aboutsummaryrefslogtreecommitdiffstats
path: root/src/vnet/buffer.h
AgeCommit message (Collapse)AuthorFilesLines
2024-02-13vnet: Don't use __unused for struct paddingTom Jones1-1/+1
__unused is a clang keyword, this struct member will trip the build when using clang. Instead call the unused padding 'pad' which should be clear to the purpose if not the usage. Type: improvement Change-Id: I0abae34841651be1ef6b7d94864f0dc8185f0733 Signed-off-by: Tom Jones <thj@freebsd.org>
2022-06-29classify: use 32 bits hashBenoît Ganne1-2/+2
classify hash used to be stored as u64 in buffer metadata, use 32 bits instead: - on almost all our supported arch (x86 and arm64) we use crc32c intrinsics to compute the final hash: we really get a 32-bits hash - the hash itself is used to compute a 32-bits bucket index by masking upper bits: we always discard the higher 32-bits - this allows to increase the l2 classify buffer metadata padding such as it does not overlap with the ip fib_index metadata anymore. This overlap is an issue when using the 'set metadata' action in the ip ACL node which updates both fields Type: fix Change-Id: I5d35bdae97b96c3cae534e859b63950fb500ff50 Signed-off-by: Benoît Ganne <bganne@cisco.com>
2022-03-23vnet: Remove the unused fields from opaque2Neale Ranns1-11/+3
Type: refactor Signed-off-by: Neale Ranns <neale@graphiant.com> Change-Id: Ibb6d19de053c306e9758dbfa827ab7bcab5de856
2021-12-31misc: Remove the unused GBP fields from the buffer meta-dataNeale Ranns1-9/+1
Type: refactor Signed-off-by: Neale Ranns <neale@graphiant.com> Change-Id: I92496501360ee073795206bde87f4731a5ce074c
2021-11-04ip: reassembly: avoid reading truncated L4 headersKlement Sekera1-1/+2
Check if L4 headers are truncated and if so, set a flag for (future) consumers instead of reading/writing garbage data. Type: fix Fixes: de34c35fc73226943538149fae9dbc5cfbdc6e75 Change-Id: I0b656ec103a11c356b98a6f36cad98536a78d1dc Signed-off-by: Klement Sekera <ksekera@cisco.com>
2021-11-02fib: allow vnet rewrite max size to be smaller the pre_dataBenoît Ganne1-1/+3
There is no reason to enforce vnet rewrite size to be equal to pre_data. Moreover, since vnet rewrite size is now saved as u8, this limits pre_data to 192 bytes. Type: fix Fixes: 7dbf9a1a4fff5c3b20ad972289e49e3f88e82f2d Change-Id: I3f848aa905ea4a794f3b4aa62c929a481261a3f1 Signed-off-by: Benoît Ganne <bganne@cisco.com>
2021-10-21ip: Add ip46-local node for local swif[rx]Nathan Skrzypczak1-2/+11
Type: improvement This adds a new ip[46]-receive node, sibling of ip[46]-local. Its goal is to set vnet_buffer (b)->ip.rx_sw_if_index to the sw_if_index of the local interface. In dependant nodes further down the line (e.g. hoststack) we then set sw_if_idx[rx] to this value. So that we know which local interface did receive the packet. The TCP issue this fixes is that : On accepts, we were setting tc->sw_if_index to the source sw_if_index. We should use the dest sw_if_index, so that packets coming back on this connection have the right source sw_if_index. And also setting it in the tx-ed packet. Change-Id: I569ed673e15c21e71f365c3ad45439b05bd14a9f Signed-off-by: Nathan Skrzypczak <nathan.skrzypczak@gmail.com>
2021-08-20vlib: add format_vnet_buffer_no_chainBenoît Ganne1-1/+1
- add format_vnet_buffer and format_vnet_buffer_no_chain to mirror format_vlib_buffer and format_vlib_buffer_no_chain - format_vnet_buffer used to be the "no chain" version, replace all of its current use with the corresponding format_vnet_buffer_no_chain - add a function to dump vnet buffer details from gdb Type: improvement Change-Id: I143ce845f80e7ef937ea33a557b6e3b5988c5b8f Signed-off-by: Benoît Ganne <bganne@cisco.com>
