[dpdk-dev] [PATCH v6 0/5] Virtio PMD RSS support & RSS fixes

[dpdk-dev] [PATCH v6 0/5] Virtio PMD RSS support & RSS fixes

[Maxime Coquelin]

This series is mainly adding support for RSS to Virtio PMD
driver. The two last patches are fixing an issue in testpmd
that could cause out of bounds access, and fix
an issue spotted in the mlx5 driver while looking for
inspiration.

The first motivation for this series is to eventually
support RSS down to the Vhost-user library, so that OVS can
benefit from it. But it will be also useful with vDPA
devices in the future.

Regarding the testing, I have tested it with qemu v5.2 from
Fedora 34. Since libvirt does not support yet enabling RSS
feature in the Qemu virtio-net device, and this feature is
disabled by default, the tester can either rebuild the qemu
package to enable it by default or use the qemu cmdline to
do the same.

The tester can use testpmd in icmpecho mode in the guest
and scapy on the host to inject random traffic on the tap
interface, e.g.:
sendp(Ether(src=RandMAC()) / IP(src=RandIP(), dst='192.168.123.9') / UDP(sport=RandShort(), dport=RandShort()), loop=True, iface='vnet7')

Then it can play with RSS config in testpmd to change the
RETA, or hash type and see traffic being steered
accordingly by checking the Rx xstats.

Changes in v6:
==============
- Rebased on top of main, including convertion to new naming of some defines
- Convert hash types definition to RTE_BIT32() (Andrew)
- Add build check for RETA size being a pow2 (Andrew)
- Restore RSS config if device update fails (Andrew)
- Prevent RSS config if RXq number == 0 (Andrew)
- Reconfigure the RETA if the number of RXq changed (Andrew)
- Apply rss_conf even if the RSS key had previously been configured (Andrew)

Changes in v5:
==============
- Remove unneeded index init (Chenbo)
- Improve error print (Chenbo)
- Add missed comment on RSS ctrl message

Changes in v4:
==============
- s/GPTU/GTPU/ (Xiaoyun)

Changes in v3:
==============
- Add applying user-specified RSS conf a device config time (Andrew)
- Remove useless checks (Chenbo)
- Clean control message payload dlen variable (Chenbo)
- Add GTPU offload type (Xiaoyun)
- Add missing types to str2flowtype() (Xiaoyun)

Changes in v2:
==============
- Rework patch 2 to keep old behaviour, but fix possible out of bounds due to key length (Andrew/Nelio/Xiaoyun)
- s/reta/RETA/ (Andrew)
- Applied A-by on patch 3 (Slava)
- Fix display of configured hash types
- Add missing flow types definition to testpmd's port info command

Maxime Coquelin (5):
  net/virtio: add initial RSS support
  app/testpmd: fix RSS key length
  app/testpmd: fix RSS type display
  net/mlx5: fix RSS RETA update
  app/testpmd: add missing flow types in port info

 app/test-pmd/cmdline.c                 |   4 +
 app/test-pmd/config.c                  |  11 +-
 doc/guides/nics/features/virtio.ini    |   3 +
 doc/guides/nics/virtio.rst             |   3 +
 doc/guides/rel_notes/release_21_11.rst |   6 +
 drivers/net/mlx5/mlx5_rss.c            |   2 +-
 drivers/net/virtio/virtio.h            |  32 +-
 drivers/net/virtio/virtio_ethdev.c     | 436 ++++++++++++++++++++++++-
 drivers/net/virtio/virtio_ethdev.h     |   3 +-
 drivers/net/virtio/virtqueue.h         |  25 ++
 10 files changed, 513 insertions(+), 12 deletions(-)

-- 
2.31.1

[dpdk-dev] [PATCH v6 1/5] net/virtio: add initial RSS support

[Maxime Coquelin]

Provide the capability to update the hash key, hash types
and RETA table on the fly (without needing to stop/start
the device). However, the key length and the number of RETA
entries are fixed to 40B and 128 entries respectively. This
is done in order to simplify the design, but may be
revisited later as the Virtio spec provides this
flexibility.

Note that only VIRTIO_NET_F_RSS support is implemented,
VIRTIO_NET_F_HASH_REPORT, which would enable reporting the
packet RSS hash calculated by the device into mbuf.rss, is
not yet supported.

Regarding the default RSS configuration, it has been
chosen to use the default Intel ixgbe key as default key,
and default RETA is a simple modulo between the hash and
the number of Rx queues.

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
---
 doc/guides/nics/features/virtio.ini    |   3 +
 doc/guides/nics/virtio.rst             |   3 +
 doc/guides/rel_notes/release_21_11.rst |   6 +
 drivers/net/virtio/virtio.h            |  32 +-
 drivers/net/virtio/virtio_ethdev.c     | 436 ++++++++++++++++++++++++-
 drivers/net/virtio/virtio_ethdev.h     |   3 +-
 drivers/net/virtio/virtqueue.h         |  25 ++
 7 files changed, 499 insertions(+), 9 deletions(-)

diff --git a/doc/guides/nics/features/virtio.ini b/doc/guides/nics/features/virtio.ini
index 48f6f393b1..a5eab4932f 100644
--- a/doc/guides/nics/features/virtio.ini
+++ b/doc/guides/nics/features/virtio.ini
@@ -14,6 +14,9 @@ Promiscuous mode     = Y
 Allmulticast mode    = Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
+RSS hash             = P
+RSS key update       = Y
+RSS reta update      = Y
 VLAN filter          = Y
 Basic stats          = Y
 Stats per queue      = Y
diff --git a/doc/guides/nics/virtio.rst b/doc/guides/nics/virtio.rst
index 82ce7399ce..98e0d012b7 100644
--- a/doc/guides/nics/virtio.rst
+++ b/doc/guides/nics/virtio.rst
@@ -73,6 +73,9 @@ In this release, the virtio PMD driver provides the basic functionality of packe
 
 *   Virtio supports using port IO to get PCI resource when UIO module is not available.
 
+*   Virtio supports RSS Rx mode with 40B configurable hash key length, 128
+    configurable RETA entries and configurable hash types.
+
 Prerequisites
 -------------
 
diff --git a/doc/guides/rel_notes/release_21_11.rst b/doc/guides/rel_notes/release_21_11.rst
index 1ccac87b73..bfdd0fe0e2 100644
--- a/doc/guides/rel_notes/release_21_11.rst
+++ b/doc/guides/rel_notes/release_21_11.rst
@@ -305,6 +305,12 @@ New Features
     * Pcapng format with timestamps and meta-data.
     * Fixes packet capture with stripped VLAN tags.
 
