summaryrefslogtreecommitdiffstats
path: root/src/plugins/dpdk
AgeCommit message (Collapse)AuthorFilesLines
2018-05-22CSIT-928 dpdk/ipsec: performance improvementRadu Nicolau3-112/+114
Replace hash with a vector to improve performance. Plus other minor performance improvements. Change-Id: I3f0ebd909782ce3727f6360ce5ff5ddd131f8574 Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
2018-05-21dpdk:enable flow director perfect modeEyal Bari3-4/+25
when flows are enabled on the device Change-Id: I971764988d5a9e7078468f627205b3fa60736263 Signed-off-by: Eyal Bari <ebari@cisco.com>
2018-05-18Add vlib_buffer_enqueue_to_next inline functionDamjan Marion1-75/+3
Change-Id: I1042c0fe179b57a00ce99c8d62cb1bdbe24d9184 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-17Add buffer pointer-to-index and index-to-pointer array functionsDamjan Marion2-106/+7
Change-Id: Ib3fcc3ceb7f315389bcdecbb7d9632540a5dd6ba Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-17flow:redirect to nodeEyal Bari1-2/+1
Change-Id: I4b6577b496c56f27f07dd0066fcfdfd0cebb6f1a Signed-off-by: Eyal Bari <ebari@cisco.com>
2018-05-16dpdk: fix rte_eth_dev_set_mtu callsites to use same mtu valueRui Cai1-1/+1
During dpdk_lib_init, it calculates MRU and MTU and later calls rte_eth_dev_set_mtu with calculated MTU value. However, dpdk_device_setup calls rte_eth_dev_set_mtu with hi->max_packet_bytes, which is set to be MRU value in dpdk_lib_init earlier. Most of the time, MRU != MTU in dpdk_lib_init and it looks like hi->max_packet_bytes is treated as MTU in other parts of vpp codebase. Therefore, dpdk_lib_init should be consistent and use MTU instead of MRU for hi->max_packet_bytes. Change-Id: I23ff2a6cd45d6bc819b6f64d5f0fc0490b8a44de Signed-off-by: Rui Cai <rucai@microsoft.com>
2018-05-12dpdk: Add constants for failsafe PMDRui Cai2-2/+13
Adding name, enum constants and formatting code for failsafe PMD. This is part of initial effort to enable vpp running over dpdk on failsafe PMD in Microsoft Azure(2/4). Change-Id: I4eb0093db9f666e2635f7ddff451e3c9064bd0c4 Signed-off-by: Rui Cai <rucai@microsoft.com>
2018-05-11dpdk: fix Unknown interface with Mellanox NICSteve Shin1-1/+1
When port_type_from_speed_capa() is called before the port link update isn't completed, xd->port_type becomes VNET_DPDK_PORT_TYPE_UNKNOWN. This happens with Mellanox NIC without lsc interrupt. Calling rte_eth_link_get before getting dev_info will ensure the link state is up-to-date. Change-Id: I83a59654778eb4bf0c65a4a4e225a326227b9641 Signed-off-by: Steve Shin <jonshin@cisco.com>
2018-05-10vppinfra: use count_trailing_zeros in sparse_vec_indexDamjan Marion2-8/+8
It is much cheaper to use ctzll than to do shift,subtract and mask in likely case when we are looking for 1st set bit in the uword. Change-Id: I31954081571978878c7098bafad0c85a91755fa2 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-10vnet: device flow offload infraDamjan Marion5-0/+362
Change-Id: Ibea4a96bdec5e368301a03d8b11a0712fa0265e0 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-10Change the way IP header pointer is calculated in esp_decrypt nodesSzymon Sliwa1-1/+7
The pointer to IP header was derived from l3_hdr_offset, which would be ok, if l3_hdr_offset was valid. But it does not have to be, so it was a bad solution. Now the previous nodes mark whether it is a IPv6 or IPv4 packet tyle, and in esp_decrypt we count get ip header pointer by substracting the size of the ip header from the pointer to esp header (which lies in front of the ip header). Change-Id: I6d425b90931053711e8ce9126811b77ae6002a16 Signed-off-by: Szymon Sliwa <szs@semihalf.com>
2018-05-10dpdk:fix tx countEyal Bari1-1/+2
Change-Id: I921465ea64b59d42674cc8f19069ed04e3b25026 Signed-off-by: Eyal Bari <ebari@cisco.com>