2021-07-16vlib: add tunnel offload flags for vlib_buffer_tMohsin Kazmi1-2/+8
Type: improvement Change-Id: Iaad50b2044702c46eff287708dfcb24e61022104 Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2021-05-06vlib: fix the offload flags sizeMohsin Kazmi1-18/+18
Type: fix Change-Id: I433fe3799975fe3ba00fa30226f6e8dae34e88fc Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2021-05-05vlib: move offload flags to 1st cacheline in vlib_buffer_tMohsin Kazmi1-8/+6
Type: improvement Some tests i.e. ipsec see performance regression when offload flags are moved to 2nd cacheline. This patch moves them back to 1st cacheline. Change-Id: I6ead45ff6d2c467b0d248f409e27c2ba31758741 Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2021-05-01vlib: refactor trajectory trace debug featureBenoît Ganne1-7/+0
trajectory trace has been broken for a while because we used to save the buffer trajectory in a vector pointed to in opaque2. This does not work well when opaque2 is copied (eg. because of a clone) as 2 buffers end up sharing the same vector. This dedicates a full cacheline in the buffer metadata instead when trajectory is compiled in. No dynamic allocation, no sharing, no tears. Type: refactor Change-Id: I6a028ca1b48d38f393a36979e5e452c2dd48ad3f Signed-off-by: Benoît Ganne <bganne@cisco.com>
2021-03-09nat: fix dst nat thread issuesKlement Sekera1-2/+5
Make sure packet lands on the right thread for dst nat case. Type: fix Signed-off-by: Klement Sekera <ksekera@cisco.com> Change-Id: I0ec4e4c2bb3fa80ff73fac588c36d36420ba68fa
2021-03-07dispatch-trace: move dispatch trace pcap code to pluginDamjan Marion1-0/+3
Type: refactor Change-Id: I02a527f57853ebff797f0d85761b71127916d6ce Signed-off-by: Damjan Marion <damarion@cisco.com>
2021-02-23vlib: fix offload flags value resetBenoît Ganne1-2/+11
When a buffer is freed and re-allocated for a new packet, opaque2 is not reset, so the offload flags can be set to a stale value. Make sure the offload flags are reset to the current value on 1st set. Type: fix Fixes: 6809538e646bf86c000dc1faba60b0a4157ad898 Change-Id: I4048febedf25b9995dbd080a11495ee7dbe59153 Signed-off-by: Benoît Ganne <bganne@cisco.com>
2021-02-15vlib: refactor checksum offload supportMohsin Kazmi1-37/+71
Type: refactor This patch refactors the offload flags in vlib_buffer_t. There are two main reasons behind this refactoring. First, offload flags are insufficient to represent outer and inner headers offloads. Second, room for these flags in first cacheline of vlib_buffer_t is also limited. This patch introduces a generic offload flag in first cacheline. And detailed offload flags in 2nd cacheline of the structure for performance optimization. Change-Id: Icc363a142fb9208ec7113ab5bbfc8230181f6004 Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2021-02-12ipsec: Store thread-index in buffer meta-data during SA handoffNeale Ranns1-0/+3
Type: improvement negates the need to load the SA in the handoff node. don't prefetch the packet data, it's not needed. Signed-off-by: Neale Ranns <neale@graphiant.com> Change-Id: I340472dc437f050cc1c3c11dfeb47ab09c609624
2021-02-10nat: fix EI hairpinning thread safetyKlement Sekera1-0/+1
