freebsd-nq/sys/netpfil/ipfw/dn_sched_prio.c

241 lines
6.4 KiB
C
Raw Normal View History

/*-
* SPDX-License-Identifier: BSD-2-Clause-FreeBSD
*
* Copyright (c) 2010 Riccardo Panicucci, Universita` di Pisa
* 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.
*/
/*
* $FreeBSD$
*/
#ifdef _KERNEL
#include <sys/malloc.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/mbuf.h>
#include <sys/module.h>
#include <sys/rwlock.h>
#include <net/if.h> /* IFNAMSIZ */
#include <netinet/in.h>
#include <netinet/ip_var.h> /* ipfw_rule_ref */
#include <netinet/ip_fw.h> /* flow_id */
#include <netinet/ip_dummynet.h>
#include <netpfil/ipfw/ip_fw_private.h>
#include <netpfil/ipfw/dn_heap.h>
#include <netpfil/ipfw/ip_dn_private.h>
Import Dummynet AQM version 0.2.1 (CoDel, FQ-CoDel, PIE and FQ-PIE). Centre for Advanced Internet Architectures Implementing AQM in FreeBSD * Overview <http://caia.swin.edu.au/freebsd/aqm/index.html> * Articles, Papers and Presentations <http://caia.swin.edu.au/freebsd/aqm/papers.html> * Patches and Tools <http://caia.swin.edu.au/freebsd/aqm/downloads.html> Overview Recent years have seen a resurgence of interest in better managing the depth of bottleneck queues in routers, switches and other places that get congested. Solutions include transport protocol enhancements at the end-hosts (such as delay-based or hybrid congestion control schemes) and active queue management (AQM) schemes applied within bottleneck queues. The notion of AQM has been around since at least the late 1990s (e.g. RFC 2309). In recent years the proliferation of oversized buffers in all sorts of network devices (aka bufferbloat) has stimulated keen community interest in four new AQM schemes -- CoDel, FQ-CoDel, PIE and FQ-PIE. The IETF AQM working group is looking to document these schemes, and independent implementations are a corner-stone of the IETF's process for confirming the clarity of publicly available protocol descriptions. While significant development work on all three schemes has occured in the Linux kernel, there is very little in FreeBSD. Project Goals This project began in late 2015, and aims to design and implement functionally-correct versions of CoDel, FQ-CoDel, PIE and FQ_PIE in FreeBSD (with code BSD-licensed as much as practical). We have chosen to do this as extensions to FreeBSD's ipfw/dummynet firewall and traffic shaper. Implementation of these AQM schemes in FreeBSD will: * Demonstrate whether the publicly available documentation is sufficient to enable independent, functionally equivalent implementations * Provide a broader suite of AQM options for sections the networking community that rely on FreeBSD platforms Program Members: * Rasool Al Saadi (developer) * Grenville Armitage (project lead) Acknowledgements: This project has been made possible in part by a gift from the Comcast Innovation Fund. Submitted by: Rasool Al-Saadi <ralsaadi@swin.edu.au> X-No objection: core MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6388
2016-05-26 21:40:13 +00:00
#ifdef NEW_AQM
#include <netpfil/ipfw/dn_aqm.h>
#endif
#include <netpfil/ipfw/dn_sched.h>
#else
#include <dn_test.h>
#endif
#define DN_SCHED_PRIO 5 //XXX
#if !defined(_KERNEL) || !defined(__linux__)
#define test_bit(ix, pData) ((*pData) & (1<<(ix)))
#define __set_bit(ix, pData) (*pData) |= (1<<(ix))
#define __clear_bit(ix, pData) (*pData) &= ~(1<<(ix))
#endif
#ifdef __MIPSEL__
#define __clear_bit(ix, pData) (*pData) &= ~(1<<(ix))
#endif
/* Size of the array of queues pointers. */
#define BITMAP_T unsigned long
#define MAXPRIO (sizeof(BITMAP_T) * 8)
/*
* The scheduler instance contains an array of pointers to queues,
* one for each priority, and a bitmap listing backlogged queues.
*/
struct prio_si {
BITMAP_T bitmap; /* array bitmap */
struct dn_queue *q_array[MAXPRIO]; /* Array of queues pointers */
};
/*
* If a queue with the same priority is already backlogged, use
* that one instead of the queue passed as argument.
*/
static int
prio_enqueue(struct dn_sch_inst *_si, struct dn_queue *q, struct mbuf *m)
{
struct prio_si *si = (struct prio_si *)(_si + 1);
int prio = q->fs->fs.par[0];
if (test_bit(prio, &si->bitmap) == 0) {
/* No queue with this priority, insert */
__set_bit(prio, &si->bitmap);
si->q_array[prio] = q;
} else { /* use the existing queue */
q = si->q_array[prio];
}
if (dn_enqueue(q, m, 0))
return 1;
return 0;
}
/*
* Packets are dequeued only from the highest priority queue.
* The function ffs() return the lowest bit in the bitmap that rapresent
* the array index (-1) which contains the pointer to the highest priority
* queue.
* After the dequeue, if this queue become empty, it is index is removed
* from the bitmap.