+* **Added initial RSS support to Virtio PMD.**
+
+  Initial support for RSS receive mode has been added to the Virtio PMD,
+  with the capability for the application to configure the hash key, the
+  RETA and the hash types. Virtio hash reporting is yet to be added.
+
 
 Removed Items
 -------------
diff --git a/drivers/net/virtio/virtio.h b/drivers/net/virtio/virtio.h
index e78b2e429e..9a2ab2caea 100644
--- a/drivers/net/virtio/virtio.h
+++ b/drivers/net/virtio/virtio.h
@@ -30,6 +30,7 @@
 #define VIRTIO_NET_F_GUEST_ANNOUNCE 21	/* Guest can announce device on the network */
 #define VIRTIO_NET_F_MQ		22	/* Device supports Receive Flow Steering */
 #define VIRTIO_NET_F_CTRL_MAC_ADDR 23	/* Set MAC address */
+#define VIRTIO_NET_F_RSS	60	/* RSS supported */
 
 /*
  * Do we get callbacks when the ring is completely used,
@@ -100,6 +101,29 @@
  */
 #define VIRTIO_MAX_INDIRECT ((int)(rte_mem_page_size() / 16))
 
+/*  Virtio RSS hash types */
+#define VIRTIO_NET_HASH_TYPE_IPV4	RTE_BIT32(0)
+#define VIRTIO_NET_HASH_TYPE_TCPV4	RTE_BIT32(1)
+#define VIRTIO_NET_HASH_TYPE_UDPV4	RTE_BIT32(2)
+#define VIRTIO_NET_HASH_TYPE_IPV6	RTE_BIT32(3)
+#define VIRTIO_NET_HASH_TYPE_TCPV6	RTE_BIT32(4)
+#define VIRTIO_NET_HASH_TYPE_UDPV6	RTE_BIT32(5)
+#define VIRTIO_NET_HASH_TYPE_IP_EX	RTE_BIT32(6)
+#define VIRTIO_NET_HASH_TYPE_TCP_EX	RTE_BIT32(7)
+#define VIRTIO_NET_HASH_TYPE_UDP_EX	RTE_BIT32(8)
+
+#define VIRTIO_NET_HASH_TYPE_MASK ( \
+	VIRTIO_NET_HASH_TYPE_IPV4 | \
+	VIRTIO_NET_HASH_TYPE_TCPV4 | \
+	VIRTIO_NET_HASH_TYPE_UDPV4 | \
+	VIRTIO_NET_HASH_TYPE_IPV6 | \
+	VIRTIO_NET_HASH_TYPE_TCPV6 | \
+	VIRTIO_NET_HASH_TYPE_UDPV6 | \
+	VIRTIO_NET_HASH_TYPE_IP_EX | \
+	VIRTIO_NET_HASH_TYPE_TCP_EX | \
+	VIRTIO_NET_HASH_TYPE_UDP_EX)
+
+
 /*
  * Maximum number of virtqueues per device.
  */
@@ -157,7 +181,9 @@ struct virtio_net_config {
 	 * Any other value stands for unknown.
 	 */
 	uint8_t duplex;
-
+	uint8_t rss_max_key_size;
+	uint16_t rss_max_indirection_table_length;
+	uint32_t supported_hash_types;
 } __rte_packed;
 
 struct virtio_hw {
@@ -190,6 +216,10 @@ struct virtio_hw {
 	rte_spinlock_t state_lock;
 	struct rte_mbuf **inject_pkts;
 	uint16_t max_queue_pairs;
+	uint16_t rss_rx_queues;
+	uint32_t rss_hash_types;
+	uint16_t *rss_reta;
+	uint8_t *rss_key;
 	uint64_t req_guest_features;
 	struct virtnet_ctl *cvq;
 	bool use_va;
diff --git a/drivers/net/virtio/virtio_ethdev.c b/drivers/net/virtio/virtio_ethdev.c
index 26de006c77..08e4c28946 100644
--- a/drivers/net/virtio/virtio_ethdev.c
+++ b/drivers/net/virtio/virtio_ethdev.c
@@ -51,6 +51,16 @@ static int virtio_dev_info_get(struct rte_eth_dev *dev,
 static int virtio_dev_link_update(struct rte_eth_dev *dev,
 	int wait_to_complete);
 static int virtio_dev_vlan_offload_set(struct rte_eth_dev *dev, int mask);
+static int virtio_dev_rss_hash_update(struct rte_eth_dev *dev,
+		struct rte_eth_rss_conf *rss_conf);
+static int virtio_dev_rss_hash_conf_get(struct rte_eth_dev *dev,
+		struct rte_eth_rss_conf *rss_conf);
+static int virtio_dev_rss_reta_update(struct rte_eth_dev *dev,
+			 struct rte_eth_rss_reta_entry64 *reta_conf,
+			 uint16_t reta_size);
+static int virtio_dev_rss_reta_query(struct rte_eth_dev *dev,
+			 struct rte_eth_rss_reta_entry64 *reta_conf,
+			 uint16_t reta_size);
 
 static void virtio_set_hwaddr(struct virtio_hw *hw);
 static void virtio_get_hwaddr(struct virtio_hw *hw);
@@ -347,20 +357,52 @@ virtio_send_command(struct virtnet_ctl *cvq, struct virtio_pmd_ctrl *ctrl,
 }
 
 static int
-virtio_set_multiple_queues(struct rte_eth_dev *dev, uint16_t nb_queues)
+virtio_set_multiple_queues_rss(struct rte_eth_dev *dev, uint16_t nb_queues)
 {
 	struct virtio_hw *hw = dev->data->dev_private;
 	struct virtio_pmd_ctrl ctrl;
-	int dlen[1];
+	struct virtio_net_ctrl_rss rss;
+	int dlen, ret;
+
+	rss.hash_types = hw->rss_hash_types & VIRTIO_NET_HASH_TYPE_MASK;
+	RTE_BUILD_BUG_ON(!RTE_IS_POWER_OF_2(VIRTIO_NET_RSS_RETA_SIZE));
+	rss.indirection_table_mask = VIRTIO_NET_RSS_RETA_SIZE - 1;
+	rss.unclassified_queue = 0;
+	memcpy(rss.indirection_table, hw->rss_reta, VIRTIO_NET_RSS_RETA_SIZE * sizeof(uint16_t));
+	rss.max_tx_vq = nb_queues;
+	rss.hash_key_length = VIRTIO_NET_RSS_KEY_SIZE;
+	memcpy(rss.hash_key_data, hw->rss_key, VIRTIO_NET_RSS_KEY_SIZE);
+
+	ctrl.hdr.class = VIRTIO_NET_CTRL_MQ;
+	ctrl.hdr.cmd = VIRTIO_NET_CTRL_MQ_RSS_CONFIG;
+	memcpy(ctrl.data, &rss, sizeof(rss));
+
+	dlen = sizeof(rss);
+
+	ret = virtio_send_command(hw->cvq, &ctrl, &dlen, 1);
+	if (ret) {
+		PMD_INIT_LOG(ERR, "RSS multiqueue configured but send command failed");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int
+virtio_set_multiple_queues_auto(struct rte_eth_dev *dev, uint16_t nb_queues)
+{
+	struct virtio_hw *hw = dev->data->dev_private;
+	struct virtio_pmd_ctrl ctrl;
+	int dlen;
 	int ret;
 
 	ctrl.hdr.class = VIRTIO_NET_CTRL_MQ;
 	ctrl.hdr.cmd = VIRTIO_NET_CTRL_MQ_VQ_PAIRS_SET;
 	memcpy(ctrl.data, &nb_queues, sizeof(uint16_t));
 
