2005-04-17 05:20:36 +07:00
|
|
|
/*
|
|
|
|
* net/sched/sch_cbq.c Class-Based Queueing discipline.
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version
|
|
|
|
* 2 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/module.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 15:04:11 +07:00
|
|
|
#include <linux/slab.h>
|
2005-04-17 05:20:36 +07:00
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/string.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/skbuff.h>
|
2007-07-03 12:49:07 +07:00
|
|
|
#include <net/netlink.h>
|
2005-04-17 05:20:36 +07:00
|
|
|
#include <net/pkt_sched.h>
|
2017-02-09 20:38:56 +07:00
|
|
|
#include <net/pkt_cls.h>
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
|
|
|
|
/* Class-Based Queueing (CBQ) algorithm.
|
|
|
|
=======================================
|
|
|
|
|
|
|
|
Sources: [1] Sally Floyd and Van Jacobson, "Link-sharing and Resource
|
2007-02-09 21:25:16 +07:00
|
|
|
Management Models for Packet Networks",
|
2005-04-17 05:20:36 +07:00
|
|
|
IEEE/ACM Transactions on Networking, Vol.3, No.4, 1995
|
|
|
|
|
2007-02-09 21:25:16 +07:00
|
|
|
[2] Sally Floyd, "Notes on CBQ and Guaranteed Service", 1995
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2007-02-09 21:25:16 +07:00
|
|
|
[3] Sally Floyd, "Notes on Class-Based Queueing: Setting
|
2005-04-17 05:20:36 +07:00
|
|
|
Parameters", 1996
|
|
|
|
|
|
|
|
[4] Sally Floyd and Michael Speer, "Experimental Results
|
|
|
|
for Class-Based Queueing", 1998, not published.
|
|
|
|
|
|
|
|
-----------------------------------------------------------------------
|
|
|
|
|
|
|
|
Algorithm skeleton was taken from NS simulator cbq.cc.
|
|
|
|
If someone wants to check this code against the LBL version,
|
|
|
|
he should take into account that ONLY the skeleton was borrowed,
|
|
|
|
the implementation is different. Particularly:
|
|
|
|
|
|
|
|
--- The WRR algorithm is different. Our version looks more
|
2007-02-09 21:25:16 +07:00
|
|
|
reasonable (I hope) and works when quanta are allowed to be
|
|
|
|
less than MTU, which is always the case when real time classes
|
|
|
|
have small rates. Note, that the statement of [3] is
|
|
|
|
incomplete, delay may actually be estimated even if class
|
|
|
|
per-round allotment is less than MTU. Namely, if per-round
|
|
|
|
allotment is W*r_i, and r_1+...+r_k = r < 1
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
delay_i <= ([MTU/(W*r_i)]*W*r + W*r + k*MTU)/B
|
|
|
|
|
|
|
|
In the worst case we have IntServ estimate with D = W*r+k*MTU
|
|
|
|
and C = MTU*r. The proof (if correct at all) is trivial.
|
|
|
|
|
|
|
|
|
|
|
|
--- It seems that cbq-2.0 is not very accurate. At least, I cannot
|
|
|
|
interpret some places, which look like wrong translations
|
|
|
|
from NS. Anyone is advised to find these differences
|
|
|
|
and explain to me, why I am wrong 8).
|
|
|
|
|
|
|
|
--- Linux has no EOI event, so that we cannot estimate true class
|
|
|
|
idle time. Workaround is to consider the next dequeue event
|
|
|
|
as sign that previous packet is finished. This is wrong because of
|
|
|
|
internal device queueing, but on a permanently loaded link it is true.