2018-05-09dpdk: fix free of tx dropped packetsFlorin Coras1-1/+1
Change-Id: I3669068f694614f8555b33bf0b703c41e45363ef Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-05-09dpdk: tx code reworkDamjan Marion3-261/+158
Change-Id: Ifea9c772e8784642433b92091f5769eb9ec06890 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-09dpdk:fix mbuf index typo'sEyal Bari1-4/+4
Change-Id: I387b22427b3f322969bcf32fcfc189123c8ed6ae Signed-off-by: Eyal Bari <ebari@cisco.com>
2018-05-07dpdk: improve perf of buffer indices calc in the input nodeDamjan Marion1-34/+53
Change-Id: I16557189aa4a763ec496cb4a45f6e12f2d46971f Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-04ipsec: allow null/null for crypto/integ algorithms pairRadu Nicolau1-0/+1
Change-Id: Ic1e189c22e3d344d165e0eab05ccb667eef088a9 Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
2018-05-04Harmonize vec/pool_get_aligned object sizes and alignment requestsDave Barach2-0/+7
Object sizes must evenly divide alignment requests, or vice versa. Otherwise, only the first object will be aligned as requested. Three choices: add CLIB_CACHE_LINE_ALIGN_MARK(align_me) at the end of structures, manually pad to an even divisor or multiple of the alignment request, or use plain vectors/pools. static assert for enforcement. Change-Id: I41aa6ff1a58267301d32aaf4b9cd24678ac1c147 Signed-off-by: Dave Barach <dbarach@cisco.com>
2018-04-30plugins: dpdk: fix check which makes not sense, likely a typoSzymon Sliwa1-1/+1
Change-Id: If33854f9c32736edf571fb66cdfa759db1c9de25 Signed-off-by: Szymon Sliwa <szs@semihalf.com>
2018-04-27A bit of buffer metadata reshuffling to accommodate flow_idDamjan Marion1-1/+1
Change-Id: I2794384557c6272fe217269b14a9db09eda19220 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-25dpdk: complete rework of the dpdk-input nodeDamjan Marion7-465/+556
Change-Id: If174d189de40e6f9ffae99997bba93a2519d9fda Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-25ipsec: make crypto_worker_main_t a full cache line in sizeFlorin Coras1-0/+1
Change-Id: I927c9358915e03187cf7d3098c00b85b5ea2f92d Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-04-18dpdk: improve loggingDamjan Marion4-35/+187
- use of vlib_log for non-dataplane logging - redirect of dpdk logs trough unix pipe into vlib_log - "show dpdk physmem" cli Change-Id: I5da70f9c130273072a8cc80d169df31fc216b2c2 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-18dpdk: resurrect removed code [VPP-1245]Steven1-0/+4
https://gerrit.fd.io/r/#/c/7701/ accidentally removed 3 lines of code in dpdk/device/init.c which were added by https://gerrit.fd.io/r/#/c/7826/ case VNET_DPDK_PMD_VHOST_ETHER: xd->port_type = VNET_DPDK_PORT_TYPE_VHOST_ETHER; break; Those lines were needed to recognize vhost-user interface which is created via vdev command in dpdk and display VhostEthernet. Without them, UnknownEthernet is displayed. Change-Id: I6d7ee6aecc6a415fbb7308595d515649475bcd5f Signed-off-by: Steven <sluong@cisco.com>
2018-04-17dpdk: add additional fields to rte_mbuf trace printDamjan Marion1-2/+5
Change-Id: I3792c524323ce9e300648abe85454cff5d19aba7 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-17dpdk: print device flags in the 'show hardware' outputDamjan Marion2-11/+37
Change-Id: I0d10f13a56420b119fdfad97dcc135b245c269e1 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-13Revert "MTU: Setting of MTU on software interface (instead of hardware ↵Damjan Marion1-14/+14
interface)" This reverts commit 70083ee74c3141bbefb185525315f1b34497dcaa. Reverting as this patch is causing following crash: 0: /home/damarion/cisco/vpp3/build-data/../src/vnet/devices/devices.h:131 (vnet_get_device_input_thread_index) assertion `queue_id < vec_len (hw->input_node_thread_index_by_queue)' fails Aborted Change-Id: Ie2a365032110b1f67be7a9d832885b9899813d39 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-13MTU: Setting of MTU on software interface (instead of hardware interface)Ole Troan1-14/+14