-	dlen[0] = sizeof(uint16_t);
+	dlen = sizeof(uint16_t);
 
-	ret = virtio_send_command(hw->cvq, &ctrl, dlen, 1);
+	ret = virtio_send_command(hw->cvq, &ctrl, &dlen, 1);
 	if (ret) {
 		PMD_INIT_LOG(ERR, "Multiqueue configured but send command "
 			  "failed, this is too late now...");
@@ -370,6 +412,17 @@ virtio_set_multiple_queues(struct rte_eth_dev *dev, uint16_t nb_queues)
 	return 0;
 }
 
+static int
+virtio_set_multiple_queues(struct rte_eth_dev *dev, uint16_t nb_queues)
+{
+	struct virtio_hw *hw = dev->data->dev_private;
+
+	if (virtio_with_feature(hw, VIRTIO_NET_F_RSS))
+		return virtio_set_multiple_queues_rss(dev, nb_queues);
+	else
+		return virtio_set_multiple_queues_auto(dev, nb_queues);
+}
+
 static uint16_t
 virtio_get_nr_vq(struct virtio_hw *hw)
 {
@@ -708,6 +761,16 @@ virtio_alloc_queues(struct rte_eth_dev *dev)
 
 static void virtio_queues_unbind_intr(struct rte_eth_dev *dev);
 
+static void
+virtio_free_rss(struct virtio_hw *hw)
+{
+	rte_free(hw->rss_key);
+	hw->rss_key = NULL;
+
+	rte_free(hw->rss_reta);
+	hw->rss_reta = NULL;
+}
+
 int
 virtio_dev_close(struct rte_eth_dev *dev)
 {
@@ -737,6 +800,7 @@ virtio_dev_close(struct rte_eth_dev *dev)
 	virtio_reset(hw);
 	virtio_dev_free_mbufs(dev);
 	virtio_free_queues(hw);
+	virtio_free_rss(hw);
 
 	return VIRTIO_OPS(hw)->dev_close(hw);
 }
@@ -977,6 +1041,10 @@ static const struct eth_dev_ops virtio_eth_dev_ops = {
 	.rx_queue_intr_enable    = virtio_dev_rx_queue_intr_enable,
 	.rx_queue_intr_disable   = virtio_dev_rx_queue_intr_disable,
 	.tx_queue_setup          = virtio_dev_tx_queue_setup,
+	.rss_hash_update         = virtio_dev_rss_hash_update,
+	.rss_hash_conf_get       = virtio_dev_rss_hash_conf_get,
+	.reta_update             = virtio_dev_rss_reta_update,
+	.reta_query              = virtio_dev_rss_reta_query,
 	/* collect stats per queue */
 	.queue_stats_mapping_set = virtio_dev_queue_stats_mapping_set,
 	.vlan_filter_set         = virtio_vlan_filter_set,
@@ -1718,6 +1786,332 @@ virtio_configure_intr(struct rte_eth_dev *dev)
 
