[dpdk-dev] MIT Fastpass: A Centralized "Zero-Queue" Datacenter Network
Good paper to read using DPDK: http://fastpass.mit.edu/
[dpdk-dev] DPDK supported processor
Dear Cao Min Thank you for the information. Best regards Byeong-Gi KIM 2014-08-22 14:30 GMT+09:00 Cao, Min : > Hi Byeong-Gi KIM, > > We tested with platform as below. Pls take information below as reference. > > Cpu: Genuine Intel(R) CPU 4000 @ 2.41GHz > NIC: Niantic 82599 > Bios: EDVLCRB1.86B.0008.R00.1303061138 > OS: 3.4.18-yocto-standard(32/64 bit) or FC20 > > We validated the test cases list below: >pmd >Checksum >Ipv4frag >Hello_world >Flow director >Jumboframe and scatter >Ieee1588 >Multiprocess >Flow control >Whitelist >blacklist >vlan > > Regards, > Cao Min > > -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of BYEONG-GI KIM > Sent: Thursday, August 21, 2014 2:04 PM > To: dev at dpdk.org > Subject: [dpdk-dev] DPDK supported processor > > Hello. > > I'm now using Intel Atom processor C2758 on the DPDK testing machines, but > it seems not working. > > Which sample application would be recommended to test the NIC actually > works well with DPDK? > > Best regards > > Byeong-Gi KIM > >
[dpdk-dev] ixgbe network card has dev_info.max_rx_queues == 0
Thanks, I understand that. I see in the ixgbe's code, that
dev_info.max_rx_queues is initialized from the hw->mac.max_rx_queues
value
dev_info->max_rx_queues = (uint16_t)hw->mac.max_rx_queues;
dev_info->max_tx_queues = (uint16_t)hw->mac.max_tx_queues;
but I can't figure out how this hw->mac is produced. Is it mapped to
the memory region controlled by the network adapter directly?
2014-08-21 16:52 GMT+04:00 Alex Markuze :
> RX and TX Are short hand for Receive and Transmit Queues.
> These Queues Store the in/egress packets.
>
> Just looking at the info you've sent it tells you that max_rx_queues
> for this dev is 0 (Clearly something is wrong here) so the nb_rx_q
> which is 3 is an Invalid Value -EINVAL == -22.
>
> On Thu, Aug 21, 2014 at 3:26 PM, Sergey Mironov wrote:
>> Hi. I have face a strange error on one of my network cards. Call to
>> rte_eth_dev_configure returns with error code -22. Increaing the
>> verbosity level shows the following:
>>
>>
>> PMD: rte_eth_dev_configure: ethdev port_id=2 nb_rx_queues=3 > 0
>> EAL: Error - exiting with code: 1
>>
>> here is the snippet of code which returns the error
>>
>>
>> ./lib/librte_ether/rte_ethdev.c : 513
>>
>> (*dev->dev_ops->dev_infos_get)(dev, &dev_info);
>> if (nb_rx_q > dev_info.max_rx_queues) {
>> PMD_DEBUG_TRACE("ethdev port_id=%d nb_rx_queues=%d > %d\n",
>> port_id, nb_rx_q, dev_info.max_rx_queues);
>> return (-EINVAL);
>> }
>>
>> What does this error means (what is rx queues of an adapter?) What may
>> cause such a problem? I am using dpdk 1.5.1r1.
>>
>> Thanks in advance,
>> Sergey
[dpdk-dev] [PATCH 0/6] link bonding
Tested-by: Jiajia, SunX < sunx.jiajia at intel.com> These patches have be verified, and are ready to integrate into DPDK.org. There are 18 test cases in this feature now. All cases are passed. Please see the test result as following: active_backup_all_slave_downPASSED active_backup_change_primaryPASSED active_backup_one_slave_downPASSED active_backup_rx_tx PASSED bonded_mac_address PASSED broadcast_rx_tx PASSED broadcast_tx_all_slave_down PASSED broadcast_tx_one_slave_down PASSED create_bonded_devices_and_slavesPASSED device_promiscuous_mode PASSED round_robin_all_slave_down PASSED round_robin_one_slave_down PASSED round_robin_rx_tx PASSED xor_l34_forward PASSED xor_rx PASSED xor_tx PASSED xor_tx_all_slave_down PASSED xor_tx_one_slave_down PASSED Test environment: Fedora 20, 3.11.10-301.fc20.x86_64, GCC 4.8.2, Intel Xeon processor E5-2680 v2, with Intel Niantic 82599. > -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Declan Doherty > Sent: Tuesday, August 19, 2014 9:52 PM > To: dev at dpdk.org > Subject: [dpdk-dev] [PATCH 0/6] link bonding > > This patch set adds support for link status interrupt in the link > bonding > pmd. It also contains some patches to tidy up the code structure and to > of the link bonding code and to fix bugs relating to transmission > failures in the under lying slave pmd which could lead to leaked mbufs. > > > Declan Doherty (6): > bond: link status interrupt support > bond: removing switch statement from rx burst method > bond: fix naming inconsistency in tx_burst_round_robin > bond: free mbufs if transmission fails in bonding tx_burst functions > test app: adding support for generating variable sized packets > testpmd: adding parameter to reconfig method to set socket_id when > adding new port to portlist > > app/test-pmd/cmdline.c |2 +- > app/test-pmd/testpmd.c |3 +- > app/test-pmd/testpmd.h |2 +- > app/test/packet_burst_generator.c | 22 +-- > app/test/packet_burst_generator.h |6 +- > app/test/test_link_bonding.c | 234 > ++-- > lib/librte_pmd_bond/rte_eth_bond_api.c |4 + > lib/librte_pmd_bond/rte_eth_bond_pmd.c | 124 +++-- > 8 files changed, 295 insertions(+), 102 deletions(-)
[dpdk-dev] [PATCH 1/5] ethdev: support of multiple sizes of redirection table
To support different sizes of redirection table,
* 'struct rte_eth_rss_reta' has been redefined.
