +eth_queue_release(void *q __rte_unused) { ; }
> +static int
> +eth_link_update(struct rte_eth_dev *dev __rte_unused,
> +		int wait_to_complete __rte_unused) { return 0; }
> +
> +static struct eth_dev_ops ops = {
> +		.dev_start = eth_dev_start,
> +		.dev_stop = eth_dev_stop,
> +		.dev_configure = eth_dev_configure,
> +		.dev_infos_get = eth_dev_info,
> +		.rx_queue_setup = eth_rx_queue_setup,
> +		.tx_queue_setup = eth_tx_queue_setup,
> +		.rx_queue_release = eth_queue_release,
> +		.tx_queue_release = eth_queue_release,
> +		.link_update = eth_link_update,
> +		.stats_get = eth_stats_get,
> +		.stats_reset = eth_stats_reset,
> +};
> +
> +static int
> +eth_dev_null_create(const char *name __rte_unused,
> +		const unsigned numa_node,
> +		unsigned packet_size,
> +		unsigned packet_copy)
> +{
> +	const unsigned nb_rx_queues = 1;
> +	const unsigned nb_tx_queues = 1;
> +	struct rte_eth_dev_data *data = NULL;
> +	struct rte_pci_device *pci_dev = NULL;
> +	struct pmd_internals *internals = NULL;
> +	struct rte_eth_dev *eth_dev = NULL;
> +
> +	RTE_LOG(INFO, PMD, "Creating null ethdev on numa socket %u\n",
> +			numa_node);
> +
> +	/* now do all data allocation - for eth_dev structure, dummy pci driver
> +	 * and internal (private) data
> +	 */
> +	data = rte_zmalloc_socket(name, sizeof(*data), 0, numa_node);
> +	if (data == NULL)
> +		goto error;
> +
> +	pci_dev = rte_zmalloc_socket(name, sizeof(*pci_dev), 0, numa_node);
> +	if (pci_dev == NULL)
> +		goto error;
> +
> +	internals = rte_zmalloc_socket(name, sizeof(*internals), 0, numa_node);
> +	if (internals == NULL)
> +		goto error;
> +
> +	/* reserve an ethdev entry */
> +	eth_dev = rte_eth_dev_allocate(name);
> +	if (eth_dev == NULL)
> +		goto error;
> +
> +	/* now put it all together
> +	 * - store queue data in internals,
> +	 * - store numa_node info in pci_driver
> +	 * - point eth_dev_data to internals and pci_driver
> +	 * - and point eth_dev structure to new eth_dev_data structure
> +	 */
> +	/* NOTE: we'll replace the data element, of originally allocated eth_dev
> +	 * so the nulls are local per-process */
> +
> +	internals->nb_rx_queues = nb_rx_queues;
> +	internals->nb_tx_queues = nb_tx_queues;
> +	internals->packet_size = packet_size;
> +	internals->packet_copy = packet_copy;
> +	internals->numa_node = numa_node;
> +
> +	pci_dev->numa_node = numa_node;
> +
> +	data->dev_private = internals;
> +	data->port_id = eth_dev->data->port_id;
> +	data->nb_rx_queues = (uint16_t)nb_rx_queues;
> +	data->nb_tx_queues = (uint16_t)nb_tx_queues;
> +	data->dev_link = pmd_link;
> +	data->mac_addrs = &eth_addr;
> +
> +	eth_dev->data = data;
> +	eth_dev->dev_ops = &ops;
> +	eth_dev->pci_dev = pci_dev;
> +
> +	/* finally assign rx and tx ops */
> +	if (packet_copy) {
> +		eth_dev->rx_pkt_burst = eth_null_copy_rx;
> +		eth_dev->tx_pkt_burst = eth_null_copy_tx;
> +	} else {
> +		eth_dev->rx_pkt_burst = eth_null_rx;