 	return 0;
 }
+
+static uint64_t
+ethdev_to_virtio_rss_offloads(uint64_t ethdev_hash_types)
+{
+	uint64_t virtio_hash_types = 0;
+
+	if (ethdev_hash_types & (ETH_RSS_IPV4 | ETH_RSS_FRAG_IPV4 | ETH_RSS_NONFRAG_IPV4_OTHER))
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IPV4;
+
+	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV4_TCP)
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCPV4;
+
+	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV4_UDP)
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDPV4;
+
+	if (ethdev_hash_types & (ETH_RSS_IPV6 | ETH_RSS_FRAG_IPV6 | ETH_RSS_NONFRAG_IPV6_OTHER))
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IPV6;
+
+	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV6_TCP)
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCPV6;
+
+	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV6_UDP)
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDPV6;
+
+	if (ethdev_hash_types & ETH_RSS_IPV6_EX)
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IP_EX;
+
+	if (ethdev_hash_types & ETH_RSS_IPV6_TCP_EX)
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCP_EX;
+
+	if (ethdev_hash_types & ETH_RSS_IPV6_UDP_EX)
+		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDP_EX;
+
+	return virtio_hash_types;
+}
+
+static uint64_t
+virtio_to_ethdev_rss_offloads(uint64_t virtio_hash_types)
+{
+	uint64_t rss_offloads = 0;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_IPV4)
+		rss_offloads |= ETH_RSS_IPV4 | ETH_RSS_FRAG_IPV4 | ETH_RSS_NONFRAG_IPV4_OTHER;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_TCPV4)
+		rss_offloads |= ETH_RSS_NONFRAG_IPV4_TCP;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_UDPV4)
+		rss_offloads |= ETH_RSS_NONFRAG_IPV4_UDP;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_IPV6)
+		rss_offloads |= ETH_RSS_IPV6 | ETH_RSS_FRAG_IPV6 | ETH_RSS_NONFRAG_IPV6_OTHER;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_TCPV6)
+		rss_offloads |= ETH_RSS_NONFRAG_IPV6_TCP;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_UDPV6)
+		rss_offloads |= ETH_RSS_NONFRAG_IPV6_UDP;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_IP_EX)
+		rss_offloads |= ETH_RSS_IPV6_EX;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_TCP_EX)
+		rss_offloads |= ETH_RSS_IPV6_TCP_EX;
+
+	if (virtio_hash_types & VIRTIO_NET_HASH_TYPE_UDP_EX)
+		rss_offloads |= ETH_RSS_IPV6_UDP_EX;
+
+	return rss_offloads;
+}
+
+static int
+virtio_dev_get_rss_config(struct virtio_hw *hw, uint32_t *rss_hash_types)
+{
+	struct virtio_net_config local_config;
+	struct virtio_net_config *config = &local_config;
+
+	virtio_read_dev_config(hw,
+			offsetof(struct virtio_net_config, rss_max_key_size),
+			&config->rss_max_key_size,
+			sizeof(config->rss_max_key_size));
+	if (config->rss_max_key_size < VIRTIO_NET_RSS_KEY_SIZE) {
+		PMD_INIT_LOG(ERR, "Invalid device RSS max key size (%u)",
+				config->rss_max_key_size);
+		return -EINVAL;
+	}
+
+	virtio_read_dev_config(hw,
+			offsetof(struct virtio_net_config,
+				rss_max_indirection_table_length),
+			&config->rss_max_indirection_table_length,
+			sizeof(config->rss_max_indirection_table_length));
+	if (config->rss_max_indirection_table_length < VIRTIO_NET_RSS_RETA_SIZE) {
+		PMD_INIT_LOG(ERR, "Invalid device RSS max reta size (%u)",
+				config->rss_max_indirection_table_length);
+		return -EINVAL;
+	}
+
+	virtio_read_dev_config(hw,
+			offsetof(struct virtio_net_config, supported_hash_types),
+			&config->supported_hash_types,
+			sizeof(config->supported_hash_types));
+	if ((config->supported_hash_types & VIRTIO_NET_HASH_TYPE_MASK) == 0) {
+		PMD_INIT_LOG(ERR, "Invalid device RSS hash types (0x%x)",
+				config->supported_hash_types);
+		return -EINVAL;
+	}
+
+	*rss_hash_types = config->supported_hash_types & VIRTIO_NET_HASH_TYPE_MASK;
+
+	PMD_INIT_LOG(DEBUG, "Device RSS config:");
+	PMD_INIT_LOG(DEBUG, "\t-Max key size: %u", config->rss_max_key_size);
+	PMD_INIT_LOG(DEBUG, "\t-Max reta size: %u", config->rss_max_indirection_table_length);
+	PMD_INIT_LOG(DEBUG, "\t-Supported hash types: 0x%x", *rss_hash_types);
+
+	return 0;
+}
+
+static int
+virtio_dev_rss_hash_update(struct rte_eth_dev *dev,
+		struct rte_eth_rss_conf *rss_conf)
+{
+	struct virtio_hw *hw = dev->data->dev_private;
+	char old_rss_key[VIRTIO_NET_RSS_KEY_SIZE];
+	uint32_t old_hash_types;
+	uint16_t nb_queues;
+	int ret;
+
+	if (!virtio_with_feature(hw, VIRTIO_NET_F_RSS))
+		return -ENOTSUP;
+
+	if (rss_conf->rss_hf & ~virtio_to_ethdev_rss_offloads(VIRTIO_NET_HASH_TYPE_MASK))
+		return -EINVAL;
+
+	old_hash_types = hw->rss_hash_types;
+	hw->rss_hash_types = ethdev_to_virtio_rss_offloads(rss_conf->rss_hf);
+
+	if (rss_conf->rss_key && rss_conf->rss_key_len) {
+		if (rss_conf->rss_key_len != VIRTIO_NET_RSS_KEY_SIZE) {
+			PMD_INIT_LOG(ERR, "Driver only supports %u RSS key length",
+					VIRTIO_NET_RSS_KEY_SIZE);
+			ret = -EINVAL;
+			goto restore_types;
+		}
+		memcpy(old_rss_key, hw->rss_key, VIRTIO_NET_RSS_KEY_SIZE);
+		memcpy(hw->rss_key, rss_conf->rss_key, VIRTIO_NET_RSS_KEY_SIZE);
+	}
+
+	nb_queues = RTE_MAX(dev->data->nb_rx_queues, dev->data->nb_tx_queues);
+	ret = virtio_set_multiple_queues_rss(dev, nb_queues);
+	if (ret < 0) {
+		PMD_INIT_LOG(ERR, "Failed to apply new RSS config to the device");
+		goto restore_key;
+	}
+
+	return 0;
+restore_key:
+	memcpy(hw->rss_key, old_rss_key, VIRTIO_NET_RSS_KEY_SIZE);
+restore_types:
+	hw->rss_hash_types = old_hash_types;
+
+	return ret;
+}
+
+static int
+virtio_dev_rss_hash_conf_get(struct rte_eth_dev *dev,
+		struct rte_eth_rss_conf *rss_conf)
+{
+	struct virtio_hw *hw = dev->data->dev_private;
+
+	if (!virtio_with_feature(hw, VIRTIO_NET_F_RSS))
+		return -ENOTSUP;
+
+	if (rss_conf->rss_key && rss_conf->rss_key_len >= VIRTIO_NET_RSS_KEY_SIZE)
+		memcpy(rss_conf->rss_key, hw->rss_key, VIRTIO_NET_RSS_KEY_SIZE);
+	rss_conf->rss_key_len = VIRTIO_NET_RSS_KEY_SIZE;
+	rss_conf->rss_hf = virtio_to_ethdev_rss_offloads(hw->rss_hash_types);
+
+	return 0;
+}
+
+static int virtio_dev_rss_reta_update(struct rte_eth_dev *dev,
+			 struct rte_eth_rss_reta_entry64 *reta_conf,
+			 uint16_t reta_size)
+{
+	struct virtio_hw *hw = dev->data->dev_private;
+	uint16_t nb_queues;
+	uint16_t old_reta[VIRTIO_NET_RSS_RETA_SIZE];
+	int idx, pos, i, ret;
+
+	if (!virtio_with_feature(hw, VIRTIO_NET_F_RSS))
+		return -ENOTSUP;
+
+	if (reta_size != VIRTIO_NET_RSS_RETA_SIZE)
+		return -EINVAL;
+
+	memcpy(old_reta, hw->rss_reta, sizeof(old_reta));
+
+	for (i = 0; i < reta_size; i++) {
+		idx = i / RTE_ETH_RETA_GROUP_SIZE;
+		pos = i % RTE_ETH_RETA_GROUP_SIZE;
+
+		if (((reta_conf[idx].mask >> pos) & 0x1) == 0)
+			continue;
+
+		hw->rss_reta[i] = reta_conf[idx].reta[pos];
+	}
+
+	nb_queues = RTE_MAX(dev->data->nb_rx_queues, dev->data->nb_tx_queues);
+	ret = virtio_set_multiple_queues_rss(dev, nb_queues);
+	if (ret < 0) {
+		PMD_INIT_LOG(ERR, "Failed to apply new RETA to the device");
+		memcpy(hw->rss_reta, old_reta, sizeof(old_reta));
+	}
+
+	hw->rss_rx_queues = dev->data->nb_rx_queues;
+
+	return ret;
+}
+
+static int virtio_dev_rss_reta_query(struct rte_eth_dev *dev,
+			 struct rte_eth_rss_reta_entry64 *reta_conf,
+			 uint16_t reta_size)
+{
+	struct virtio_hw *hw = dev->data->dev_private;
+	int idx, i;
+
+	if (!virtio_with_feature(hw, VIRTIO_NET_F_RSS))
+		return -ENOTSUP;
+
+	if (reta_size != VIRTIO_NET_RSS_RETA_SIZE)
+		return -EINVAL;
+
+	for (i = 0; i < reta_size; i++) {
+		idx = i / RTE_ETH_RETA_GROUP_SIZE;
+		reta_conf[idx].reta[i % RTE_ETH_RETA_GROUP_SIZE] = hw->rss_reta[i];
+	}
+
+	return 0;
+}
+
+/*
+ * As default RSS hash key, it uses the default key of the
+ * Intel IXGBE devices. It can be updated by the application
+ * with any 40B key value.
+ */
+static uint8_t rss_intel_key[VIRTIO_NET_RSS_KEY_SIZE] = {
+	0x6D, 0x5A, 0x56, 0xDA, 0x25, 0x5B, 0x0E, 0xC2,
+	0x41, 0x67, 0x25, 0x3D, 0x43, 0xA3, 0x8F, 0xB0,
+	0xD0, 0xCA, 0x2B, 0xCB, 0xAE, 0x7B, 0x30, 0xB4,
+	0x77, 0xCB, 0x2D, 0xA3, 0x80, 0x30, 0xF2, 0x0C,
+	0x6A, 0x42, 0xB7, 0x3B, 0xBE, 0xAC, 0x01, 0xFA,
+};
+
+static int
+virtio_dev_rss_init(struct rte_eth_dev *eth_dev)
+{
+	struct virtio_hw *hw = eth_dev->data->dev_private;
+	uint16_t nb_rx_queues = eth_dev->data->nb_rx_queues;
+	struct rte_eth_rss_conf *rss_conf;
+	int ret, i;
+
+	if (!nb_rx_queues) {
+		PMD_INIT_LOG(ERR, "Cannot init RSS if no Rx queues");
+		return -EINVAL;
+	}
+
+	rss_conf = &eth_dev->data->dev_conf.rx_adv_conf.rss_conf;
+
+	ret = virtio_dev_get_rss_config(hw, &hw->rss_hash_types);
+	if (ret)
+		return ret;
+
+	if (rss_conf->rss_hf) {
+		/*  Ensure requested hash types are supported by the device */
+		if (rss_conf->rss_hf & ~virtio_to_ethdev_rss_offloads(hw->rss_hash_types))
+			return -EINVAL;
+
+		hw->rss_hash_types = ethdev_to_virtio_rss_offloads(rss_conf->rss_hf);
+	}
+
+	if (!hw->rss_key) {
+		/* Setup default RSS key if not already setup by the user */
+		hw->rss_key = rte_malloc_socket("rss_key",
+				VIRTIO_NET_RSS_KEY_SIZE, 0,
+				eth_dev->device->numa_node);
+		if (!hw->rss_key) {
+			PMD_INIT_LOG(ERR, "Failed to allocate RSS key");
+			return -1;
+		}
+	}
+
+	if (rss_conf->rss_key && rss_conf->rss_key_len) {
+		if (rss_conf->rss_key_len != VIRTIO_NET_RSS_KEY_SIZE) {
+			PMD_INIT_LOG(ERR, "Driver only supports %u RSS key length",
+					VIRTIO_NET_RSS_KEY_SIZE);
+			return -EINVAL;
+		}
+		memcpy(hw->rss_key, rss_conf->rss_key, VIRTIO_NET_RSS_KEY_SIZE);
+	} else {
+		memcpy(hw->rss_key, rss_intel_key, VIRTIO_NET_RSS_KEY_SIZE);
+	}
+
+	if (!hw->rss_reta) {
+		/* Setup default RSS reta if not already setup by the user */
+		hw->rss_reta = rte_zmalloc_socket("rss_reta",
+				VIRTIO_NET_RSS_RETA_SIZE * sizeof(uint16_t), 0,
+				eth_dev->device->numa_node);
+		if (!hw->rss_reta) {
+			PMD_INIT_LOG(ERR, "Failed to allocate RSS reta");
+			return -1;
+		}
+
+		hw->rss_rx_queues = 0;
+	}
+
+	/* Re-initialize the RSS reta if the number of RX queues has changed */
+	if (hw->rss_rx_queues != nb_rx_queues) {
+		for (i = 0; i < VIRTIO_NET_RSS_RETA_SIZE; i++)
+			hw->rss_reta[i] = i % nb_rx_queues;
+		hw->rss_rx_queues = nb_rx_queues;
+	}
+
+	return 0;
+}
+
 #define DUPLEX_UNKNOWN   0xff
 /* reset device and renegotiate features if needed */
 static int
@@ -1805,14 +2199,15 @@ virtio_init_device(struct rte_eth_dev *eth_dev, uint64_t req_features)
 			config->status = 0;
 		}
 