* New element of 'reta_size' is added into
'struct rte_eth_dev_info'.
* New parameter of 'redirection table size' is
required for updating/querying redirection
table entries.
* Interfaces of updating/querying redirection
table entries have been reworked to support
multiple sizes.
Currently the supported redirection table sizes
are 64, 128 and 512 entries, according to the
hardware specifications.
Signed-off-by: Helin Zhang
Reviewed-by: Jijiang Liu
Reviewed-by: Cunming Liang
Reviewed-by: Jingjing Wu
---
lib/librte_ether/rte_ethdev.c | 116 ++
lib/librte_ether/rte_ethdev.h | 43 ++--
2 files changed, 99 insertions(+), 60 deletions(-)
diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c
index fd1010a..3b483c4 100644
--- a/lib/librte_ether/rte_ethdev.c
+++ b/lib/librte_ether/rte_ethdev.c
@@ -1780,78 +1780,104 @@ rte_eth_dev_priority_flow_ctrl_set(uint8_t port_id,
struct rte_eth_pfc_conf *pfc
return (-ENOTSUP);
}
-int
-rte_eth_dev_rss_reta_update(uint8_t port_id, struct rte_eth_rss_reta
*reta_conf)
+static inline int
+rte_eth_check_reta_mask(struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size)
{
- struct rte_eth_dev *dev;
- uint16_t max_rxq;
- uint8_t i,j;
+ uint16_t i, num = reta_size / RTE_BIT_WIDTH_64;
- if (port_id >= nb_ports) {
- PMD_DEBUG_TRACE("Invalid port_id=%d\n", port_id);
- return (-ENODEV);
- }
+ if (!reta_conf)
+ return -EINVAL;
- /* Invalid mask bit(s) setting */
- if ((reta_conf->mask_lo == 0) && (reta_conf->mask_hi == 0)) {
- PMD_DEBUG_TRACE("Invalid update mask bits for
port=%d\n",port_id);
- return (-EINVAL);
+ for (i = 0; i < num; i++) {
+ if (reta_conf[i].mask)
+ return 0;
}
- dev = &rte_eth_devices[port_id];
- max_rxq = (dev->data->nb_rx_queues <= ETH_RSS_RETA_MAX_QUEUE) ?