Change-Id: I98bd454a761a1032738a21edeb0fe847e801f901 Signed-off-by: Ole Troan <ot@cisco.com>
2018-04-11dpdk: fail in early init if we cannot alloc hugepagesDamjan Marion1-4/+19
Change-Id: If536ae142dc0109b587d92981d337bc6f15e070a Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-10CSIT-895 dpdk/ipsec: add locks on session data hash updatesRadu Nicolau2-33/+40
Change-Id: I6400b77de388c01e85209e5dc5f11ccafb79a459 Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
2018-04-09plugins: unload plugin if early init failsDamjan Marion1-0/+43
Change-Id: I32f68e2ee8f5d32962acdefb0193583f71d342b3 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-04dpdk:fix checksum handling of l2 interfacesEyal Bari1-17/+19
dpdk-input was dropping packets with bad ip-checksum on l2 interfaces Change-Id: Ife5b52766bb71e878b1da6e94ae7b8a1e59fc478 Signed-off-by: Eyal Bari <ebari@cisco.com>
2018-03-30dpdk: fix crash due to incorrect xd->flags value with slave's link togglingSteve Shin1-2/+5
xd->flags is set incorrectly when a slave link is down in bonded interface mode. This can result in VPP crash when data traffic flows to the interface. Change-Id: Ideb9f5231db1211e8452c52fde646d681310c951 Signed-off-by: Steve Shin <jonshin@cisco.com>
2018-03-26plugins: dpdk: ipsec: fix l3 offsetSzymon Sliwa1-2/+1
Changes the source of the l3 offset to a more proper one, same as I5d9f41599ba8d8eb14ce2d9d523f82ea6e0fd10d. Change-Id: I5ff05d7d89507ecb378a2bd62f5b149189ca9e99 Signed-off-by: Szymon Sliwa <szs@semihalf.com>
2018-03-22VPP-1204: Fix coverity warningDave Barach1-2/+5
Change-Id: Iacb32e6e855f7b77108154d956ef27ee141bbde0 Signed-off-by: Dave Barach <dave@barachs.net>
2018-03-14vlib: internal buffer manager reworkDamjan Marion1-30/+18
- buffer_main is no longer part of vlib_main_t - pool of free lists is still part of vlib_main_t - mheap is not used anymore for buffer allocation - simple bitmap bassed buffer alloc scheme is introduced Change-Id: I3e1e6d00e2c8122293ed0a741245eb841315a1ff Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-03-14Prevent calling rte_eth_xstats_get not initialized devSzymon Sliwa1-0/+2
DPDK API forbids calling rte_eth_xstats_get on a port which has not been setup up yet. Citing the DPDK docs (reformated): " The functions exported by the application Ethernet API to setup a device designated by its port identifier must be invoked in the following order: rte_eth_dev_configure() rte_eth_tx_queue_setup() rte_eth_rx_queue_setup() rte_eth_dev_start() Then, the network application can invoke, in any order, the functions exported by the Ethernet API to get the MAC address of a given device, to get the speed and the status of a device physical link, to receive/transmit [burst of] packets, and so on. " original can be found here: http://dpdk.org/doc/api/rte__ethdev_8h.html#a36ba70a5a6fce2c2c1f774828ba78f8d Change-Id: I91854b8b0dd12dd028b4b36665cca49f16eac24c Signed-off-by: Szymon Sliwa <szs@semihalf.com>
2018-03-11dpdk-input node packet trace intermittent on IP forwading pathJohn Lo1-36/+37
Packet trace initiated by dpdk-input node would occasionally not fully function if next node is ip4-input. Change packet trace saving order in the quad-loop so "sho trace" will display trace in packet receive order. Fold calling of vlib_trace_buffer() into dpdk_add_trace(). Change-Id: I9d7a9bf3b9391f95590e66150b26b0b15912d803 Signed-off-by: John Lo <loj@cisco.com>
2018-03-09Coordinate known Ethernet speeds with Linux kernel and DPDKLee Roberts3-0/+43