-		if (virtio_with_feature(hw, VIRTIO_NET_F_MQ)) {
+		if (virtio_with_feature(hw, VIRTIO_NET_F_MQ) ||
+				virtio_with_feature(hw, VIRTIO_NET_F_RSS)) {
 			virtio_read_dev_config(hw,
 				offsetof(struct virtio_net_config, max_virtqueue_pairs),
 				&config->max_virtqueue_pairs,
 				sizeof(config->max_virtqueue_pairs));
 		} else {
 			PMD_INIT_LOG(DEBUG,
-				     "VIRTIO_NET_F_MQ is not supported");
+				     "Neither VIRTIO_NET_F_MQ nor VIRTIO_NET_F_RSS are supported");
 			config->max_virtqueue_pairs = 1;
 		}
 
@@ -1844,6 +2239,11 @@ virtio_init_device(struct rte_eth_dev *eth_dev, uint64_t req_features)
 				VLAN_TAG_LEN - hw->vtnet_hdr_size;
 		}
 
+		hw->rss_hash_types = 0;
+		if (virtio_with_feature(hw, VIRTIO_NET_F_RSS))
+			if (virtio_dev_rss_init(eth_dev))
+				return -1;
+
 		PMD_INIT_LOG(DEBUG, "config->max_virtqueue_pairs=%d",
 				config->max_virtqueue_pairs);
 		PMD_INIT_LOG(DEBUG, "config->status=%d", config->status);
@@ -2086,7 +2486,7 @@ virtio_dev_configure(struct rte_eth_dev *dev)
 	PMD_INIT_LOG(DEBUG, "configure");
 	req_features = VIRTIO_PMD_DEFAULT_GUEST_FEATURES;
 
-	if (rxmode->mq_mode != RTE_ETH_MQ_RX_NONE) {
+	if (rxmode->mq_mode != RTE_ETH_MQ_RX_NONE && rxmode->mq_mode != RTE_ETH_MQ_RX_RSS) {
 		PMD_DRV_LOG(ERR,
 			"Unsupported Rx multi queue mode %d",
 			rxmode->mq_mode);
@@ -2106,6 +2506,9 @@ virtio_dev_configure(struct rte_eth_dev *dev)
 			return ret;
 	}
 
+	if (rxmode->mq_mode == RTE_ETH_MQ_RX_RSS)
+		req_features |= (1ULL << VIRTIO_NET_F_RSS);
+
 	if (rxmode->mtu > hw->max_mtu)
 		req_features &= ~(1ULL << VIRTIO_NET_F_MTU);
 
@@ -2136,6 +2539,12 @@ virtio_dev_configure(struct rte_eth_dev *dev)
 			return ret;
 	}
 
+	if ((rxmode->mq_mode & RTE_ETH_MQ_RX_RSS_FLAG) &&
+			!virtio_with_feature(hw, VIRTIO_NET_F_RSS)) {
+		PMD_DRV_LOG(ERR, "RSS support requested but not supported by the device");
+		return -ENOTSUP;
+	}
+
 	if ((rx_offloads & (RTE_ETH_RX_OFFLOAD_UDP_CKSUM |
 			    RTE_ETH_RX_OFFLOAD_TCP_CKSUM)) &&
 		!virtio_with_feature(hw, VIRTIO_NET_F_GUEST_CSUM)) {
@@ -2533,6 +2942,7 @@ static int
 virtio_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
 {
 	uint64_t tso_mask, host_features;
+	uint32_t rss_hash_types = 0;
 	struct virtio_hw *hw = dev->data->dev_private;
 	dev_info->speed_capa = virtio_dev_speed_capa_get(hw->speed);
 