- dev->data->nb_rx_queues : ETH_RSS_RETA_MAX_QUEUE;
- if (reta_conf->mask_lo != 0) {
- for (i = 0; i < ETH_RSS_RETA_NUM_ENTRIES/2; i++) {
- if ((reta_conf->mask_lo & (1ULL << i)) &&
- (reta_conf->reta[i] >= max_rxq)) {
- PMD_DEBUG_TRACE("RETA hash index output"
- "configration for port=%d,invalid"
-
"queue=%d\n",port_id,reta_conf->reta[i]);
+ return -EINVAL;
+}
- return (-EINVAL);
- }
+static inline int
+rte_eth_check_reta_entry(struct rte_eth_rss_reta_entry64 *reta_conf,
+uint16_t reta_size,
+uint8_t max_rxq)
+{
+ uint16_t i, idx, shift;
+
+ if (!reta_conf)
+ return -EINVAL;
+
+ if (max_rxq == 0) {
+ PMD_DEBUG_TRACE("No receive queue is available\n");
+ return -EINVAL;
+ }
+
+ for (i = 0; i < reta_size; i++) {
+ idx = i / RTE_BIT_WIDTH_64;
+ shift = i % RTE_BIT_WIDTH_64;
+ if ((reta_conf[idx].mask & (0x1 << shift)) &&
+ (reta_conf[idx].reta[shift] >= max_rxq)) {
+ PMD_DEBUG_TRACE("reta_conf[%u]->reta[%u]: %u exceeds "
+ "the maximum rxq index: %u\n", idx, shift,
+ reta_conf[idx].reta[shift], max_rxq);
+ return -EINVAL;
}
}
- if (reta_conf->mask_hi != 0) {
- for (i = 0; i< ETH_RSS_RETA_NUM_ENTRIES/2; i++) {
- j = (uint8_t)(i + ETH_RSS_RETA_NUM_ENTRIES/2);
+ return 0;
+}
- /* Check if the max entry >= 128 */
- if ((reta_conf->mask_hi & (1ULL << i)) &&
- (reta_conf->reta[j] >= max_rxq)) {
- PMD_DEBUG_TRACE("RETA hash index output"
- "configration for port=%d,invalid"
-
"queue=%d\n",port_id,reta_conf->reta[j]);
+int
+rte_eth_dev_rss_reta_update(uint8_t port_id,
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size)
+{
+ struct rte_eth_dev *dev;
+ int ret;
- return (-EINVAL);
- }
- }
+ if (port_id >= nb_ports) {
+ PMD_DEBUG_TRACE("Invalid port_id=%d\n", port_id);
+ return -ENODEV;
}
+ /* Check mask bits */
+ ret = rte_eth_check_reta_mask(reta_conf, reta_size);
+ if (ret < 0)
+ return ret;
+
+ dev = &rte_e
[dpdk-dev] [PATCH 2/5] e1000: rework of updating/querying redirection table
As ethdev has been changed to support multiple sizes
of redirection table, the functions of updating/querying
redirection table need to be reworked. In addition,
getting the redirection table size is supported in ops
of 'dev_infos_get'.
Signed-off-by: Helin Zhang
Reviewed-by: Jijiang Liu
Reviewed-by: Cunming Liang
Reviewed-by: Jingjing Wu
---
lib/librte_pmd_e1000/igb_ethdev.c | 184 ++
1 file changed, 127 insertions(+), 57 deletions(-)
diff --git a/lib/librte_pmd_e1000/igb_ethdev.c
b/lib/librte_pmd_e1000/igb_ethdev.c
index 3187d92..47438ad 100644
--- a/lib/librte_pmd_e1000/igb_ethdev.c
+++ b/lib/librte_pmd_e1000/igb_ethdev.c
@@ -71,6 +71,8 @@ static void eth_igb_stats_get(struct rte_eth_dev *dev,
struct rte_eth_stats *rte_stats);
static void eth_igb_stats_reset(struct rte_eth_dev *dev);
static void eth_igb_infos_get(struct rte_eth_dev *dev,
+ struct rte_eth_dev_info *dev_info);
+static void eth_igbvf_infos_get(struct rte_eth_dev *dev,
struct rte_eth_dev_info *dev_info);
static int eth_igb_flow_ctrl_get(struct rte_eth_dev *dev,
struct rte_eth_fc_conf *fc_conf);
@@ -124,10 +126,11 @@ static int igbvf_vlan_filter_set(struct rte_eth_dev *dev,
static int igbvf_set_vfta(struct e1000_hw *hw, uint16_t vid, bool on);
static void igbvf_set_vfta_all(struct rte_eth_dev *dev, bool on);
static int eth_igb_rss_reta_update(struct rte_eth_dev *dev,
-struct rte_eth_rss_reta *reta_conf);
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size);
static int eth_igb_rss_reta_query(struct rte_eth_dev *dev,
