HardenedBSD/share/man/man9/drbr.9

141 lines
4.0 KiB
Groff
Raw Normal View History

.\" Copyright (c) 2009 Bitgravity Inc
.\" Written by: Kip Macy <kmacy@@FreeBSD.org>
.\" 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, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
The drbr(9) API appeared to be so unclear, that most drivers in tree used it incorrectly, which lead to inaccurate overrated if_obytes accounting. The drbr(9) used to update ifnet stats on drbr_enqueue(), which is not accurate since enqueuing doesn't imply successful processing by driver. Dequeuing neither mean that. Most drivers also called drbr_stats_update() which did accounting again, leading to doubled if_obytes statistics. And in case of severe transmitting, when a packet could be several times enqueued and dequeued it could have been accounted several times. o Thus, make drbr(9) API thinner. Now drbr(9) merely chooses between ALTQ queueing or buf_ring(9) queueing. - It doesn't touch the buf_ring stats any more. - It doesn't touch ifnet stats anymore. - drbr_stats_update() no longer exists. o buf_ring(9) handles its stats itself: - It handles br_drops itself. - br_prod_bytes stats are dropped. Rationale: no one ever reads them but update of a common counter on every packet negatively affects performance due to excessive cache invalidation. - buf_ring_enqueue_bytes() reduced to buf_ring_enqueue(), since we no longer account bytes. o Drivers handle their stats theirselves: if_obytes, if_omcasts. o mlx4(4), igb(4), em(4), vxge(4), oce(4) and ixv(4) no longer use drbr_stats_update(), and update ifnet stats theirselves. o bxe(4) was the most correct driver, it didn't call drbr_stats_update(), thus it was the only driver accurate under moderate load. Now it also maintains stats itself. o ixgbe(4) had already taken stats from hardware, so just - drop software stats updating. - take multicast packet count from hardware as well. o mxge(4) just no longer needs NO_SLOW_STATS define. o cxgb(4), cxgbe(4) need no change, since they obtain stats from hardware. Reviewed by: jfv, gnn
2012-09-28 20:28:27 +02:00
.Dd September 27, 2012
.Dt DRBR 9
.Os
.Sh NAME
.Nm drbr ,
.Nm drbr_free ,
.Nm drbr_enqueue ,
.Nm drbr_dequeue ,
.Nm drbr_dequeue_cond ,
.Nm drbr_flush ,
.Nm drbr_empty ,
.Nm drbr_inuse ,
.Nd network driver interface to buf_ring
.Sh SYNOPSIS
.In sys/param.h
.In net/if.h
.In net/if_var.h
.Ft void
.Fn drbr_free "struct buf_ring *br" "struct malloc_type *type"
.Ft int
.Fn drbr_enqueue "struct ifnet *ifp" "struct buf_ring *br" "struct mbuf *m"
.Ft struct mbuf *
.Fn drbr_dequeue "struct ifnet *ifp" "struct buf_ring *br"
.Ft struct mbuf *
.Fn drbr_dequeue_cond "struct ifnet *ifp" "struct buf_ring *br" "int (*func) (struct mbuf *, void *)" "void *arg"
.Ft void
.Fn drbr_flush "struct ifnet *ifp" "struct buf_ring *br"
.Ft int
.Fn drbr_empty "struct ifnet *ifp" "struct buf_ring *br"
.Ft int
.Fn drbr_inuse "struct ifnet *ifp" "struct buf_ring *br"
.Sh DESCRIPTION
The
.Nm
functions provide an API to network drivers for using
.Xr buf_ring 9
for enqueueing and dequeueing packets.
This is meant as a replacement for the IFQ interface for packet queuing.
It allows a packet to be enqueued with a single atomic and packet
dequeue to be done without any per-packet atomics as it is protected
by the driver's tx queue lock.
If
.Dv INVARIANTS
is enabled,
.Fn drbr_dequeue
will assert that the tx queue lock is held when it is called.
.Pp
The
.Fn drbr_free
function frees all the enqueued mbufs and then frees the buf_ring.
.Pp
The
.Fn drbr_enqueue
function is used to enqueue an mbuf to a buf_ring, falling back to the
ifnet's IFQ if
.Xr ALTQ 4
is enabled.
.Pp
The
.Fn drbr_dequeue
function is used to dequeue an mbuf from a buf_ring or, if
.Xr ALTQ 4
is enabled, from the ifnet's IFQ.
.Pp
The
.Fn drbr_dequeue_cond
function is used to conditionally dequeue an mbuf from a buf_ring based
on whether
.Fa func
returns
.Dv TRUE
or
.Dv FALSE .
.Pp
The
.Fn drbr_flush
function frees all mbufs enqueued in the buf_ring and the ifnet's IFQ.
.Pp
The
.Fn drbr_empty
function returns
.Dv TRUE
if there are no mbufs enqueued,
.Dv FALSE
otherwise.
.Pp
The
.Fn drbr_inuse
function returns the number of mbufs enqueued.
Note to users that this is intrinsically racy as there is no guarantee that
there will not be more mbufs when
.Fn drbr_dequeue
is actually called.
Provided the tx queue lock is held there will not be less.
.Sh RETURN VALUES
The
.Fn drbr_enqueue
function returns
.Er ENOBUFS
if there are no slots available in the buf_ring and
.Dv 0
on success.
.Pp
The
.Fn drbr_dequeue
and
.Fn drbr_dequeue_cond
functions return an mbuf on success and
.Dv NULL
if the buf_ring is empty.
.Sh HISTORY
These functions were introduced in
.Fx 8.0 .