Copyright (c) 2009 Robert N. M. Watson All rights reserved. Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met: 1. Redistributions of source code must retain the above copyright notice(s), this list of conditions and the following disclaimer as the first lines of this file unmodified other than the possible addition of one or more copyright notices. 2. Redistributions in binary form must repro...
NAMEnetisr - Kernel network dispatch service
SYNOPSISIn net/netisr.h Ft void Fn netisr_register const struct netisr_handler *nhp Ft void Fn netisr_unregister const struct netisr_handler *nhp Ft int Fn netisr_dispatch u_int proto struct mbuf *m Ft int Fn netisr_dispatch_src u_int proto uintptr_t source struct mbuf *m Ft int Fn netisr_queue u_int proto struct mbuf *m Ft int Fn netisr_queue_src u_int proto uintptr_t source struct mbuf *m Ft void Fn netisr_clearqdrops const struct netisr_handler *nhp Ft void Fn netisr_getqdrops const struct netisr_handler *nhp uint64_t *qdropsp Ft void Fn netisr_getqlimit const struct netisr_handler *nhp u_int *qlimitp Ft int Fn netisr_setqlimit const struct netisr_handler *nhp u_int qlimit Ft u_int Fn netisr_default_flow2cpu u_int flowid Ft u_int Fn netisr_get_cpucount void Ft u_int Fn netisr_get_cpuid u_int cpunumber
With optional virtual network stack support enabled via the following kernel compile option:
options VIMAGEFt void Fn netisr_register_vnet const struct netisr_handler *nhp Ft void Fn netisr_unregister_vnet const struct netisr_handler *nhp
DESCRIPTIONThe sysctl Cm net.inet.tcp.syncookies kernel interface suite allows device drivers (and other packet sources) to direct packets to protocols for directly dispatched or deferred processing. Protocol registration and work stream statistics may be monitored using netstat(1).
Protocol registrationProtocols register and unregister handlers using Fn netisr_register and Fn netisr_unregister , and may also manage queue limits and statistics using the Fn netisr_clearqdrops , Fn netisr_getqdrops , Fn netisr_getqlimit , and Fn netisr_setqlimit .
In case of VIMAGE kernels each virtual network stack (vnet), that is not the default base system network stack, calls Fn netisr_register_vnet and Fn netisr_unregister_vnet to enable or disable packet processing by the sysctl Cm net.inet.tcp.syncookies for each protocol. Disabling will also purge any outstanding packet from the protocol queue.
sysctl Cm net.inet.tcp.syncookies supports multi-processor execution of handlers, and relies on a combination of source ordering and protocol-specific ordering and work-placement policies to decide how to distribute work across one or more worker threads. Registering protocols will declare one of three policies:
- sysctl Cm net.inet.tcp.syncookies should maintain source ordering without advice from the protocol. sysctl Cm net.inet.tcp.syncookies will ignore any flow IDs present on Vt mbuf headers for the purposes of work placement.
- sysctl Cm net.inet.tcp.syncookies should maintain flow ordering as defined by the Vt mbuf header flow ID field. If the protocol implements nh_m2flow then sysctl Cm net.inet.tcp.syncookies will query the protocol in the event that the Vt mbuf doesn't have a flow ID, falling back on source ordering.
- sysctl Cm net.inet.tcp.syncookies will entirely delegate all work placement decisions to the protocol, querying nh_m2cpuid for each packet.
Registration is declared using Vt struct netisr_handler , whose fields are defined as follows:
- Vt const char * nh_name
- Unique character string name of the protocol, which may be included in sysctl(3) MIB names, so should not contain whitespace.
- Vt netisr_handler_t nh_handler
- Protocol handler function that will be invoked on each packet received for the protocol.
- Vt netisr_m2flow_t nh_m2flow
- Optional protocol function to generate a flow ID and set a valid hashtype for packets that enter the sysctl Cm net.inet.tcp.syncookies with M_HASHTYPE_GET(m) equal to M_HASHTYPE_NONE Will be used only with NETISR_POLICY_FLOW
- Vt netisr_m2cpuid_t nh_m2cpuid
- Protocol function to determine what CPU a packet should be processed on. Will be used only with NETISR_POLICY_CPU
- Vt netisr_drainedcpu_t nh_drainedcpu
- Optional callback function that will be invoked when a per-CPU queue was drained. It will never fire for directly dispatched packets. Unless fully understood, this special-purpose function should not be used.
- Vt u_int nh_proto
- Protocol number used by both protocols to identify themselves to , and by packet sources to select what handler will be used to process packets. A table of supported protocol numbers appears below. For implementation reasons, protocol numbers great than 15 are currently unsupported.
- Vt u_int nh_qlimit
- The maximum per-CPU queue depth for the protocol; due to internal implementation details, the effective queue depth may be as much as twice this number.
- Vt u_int nh_policy
- The ordering and work placement policy for the protocol, as described earlier.
Packet source interfacePacket sources, such as network interfaces, may request protocol processing using the Fn netisr_dispatch and Fn netisr_queue interfaces. Both accept a protocol number and Vt mbuf argument, but while Fn netisr_queue will always execute the protocol handler asynchronously in a deferred context, Fn netisr_dispatch will optionally direct dispatch if permitted by global and per-protocol policy.
In order to provide additional load balancing and flow information, packet sources may also specify an opaque source identifier, which in practice might be a network interface number or socket pointer, using the Fn netisr_dispatch_src and Fn netisr_queue_src variants.
Protocol number constantsThe follow protocol numbers are currently defined:
- IGMPv3 loopback
- Routing socket loopback
- netstat(1), epair(4)