> +		eth_dev->tx_pkt_burst = eth_null_tx;
> +	}
> +
> +	return 0;
> +
> +error:
> +	if (data)
> +		rte_free(data);
> +	if (pci_dev)
> +		rte_free(pci_dev);
> +	if (internals)
> +		rte_free(internals);
> +	return -1;
> +}
> +
> +static inline int
> +get_packet_size_arg(const char *key __rte_unused,
> +		const char *value, void *extra_args)
> +{
> +	const char *a = value;
> +	unsigned *packet_size = extra_args;
> +
> +	*packet_size = (unsigned)strtoul(a, NULL, 0);
> +	if (*packet_size == UINT_MAX)
> +		return -1;
> +
> +	return 0;
> +}
> +
> +static inline int
> +get_packet_copy_arg(const char *key __rte_unused,
> +		const char *value, void *extra_args)
> +{
> +	const char *a = value;
> +	unsigned *packet_copy = extra_args;
> +
> +	*packet_copy = (unsigned)strtoul(a, NULL, 0);
> +	if (*packet_copy == UINT_MAX)
> +		return -1;
> +
> +	return 0;
> +}
> +
> +static int
> +rte_pmd_null_devinit(const char *name, const char *params)
> +{
> +	unsigned numa_node;
> +	unsigned packet_size = default_packet_size;
> +	unsigned packet_copy = default_packet_copy;
> +	struct rte_kvargs *kvlist;
> +	int ret;
> +
> +	RTE_LOG(INFO, PMD, "Initializing pmd_null for %s\n", name);
> +
> +	numa_node = rte_socket_id();
> +
> +	kvlist = rte_kvargs_parse(params, valid_arguments);
> +	if (kvlist == NULL)
> +		return -1;
> +
> +	if (rte_kvargs_count(kvlist, ETH_NULL_PACKET_SIZE_ARG) == 1) {
> +
> +		ret = rte_kvargs_process(kvlist, ETH_NULL_PACKET_SIZE_ARG,
> +				&get_packet_size_arg, &packet_size);
> +		if (ret < 0)
> +			return -1;
> +	}
> +
> +	if (rte_kvargs_count(kvlist, ETH_NULL_PACKET_COPY_ARG) == 1) {
> +
> +		ret = rte_kvargs_process(kvlist, ETH_NULL_PACKET_COPY_ARG,
> +				&get_packet_copy_arg, &packet_copy);
> +		if (ret < 0)
> +			return -1;
> +	}
> +
> +	RTE_LOG(INFO, PMD, "Configure pmd_null: packet size is %d, "
> +			"packet copy is %s\n", packet_size,
> +			packet_copy ? "enabled" : "disabled");
> +
> +	return eth_dev_null_create(name, numa_node, packet_size, packet_copy);
> +}
> +
> +static struct rte_driver pmd_null_drv = {
> +	.name = "eth_null",
> +	.type = PMD_VDEV,
> +	.init = rte_pmd_null_devinit,
> +};
> +
> +PMD_REGISTER_DRIVER(pmd_null_drv);
>


-- 
Ivan Boule
6WIND Development Engineer

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [dpdk-dev] [RFC] librte_pmd_null: Add null PMD
  2014-09-22  7:46   ` Ivan Boule
@ 2014-09-22  8:07     ` Tetsuya Mukawa
  0 siblings, 0 replies; 4+ messages in thread
From: Tetsuya Mukawa @ 2014-09-22  8:07 UTC (permalink / raw)
  To: Ivan Boule, dev; +Cc: nakajima.yoshihiro, masutani.hitoshi

Hi Ivan

(2014/09/22 16:46), Ivan Boule wrote:
>
> This is a "nice to have" PMD.
>
> Just a minor comment.
> I guess it should be NTT instead of Intel in the following comment
> "#     * Neither the name of Intel Corporation nor the names of its".
>
I appreciate your comment. I will rewrite and post it again.