Linux kernel and DPDK recognize the following Ethernet speeds: 10M, 100M, 1G, 2.5G, 5G, 10G, 20G, 25G, 40G, 50G, 56G and 100G. Add consistent Ethernet speeds to VPP. Change-Id: I4cfcf378fb34425c1206db5aa2f6bdcc66e0a6ab Signed-off-by: Lee Roberts <lee.roberts@hpe.com>
2018-03-09dpdk: move DPDK vfio hack to dpdk pluginDamjan Marion1-2/+67
Change-Id: I806cbf8c6c49643fe6c317bcceab93c1b9d441ab Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-03-05vlib: vfio code reworkDamjan Marion1-1/+3
Change-Id: I99cf3e7cc991aa7d32385a155c707a6516516117 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-02-26vnet: add 25G interface speed flagDamjan Marion1-0/+6
Change-Id: I1d3ede2b043e1fd4abc54f540bb1d3ac9863016e Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-02-25Fix bug in dpdk_crypto_session_disposal()Matthew Smith1-0/+2
The expression to determine whether to delete a session from the disposal list only evaluates true if some, but not all, of the sessions in the list were freed. When all sessions in the list are freed, it evaluates false and the sessions are left in the list to be freed again later, which can result in a session pool element that was reallocated to a different SA being freed, breaking crypto for the newer SA. Add an 'else' that handles the case where all sessions were freed. Change-Id: I3ae54d5b3bfc3658bf406caa50646924baaae589 Signed-off-by: Matthew Smith <mgsmith@netgate.com>
2018-02-24Fix crypto session deletion crashMatthew Smith1-6/+9
When using a DPDK cryptodev with IPsec, deleting a session often results in a SEGV. A bad pointer is being passed to rte_cryptodev_sym_session_free(). Put the correct value on the crypto disposal list and add a check to determine whether the call to free the session is going to result in a crash before doing it. Change-Id: I8a6b0a594585ebcfa56b555ede7ef7d67e5e2b33 Signed-off-by: Matthew Smith <mgsmith@netgate.com>
2018-02-23Disable scatter-gather for ENAMatthew Smith1-0/+1
ENA doesn't support scatter-gather. The PMD started failing rte_eth_dev_rx_queue_config() in DPDK 18.02 if the flag to enable it is set. Turn the flag off in dpdk_lib_init(). Change-Id: Ifdd9f188c89b46efe82412c75fb935a92436da1c Signed-off-by: Matthew Smith <mgsmith@netgate.com>
2018-02-19dpdk: Fix MTU calc for NICs that support mtu<9216Nitin Saxena1-34/+66
Problem: rte_eth_dev_set_mtu() returns with failure from ThunderX NICVF DPDK PMD driver which supports MTU less than ETHERNET_MAX_PACKET_BYTES. rte_eth_dev_set_mtu() being called twice from dpdk_lib_init(): one via dpdk_device_setup() and second in dpdk_lib_init() itself. Currently dpdk_lib_init() passes vnet_hardware_interface->max_packet_bytes as an argument to rte_eth_dev_set_mtu() without consulting dev_info.max_rx_pktlen. NICs like i4oe, ixgbe can support MTU much greater than 9216 hence its not a problem for those NICS. Fix: This patch calculates dpdk_device->port_conf.rxmode.max_rx_pkt_len, vnet_hardware_interface->max_packet_bytes and MTU by consulting dev_info.max_rx_pktlen. Change-Id: If04bbfae49ee971dac0063ff1835e4a9c3087865 Signed-off-by: Nitin Saxena <nitin.saxena@cavium.com>
2018-02-15dpdk: add support for DPDK 18.02, deprecate 17.08Damjan Marion4-24/+0
17.11 is still default. Change-Id: I524d232579db8a59c717c5d760398b6b7f811d03 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-02-12dpdk:fix trace to follow feature arcEyal Bari3-87/+66
Change-Id: Icd2184dec16d30cdcc689ca37c834b5df2f0a1a3 Signed-off-by: Eyal Bari <ebari@cisco.com>
2018-02-07Refactor vlib_buffer flagsDamjan Marion3-16/+6
Change-Id: I853386aebfe488ebb10328435b81b6e3403c5dd0 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-02-01dpdk:fix typo in prefetch sequenceEyal Bari1-2/+2
Change-Id: I7110436626352d45ffe0ca71fb88dea2c77ab639 Signed-off-by: Eyal Bari <ebari@cisco.com>