|
|
|
|
Moreover, combined with clock integrator, this scheme looks
|
|
|
|
very close to an ideal solution. */
|
|
|
|
|
|
|
|
struct cbq_sched_data;
|
|
|
|
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class {
|
2008-07-06 13:22:05 +07:00
|
|
|
struct Qdisc_class_common common;
|
2005-04-17 05:20:36 +07:00
|
|
|
struct cbq_class *next_alive; /* next class with backlog in this priority band */
|
|
|
|
|
|
|
|
/* Parameters */
|
|
|
|
unsigned char priority; /* class priority */
|
|
|
|
unsigned char priority2; /* priority to be used after overlimit */
|
|
|
|
unsigned char ewma_log; /* time constant for idle time calculation */
|
|
|
|
|
|
|
|
u32 defmap;
|
|
|
|
|
|
|
|
/* Link-sharing scheduler parameters */
|
|
|
|
long maxidle; /* Class parameters: see below. */
|
|
|
|
long offtime;
|
|
|
|
long minidle;
|
|
|
|
u32 avpkt;
|
|
|
|
struct qdisc_rate_table *R_tab;
|
|
|
|
|
|
|
|
/* General scheduler (WRR) parameters */
|
|
|
|
long allot;
|
|
|
|
long quantum; /* Allotment per WRR round */
|
|
|
|
long weight; /* Relative allotment: see below */
|
|
|
|
|
|
|
|
struct Qdisc *qdisc; /* Ptr to CBQ discipline */
|
|
|
|
struct cbq_class *split; /* Ptr to split node */
|
|
|
|
struct cbq_class *share; /* Ptr to LS parent in the class tree */
|
|
|
|
struct cbq_class *tparent; /* Ptr to tree parent in the class tree */
|
|
|
|
struct cbq_class *borrow; /* NULL if class is bandwidth limited;
|
|
|
|
parent otherwise */
|
|
|
|
struct cbq_class *sibling; /* Sibling chain */
|
|
|
|
struct cbq_class *children; /* Pointer to children chain */
|
|
|
|
|
|
|
|
struct Qdisc *q; /* Elementary queueing discipline */
|
|
|
|
|
|
|
|
|
|
|
|
/* Variables */
|
|
|
|
unsigned char cpriority; /* Effective priority */
|
|
|
|
unsigned char delayed;
|
|
|
|
unsigned char level; /* level of the class in hierarchy:
|
|
|
|
0 for leaf classes, and maximal
|
|
|
|
level of children + 1 for nodes.
|
|
|
|
*/
|
|
|
|
|
|
|
|
psched_time_t last; /* Last end of service */
|
|
|
|
psched_time_t undertime;
|
|
|
|
long avgidle;
|
|
|
|
long deficit; /* Saved deficit for WRR */
|
2007-03-16 15:22:20 +07:00
|
|
|
psched_time_t penalized;
|
2009-08-16 16:36:49 +07:00
|
|
|
struct gnet_stats_basic_packed bstats;
|
2005-04-17 05:20:36 +07:00
|
|
|
struct gnet_stats_queue qstats;
|
2016-12-05 00:48:16 +07:00
|
|
|
struct net_rate_estimator __rcu *rate_est;
|
2005-04-17 05:20:36 +07:00
|
|
|
struct tc_cbq_xstats xstats;
|
|
|
|
|
2014-09-13 10:05:27 +07:00
|
|
|
struct tcf_proto __rcu *filter_list;
|
2017-05-17 16:07:55 +07:00
|
|
|
struct tcf_block *block;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
int filters;
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *defaults[TC_PRIO_MAX + 1];
|
2005-04-17 05:20:36 +07:00
|
|
|
};
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_sched_data {
|
2008-07-06 13:22:05 +07:00
|
|
|
struct Qdisc_class_hash clhash; /* Hash table of all classes */
|
2011-01-20 02:26:56 +07:00
|
|
|
int nclasses[TC_CBQ_MAXPRIO + 1];
|
|
|
|
unsigned int quanta[TC_CBQ_MAXPRIO + 1];
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
struct cbq_class link;
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int activemask;
|
|
|
|
struct cbq_class *active[TC_CBQ_MAXPRIO + 1]; /* List of all classes
|
2005-04-17 05:20:36 +07:00
|
|
|
with backlog */
|
|
|
|
|
2007-07-15 14:03:05 +07:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
2005-04-17 05:20:36 +07:00
|
|
|
struct cbq_class *rx_class;