Regards,
Tetsuya


> Regards,
> Ivan
>
>>
>> Signed-off-by: Tetsuya Mukawa <mukawa@igel.co.jp>
>> ---
>>   config/common_bsdapp               |   5 +
>>   config/common_linuxapp             |   5 +
>>   lib/Makefile                       |   1 +
>>   lib/librte_pmd_null/Makefile       |  58 +++++
>>   lib/librte_pmd_null/rte_eth_null.c | 474
>> +++++++++++++++++++++++++++++++++++++
>>   5 files changed, 543 insertions(+)
>>   create mode 100644 lib/librte_pmd_null/Makefile
>>   create mode 100644 lib/librte_pmd_null/rte_eth_null.c
>>
>> diff --git a/config/common_bsdapp b/config/common_bsdapp
>> index 645949f..a86321f 100644
>> --- a/config/common_bsdapp
>> +++ b/config/common_bsdapp
>> @@ -226,6 +226,11 @@ CONFIG_RTE_LIBRTE_PMD_PCAP=y
>>   CONFIG_RTE_LIBRTE_PMD_BOND=y
>>
>>   #
>> +# Compile null PMD
>> +#
>> +CONFIG_RTE_LIBRTE_PMD_NULL=y
>> +
>> +#
>>   # Do prefetch of packet data within PMD driver receive function
>>   #
>>   CONFIG_RTE_PMD_PACKET_PREFETCH=y
>> diff --git a/config/common_linuxapp b/config/common_linuxapp
>> index 5bee910..e3bd8c0 100644
>> --- a/config/common_linuxapp
>> +++ b/config/common_linuxapp
>> @@ -254,6 +254,11 @@ CONFIG_RTE_LIBRTE_PMD_BOND=y
>>   CONFIG_RTE_LIBRTE_PMD_XENVIRT=n
>>
>>   #
>> +# Compile null PMD
>> +#
>> +CONFIG_RTE_LIBRTE_PMD_NULL=y
>> +
>> +#
>>   # Do prefetch of packet data within PMD driver receive function
>>   #
>>   CONFIG_RTE_PMD_PACKET_PREFETCH=y
>> diff --git a/lib/Makefile b/lib/Makefile
>> index 10c5bb3..61d6ed1 100644
>> --- a/lib/Makefile
>> +++ b/lib/Makefile
>> @@ -50,6 +50,7 @@ DIRS-$(CONFIG_RTE_LIBRTE_PMD_PCAP) += librte_pmd_pcap
>>   DIRS-$(CONFIG_RTE_LIBRTE_VIRTIO_PMD) += librte_pmd_virtio
>>   DIRS-$(CONFIG_RTE_LIBRTE_VMXNET3_PMD) += librte_pmd_vmxnet3
>>   DIRS-$(CONFIG_RTE_LIBRTE_PMD_XENVIRT) += librte_pmd_xenvirt
>> +DIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += librte_pmd_null
>>   DIRS-$(CONFIG_RTE_LIBRTE_HASH) += librte_hash
>>   DIRS-$(CONFIG_RTE_LIBRTE_LPM) += librte_lpm
>>   DIRS-$(CONFIG_RTE_LIBRTE_ACL) += librte_acl
>> diff --git a/lib/librte_pmd_null/Makefile b/lib/librte_pmd_null/Makefile
>> new file mode 100644
>> index 0000000..e017918
>> --- /dev/null
>> +++ b/lib/librte_pmd_null/Makefile
>> @@ -0,0 +1,58 @@
>> +#   BSD LICENSE
>> +#
>> +#   Copyright (C) 2014 Nippon Telegraph and Telephone Corporation.
