netisr(9) Kernel network dispatch service


In 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


The 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 registration

Protocols 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 .

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:

should maintain source ordering without advice from the protocol. will ignore any flow IDs present on Vt mbuf headers for the purposes of work placement.
should maintain flow ordering as defined by the Vt mbuf header flow ID field. If the protocol implements nh_m2flow then will query the protocol in the event that the Vt mbuf doesn't have a flow ID, falling back on source ordering.
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(2) 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 M_FLOWID for packets that do not enter with M_FLOWID defined. 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 interface

Packet 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 constants

The follow protocol numbers are currently defined:

IGMPv3 loopback
Routing socket loopback
Appletalk AARP
Appletalk phase 1
Appletalk phase 2
netstat(1), epair(4)


This manual page and the implementation were written by An Robert N. M. Watson .