|
|
|
|
#endif
|
|
|
|
struct cbq_class *tx_class;
|
|
|
|
struct cbq_class *tx_borrowed;
|
|
|
|
int tx_len;
|
|
|
|
psched_time_t now; /* Cached timestamp */
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int pmask;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2009-09-02 07:59:25 +07:00
|
|
|
struct hrtimer delay_timer;
|
2007-03-16 15:21:11 +07:00
|
|
|
struct qdisc_watchdog watchdog; /* Watchdog timer,
|
2005-04-17 05:20:36 +07:00
|
|
|
started when CBQ has
|
|
|
|
backlog, but cannot
|
|
|
|
transmit just now */
|
2007-03-16 15:21:11 +07:00
|
|
|
psched_tdiff_t wd_expires;
|
2005-04-17 05:20:36 +07:00
|
|
|
int toplevel;
|
|
|
|
u32 hgenerator;
|
|
|
|
};
|
|
|
|
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
#define L2T(cl, len) qdisc_l2t((cl)->R_tab, len)
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static inline struct cbq_class *
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_class_lookup(struct cbq_sched_data *q, u32 classid)
|
|
|
|
{
|
2008-07-06 13:22:05 +07:00
|
|
|
struct Qdisc_class_common *clc;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
clc = qdisc_class_find(&q->clhash, classid);
|
|
|
|
if (clc == NULL)
|
|
|
|
return NULL;
|
|
|
|
return container_of(clc, struct cbq_class, common);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2007-07-15 14:03:05 +07:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
static struct cbq_class *
|
|
|
|
cbq_reclassify(struct sk_buff *skb, struct cbq_class *this)
|
|
|
|
{
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
for (cl = this->tparent; cl; cl = cl->tparent) {
|
|
|
|
struct cbq_class *new = cl->defaults[TC_PRIO_BESTEFFORT];
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
if (new != NULL && new != this)
|
|
|
|
return new;
|
|
|
|
}
|
2005-04-17 05:20:36 +07:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/* Classify packet. The procedure is pretty complicated, but
|
2011-01-20 02:26:56 +07:00
|
|
|
* it allows us to combine link sharing and priority scheduling
|
|
|
|
* transparently.
|
|
|
|
*
|
|
|
|
* Namely, you can put link sharing rules (f.e. route based) at root of CBQ,
|
|
|
|
* so that it resolves to split nodes. Then packets are classified
|
|
|
|
* by logical priority, or a more specific classifier may be attached
|
|
|
|
* to the split node.
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
|
|
|
|
static struct cbq_class *
|
|
|
|
cbq_classify(struct sk_buff *skb, struct Qdisc *sch, int *qerr)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
struct cbq_class *head = &q->link;
|
|
|
|
struct cbq_class **defmap;
|
|
|
|
struct cbq_class *cl = NULL;
|
|
|
|
u32 prio = skb->priority;
|
2014-09-13 10:05:27 +07:00
|
|
|
struct tcf_proto *fl;
|
2005-04-17 05:20:36 +07:00
|
|
|
struct tcf_result res;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Step 1. If skb->priority points to one of our classes, use it.
|
|
|
|
*/
|
2011-01-20 02:26:56 +07:00
|
|
|
if (TC_H_MAJ(prio ^ sch->handle) == 0 &&
|
2005-04-17 05:20:36 +07:00
|
|
|
(cl = cbq_class_lookup(q, prio)) != NULL)
|
|
|
|
return cl;
|
|
|
|
|
2008-08-05 12:39:11 +07:00
|
|
|
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS;
|
2005-04-17 05:20:36 +07:00
|
|
|
for (;;) {
|
|
|
|
int result = 0;
|
|
|
|
defmap = head->defaults;
|
|
|
|
|
2014-09-13 10:05:27 +07:00
|
|
|
fl = rcu_dereference_bh(head->filter_list);
|
2005-04-17 05:20:36 +07:00
|
|
|
/*
|
|
|
|
* Step 2+n. Apply classifier.