>> +#   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 $(RTE_SDK)/mk/rte.vars.mk
>> +
>> +#
>> +# library name
>> +#
>> +LIB = librte_pmd_null.a
>> +
>> +CFLAGS += -O3
>> +CFLAGS += $(WERROR_FLAGS)
>> +
>> +#
>> +# all source are stored in SRCS-y
>> +#
>> +SRCS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += rte_eth_null.c
>> +
>> +#
>> +# Export include files
>> +#
>> +SYMLINK-y-include +=
>> +
>> +# this lib depends upon:
>> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_mbuf
>> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_ether
>> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_malloc
>> +DEPDIRS-$(CONFIG_RTE_LIBRTE_PMD_NULL) += lib/librte_kvargs
>> +
>> +include $(RTE_SDK)/mk/rte.lib.mk
>> diff --git a/lib/librte_pmd_null/rte_eth_null.c
>> b/lib/librte_pmd_null/rte_eth_null.c
>> new file mode 100644
>> index 0000000..1a81843
>> --- /dev/null
>> +++ b/lib/librte_pmd_null/rte_eth_null.c
>> @@ -0,0 +1,474 @@
>> +/*-
>> + *   BSD LICENSE
>> + *
>> + *   Copyright (C) 2014 Nippon Telegraph and Telephone Corporation.
>> + *   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 <rte_mbuf.h>
>> +#include <rte_ethdev.h>
>> +#include <rte_malloc.h>
>> +#include <rte_memcpy.h>
>> +#include <rte_dev.h>
>> +#include <rte_kvargs.h>
>> +
>> +#define ETH_NULL_PACKET_SIZE_ARG    "size"
>> +#define ETH_NULL_PACKET_COPY_ARG    "copy"
>> +
>> +static unsigned default_packet_size = 64;
>> +static unsigned default_packet_copy;
>> +
>> +static const char const *valid_arguments[] = {
>> +    ETH_NULL_PACKET_SIZE_ARG,
>> +    ETH_NULL_PACKET_COPY_ARG,
>> +    NULL
>> +};
>> +
>> +struct pmd_internals;
>> +
>> +struct null_queue {
>> +    struct pmd_internals *internals;
>> +
>> +    struct rte_mempool *mb_pool;
>> +    struct rte_mbuf *dummy_packet;
>> +
>> +    rte_atomic64_t rx_pkts;
>> +    rte_atomic64_t tx_pkts;
>> +    rte_atomic64_t err_pkts;
>> +};
>> +
>> +struct pmd_internals {
>> +    unsigned packet_size;
>> +    unsigned packet_copy;
>> +    unsigned numa_node;
>> +
>> +    unsigned nb_rx_queues;
>> +    unsigned nb_tx_queues;
>> +
>> +    struct null_queue rx_null_queues[1];
>> +    struct null_queue tx_null_queues[1];
>> +};
>> +
>> +
>> +static struct ether_addr eth_addr = { .addr_bytes = {0} };
>> +static const char *drivername = "Null PMD";
>> +static struct rte_eth_link pmd_link = {
>> +    .link_speed = 10000,
>> +    .link_duplex = ETH_LINK_FULL_DUPLEX,
>> +    .link_status = 0
>> +};
>> +
>> +static uint16_t
>> +eth_null_rx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
>> +{
>> +    int i;
>> +    struct null_queue *h = q;
>> +    unsigned packet_size = h->internals->packet_size;
>> +
>> +    for (i = 0; i < nb_bufs; i++) {
>> +        bufs[i] = rte_pktmbuf_alloc(h->mb_pool);
>> +        if (!bufs[i])
>> +            break;
>> +        bufs[i]->data_len = (uint16_t)packet_size;
>> +        bufs[i]->pkt_len = packet_size;
>> +        bufs[i]->nb_segs = 1;
>> +        bufs[i]->next = NULL;
>> +    }
>> +
>> +    rte_atomic64_add(&(h->rx_pkts), i);
>> +