@@ -2573,6 +2983,18 @@ virtio_dev_info_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
 	if ((host_features & tso_mask) == tso_mask)
 		dev_info->tx_offload_capa |= RTE_ETH_TX_OFFLOAD_TCP_TSO;
 
+	if (host_features & (1ULL << VIRTIO_NET_F_RSS)) {
+		virtio_dev_get_rss_config(hw, &rss_hash_types);
+		dev_info->hash_key_size = VIRTIO_NET_RSS_KEY_SIZE;
+		dev_info->reta_size = VIRTIO_NET_RSS_RETA_SIZE;
+		dev_info->flow_type_rss_offloads =
+			virtio_to_ethdev_rss_offloads(rss_hash_types);
+	} else {
+		dev_info->hash_key_size = 0;
+		dev_info->reta_size = 0;
+		dev_info->flow_type_rss_offloads = 0;
+	}
+
 	if (host_features & (1ULL << VIRTIO_F_RING_PACKED)) {
 		/*
 		 * According to 2.7 Packed Virtqueues,
diff --git a/drivers/net/virtio/virtio_ethdev.h b/drivers/net/virtio/virtio_ethdev.h
index 40be484218..c08f382791 100644
--- a/drivers/net/virtio/virtio_ethdev.h
+++ b/drivers/net/virtio/virtio_ethdev.h
@@ -45,7 +45,8 @@
 	 1u << VIRTIO_NET_F_GUEST_TSO6     |	\
 	 1u << VIRTIO_NET_F_CSUM           |	\
 	 1u << VIRTIO_NET_F_HOST_TSO4      |	\
-	 1u << VIRTIO_NET_F_HOST_TSO6)
+	 1u << VIRTIO_NET_F_HOST_TSO6      |	\
+	 1ULL << VIRTIO_NET_F_RSS)
 
 extern const struct eth_dev_ops virtio_user_secondary_eth_dev_ops;
 
diff --git a/drivers/net/virtio/virtqueue.h b/drivers/net/virtio/virtqueue.h
index 97fc264e46..855f57a956 100644
--- a/drivers/net/virtio/virtqueue.h
+++ b/drivers/net/virtio/virtqueue.h
@@ -201,6 +201,28 @@ struct virtio_net_ctrl_mac {
 #define VIRTIO_NET_CTRL_VLAN_ADD 0
 #define VIRTIO_NET_CTRL_VLAN_DEL 1
 
+/**
+ * RSS control
+ *
+ * The RSS feature configuration message is sent by the driver when
+ * VIRTIO_NET_F_RSS has been negotiated. It provides the device with
+ * hash types to use, hash key and indirection table. In this
+ * implementation, the driver only supports fixed key length (40B)
+ * and indirection table size (128 entries).
+ */
+#define VIRTIO_NET_RSS_RETA_SIZE 128
+#define VIRTIO_NET_RSS_KEY_SIZE 40
+
+struct virtio_net_ctrl_rss {
+	uint32_t hash_types;
+	uint16_t indirection_table_mask;
+	uint16_t unclassified_queue;
+	uint16_t indirection_table[VIRTIO_NET_RSS_RETA_SIZE];
+	uint16_t max_tx_vq;
+	uint8_t hash_key_length;
+	uint8_t hash_key_data[VIRTIO_NET_RSS_KEY_SIZE];
+};
+
 /*
  * Control link announce acknowledgement
  *
@@ -292,7 +314,10 @@ struct virtqueue {
 
 /* If multiqueue is provided by host, then we suppport it. */
 #define VIRTIO_NET_CTRL_MQ   4
+
 #define VIRTIO_NET_CTRL_MQ_VQ_PAIRS_SET        0
+#define VIRTIO_NET_CTRL_MQ_RSS_CONFIG          1
+
 #define VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MIN        1
 #define VIRTIO_NET_CTRL_MQ_VQ_PAIRS_MAX        0x8000
 
-- 
2.31.1

[dpdk-dev] [PATCH v6 2/5] app/testpmd: fix RSS key length

[Maxime Coquelin]

port_rss_hash_key_update() initializes rss_conf with the
RSS key configuration provided  by the user, but it calls
rte_eth_dev_rss_hash_conf_get() before calling
rte_eth_dev_rss_hash_update(), which overrides the parsed
RSS config.

While the RSS key value is set again after, this is not
the case of the key length. It could cause out of bounds
access if the key length parsed is smaller than the one
read from rte_eth_dev_rss_hash_conf_get().

This patch restores the key length before the
rte_eth_dev_rss_hash_update() call to ensure the RSS key
value/length pair is consistent.

Fixes: 8205e241b2b0 ("app/testpmd: add missing type to RSS hash commands")
Cc: stable@dpdk.org

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Xiaoyun Li <xiaoyun.li@intel.com>
Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
---
 app/test-pmd/config.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index a18871d461..e5c58ceb37 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -3040,7 +3040,7 @@ port_rss_hash_key_update(portid_t port_id, char rss_type[], uint8_t *hash_key,
 	unsigned int i;
 
 	rss_conf.rss_key = NULL;
-	rss_conf.rss_key_len = hash_key_len;
+	rss_conf.rss_key_len = 0;
 	rss_conf.rss_hf = 0;
 	for (i = 0; rss_type_table[i].str; i++) {
 		if (!strcmp(rss_type_table[i].str, rss_type))
@@ -3049,6 +3049,7 @@ port_rss_hash_key_update(portid_t port_id, char rss_type[], uint8_t *hash_key,
 	diag = rte_eth_dev_rss_hash_conf_get(port_id, &rss_conf);
 	if (diag == 0) {
 		rss_conf.rss_key = hash_key;
+		rss_conf.rss_key_len = hash_key_len;
 		diag = rte_eth_dev_rss_hash_update(port_id, &rss_conf);
 	}
 	if (diag == 0)
-- 
2.31.1

[dpdk-dev] [PATCH v6 3/5] app/testpmd: fix RSS type display

[Maxime Coquelin]

This patch fixes the display of the RSS hash types
configured in the port, which displayed "all" even
if only a single type was configured

Fixes: 3c90743dd3b9 ("app/testpmd: support more types for flow RSS")
Cc: stable@dpdk.org

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Xiaoyun Li <xiaoyun.li@intel.com>
Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
---
 app/test-pmd/config.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index e5c58ceb37..2e52664964 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -3019,7 +3019,9 @@ port_rss_hash_conf_show(portid_t port_id, int show_rss_key)
 	}
 	printf("RSS functions:\n ");
 	for (i = 0; rss_type_table[i].str; i++) {
-		if (rss_hf & rss_type_table[i].rss_type)
+		if (rss_type_table[i].rss_type == 0)
+			continue;
+		if ((rss_hf & rss_type_table[i].rss_type) == rss_type_table[i].rss_type)
 			printf("%s ", rss_type_table[i].str);
 	}
 	printf("\n");
-- 
2.31.1

[dpdk-dev] [PATCH v6 4/5] net/mlx5: fix RSS RETA update

[Maxime Coquelin]

This patch fixes RETA updating for entries above 64.
Without ithat, these entries are never updated as
calculated mask value will always be 0.

