2012-01-30 21:02:25 +00:00
|
|
|
.\" 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 18:28:27 +00:00
|
|
|
.Dd September 27, 2012
|
2012-01-30 21:02:25 +00:00
|
|
|
.Dt BUF_RING 9
|
|
|
|
.Os
|
|
|
|
.Sh NAME
|
|
|
|
.Nm buf_ring ,
|
|
|
|
.Nm buf_ring_alloc ,
|
|
|
|
.Nm buf_ring_free ,
|
|
|
|
.Nm buf_ring_enqueue ,
|
|
|
|
.Nm buf_ring_dequeue_mc ,
|
|
|
|
.Nm buf_ring_dequeue_sc ,
|
|
|
|
.Nm buf_ring_count ,
|
|
|
|
.Nm buf_ring_empty ,
|
|
|
|
.Nm buf_ring_full ,
|
|
|
|
.Nm buf_ring_peek ,
|
|
|
|
.Nd multi-producer, {single, multi}-consumer lock-less ring buffer
|
|
|
|
.Sh SYNOPSIS
|
|
|
|
.In sys/param.h
|
|
|
|
.In sys/buf_ring.h
|
|
|
|
.Ft struct buf_ring *
|
|
|
|
.Fn buf_ring_alloc "int count" "struct malloc_type *type" "int flags" "struct mtx *sc_lock"
|
|
|
|
.Ft void
|
|
|
|
.Fn buf_ring_free "struct buf_ring *br" "struct malloc_type *type"
|
|
|
|
.Ft int
|
|
|
|
.Fn buf_ring_enqueue "struct buf_ring *br" "void *buf"
|
|
|
|
.Ft void *
|
|
|
|
.Fn buf_ring_dequeue_mc "struct buf_ring *br"
|
|
|
|
.Ft void *
|
|
|
|
.Fn buf_ring_dequeue_sc "struct buf_ring *br"
|
|
|
|
.Ft int
|
|
|
|
.Fn buf_ring_count "struct buf_ring *br"
|
|
|
|
.Ft int
|
|
|
|
.Fn buf_ring_empty "struct buf_ring *br"
|
|
|
|
.Ft int
|
|
|
|
.Fn buf_ring_full "struct buf_ring *br"
|
|
|
|
.Ft void *
|
|
|
|
.Fn buf_ring_peek "struct buf_ring *br"
|
|
|
|
.Sh DESCRIPTION
|
|
|
|
The
|
|
|
|
.Nm
|
|
|
|
functions provide a lock-less multi-producer and lock-less multi-consumer as
|
2012-03-29 05:02:12 +00:00
|
|
|
well as single-consumer ring buffer.
|
2012-01-30 21:02:25 +00:00
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_alloc
|
|
|
|
function is used to allocate a buf_ring ring buffer with
|
|
|
|
.Fa count
|
|
|
|
slots using malloc_type
|
|
|
|
.Fa type
|
|
|
|
and memory flags
|
|
|
|
.Fa flags .
|
|
|
|
The single consumer interface is protected by
|
|
|
|
.Fa sc_lock .
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_free
|
|
|
|
function is used to free a buf_ring.
|
|
|
|
The user is responsible for freeing any enqueued items.
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_enqueue
|
|
|
|
function is used to enqueue a buffer to a buf_ring.
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_dequeue_mc
|
|
|
|
function is a multi-consumer safe way of dequeueing elements from a buf_ring.
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_dequeue_sc
|
|
|
|
function is a single-consumer interface to dequeue elements - requiring
|
|
|
|
the user to serialize accesses with a lock.
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_count
|
|
|
|
function returns the number of elements in a buf_ring.
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_empty
|
|
|
|
function returns
|
|
|
|
.Dv TRUE
|
|
|
|
if the buf_ring is empty,
|
|
|
|
.Dv FALSE
|
|
|
|
otherwise.
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_full
|
|
|
|
function returns
|
|
|
|
.Dv TRUE
|
|
|
|
if no more items can be enqueued,
|
|
|
|
.Dv FALSE
|
|
|
|
otherwise.
|
|
|
|
.Pp
|
|
|
|
The
|
|
|
|
.Fn buf_ring_peek
|
|
|
|
function returns a pointer to the last element in the buf_ring if the
|
|
|
|
buf_ring is not empty,
|
|
|
|
.Dv NULL
|
|
|
|
otherwise.
|
|
|
|
.Sh RETURN VALUES
|
|
|
|
The
|
|
|
|
.Fn buf_ring_enqueue
|
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 18:28:27 +00:00
|
|
|
function return
|
2012-01-30 21:02:25 +00:00
|
|
|
.Er ENOBUFS
|
|
|
|
if there are no available slots in the buf_ring.
|
|
|
|
.Sh HISTORY
|
|
|
|
These functions were introduced in
|
|
|
|
.Fx 8.0 .
|