>> +    return i;
>> +}
>> +
>> +static uint16_t
>> +eth_null_copy_rx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
>> +{
>> +    int i;
>> +    struct null_queue *h = q;
>> +    unsigned packet_size = h->internals->packet_size;
>> +
>> +    for (i = 0; i < nb_bufs; i++) {
>> +        bufs[i] = rte_pktmbuf_alloc(h->mb_pool);
>> +        if (!bufs[i])
>> +            break;
>> +        rte_memcpy(rte_pktmbuf_mtod(bufs[i], void *), h->dummy_packet,
>> +                    packet_size);
>> +        bufs[i]->data_len = (uint16_t)packet_size;
>> +        bufs[i]->pkt_len = packet_size;
>> +        bufs[i]->nb_segs = 1;
>> +        bufs[i]->next = NULL;
>> +    }
>> +
>> +    rte_atomic64_add(&(h->rx_pkts), i);
>> +
>> +    return i;
>> +}
>> +
>> +static uint16_t
>> +eth_null_tx(void *q, struct rte_mbuf **bufs __rte_unused, uint16_t
>> nb_bufs)
>> +{
>> +    int i;
>> +    struct null_queue *h = q;
>> +
>> +    for (i = 0; i < nb_bufs; i++)
>> +        rte_pktmbuf_free(bufs[i]);
>> +
>> +    rte_atomic64_add(&(h->tx_pkts), i);
>> +
>> +    return i;
>> +}
>> +
>> +static uint16_t
>> +eth_null_copy_tx(void *q, struct rte_mbuf **bufs, uint16_t nb_bufs)
>> +{
>> +    int i;
>> +    struct null_queue *h = q;
>> +    unsigned packet_size = h->internals->packet_size;
>> +
>> +    for (i = 0; i < nb_bufs; i++) {
>> +        rte_memcpy(h->dummy_packet, rte_pktmbuf_mtod(bufs[i], void *),
>> +                    packet_size);
>> +        rte_pktmbuf_free(bufs[i]);
>> +    }
>> +
>> +    rte_atomic64_add(&(h->tx_pkts), i);
>> +
>> +    return i;
>> +}
>> +
>> +static int
>> +eth_dev_configure(struct rte_eth_dev *dev __rte_unused) { return 0; }
>> +
>> +static int
>> +eth_dev_start(struct rte_eth_dev *dev)
>> +{
>> +    dev->data->dev_link.link_status = 1;
>> +    return 0;
>> +}
>> +
>> +static void
>> +eth_dev_stop(struct rte_eth_dev *dev)
>> +{
>> +    dev->data->dev_link.link_status = 0;
>> +}
>> +
>> +static int
>> +eth_rx_queue_setup(struct rte_eth_dev *dev, uint16_t rx_queue_id,
>> +        uint16_t nb_rx_desc __rte_unused,
>> +        unsigned int socket_id __rte_unused,
>> +        const struct rte_eth_rxconf *rx_conf __rte_unused,
>> +        struct rte_mempool *mb_pool __rte_unused)
>> +{
>> +    struct rte_mbuf *dummy_packet;
>> +    struct pmd_internals *internals = dev->data->dev_private;
>> +    unsigned packet_size = internals->packet_size;
>> +
>> +    if (rx_queue_id != 0)
>> +            return -ENODEV;
>> +
>> +    internals->rx_null_queues[rx_queue_id].mb_pool = mb_pool;
>> +    dev->data->rx_queues[rx_queue_id] =
>> +        &internals->rx_null_queues[rx_queue_id];
>> +    dummy_packet = rte_zmalloc_socket(NULL,
>> +            packet_size, 0, internals->numa_node);
>> +    if (dummy_packet == NULL)
>> +        return -ENOMEM;
>> +
>> +    internals->rx_null_queues[rx_queue_id].internals = internals;
>> +    internals->rx_null_queues[rx_queue_id].dummy_packet = dummy_packet;
>> +
>> +    return 0;
>> +}
>> +
>> +static int
>> +eth_tx_queue_setup(struct rte_eth_dev *dev, uint16_t tx_queue_id,