- struct rte_eth_rss_reta *reta_conf);
-
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size);
static int eth_igb_add_syn_filter(struct rte_eth_dev *dev,
struct rte_syn_filter *filter, uint16_t rx_queue);
static int eth_igb_remove_syn_filter(struct rte_eth_dev *dev);
@@ -270,7 +273,7 @@ static struct eth_dev_ops igbvf_eth_dev_ops = {
.stats_get= eth_igbvf_stats_get,
.stats_reset = eth_igbvf_stats_reset,
.vlan_filter_set = igbvf_vlan_filter_set,
- .dev_infos_get= eth_igb_infos_get,
+ .dev_infos_get= eth_igbvf_infos_get,
.rx_queue_setup = eth_igb_rx_queue_setup,
.rx_queue_release = eth_igb_rx_queue_release,
.tx_queue_setup = eth_igb_tx_queue_setup,
@@ -1257,8 +1260,7 @@ eth_igbvf_stats_reset(struct rte_eth_dev *dev)
}
static void
-eth_igb_infos_get(struct rte_eth_dev *dev,
- struct rte_eth_dev_info *dev_info)
+eth_igb_infos_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
{
struct e1000_hw *hw = E1000_DEV_PRIVATE_TO_HW(dev->data->dev_private);
@@ -1331,6 +1333,72 @@ eth_igb_infos_get(struct rte_eth_dev *dev,
dev_info->max_tx_queues = 0;
dev_info->max_vmdq_pools = 0;
}
+ dev_info->reta_size = ETH_RSS_RETA_SIZE_128;
+}
+
+static void
+eth_igbvf_infos_get(struct rte_eth_dev *dev, struct rte_eth_dev_info *dev_info)
+{
+ struct e1000_hw *hw = E1000_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+
+ dev_info->min_rx_bufsize = 256; /* See BSIZE field of RCTL register. */
+ dev_info->max_rx_pktlen = 0x3FFF; /* See RLPML register. */
+ dev_info->max_mac_addrs = hw->mac.rar_entry_count;
+ dev_info->rx_offload_capa = DEV_RX_OFFLOAD_VLAN_STRIP |
+ DEV_RX_OFFLOAD_IPV4_CKSUM |
+ DEV_RX_OFFLOAD_UDP_CKSUM |
+ DEV_RX_OFFLOAD_TCP_CKSUM;
+ dev_info->tx_offload_capa = DEV_TX_OFFLOAD_VLAN_INSERT |
+ DEV_TX_OFFLOAD_IPV4_CKSUM |
+ DEV_TX_OFFLOAD_UDP_CKSUM |
+ DEV_TX_OFFLOAD_TCP_CKSUM |
+ DEV_TX_OFFLOAD_SCTP_CKSUM;
+ switch (hw->mac.type) {
+ case e1000_82575:
+ dev_info->max_rx_queues = 4;
+ dev_info->max_tx_queues = 4;
+ dev_info->max_vmdq_pools = 0;
+ break;
+ case e1000_82576:
+ dev_info->max_rx_queues = 16;
+ dev_info->max_tx_queues = 16;
+ dev_info->max_vmdq_pools = ETH_8_POOLS;
+ break;
+ case e1000_82580:
+ dev_info->max_rx_queues = 8;
+ dev_info->max_tx_queues = 8;
+ dev_info->max_vmdq_pools = ETH_8_POOLS;
+ break;
+ case e1000_i350:
+ dev_info->max_rx_queues = 8;
+ dev_info->max_tx_queues = 8;
+ dev_info->max_vmdq_pools = ETH_8_POOLS;
+ break;
+ case e1000_
[dpdk-dev] [PATCH 3/5] ixgbe: rework of updating/querying redirection table
As ethdev has been changed to support multiple sizes
of redirection table, the functions of updating/querying
redirection table need to be reworked. In addition,
getting the redirection table size is supported in ops
of 'dev_infos_get'.
Signed-off-by: Helin Zhang
Reviewed-by: Jijiang Liu
Reviewed-by: Cunming Liang
Reviewed-by: Jingjing Wu
---
lib/librte_pmd_ixgbe/ixgbe_ethdev.c | 136 +++-
1 file changed, 87 insertions(+), 49 deletions(-)
diff --git a/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
b/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
index 59122a1..4f036ec 100644
--- a/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
+++ b/lib/librte_pmd_ixgbe/ixgbe_ethdev.c
@@ -118,8 +118,9 @@ static int ixgbe_dev_queue_stats_mapping_set(struct
rte_eth_dev *eth_dev,
uint8_t stat_idx,
uint8_t is_rx);
static void ixgbe_dev_info_get(struct rte_eth_dev *dev,
- struct rte_eth_dev_info *dev_info);
-
+ struct rte_eth_dev_info *dev_info);