Avoid doing inter-thread reads without locks by doing a handoff before destination address rewrite. Destination address is read from a session which is possibly owned by a different thread. By splitting the work in two parts with a handoff in the middle, we can do both in a thread safe way. Type: improvement Signed-off-by: Klement Sekera <ksekera@cisco.com> Change-Id: I1c50d188393a610f5564fa230c75771a8065f273
2021-01-18nat: deal with flows instead of sessionsKlement Sekera1-3/+5
This change introduces flow concept to endpoint-dependent NAT. Instead of having a session and a plethora of special cases in code for e.g. hairpinning, twice-nat and others, figure all this out and store it in flow logic. Every flow has a match and a rewrite part. This unifies all the NAT packet processing cases into one - match a flow and rewrite the packet based on that flow. It also provides a cure for hairpinning dilemma where one part of the flow is on one worker and another on a different one. These cases are also sped up by not requiring destination adress lookup every single time to be able to rewrite source nat as this is now part of flow rewrite logic. Type: improvement Change-Id: Ib60c992e16792ea4d4129bc10202ebb99a73b5be Signed-off-by: Klement Sekera <ksekera@cisco.com>
2020-12-03nat: fix forwarding handoff workaroundFilip Varga1-3/+2
A special case when out2in packet needs to be handoffed to other worker thread. We are not able to determine which thread they belong to in the first processing of nat handoff node. These packets needs to go through out2in slowpath before we are able to tell where to handoff them. Type: fix Ticket: VPP-1941 Change-Id: I7173bda970ce6a91d81f48fc72aa2457586a076f Signed-off-by: Filip Varga <fivarga@cisco.com>
2020-05-14nat: ED: store both thread&session idx in hashKlement Sekera1-1/+1
By storing thread and session index in hash table we are able to skip multiple hash lookups in multi-worker scenario, which were used for handoff before. Also, by storing sesion index in vnet_buffer2, we can avoid repeating the lookup after handoff. Type: improvement Signed-off-by: Klement Sekera <ksekera@cisco.com> Change-Id: I406fb12f4e2dd8f4a5ca5d83d59dbc37e1af9abf
2020-04-24nat: make usage of vnet_buffer2 transparentKlement Sekera1-0/+6
Type: fix Signed-off-by: Klement Sekera <ksekera@cisco.com> Change-Id: I50df248afb3f6b46c49e6695b3f124cfd584f016
2020-01-15ip: align reass.save_rewrite_lengthv20.05-rc0Klement Sekera1-4/+13
By aligning vnet_buffer_opaque.ip.save_rewrite_length and vnet_buffer_opaque.ip.reass.save_rewrite_length we prevent shallow virtual reassembly code from overwrite save_rewrite_length, allowing other features down the pipe to rely on this value. A static assert is added to guard this alignment. Type: fix Fixes: f126e746fc01c75bc99329d10ce9127b26b23814 Change-Id: Ie7c7f3abc2a221bbcf2830c0f006a4368088b342 Signed-off-by: Klement Sekera <ksekera@cisco.com>
2020-01-03nat: use SVRKlement Sekera1-19/+26
Remove NAT's implementation of shallow virtual reassembly with corresponding CLIs, APIs & tests. Replace with standalone shallow virtual reassembly provided by ipX-sv-reass* nodes. Type: refactor Change-Id: I7e6c7487a5a500d591f6871474a359e0993e59b6 Signed-off-by: Klement Sekera <ksekera@cisco.com>
2019-11-26fib: reduce save_rewrite_length to u8Klement Sekera1-2/+12
This is a preparation step for migrating NAT to use SVR (shallow virtual reassembly) to conserve space in vnet_buffer. Since max rewrite length is currently pre-data size (128), u8 is sufficient to hold that value. Type: refactor Change-Id: I5374bb396e178245b870cb0bbf1370d2a54230bc Signed-off-by: Klement Sekera <ksekera@cisco.com>