|
|
|
|
*/
|
2017-05-17 16:07:54 +07:00
|
|
|
result = tcf_classify(skb, fl, &res, true);
|
2014-09-13 10:05:27 +07:00
|
|
|
if (!fl || result < 0)
|
2005-04-17 05:20:36 +07:00
|
|
|
goto fallback;
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
cl = (void *)res.class;
|
|
|
|
if (!cl) {
|
2005-04-17 05:20:36 +07:00
|
|
|
if (TC_H_MAJ(res.classid))
|
|
|
|
cl = cbq_class_lookup(q, res.classid);
|
2011-01-20 02:26:56 +07:00
|
|
|
else if ((cl = defmap[res.classid & TC_PRIO_MAX]) == NULL)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl = defmap[TC_PRIO_BESTEFFORT];
|
|
|
|
|
2012-09-11 20:11:12 +07:00
|
|
|
if (cl == NULL)
|
2005-04-17 05:20:36 +07:00
|
|
|
goto fallback;
|
|
|
|
}
|
2012-09-11 20:11:12 +07:00
|
|
|
if (cl->level >= head->level)
|
|
|
|
goto fallback;
|
2005-04-17 05:20:36 +07:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
|
|
|
switch (result) {
|
|
|
|
case TC_ACT_QUEUED:
|
2007-02-09 21:25:16 +07:00
|
|
|
case TC_ACT_STOLEN:
|
2017-06-06 19:12:02 +07:00
|
|
|
case TC_ACT_TRAP:
|
2008-08-05 12:31:03 +07:00
|
|
|
*qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN;
|
2017-10-20 04:28:24 +07:00
|
|
|
/* fall through */
|
2005-04-17 05:20:36 +07:00
|
|
|
case TC_ACT_SHOT:
|
|
|
|
return NULL;
|
2007-07-15 14:02:31 +07:00
|
|
|
case TC_ACT_RECLASSIFY:
|
|
|
|
return cbq_reclassify(skb, cl);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
if (cl->level == 0)
|
|
|
|
return cl;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Step 3+n. If classifier selected a link sharing class,
|
|
|
|
* apply agency specific classifier.
|
|
|
|
* Repeat this procdure until we hit a leaf node.
|
|
|
|
*/
|
|
|
|
head = cl;
|
|
|
|
}
|
|
|
|
|
|
|
|
fallback:
|
|
|
|
cl = head;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Step 4. No success...
|
|
|
|
*/
|
|
|
|
if (TC_H_MAJ(prio) == 0 &&
|
2011-01-20 02:26:56 +07:00
|
|
|
!(cl = head->defaults[prio & TC_PRIO_MAX]) &&
|
2005-04-17 05:20:36 +07:00
|
|
|
!(cl = head->defaults[TC_PRIO_BESTEFFORT]))
|
|
|
|
return head;
|
|
|
|
|
|
|
|
return cl;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2011-01-20 02:26:56 +07:00
|
|
|
* A packet has just been enqueued on the empty class.
|
|
|
|
* cbq_activate_class adds it to the tail of active class list
|
|
|
|
* of its priority band.
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static inline void cbq_activate_class(struct cbq_class *cl)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(cl->qdisc);
|
|
|
|
int prio = cl->cpriority;
|
|
|
|
struct cbq_class *cl_tail;
|
|
|
|
|
|
|
|
cl_tail = q->active[prio];
|
|
|
|
q->active[prio] = cl;
|
|
|
|
|
|
|
|
if (cl_tail != NULL) {
|
|
|
|
cl->next_alive = cl_tail->next_alive;
|
|
|
|
cl_tail->next_alive = cl;
|
|
|
|
} else {
|
|
|
|
cl->next_alive = cl;
|
|
|
|
q->activemask |= (1<<prio);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2011-01-20 02:26:56 +07:00
|
|
|
* Unlink class from active chain.
|
|
|
|
* Note that this same procedure is done directly in cbq_dequeue*
|
|
|
|
* during round-robin procedure.