+static void ixgbevf_dev_info_get(struct rte_eth_dev *dev,
+struct rte_eth_dev_info *dev_info);
static int ixgbe_dev_mtu_set(struct rte_eth_dev *dev, uint16_t mtu);
static int ixgbe_vlan_filter_set(struct rte_eth_dev *dev,
@@ -144,9 +145,11 @@ static int ixgbe_flow_ctrl_set(struct rte_eth_dev *dev,
static int ixgbe_priority_flow_ctrl_set(struct rte_eth_dev *dev,
struct rte_eth_pfc_conf *pfc_conf);
static int ixgbe_dev_rss_reta_update(struct rte_eth_dev *dev,
- struct rte_eth_rss_reta *reta_conf);
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size);
static int ixgbe_dev_rss_reta_query(struct rte_eth_dev *dev,
- struct rte_eth_rss_reta *reta_conf);
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size);
static void ixgbe_dev_link_status_print(struct rte_eth_dev *dev);
static int ixgbe_dev_lsc_interrupt_setup(struct rte_eth_dev *dev);
static int ixgbe_dev_interrupt_get_status(struct rte_eth_dev *dev);
@@ -377,7 +380,7 @@ static struct eth_dev_ops ixgbevf_eth_dev_ops = {
.stats_get= ixgbevf_dev_stats_get,
.stats_reset = ixgbevf_dev_stats_reset,
.dev_close= ixgbevf_dev_close,
- .dev_infos_get= ixgbe_dev_info_get,
+ .dev_infos_get= ixgbevf_dev_info_get,
.mtu_set = ixgbevf_dev_set_mtu,
.vlan_filter_set = ixgbevf_vlan_filter_set,
.vlan_strip_queue_set = ixgbevf_vlan_strip_queue_set,
@@ -1948,6 +1951,35 @@ ixgbe_dev_info_get(struct rte_eth_dev *dev, struct
rte_eth_dev_info *dev_info)
DEV_TX_OFFLOAD_UDP_CKSUM |
DEV_TX_OFFLOAD_TCP_CKSUM |
DEV_TX_OFFLOAD_SCTP_CKSUM;
+ dev_info->reta_size = ETH_RSS_RETA_SIZE_128;
+}
+
+static void
+ixgbevf_dev_info_get(struct rte_eth_dev *dev,
+struct rte_eth_dev_info *dev_info)
+{
+ struct ixgbe_hw *hw = IXGBE_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+
+ dev_info->max_rx_queues = (uint16_t)hw->mac.max_rx_queues;
+ dev_info->max_tx_queues = (uint16_t)hw->mac.max_tx_queues;
+ dev_info->min_rx_bufsize = 1024; /* cf BSIZEPACKET in SRRCTL reg */
+ dev_info->max_rx_pktlen = 15872; /* includes CRC, cf MAXFRS reg */
+ dev_info->max_mac_addrs = hw->mac.num_rar_entries;
+ dev_info->max_hash_mac_addrs = IXGBE_VMDQ_NUM_UC_MAC;
+ dev_info->max_vfs = dev->pci_dev->max_vfs;
+ if (hw->mac.type == ixgbe_mac_82598EB)
+ dev_info->max_vmdq_pools = ETH_16_POOLS;
+ else
+ dev_info->max_vmdq_pools = ETH_64_POOLS;
+ dev_info->rx_offload_capa = DEV_RX_OFFLOAD_VLAN_STRIP |
+ DEV_RX_OFFLOAD_IPV4_CKSUM |
+ DEV_RX_OFFLOAD_UDP_CKSUM |
+ DEV_RX_OFFLOAD_TCP_CKSUM;
+ dev_info->tx_offload_capa = DEV_TX_OFFLOAD_VLAN_INSERT |
+ DEV_TX_OFFLOAD_IPV4_CKSUM |
+ DEV_TX_OFFLOAD_UDP_CKSUM |
+ DEV_TX_OFFLOAD_TCP_CKSUM |
+ DEV_TX_OFFLOAD_SCTP_CKSUM;
}
/* return 0 means link status changed, -1 means not changed */
@@ -2624,38 +2656,41 @@ ixgbe_priority_flow_ctrl_set(struct rte_eth_dev *dev,
struct rte_eth_pfc_conf *p
static int
ixgbe_dev_rss_reta_update(struct rte_eth_dev *dev,
- struct rte_eth_rss_reta *reta_conf)
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size)
{
uint8_t i,j,mask;
- uint32_t reta;
+ uint32_t reta, r;
+ uint16_t idx, shift;
struct ixgbe_hw *hw =
IXGBE_DEV_PRIVATE_T
[dpdk-dev] [PATCH 4/5] i40e: rework of updating/querying redirection table
i40e can support 128 or 512 entries of redirection
table entries, according to the firmware configuration.
In addition, as ethdev has been changed to support
multiple sizes of redirection table, the functions of
updating/querying redirection table need to be reworked.
Getting the redirection table size is supported in ops
of 'dev_infos_get'.