2019-10-22ip: reassembly: avoid corruption in vnet bufferKlement Sekera1-17/+18
Avoid corrupt next_index in vnet_buffer by moving input and output variables into different memory places instead of sharing a common space. Type: fix Fixes: de34c35fc73226943538149fae9dbc5cfbdc6e75 Change-Id: I34471fc6d0c8487535fac21349e688f398934f6d Signed-off-by: Klement Sekera <ksekera@cisco.com>
2019-10-09mpls: support fragmentation of mpls output packetRajesh Goel1-0/+4
Type: fix Signed-off-by: Rajesh Goel <rajegoel@cisco.com> Change-Id: Ie4372c5cf58ab215cdec5ce56f8a994daaba2844
2019-09-26ip: add shallow virtual reassembly functionalityKlement Sekera1-7/+32
Type: feature Change-Id: Ibc8334e26c7e6f6120696c3e313b6e11d73dab99 Signed-off-by: Klement Sekera <ksekera@cisco.com>
2019-07-22gso: Fix the l3 packet gso segment sizeMohsin Kazmi1-1/+4
Type: fix Ticket: VPP-1721 Change-Id: I7a5d4f1440048ddc9f599ac11d06e5a7df20440e Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2019-07-05sctp: move to plugins, disabled by defaultFlorin Coras1-14/+0
Removed sctp buffer metadata from vnet/buffer.h, added it to the plugin. Add registration APIs for plugin-based vlib_buffer_opaque / opaque2 decoders, used by "pcap dispatch trace ..." for display in the wireshark dissector. Type:refactor Not actively maintained. Change-Id: Ie4cb6ba66f68b3b3a7d7d2c63c917fdccf994371 Signed-off-by: Florin Coras <fcoras@cisco.com> Signed-off-by: Dave Barach <dave@barachs.net>
2019-06-21tcp: add option to pass opaque to next nodeFlorin Coras1-1/+5
Type:feature Change-Id: I0b72954a6ae6a05abe0761cb4f227072863f127b Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-06-18ipsec: ipsec-tun protectNeale Ranns1-0/+1
please consult the new tunnel proposal at: https://wiki.fd.io/view/VPP/IPSec Type: feature Change-Id: I52857fc92ae068b85f59be08bdbea1bd5932e291 Signed-off-by: Neale Ranns <nranns@cisco.com>
2019-06-18ip: reassembly-separate feature and custom codeKlement Sekera1-2/+2
This change is made fix a crash, because is_feature flag semantics turn out to be different from "custom app code" semantics. Introduce a flag which custom plugins/apps can use to instead of tying that code to is_feature flag. Change-Id: Ief5898711e68529f9306cfac54c4dc9b3650f9e3 Ticket: N/A Type: fix Fixes: 21aa8f1022590b8b5caf819b4bbd485de0f1dfe5 Signed-off-by: Klement Sekera <ksekera@cisco.com> Signed-off-by: Ole Troan <ot@cisco.com>
2019-06-06DVR: Control the reinject as L2 or L3 based on the output interface typeNeale Ranns1-1/+1
Change-Id: Ib4cdbe8a6a1d10a643941c13aa0acbed410f876c Type: Feature Signed-off-by: Neale Ranns <nranns@cisco.com>
2019-06-06reassembly: support more custom options for appsKlement Sekera1-0/+1
Change-Id: Ib9f98fba5a724480ca95f11a762002c53e08df70 Signed-off-by: Klement Sekera <ksekera@cisco.com>
2019-03-29IPSEC-GRE: fixes and API update to common types.Neale Ranns1-1/+0
Change-Id: Icdcbac7453baa837a9c0c4a2401dff4a6aa6cba0 Signed-off-by: Neale Ranns <nranns@cisco.com>
2019-03-06GBP: use sclass in the DP for policyNeale Ranns1-5/+1
Change-Id: I154e18f22ec7708127b8ade98e80546ab1dcd05b Signed-off-by: Neale Ranns <nranns@cisco.com>
2019-03-04Hash and handoff reassembly fragmentsVijayabhaskar Katamreddy1-0/+2