>> +        uint16_t nb_tx_desc __rte_unused,
>> +        unsigned int socket_id __rte_unused,
>> +        const struct rte_eth_txconf *tx_conf __rte_unused)
>> +{
>> +    struct rte_mbuf *dummy_packet;
>> +    struct pmd_internals *internals = dev->data->dev_private;
>> +    unsigned packet_size = internals->packet_size;
>> +
>> +    if (tx_queue_id != 0)
>> +            return -ENODEV;
>> +
>> +    dev->data->tx_queues[tx_queue_id] =
>> +        &internals->tx_null_queues[tx_queue_id];
>> +    dummy_packet = rte_zmalloc_socket(NULL,
>> +            packet_size, 0, internals->numa_node);
>> +    if (dummy_packet == NULL)
>> +        return -ENOMEM;
>> +
>> +    internals->tx_null_queues[tx_queue_id].internals = internals;
>> +    internals->tx_null_queues[tx_queue_id].dummy_packet = dummy_packet;
>> +
>> +    return 0;
>> +}
>> +
>> +
>> +static void
>> +eth_dev_info(struct rte_eth_dev *dev,
>> +        struct rte_eth_dev_info *dev_info)
>> +{
>> +    struct pmd_internals *internals = dev->data->dev_private;
>> +
>> +    dev_info->driver_name = drivername;
>> +    dev_info->max_mac_addrs = 1;
>> +    dev_info->max_rx_pktlen = (uint32_t)-1;
>> +    dev_info->max_rx_queues = (uint16_t)internals->nb_rx_queues;
>> +    dev_info->max_tx_queues = (uint16_t)internals->nb_tx_queues;
>> +    dev_info->min_rx_bufsize = 0;
>> +    dev_info->pci_dev = NULL;
>> +}
>> +
>> +static void
>> +eth_stats_get(struct rte_eth_dev *dev, struct rte_eth_stats *igb_stats)
>> +{
>> +    unsigned i;
>> +    unsigned long rx_total = 0, tx_total = 0, tx_err_total = 0;
>> +    const struct pmd_internals *internal = dev->data->dev_private;
>> +
>> +    memset(igb_stats, 0, sizeof(*igb_stats));
>> +    for (i = 0; i < RTE_ETHDEV_QUEUE_STAT_CNTRS &&
>> +            i < internal->nb_rx_queues; i++) {
>> +        igb_stats->q_ipackets[i] =
>> +            internal->rx_null_queues[i].rx_pkts.cnt;
>> +        rx_total += igb_stats->q_ipackets[i];
>> +    }
>> +
>> +    for (i = 0; i < RTE_ETHDEV_QUEUE_STAT_CNTRS &&
>> +            i < internal->nb_tx_queues; i++) {
>> +        igb_stats->q_opackets[i] =
>> +            internal->tx_null_queues[i].tx_pkts.cnt;
>> +        igb_stats->q_errors[i] =
>> +            internal->tx_null_queues[i].err_pkts.cnt;
>> +        tx_total += igb_stats->q_opackets[i];
>> +        tx_err_total += igb_stats->q_errors[i];
>> +    }
>> +
>> +    igb_stats->ipackets = rx_total;
>> +    igb_stats->opackets = tx_total;
>> +    igb_stats->oerrors = tx_err_total;
>> +}
>> +
>> +static void
>> +eth_stats_reset(struct rte_eth_dev *dev)
>> +{
>> +    unsigned i;
>> +    struct pmd_internals *internal = dev->data->dev_private;
>> +
>> +    for (i = 0; i < internal->nb_rx_queues; i++)
>> +        internal->rx_null_queues[i].rx_pkts.cnt = 0;
>> +    for (i = 0; i < internal->nb_tx_queues; i++) {
>> +        internal->tx_null_queues[i].tx_pkts.cnt = 0;
>> +        internal->tx_null_queues[i].err_pkts.cnt = 0;
>> +    }
>> +}
>> +
>> +static void
>> +eth_queue_release(void *q __rte_unused) { ; }
>> +static int