Signed-off-by: Helin Zhang
Reviewed-by: Jijiang Liu
Reviewed-by: Cunming Liang
Reviewed-by: Jingjing Wu
---
lib/librte_pmd_i40e/i40e_ethdev.c| 88
lib/librte_pmd_i40e/i40e_ethdev.h| 1 +
lib/librte_pmd_i40e/i40e_ethdev_vf.c | 1 +
3 files changed, 62 insertions(+), 28 deletions(-)
diff --git a/lib/librte_pmd_i40e/i40e_ethdev.c
b/lib/librte_pmd_i40e/i40e_ethdev.c
index 9ed31b5..7289f1a 100644
--- a/lib/librte_pmd_i40e/i40e_ethdev.c
+++ b/lib/librte_pmd_i40e/i40e_ethdev.c
@@ -158,9 +158,11 @@ static void i40e_macaddr_add(struct rte_eth_dev *dev,
uint32_t pool);
static void i40e_macaddr_remove(struct rte_eth_dev *dev, uint32_t index);
static int i40e_dev_rss_reta_update(struct rte_eth_dev *dev,
- struct rte_eth_rss_reta *reta_conf);
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size);
static int i40e_dev_rss_reta_query(struct rte_eth_dev *dev,
- struct rte_eth_rss_reta *reta_conf);
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size);
static int i40e_get_cap(struct i40e_hw *hw);
static int i40e_pf_parameter_init(struct rte_eth_dev *dev);
@@ -1231,6 +1233,7 @@ i40e_dev_info_get(struct rte_eth_dev *dev, struct
rte_eth_dev_info *dev_info)
DEV_TX_OFFLOAD_UDP_CKSUM |
DEV_TX_OFFLOAD_TCP_CKSUM |
DEV_TX_OFFLOAD_SCTP_CKSUM;
+ dev_info->reta_size = pf->hash_lut_size;
}
static int
@@ -1431,32 +1434,40 @@ i40e_macaddr_remove(struct rte_eth_dev *dev, uint32_t
index)
static int
i40e_dev_rss_reta_update(struct rte_eth_dev *dev,
-struct rte_eth_rss_reta *reta_conf)
+struct rte_eth_rss_reta_entry64 *reta_conf,
+uint16_t reta_size)
{
+ struct i40e_pf *pf = I40E_DEV_PRIVATE_TO_PF(dev->data->dev_private);
struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private);
uint32_t lut, l;
- uint8_t i, j, mask, max = ETH_RSS_RETA_NUM_ENTRIES / 2;
-
- for (i = 0; i < ETH_RSS_RETA_NUM_ENTRIES; i += 4) {
- if (i < max)
- mask = (uint8_t)((reta_conf->mask_lo >> i) & 0xF);
- else
- mask = (uint8_t)((reta_conf->mask_hi >>
- (i - max)) & 0xF);
+ uint16_t i, j, lut_size = pf->hash_lut_size;
+ uint16_t idx, shift;
+ uint8_t mask;
+
+ if (reta_size != lut_size ||
+ reta_size > ETH_RSS_RETA_SIZE_512) {
+ PMD_DRV_LOG(ERR, "The size of hash lookup table configured "
+ "(%d) doesn't match the number hardware can supported "
+ "(%d)\n", reta_size, lut_size);
+ return -EINVAL;
+ }
+ for (i = 0; i < reta_size; i += 4) {
+ idx = i / RTE_BIT_WIDTH_64;
+ shift = i % RTE_BIT_WIDTH_64;
+ mask = (uint8_t)((reta_conf[idx].mask >> shift) & 0xf);
if (!mask)
continue;
-
- if (mask == 0xF)
+ if (mask == 0xf)
l = 0;
else
l = I40E_READ_REG(hw, I40E_PFQF_HLUT(i >> 2));
-
for (j = 0, lut = 0; j < 4; j++) {
if (mask & (0x1 << j))
- lut |= reta_conf->reta[i + j] << (8 * j);
+ lut |= reta_conf[idx].reta[shift + j] <<
+ (CHAR_BIT * j);
else
- lut |= l & (0xFF << (8 * j));
+ lut |= l & (0xff << (CHAR_BIT * j));
}
I40E_WRITE_REG(hw, I40E_PFQF_HLUT(i >> 2), lut);
}
@@ -1466,27 +1477,36 @@ i40e_dev_rss_reta_update(struct rte_eth_dev *dev,
static int
i40e_dev_rss_reta_query(struct rte_eth_dev *dev,
- struct rte_eth_rss_reta *reta_conf)
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t reta_size)
{
+ struct i40e_pf *pf = I40E_DEV_PRIVATE_TO_PF(dev->data->dev_private);
struct i40e_hw *hw = I40E_DEV_PRIVATE_TO_HW(dev->data->dev_private);
uint32_t lut;
- uint8_t i, j, mask, max = ETH_RSS_RETA_NUM_ENTRIES / 2;
-
- for (i = 0; i < ETH_RSS_RETA_NUM_ENTRIES
[dpdk-dev] [PATCH 5/5] app/testpmd: rework of updating/querying redirection table
As multiple sizes (64, 128, 512) of redirection table have
been supported, the commands of updating/querying redirection
table entries have been reworked. In addition, the
redirection table size can be queried by the existing
command of 'show port info <>'.