in the following two scenarios 1. When fragments arrive in multiple interfaces and endup in different threads 2. When fragments arrive in same interafce but in different queues due to interface RSS doesnt have the ability to place fragments in the right queues Change-Id: I9f9a8a4085692055ef6823d634c8e19ff3daea05 Signed-off-by: Vijayabhaskar Katamreddy <vkatamre@cisco.com>
2019-02-19tap gso: experimental supportAndrew Yourtchenko1-9/+26
This commit adds a "gso" parameter to existing "create tap..." CLI, and a "no-gso" parameter for the compatibility with the future, when/if defaults change. It makes use of the lowest bit of the "tap_flags" field in the API call in order to allow creation of GSO interfaces via API as well. It does the necessary syscalls to enable the GSO and checksum offload support on the kernel side and sets two flags on the interface: virtio-specific virtio_if_t.gso_enabled, and vnet_hw_interface_t.flags & VNET_HW_INTERFACE_FLAG_SUPPORTS_GSO. The first one, if enabled, triggers the marking of the GSO-encapsulated packets on ingress with VNET_BUFFER_F_GSO flag, and setting vnet_buffer2(b)->gso_size to the desired L4 payload size. VNET_HW_INTERFACE_FLAG_SUPPORTS_GSO determines the egress packet processing in interface-output for such packets: When the flag is set, they are sent out almost as usual (just taking care to set the vnet header for virtio). When the flag is not enabled (the case for most interfaces), the egress path performs the re-segmentation such that the L4 payload of the transmitted packets equals gso_size. The operations in the datapath are enabled only when there is at least one GSO-compatible interface in the system - this is done by tracking the count in interface_main.gso_interface_count. This way the impact of conditional checks for the setups that do not use GSO is minimized. "show tap" CLI shows the state of the GSO flag on the interface, and the total count of GSO-enabled interfaces (which is used to enable the GSO-related processing in the packet path). This commit lacks IPv6 extension header traversal support of any kind - the L4 payload is assumed to follow the IPv6 header. Also it performs the offloads only for TCP (TSO - TCP segmentation offload). The UDP fragmentation offload (UFO) is not part of it. For debug purposes it also adds the debug CLI: "set tap gso {<interface> | sw_if_index <sw_idx>} <enable|disable>" Change-Id: Ifd562db89adcc2208094b3d1032cee8c307aaef9 Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
2019-01-22GBP: Sclass to src-epg conversionsNeale Ranns1-1/+5
Change-Id: Ica88268fd6a6ee01da7e9219bb4e81f22ed2fd4b Signed-off-by: Neale Ranns <nranns@cisco.com>
2019-01-02Revert "add ipsecmb plugin"Florin Coras1-4/+0
This reverts commit be16020c5034bc69df25a8ecd7081aec9898d93c. The arm verify job actually failed but the result was overwritten by an x86 ubuntu retry. Change-Id: Idcae7691fc575053563b8ff8bcad661c15891668 Signed-off-by: Florin Coras <fcoras@cisco.com>
2019-01-02add ipsecmb pluginKlement Sekera1-0/+4
Change-Id: I99c0737dfeeec2db267773625ddc9b55324fd237 Signed-off-by: Klement Sekera <ksekera@cisco.com>
2018-12-04Add VNET_BUFFER_F_AVAIL definitionsDave Barach1-1/+25
Add a check to make sure that the vlib and vnet buffer flag bit definitions do not overlap. The VNET_BUFFER_F_AVAIL1...8 definitions allow out-of-tree codes to: #define VNET_BUFFER_F_MY_USECASE VNET_BUFFER_F_AVAIL1 and so on. This avoids introducing irrelevant and/or proprietary bit definitions into vnet/buffer.h, and hopefully minimizes merge pain for everyone involved. Change-Id: I5be4f61dceb81b5bfca005f6d609ade074af205b Signed-off-by: Dave Barach <dave@barachs.net>