>> +eth_link_update(struct rte_eth_dev *dev __rte_unused,
>> +        int wait_to_complete __rte_unused) { return 0; }
>> +
>> +static struct eth_dev_ops ops = {
>> +        .dev_start = eth_dev_start,
>> +        .dev_stop = eth_dev_stop,
>> +        .dev_configure = eth_dev_configure,
>> +        .dev_infos_get = eth_dev_info,
>> +        .rx_queue_setup = eth_rx_queue_setup,
>> +        .tx_queue_setup = eth_tx_queue_setup,
>> +        .rx_queue_release = eth_queue_release,
>> +        .tx_queue_release = eth_queue_release,
>> +        .link_update = eth_link_update,
>> +        .stats_get = eth_stats_get,
>> +        .stats_reset = eth_stats_reset,
>> +};
>> +
>> +static int
>> +eth_dev_null_create(const char *name __rte_unused,
>> +        const unsigned numa_node,
>> +        unsigned packet_size,
>> +        unsigned packet_copy)
>> +{
>> +    const unsigned nb_rx_queues = 1;
>> +    const unsigned nb_tx_queues = 1;
>> +    struct rte_eth_dev_data *data = NULL;
>> +    struct rte_pci_device *pci_dev = NULL;
>> +    struct pmd_internals *internals = NULL;
>> +    struct rte_eth_dev *eth_dev = NULL;
>> +
>> +    RTE_LOG(INFO, PMD, "Creating null ethdev on numa socket %u\n",
>> +            numa_node);
>> +
>> +    /* now do all data allocation - for eth_dev structure, dummy pci
>> driver
>> +     * and internal (private) data
>> +     */
>> +    data = rte_zmalloc_socket(name, sizeof(*data), 0, numa_node);
>> +    if (data == NULL)
>> +        goto error;
>> +
>> +    pci_dev = rte_zmalloc_socket(name, sizeof(*pci_dev), 0, numa_node);
>> +    if (pci_dev == NULL)
>> +        goto error;
>> +
>> +    internals = rte_zmalloc_socket(name, sizeof(*internals), 0,
>> numa_node);
>> +    if (internals == NULL)
>> +        goto error;
>> +
>> +    /* reserve an ethdev entry */
>> +    eth_dev = rte_eth_dev_allocate(name);
>> +    if (eth_dev == NULL)
>> +        goto error;
>> +
>> +    /* now put it all together
>> +     * - store queue data in internals,
>> +     * - store numa_node info in pci_driver
>> +     * - point eth_dev_data to internals and pci_driver
>> +     * - and point eth_dev structure to new eth_dev_data structure
>> +     */
>> +    /* NOTE: we'll replace the data element, of originally allocated
>> eth_dev
>> +     * so the nulls are local per-process */
>> +
>> +    internals->nb_rx_queues = nb_rx_queues;
>> +    internals->nb_tx_queues = nb_tx_queues;
>> +    internals->packet_size = packet_size;
>> +    internals->packet_copy = packet_copy;
>> +    internals->numa_node = numa_node;
>> +
>> +    pci_dev->numa_node = numa_node;
>> +
>> +    data->dev_private = internals;
>> +    data->port_id = eth_dev->data->port_id;
>> +    data->nb_rx_queues = (uint16_t)nb_rx_queues;
>> +    data->nb_tx_queues = (uint16_t)nb_tx_queues;
>> +    data->dev_link = pmd_link;
>> +    data->mac_addrs = &eth_addr;
>> +
>> +    eth_dev->data = data;
>> +    eth_dev->dev_ops = &ops;
>> +    eth_dev->pci_dev = pci_dev;
>> +
>> +    /* finally assign rx and tx ops */
>> +    if (packet_copy) {