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
|
|
|
|
static void cbq_deactivate_class(struct cbq_class *this)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(this->qdisc);
|
|
|
|
int prio = this->cpriority;
|
|
|
|
struct cbq_class *cl;
|
|
|
|
struct cbq_class *cl_prev = q->active[prio];
|
|
|
|
|
|
|
|
do {
|
|
|
|
cl = cl_prev->next_alive;
|
|
|
|
if (cl == this) {
|
|
|
|
cl_prev->next_alive = cl->next_alive;
|
|
|
|
cl->next_alive = NULL;
|
|
|
|
|
|
|
|
if (cl == q->active[prio]) {
|
|
|
|
q->active[prio] = cl_prev;
|
|
|
|
if (cl == q->active[prio]) {
|
|
|
|
q->active[prio] = NULL;
|
|
|
|
q->activemask &= ~(1<<prio);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
} while ((cl_prev = cl) != q->active[prio]);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
cbq_mark_toplevel(struct cbq_sched_data *q, struct cbq_class *cl)
|
|
|
|
{
|
|
|
|
int toplevel = q->toplevel;
|
|
|
|
|
2016-06-11 06:41:37 +07:00
|
|
|
if (toplevel > cl->level) {
|
2014-08-14 15:27:59 +07:00
|
|
|
psched_time_t now = psched_get_time();
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
do {
|
2007-03-24 01:28:07 +07:00
|
|
|
if (cl->undertime < now) {
|
2005-04-17 05:20:36 +07:00
|
|
|
q->toplevel = cl->level;
|
|
|
|
return;
|
|
|
|
}
|
2011-01-20 02:26:56 +07:00
|
|
|
} while ((cl = cl->borrow) != NULL && toplevel > cl->level);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2016-06-22 13:16:49 +07:00
|
|
|
cbq_enqueue(struct sk_buff *skb, struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2007-09-17 04:54:05 +07:00
|
|
|
int uninitialized_var(ret);
|
2005-04-17 05:20:36 +07:00
|
|
|
struct cbq_class *cl = cbq_classify(skb, sch, &ret);
|
|
|
|
|
2007-07-15 14:03:05 +07:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
2005-04-17 05:20:36 +07:00
|
|
|
q->rx_class = cl;
|
|
|
|
#endif
|
|
|
|
if (cl == NULL) {
|
2008-08-05 12:39:11 +07:00
|
|
|
if (ret & __NET_XMIT_BYPASS)
|
2014-09-29 01:53:29 +07:00
|
|
|
qdisc_qstats_drop(sch);
|
2016-06-22 13:16:49 +07:00
|
|
|
__qdisc_drop(skb, to_free);
|
2005-04-17 05:20:36 +07:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-06-22 13:16:49 +07:00
|
|
|
ret = qdisc_enqueue(skb, cl->q, to_free);
|
2008-07-20 14:08:04 +07:00
|
|
|
if (ret == NET_XMIT_SUCCESS) {
|
2005-04-17 05:20:36 +07:00
|
|
|
sch->q.qlen++;
|
|
|
|
cbq_mark_toplevel(q, cl);
|
|
|
|
if (!cl->next_alive)
|
|
|
|
cbq_activate_class(cl);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2008-08-05 12:31:03 +07:00
|
|
|
if (net_xmit_drop_count(ret)) {
|
2014-09-29 01:53:29 +07:00
|
|
|
qdisc_qstats_drop(sch);
|
2008-08-05 12:31:03 +07:00
|
|
|
cbq_mark_toplevel(q, cl);
|
|
|
|
cl->qstats.drops++;
|
|
|
|
}
|
2005-04-17 05:20:36 +07:00
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2016-06-09 05:27:39 +07:00
|
|
|
/* Overlimit action: penalize leaf class by adding offtime */
|
|
|
|
static void cbq_overlimit(struct cbq_class *cl)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(cl->qdisc);
|
2007-03-24 01:28:55 +07:00
|
|
|
psched_tdiff_t delay = cl->undertime - q->now;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (!cl->delayed) {
|
|
|
|
delay += cl->offtime;
|
|
|
|
|
2007-02-09 21:25:16 +07:00
|
|
|
/*
|
2011-01-20 02:26:56 +07:00
|
|
|
* Class goes to sleep, so that it will have no
|
|
|
|
* chance to work avgidle. Let's forgive it 8)
|
|
|
|
*
|
|
|
|
* BTW cbq-2.0 has a crap in this
|
|
|
|
* place, apparently they forgot to shift it by cl->ewma_log.