Signed-off-by: Helin Zhang
Reviewed-by: Jijiang Liu
Reviewed-by: Cunming Liang
Reviewed-by: Jingjing Wu
---
app/test-pmd/cmdline.c | 159 -
app/test-pmd/config.c | 37 ++--
app/test-pmd/testpmd.h | 4 +-
3 files changed, 137 insertions(+), 63 deletions(-)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index 345be11..5ad501e 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -59,6 +59,7 @@
#include
#include
#include
+#include
#include
#include
#include
@@ -186,6 +187,11 @@ static void cmd_help_long_parsed(void *parsed_result,
"show port (info|stats|fdir|stat_qmap) (port_id|all)\n"
"Display information for port_id, or all.\n\n"
+ "show port X rss reta (size) (mask0,mask1,...)\n"
+ "Display the rss redirection table entry indicated"
+ " by masks or port X. size is used to indicate the"
+ " hardware supported reta size\n\n"
+
"show port rss-hash [key]\n"
"Display the RSS hash functions and RSS hash key"
" of port X\n\n"
@@ -1436,11 +1442,13 @@ struct cmd_config_rss_reta {
};
static int
-parse_reta_config(const char *str, struct rte_eth_rss_reta *reta_conf)
+parse_reta_config(const char *str,
+ struct rte_eth_rss_reta_entry64 *reta_conf,
+ uint16_t nb_entries)
{
int i;
unsigned size;
- uint8_t hash_index;
+ uint16_t hash_index, idx, shift;
uint8_t nb_queue;
char s[256];
const char *p, *p0 = str;
@@ -1475,17 +1483,15 @@ parse_reta_config(const char *str, struct
rte_eth_rss_reta *reta_conf)
hash_index = (uint8_t)int_fld[FLD_HASH_INDEX];
nb_queue = (uint8_t)int_fld[FLD_QUEUE];
- if (hash_index >= ETH_RSS_RETA_NUM_ENTRIES) {
+ if (hash_index >= nb_entries) {
printf("Invalid RETA hash index=%d",hash_index);
return -1;
}
- if (hash_index < ETH_RSS_RETA_NUM_ENTRIES/2)
- reta_conf->mask_lo |= (1ULL << hash_index);
- else
- reta_conf->mask_hi |= (1ULL << (hash_index -
ETH_RSS_RETA_NUM_ENTRIES/2));
-
- reta_conf->reta[hash_index] = nb_queue;
+ idx = hash_index / RTE_BIT_WIDTH_64;
+ shift = hash_index % RTE_BIT_WIDTH_64;
+ reta_conf[idx].mask |= (1ULL << shift);
+ reta_conf[idx].reta[shift] = nb_queue;
}
return 0;
@@ -1493,22 +1499,42 @@ parse_reta_config(const char *str, struct
rte_eth_rss_reta *reta_conf)
static void
cmd_set_rss_reta_parsed(void *parsed_result,
- __attribute__((unused)) struct cmdline *cl,
- __attribute__((unused)) void *data)
+ __attribute__((unused)) struct cmdline *cl,
+ __attribute__((unused)) void *data)
{
int ret;
- struct rte_eth_rss_reta reta_conf;
+ struct rte_eth_dev_info dev_info;
+ struct rte_eth_rss_reta_entry64 reta_conf[8];
struct cmd_config_rss_reta *res = parsed_result;
- memset(&reta_conf,0,sizeof(struct rte_eth_rss_reta));
+ memset(&dev_info, 0, sizeof(dev_info));
+ rte_eth_dev_info_get(res->port_id, &dev_info);
+ if (dev_info.reta_size == 0) {
+ printf("Redirection table size is 0 which is "
+ "invalid for RSS\n");
+ return;
+ } else
+ printf("The reta size of port %d is %u\n",
+ res->port_id, dev_info.reta_size);
+ if (dev_info.reta_size > ETH_RSS_RETA_SIZE_512) {
+ printf("Currently do not support more than %u entries of "
+ "redirection table\n", ETH_RSS_RETA_SIZE_512);
+ return;
+ }
+
+ memset(reta_conf, 0, sizeof(reta_conf));
if (!strcmp(res->list_name, "reta")) {
- if (parse_reta_config(res->list_of_items, &reta_conf)) {
- printf("Invalid RSS Redirection Table config
entered\n");
+ if (parse_reta_config(res->list_of_items, reta_conf,
+ dev_info.reta_size)) {
+ printf("Invalid RSS Redirection Table "
+ "config entered\n");
return;
}
- ret = rte_eth_dev_rss_reta_update(res->port_id
[dpdk-dev] [PATCH 0/5] support of multiple sizes of redirection table
As e1000, ixgbe and i40e hardware use different sizes of redirection table, ethdev needs to be reworked to support multiple sizes of that table. Also each PMD should be reworked to support this change, and can be told the table size directly. In addition, commands in testpmd have been reworked to support these changes. Helin Zhang (5): ethdev: support of multiple sizes of redirection table e1000: rework of updating/querying redirection table ixgbe: rework of updating/querying redirection table i40e: rework of updating/querying redirection table app/testpmd: rework of updating/querying redirection table app/test-pmd/cmdline.c | 159 ++ app/test-pmd/config.c| 37 --- app/test-pmd/testpmd.h | 4 +- lib/librte_ether/rte_ethdev.c| 116 +- lib/librte_ether/rte_ethdev.h| 43 +--- lib/librte_pmd_e1000/igb_ethdev.c| 184 --- lib/librte_pmd_i40e/i40e_ethdev.c| 88 +++-- lib/librte_pmd_i40e/i40e_ethdev.h| 1 + lib/librte_pmd_i40e/i40e_ethdev_vf.c | 1 + lib/librte_pmd_ixgbe/ixgbe_ethdev.c | 136 -- 10 files changed, 512 insertions(+), 257 deletions(-) -- 1.8.1.4