>> +        eth_dev->rx_pkt_burst = eth_null_copy_rx;
>> +        eth_dev->tx_pkt_burst = eth_null_copy_tx;
>> +    } else {
>> +        eth_dev->rx_pkt_burst = eth_null_rx;
>> +        eth_dev->tx_pkt_burst = eth_null_tx;
>> +    }
>> +
>> +    return 0;
>> +
>> +error:
>> +    if (data)
>> +        rte_free(data);
>> +    if (pci_dev)
>> +        rte_free(pci_dev);
>> +    if (internals)
>> +        rte_free(internals);
>> +    return -1;
>> +}
>> +
>> +static inline int
>> +get_packet_size_arg(const char *key __rte_unused,
>> +        const char *value, void *extra_args)
>> +{
>> +    const char *a = value;
>> +    unsigned *packet_size = extra_args;
>> +
>> +    *packet_size = (unsigned)strtoul(a, NULL, 0);
>> +    if (*packet_size == UINT_MAX)
>> +        return -1;
>> +
>> +    return 0;
>> +}
>> +
>> +static inline int
>> +get_packet_copy_arg(const char *key __rte_unused,
>> +        const char *value, void *extra_args)
>> +{
>> +    const char *a = value;
>> +    unsigned *packet_copy = extra_args;
>> +
>> +    *packet_copy = (unsigned)strtoul(a, NULL, 0);
>> +    if (*packet_copy == UINT_MAX)
>> +        return -1;
>> +
>> +    return 0;
>> +}
>> +
>> +static int
>> +rte_pmd_null_devinit(const char *name, const char *params)
>> +{
>> +    unsigned numa_node;
>> +    unsigned packet_size = default_packet_size;
>> +    unsigned packet_copy = default_packet_copy;
>> +    struct rte_kvargs *kvlist;
>> +    int ret;
>> +
>> +    RTE_LOG(INFO, PMD, "Initializing pmd_null for %s\n", name);
>> +
>> +    numa_node = rte_socket_id();
>> +
>> +    kvlist = rte_kvargs_parse(params, valid_arguments);
>> +    if (kvlist == NULL)
>> +        return -1;
>> +
>> +    if (rte_kvargs_count(kvlist, ETH_NULL_PACKET_SIZE_ARG) == 1) {
>> +
>> +        ret = rte_kvargs_process(kvlist, ETH_NULL_PACKET_SIZE_ARG,
>> +                &get_packet_size_arg, &packet_size);
>> +        if (ret < 0)
>> +            return -1;
>> +    }
>> +
>> +    if (rte_kvargs_count(kvlist, ETH_NULL_PACKET_COPY_ARG) == 1) {
>> +
>> +        ret = rte_kvargs_process(kvlist, ETH_NULL_PACKET_COPY_ARG,
>> +                &get_packet_copy_arg, &packet_copy);
>> +        if (ret < 0)
>> +            return -1;
>> +    }
>> +
>> +    RTE_LOG(INFO, PMD, "Configure pmd_null: packet size is %d, "
>> +            "packet copy is %s\n", packet_size,
>> +            packet_copy ? "enabled" : "disabled");
>> +
>> +    return eth_dev_null_create(name, numa_node, packet_size,
>> packet_copy);
>> +}
>> +
>> +static struct rte_driver pmd_null_drv = {
>> +    .name = "eth_null",
>> +    .type = PMD_VDEV,
>> +    .init = rte_pmd_null_devinit,
>> +};
>> +
>> +PMD_REGISTER_DRIVER(pmd_null_drv);
>>
>
>

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2014-09-22  8:01 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-09-19 12:27 [dpdk-dev] [RFC] PMD for performance measurement mukawa
2014-09-19 12:27 ` [dpdk-dev] [RFC] librte_pmd_null: Add null PMD mukawa
2014-09-22  7:46   ` Ivan Boule
2014-09-22  8:07     ` Tetsuya Mukawa

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