Fixes: 634efbc2c8c0 ("mlx5: support RETA query and update")
Cc: stable@dpdk.org
Cc: nelio.laranjeiro@6wind.com

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
---
 drivers/net/mlx5/mlx5_rss.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/mlx5/mlx5_rss.c b/drivers/net/mlx5/mlx5_rss.c
index a4f690039e..a04e22398d 100644
--- a/drivers/net/mlx5/mlx5_rss.c
+++ b/drivers/net/mlx5/mlx5_rss.c
@@ -211,7 +211,7 @@ mlx5_dev_rss_reta_update(struct rte_eth_dev *dev,
 	for (idx = 0, i = 0; (i != reta_size); ++i) {
 		idx = i / RTE_ETH_RETA_GROUP_SIZE;
 		pos = i % RTE_ETH_RETA_GROUP_SIZE;
-		if (((reta_conf[idx].mask >> i) & 0x1) == 0)
+		if (((reta_conf[idx].mask >> pos) & 0x1) == 0)
 			continue;
 		MLX5_ASSERT(reta_conf[idx].reta[pos] < priv->rxqs_n);
 		(*priv->reta_idx)[i] = reta_conf[idx].reta[pos];
-- 
2.31.1

[dpdk-dev] [PATCH v6 5/5] app/testpmd: add missing flow types in port info

[Maxime Coquelin]

This patch adds missing IPv6-Ex and GTPU flow types to port
info command. It also add the same definitions to
str2flowtype(), used to configure flow director.

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Xiaoyun Li <xiaoyun.li@intel.com>
---
 app/test-pmd/cmdline.c | 4 ++++
 app/test-pmd/config.c  | 4 ++++
 2 files changed, 8 insertions(+)

diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index ebea13f86a..d08a1c9777 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -10438,6 +10438,10 @@ str2flowtype(char *string)
 		{"ipv6-sctp", RTE_ETH_FLOW_NONFRAG_IPV6_SCTP},
 		{"ipv6-other", RTE_ETH_FLOW_NONFRAG_IPV6_OTHER},
 		{"l2_payload", RTE_ETH_FLOW_L2_PAYLOAD},
+		{"ipv6-ex", RTE_ETH_FLOW_IPV6_EX},
+		{"ipv6-tcp-ex", RTE_ETH_FLOW_IPV6_TCP_EX},
+		{"ipv6-udp-ex", RTE_ETH_FLOW_IPV6_UDP_EX},
+		{"gtpu", RTE_ETH_FLOW_GTPU},
 	};
 
 	for (i = 0; i < RTE_DIM(flowtype_str); i++) {
diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index 2e52664964..11968a160a 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -4819,11 +4819,15 @@ flowtype_to_str(uint16_t flow_type)
 		{"ipv6-sctp", RTE_ETH_FLOW_NONFRAG_IPV6_SCTP},
 		{"ipv6-other", RTE_ETH_FLOW_NONFRAG_IPV6_OTHER},
 		{"l2_payload", RTE_ETH_FLOW_L2_PAYLOAD},
+		{"ipv6-ex", RTE_ETH_FLOW_IPV6_EX},
+		{"ipv6-tcp-ex", RTE_ETH_FLOW_IPV6_TCP_EX},
+		{"ipv6-udp-ex", RTE_ETH_FLOW_IPV6_UDP_EX},
 		{"port", RTE_ETH_FLOW_PORT},
 		{"vxlan", RTE_ETH_FLOW_VXLAN},
 		{"geneve", RTE_ETH_FLOW_GENEVE},
 		{"nvgre", RTE_ETH_FLOW_NVGRE},
 		{"vxlan-gpe", RTE_ETH_FLOW_VXLAN_GPE},
+		{"gtpu", RTE_ETH_FLOW_GTPU},
 	};
 
 	for (i = 0; i < RTE_DIM(flowtype_str_table); i++) {
-- 
2.31.1

Re: [dpdk-dev] [PATCH v6 1/5] net/virtio: add initial RSS support

[Andrew Rybchenko]

On 10/27/21 5:22 PM, Maxime Coquelin wrote:
> Provide the capability to update the hash key, hash types
> and RETA table on the fly (without needing to stop/start
> the device). However, the key length and the number of RETA
> entries are fixed to 40B and 128 entries respectively. This
> is done in order to simplify the design, but may be
> revisited later as the Virtio spec provides this
> flexibility.
> 
> Note that only VIRTIO_NET_F_RSS support is implemented,
> VIRTIO_NET_F_HASH_REPORT, which would enable reporting the
> packet RSS hash calculated by the device into mbuf.rss, is
> not yet supported.
> 
> Regarding the default RSS configuration, it has been
> chosen to use the default Intel ixgbe key as default key,
> and default RETA is a simple modulo between the hash and
> the number of Rx queues.
> 
> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>

Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>

Re: [dpdk-dev] [PATCH v6 0/5] Virtio PMD RSS support & RSS fixes

[Maxime Coquelin]