2018-11-30Metadata / opaque formatting belongs in vppDave Barach1-22/+22
VPP graph dispatch trace record description: 0 1 2 3 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Major Version | Minor Version | NStrings | ProtoHint | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Buffer index (big endian) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + VPP graph node name ... ... | NULL octet | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Buffer Metadata ... ... | NULL octet | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Buffer Opaque ... ... | NULL octet | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Buffer Opaque 2 ... ... | NULL octet | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | VPP ASCII packet trace (if NStrings > 4) | NULL octet | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | Packet data (up to 16K) | +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ Graph dispatch records comprise a version stamp, an indication of how many NULL-terminated strings will follow the record header, and a protocol hint. The buffer index allows downstream consumers of these data to easily filter/track single packets as they traverse the forwarding graph. FWIW, the 32-bit buffer index is stored in big endian format. As of this writing, major version = 1, minor version = 0. Nstrings will be either 4 or 5. Here is the current set of protocol hints: typedef enum { VLIB_NODE_PROTO_HINT_NONE = 0, VLIB_NODE_PROTO_HINT_ETHERNET, VLIB_NODE_PROTO_HINT_IP4, VLIB_NODE_PROTO_HINT_IP6, VLIB_NODE_PROTO_HINT_TCP, VLIB_NODE_PROTO_HINT_UDP, VLIB_NODE_N_PROTO_HINTS, } vlib_node_proto_hint_t; Example: VLIB_NODE_PROTO_HINT_IP6 means that the first octet of packet data SHOULD be 0x60, and should begin an ipv6 packet header. Change-Id: Idf310bad80cc0e4207394c80f18db5f77c378741 Signed-off-by: Dave Barach <dave@barachs.net>
2018-09-27IPIP and IPv6 fragmentationOle Troan1-1/+0
- Error where ICMPv6 error code doesn't reset VLIB_TX = -1 Leading to crash for ICMP generated on tunnelled packets - Missed setting VNET_BUFFER_F_LOCALLY_ORIGINATED, so IP in IPv6 packets never got fragmented. - Add support for fragmentation of buffer chains. - Remove support for inner fragmentation in frag code itself. Change-Id: If9a97301b7e35ca97ffa5c0fada2b9e7e7dbfb27 Signed-off-by: Ole Troan <ot@cisco.com>
2018-09-24Trivial: Clean up some typos.Paul Vinciguerra1-4/+4
Change-Id: I085615fde1f966490f30ed5d32017b8b088cfd59 Signed-off-by: Paul Vinciguerra <pvinci@vinciconsulting.com>
2018-09-10vxlan-gbp: Add support for vxlan gbpMohsin Kazmi1-1/+3
This patch implements vxlan with extension of group based policy support. Change-Id: I70405bf7332c02867286da8958d9652837edd3c2 Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2018-08-03loop counter to prevent infiinte number of look ups per-packetNeale Ranns1-1/+3
Change-Id: I59235d11baac18785a4c90cdaf14e8f3ddf06dab Signed-off-by: Neale Ranns <neale.ranns@cisco.com>
2018-07-19Cleanup of handoff codeDamjan Marion1-15/+0
- removed handoff-dispatch node - removed some unused buffer metadata fields - enqueue to thread logic moved to inline function Change-Id: I7361e1d88f8cce74cd4fcec90d172eade1855cbd Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-02Fixing ip4-reassembly ip6-reassembly non feature node, next_node_index settingVijayabhaskar Katamreddy1-1/+0
Change-Id: Ib988d87e6758ffa31862096391f9f286b0797f2b Signed-off-by: Vijayabhaskar Katamreddy <vkatamre@cisco.com>