[dpdk-dev] [PATCH 1/2] igb_uio: fix compability on old kernel
Hi Thomas, Not that I would *like* to fix this, but I *need* to fix it. We are using CentOS 6.5, which I believe is based on RHEL. We have kernel 2.6.32-431.3.1.el6.x86_64. I realize that we need to add/change ifdefs around pci_num_vf, pci_intx_mask_supported, and pci_check_and_mask_intx in igb_uio/compat.h. Any more specific suggestions on how to (elegantly) fix it for us, but not break it for anyone else? -- Regards, Robert >2014-07-25 10:36, Stephen Hemminger: >> Add more compatibility wrappers, and split out all the wrapper >> code to a separate file. Builds on Debian Squeeze (2.6.32) which >> is oldest version of kernel current DPDK supports. >> >> Signed-off-by: Stephen Hemminger > >Acked-by: Thomas Monjalon > >Applied for version 1.7.1. > >There are still some compilation issues with RHEL: >include/linux/pci.h:1572: note: previous declaration of ?pci_num_vf? was >here >include/linux/pci.h:868: note: previous declaration of >?pci_intx_mask_supported? was here >include/linux/pci.h:869: note: previous declaration of >?pci_check_and_mask_intx? was here > >Some ifdefs are missing but I don't want to dig into RHEL kernel headers >to >find what is the first RHEL release to support these functions. >By the way, if someone knows an easy method to get all RHEL kernel headers >or to know the release where a symbol appeared, it would be very useful. > >-- >Thomas
[dpdk-dev] [PATCH 1/2] igb_uio: fix compability on old kernel
This is what we came up with. It works for us. In our kernel headers'
linux/pci.h, pci_num_vf is enclosed within "#ifdef CONFIG_PCI_IOV/#endif";
pci_intx_mask_supported and pci_check_and_mask_intx are enclosed within
"#ifdef HAVE_PCI_SET_MWI/#endif".
What do you think?
--
Thanks,
Robert
---
lib/librte_eal/linuxapp/igb_uio/compat.h |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/librte_eal/linuxapp/igb_uio/compat.h
b/lib/librte_eal/linuxapp/igb_uio/compat.h
index 2a16540..f7404d8 100644
--- a/lib/librte_eal/linuxapp/igb_uio/compat.h
+++ b/lib/librte_eal/linuxapp/igb_uio/compat.h
@@ -17,7 +17,7 @@
#define PCI_MSIX_ENTRY_CTRL_MASKBIT 1
#endif
-#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 34)
+#if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 34) &&
!defined(CONFIG_PCI_IOV)
static int pci_num_vf(struct pci_dev *dev)
{
struct iov {
@@ -38,7 +38,7 @@ static int pci_num_vf(struct pci_dev *dev)
#endif
-#if LINUX_VERSION_CODE < KERNEL_VERSION(3, 3, 0)
+#if LINUX_VERSION_CODE < KERNEL_VERSION(3, 3, 0) &&
!defined(HAVE_PCI_SET_MWI)
/* Check if INTX works to control irq's.
* Set's INTX_DISABLE flag and reads it back
--
1.7.1
On Fri, Aug 22, 2014 at 1:29 PM, Sanford, Robert
wrote:
>
> Hi Thomas,
>
> Not that I would *like* to fix this, but I *need* to fix it. We are using
> CentOS 6.5, which I believe is based on RHEL. We have kernel
> 2.6.32-431.3.1.el6.x86_64.
>
> I realize that we need to add/change ifdefs around pci_num_vf,
> pci_intx_mask_supported, and pci_check_and_mask_intx in igb_uio/compat.h.
> Any more specific suggestions on how to (elegantly) fix it for us, but not
> break it for anyone else?
>
> --
> Regards,
> Robert
>
[dpdk-dev] Issue starting DPDK-1.7 in the guest
Hi, I am trying to run a simple L2FWD in my guest using DPDK 1.7 version to test the i40e linux driver. Initially my host OS was Fedora 20 but there was some issue with the i40e linux driver so I went back to Fedora 14(2.6.35 - kernel) as per the DPDK documents to check if that works. But now when I try and start DPDK in the guest it gives me below error. Error: This system does not support AES Please check that RTE_MACHINE is set correctly I did not have this issue when my Host OS was Fedora 20. And I am using the same guest image I was using then. The host OS is now Fedora 14 and qemu version 0.40.0 as per mentioned by the document. (My guest OS is fedora 18). Can anyone tell me what is the issue here? Rahul Shah