* Scheduler is idle if the bitmap is empty
*
* NOTE: highest priority is 0, lowest is sched->max_prio_q
*/
static struct mbuf *
prio_dequeue(struct dn_sch_inst *_si)
{
struct prio_si *si = (struct prio_si *)(_si + 1);
struct mbuf *m;
struct dn_queue *q;
int prio;
if (si->bitmap == 0) /* scheduler idle */
return NULL;
prio = ffs(si->bitmap) - 1;
/* Take the highest priority queue in the scheduler */
q = si->q_array[prio];
// assert(q)
m = dn_dequeue(q);
if (q->mq.head == NULL) {
/* Queue is now empty, remove from scheduler
* and mark it
*/
si->q_array[prio] = NULL;
__clear_bit(prio, &si->bitmap);
}
return m;
}
static int
prio_new_sched(struct dn_sch_inst *_si)
{
struct prio_si *si = (struct prio_si *)(_si + 1);
bzero(si->q_array, sizeof(si->q_array));
si->bitmap = 0;
return 0;
}
static int
prio_new_fsk(struct dn_fsk *fs)
{
/* Check if the prioritiy is between 0 and MAXPRIO-1 */
ipdn_bound_var(&fs->fs.par[0], 0, 0, MAXPRIO - 1, "PRIO priority");
return 0;
}
static int
prio_new_queue(struct dn_queue *q)
{
struct prio_si *si = (struct prio_si *)(q->_si + 1);
int prio = q->fs->fs.par[0];
struct dn_queue *oldq;
q->ni.oid.subtype = DN_SCHED_PRIO;
if (q->mq.head == NULL)
return 0;
/* Queue already full, must insert in the scheduler or append
* mbufs to existing queue. This partly duplicates prio_enqueue
*/
if (test_bit(prio, &si->bitmap) == 0) {
/* No queue with this priority, insert */
__set_bit(prio, &si->bitmap);
si->q_array[prio] = q;
} else if ( (oldq = si->q_array[prio]) != q) {
/* must append to the existing queue.
* can simply append q->mq.head to q2->...
* and add the counters to those of q2
*/
oldq->mq.tail->m_nextpkt = q->mq.head;
oldq->mq.tail = q->mq.tail;
oldq->ni.length += q->ni.length;
q->ni.length = 0;
oldq->ni.len_bytes += q->ni.len_bytes;
q->ni.len_bytes = 0;
q->mq.tail = q->mq.head = NULL;
}
return 0;
}
static int
prio_free_queue(struct dn_queue *q)
{
int prio = q->fs->fs.par[0];
struct prio_si *si = (struct prio_si *)(q->_si + 1);
if (si->q_array[prio] == q) {
si->q_array[prio] = NULL;
__clear_bit(prio, &si->bitmap);
}
return 0;
}
static struct dn_alg prio_desc = {
_SI( .type = ) DN_SCHED_PRIO,
_SI( .name = ) "PRIO",
_SI( .flags = ) DN_MULTIQUEUE,
/* we need extra space in the si and the queue */
_SI( .schk_datalen = ) 0,
_SI( .si_datalen = ) sizeof(struct prio_si),
_SI( .q_datalen = ) 0,
_SI( .enqueue = ) prio_enqueue,
_SI( .dequeue = ) prio_dequeue,
_SI( .config = ) NULL,
_SI( .destroy = ) NULL,
_SI( .new_sched = ) prio_new_sched,
_SI( .free_sched = ) NULL,
_SI( .new_fsk = ) prio_new_fsk,
_SI( .free_fsk = ) NULL,
_SI( .new_queue = ) prio_new_queue,
_SI( .free_queue = ) prio_free_queue,
Import Dummynet AQM version 0.2.1 (CoDel, FQ-CoDel, PIE and FQ-PIE). Centre for Advanced Internet Architectures Implementing AQM in FreeBSD * Overview <http://caia.swin.edu.au/freebsd/aqm/index.html> * Articles, Papers and Presentations <http://caia.swin.edu.au/freebsd/aqm/papers.html> * Patches and Tools <http://caia.swin.edu.au/freebsd/aqm/downloads.html> Overview Recent years have seen a resurgence of interest in better managing the depth of bottleneck queues in routers, switches and other places that get congested. Solutions include transport protocol enhancements at the end-hosts (such as delay-based or hybrid congestion control schemes) and active queue management (AQM) schemes applied within bottleneck queues. The notion of AQM has been around since at least the late 1990s (e.g. RFC 2309). In recent years the proliferation of oversized buffers in all sorts of network devices (aka bufferbloat) has stimulated keen community interest in four new AQM schemes -- CoDel, FQ-CoDel, PIE and FQ-PIE. The IETF AQM working group is looking to document these schemes, and independent implementations are a corner-stone of the IETF's process for confirming the clarity of publicly available protocol descriptions. While significant development work on all three schemes has occured in the Linux kernel, there is very little in FreeBSD. Project Goals This project began in late 2015, and aims to design and implement functionally-correct versions of CoDel, FQ-CoDel, PIE and FQ_PIE in FreeBSD (with code BSD-licensed as much as practical). We have chosen to do this as extensions to FreeBSD's ipfw/dummynet firewall and traffic shaper. Implementation of these AQM schemes in FreeBSD will: * Demonstrate whether the publicly available documentation is sufficient to enable independent, functionally equivalent implementations * Provide a broader suite of AQM options for sections the networking community that rely on FreeBSD platforms Program Members: * Rasool Al Saadi (developer) * Grenville Armitage (project lead) Acknowledgements: This project has been made possible in part by a gift from the Comcast Innovation Fund. Submitted by: Rasool Al-Saadi <ralsaadi@swin.edu.au> X-No objection: core MFC after: 2 weeks Differential Revision: https://reviews.freebsd.org/D6388
2016-05-26 21:40:13 +00:00
#ifdef NEW_AQM
_SI( .getconfig = ) NULL,
#endif
};
DECLARE_DNSCHED_MODULE(dn_prio, &prio_desc);