On 10/27/21 16:22, Maxime Coquelin wrote:
> This series is mainly adding support for RSS to Virtio PMD
> driver. The two last patches are fixing an issue in testpmd
> that could cause out of bounds access, and fix
> an issue spotted in the mlx5 driver while looking for
> inspiration.
> 
> The first motivation for this series is to eventually
> support RSS down to the Vhost-user library, so that OVS can
> benefit from it. But it will be also useful with vDPA
> devices in the future.
> 
> Regarding the testing, I have tested it with qemu v5.2 from
> Fedora 34. Since libvirt does not support yet enabling RSS
> feature in the Qemu virtio-net device, and this feature is
> disabled by default, the tester can either rebuild the qemu
> package to enable it by default or use the qemu cmdline to
> do the same.
> 
> The tester can use testpmd in icmpecho mode in the guest
> and scapy on the host to inject random traffic on the tap
> interface, e.g.:
> sendp(Ether(src=RandMAC()) / IP(src=RandIP(), dst='192.168.123.9') / UDP(sport=RandShort(), dport=RandShort()), loop=True, iface='vnet7')
> 
> Then it can play with RSS config in testpmd to change the
> RETA, or hash type and see traffic being steered
> accordingly by checking the Rx xstats.
> 
> Changes in v6:
> ==============
> - Rebased on top of main, including convertion to new naming of some defines
> - Convert hash types definition to RTE_BIT32() (Andrew)
> - Add build check for RETA size being a pow2 (Andrew)
> - Restore RSS config if device update fails (Andrew)
> - Prevent RSS config if RXq number == 0 (Andrew)
> - Reconfigure the RETA if the number of RXq changed (Andrew)
> - Apply rss_conf even if the RSS key had previously been configured (Andrew)
> 
> Changes in v5:
> ==============
> - Remove unneeded index init (Chenbo)
> - Improve error print (Chenbo)
> - Add missed comment on RSS ctrl message
> 
> Changes in v4:
> ==============
> - s/GPTU/GTPU/ (Xiaoyun)
> 
> Changes in v3:
> ==============
> - Add applying user-specified RSS conf a device config time (Andrew)
> - Remove useless checks (Chenbo)
> - Clean control message payload dlen variable (Chenbo)
> - Add GTPU offload type (Xiaoyun)
> - Add missing types to str2flowtype() (Xiaoyun)
> 
> Changes in v2:
> ==============
> - Rework patch 2 to keep old behaviour, but fix possible out of bounds due to key length (Andrew/Nelio/Xiaoyun)
> - s/reta/RETA/ (Andrew)
> - Applied A-by on patch 3 (Slava)
> - Fix display of configured hash types
> - Add missing flow types definition to testpmd's port info command
> 
> Maxime Coquelin (5):
>    net/virtio: add initial RSS support
>    app/testpmd: fix RSS key length
>    app/testpmd: fix RSS type display
>    net/mlx5: fix RSS RETA update
>    app/testpmd: add missing flow types in port info
> 
>   app/test-pmd/cmdline.c                 |   4 +
>   app/test-pmd/config.c                  |  11 +-
>   doc/guides/nics/features/virtio.ini    |   3 +
>   doc/guides/nics/virtio.rst             |   3 +
>   doc/guides/rel_notes/release_21_11.rst |   6 +
>   drivers/net/mlx5/mlx5_rss.c            |   2 +-
>   drivers/net/virtio/virtio.h            |  32 +-
>   drivers/net/virtio/virtio_ethdev.c     | 436 ++++++++++++++++++++++++-
>   drivers/net/virtio/virtio_ethdev.h     |   3 +-
>   drivers/net/virtio/virtqueue.h         |  25 ++
>   10 files changed, 513 insertions(+), 12 deletions(-)
> 


Applied to dpdk-next-virtio/main.

Thanks,
Maxime

Re: [dpdk-dev] [PATCH v6 1/5] net/virtio: add initial RSS support

[Ferruh Yigit]

On 10/27/2021 3:22 PM, Maxime Coquelin wrote:
> Provide the capability to update the hash key, hash types
> and RETA table on the fly (without needing to stop/start
> the device). However, the key length and the number of RETA
> entries are fixed to 40B and 128 entries respectively. This
> is done in order to simplify the design, but may be
> revisited later as the Virtio spec provides this
> flexibility.
> 
> Note that only VIRTIO_NET_F_RSS support is implemented,
> VIRTIO_NET_F_HASH_REPORT, which would enable reporting the
> packet RSS hash calculated by the device into mbuf.rss, is
> not yet supported.
> 
> Regarding the default RSS configuration, it has been
> chosen to use the default Intel ixgbe key as default key,
> and default RETA is a simple modulo between the hash and
> the number of Rx queues.
> 
> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>

<...>

> +static uint64_t
> +ethdev_to_virtio_rss_offloads(uint64_t ethdev_hash_types)
> +{
> +	uint64_t virtio_hash_types = 0;
> +
> +	if (ethdev_hash_types & (ETH_RSS_IPV4 | ETH_RSS_FRAG_IPV4 | ETH_RSS_NONFRAG_IPV4_OTHER))
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IPV4;
> +
> +	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV4_TCP)
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCPV4;
> +
> +	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV4_UDP)
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDPV4;
> +
> +	if (ethdev_hash_types & (ETH_RSS_IPV6 | ETH_RSS_FRAG_IPV6 | ETH_RSS_NONFRAG_IPV6_OTHER))
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IPV6;
> +
> +	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV6_TCP)
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCPV6;
> +
> +	if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV6_UDP)
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDPV6;
> +
> +	if (ethdev_hash_types & ETH_RSS_IPV6_EX)
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IP_EX;
> +
> +	if (ethdev_hash_types & ETH_RSS_IPV6_TCP_EX)
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCP_EX;
> +
> +	if (ethdev_hash_types & ETH_RSS_IPV6_UDP_EX)
> +		virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDP_EX;

Should use RTE_ETH_RSS macros with RTE_ prefix versions, updating in next-net.
Same for below virtio_to_ethdev_rss_offloads().

Re: [dpdk-dev] [PATCH v6 1/5] net/virtio: add initial RSS support

[Maxime Coquelin]

On 10/29/21 19:40, Ferruh Yigit wrote:
> On 10/27/2021 3:22 PM, Maxime Coquelin wrote:
>> Provide the capability to update the hash key, hash types
>> and RETA table on the fly (without needing to stop/start
>> the device). However, the key length and the number of RETA
>> entries are fixed to 40B and 128 entries respectively. This
>> is done in order to simplify the design, but may be
>> revisited later as the Virtio spec provides this
>> flexibility.
>>
>> Note that only VIRTIO_NET_F_RSS support is implemented,
>> VIRTIO_NET_F_HASH_REPORT, which would enable reporting the
>> packet RSS hash calculated by the device into mbuf.rss, is
>> not yet supported.
>>
>> Regarding the default RSS configuration, it has been
>> chosen to use the default Intel ixgbe key as default key,
>> and default RETA is a simple modulo between the hash and
>> the number of Rx queues.
>>
>> Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
> 
> <...>
> 
>> +static uint64_t
>> +ethdev_to_virtio_rss_offloads(uint64_t ethdev_hash_types)
>> +{
>> +    uint64_t virtio_hash_types = 0;
>> +
>> +    if (ethdev_hash_types & (ETH_RSS_IPV4 | ETH_RSS_FRAG_IPV4 | 
>> ETH_RSS_NONFRAG_IPV4_OTHER))
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IPV4;
>> +
>> +    if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV4_TCP)
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCPV4;
>> +
>> +    if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV4_UDP)
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDPV4;
>> +
>> +    if (ethdev_hash_types & (ETH_RSS_IPV6 | ETH_RSS_FRAG_IPV6 | 
>> ETH_RSS_NONFRAG_IPV6_OTHER))
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IPV6;
>> +
>> +    if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV6_TCP)
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCPV6;
>> +
>> +    if (ethdev_hash_types & ETH_RSS_NONFRAG_IPV6_UDP)
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDPV6;
>> +
>> +    if (ethdev_hash_types & ETH_RSS_IPV6_EX)
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_IP_EX;
>> +
>> +    if (ethdev_hash_types & ETH_RSS_IPV6_TCP_EX)
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_TCP_EX;
>> +
>> +    if (ethdev_hash_types & ETH_RSS_IPV6_UDP_EX)
>> +        virtio_hash_types |= VIRTIO_NET_HASH_TYPE_UDP_EX;
> 
> Should use RTE_ETH_RSS macros with RTE_ prefix versions, updating in 
> next-net.
> Same for below virtio_to_ethdev_rss_offloads().
> 

Thanks, for doing the change.

When rebasing, I renamed most of these defines, but forgot the ones in
these two functions.

Maxime