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
if (cl->avgidle < 0)
|
|
|
|
delay -= (-cl->avgidle) - ((-cl->avgidle) >> cl->ewma_log);
|
|
|
|
if (cl->avgidle < cl->minidle)
|
|
|
|
cl->avgidle = cl->minidle;
|
|
|
|
if (delay <= 0)
|
|
|
|
delay = 1;
|
2007-03-24 01:27:45 +07:00
|
|
|
cl->undertime = q->now + delay;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
cl->xstats.overactions++;
|
|
|
|
cl->delayed = 1;
|
|
|
|
}
|
|
|
|
if (q->wd_expires == 0 || q->wd_expires > delay)
|
|
|
|
q->wd_expires = delay;
|
|
|
|
|
|
|
|
/* Dirty work! We must schedule wakeups based on
|
2011-01-20 02:26:56 +07:00
|
|
|
* real available rate, rather than leaf rate,
|
|
|
|
* which may be tiny (even zero).
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
if (q->toplevel == TC_CBQ_MAXLEVEL) {
|
|
|
|
struct cbq_class *b;
|
|
|
|
psched_tdiff_t base_delay = q->wd_expires;
|
|
|
|
|
|
|
|
for (b = cl->borrow; b; b = b->borrow) {
|
2007-03-24 01:28:55 +07:00
|
|
|
delay = b->undertime - q->now;
|
2005-04-17 05:20:36 +07:00
|
|
|
if (delay < base_delay) {
|
|
|
|
if (delay <= 0)
|
|
|
|
delay = 1;
|
|
|
|
base_delay = delay;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
q->wd_expires = base_delay;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-03-16 15:22:20 +07:00
|
|
|
static psched_tdiff_t cbq_undelay_prio(struct cbq_sched_data *q, int prio,
|
|
|
|
psched_time_t now)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_class *cl;
|
|
|
|
struct cbq_class *cl_prev = q->active[prio];
|
2007-03-16 15:22:20 +07:00
|
|
|
psched_time_t sched = now;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (cl_prev == NULL)
|
2007-03-16 15:21:40 +07:00
|
|
|
return 0;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
do {
|
|
|
|
cl = cl_prev->next_alive;
|
2007-03-16 15:22:20 +07:00
|
|
|
if (now - cl->penalized > 0) {
|
2005-04-17 05:20:36 +07:00
|
|
|
cl_prev->next_alive = cl->next_alive;
|
|
|
|
cl->next_alive = NULL;
|
|
|
|
cl->cpriority = cl->priority;
|
|
|
|
cl->delayed = 0;
|
|
|
|
cbq_activate_class(cl);
|
|
|
|
|
|
|
|
if (cl == q->active[prio]) {
|
|
|
|
q->active[prio] = cl_prev;
|
|
|
|
if (cl == q->active[prio]) {
|
|
|
|
q->active[prio] = NULL;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
cl = cl_prev->next_alive;
|
2007-03-16 15:22:20 +07:00
|
|
|
} else if (sched - cl->penalized > 0)
|
2005-04-17 05:20:36 +07:00
|
|
|
sched = cl->penalized;
|
|
|
|
} while ((cl_prev = cl) != q->active[prio]);
|
|
|
|
|
2007-03-16 15:22:20 +07:00
|
|
|
return sched - now;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2007-03-16 15:22:20 +07:00
|
|
|
static enum hrtimer_restart cbq_undelay(struct hrtimer *timer)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
2007-03-16 15:22:20 +07:00
|
|
|
struct cbq_sched_data *q = container_of(timer, struct cbq_sched_data,
|
2009-09-02 07:59:25 +07:00
|
|
|
delay_timer);
|
2007-03-16 15:22:20 +07:00
|
|
|
struct Qdisc *sch = q->watchdog.qdisc;
|
|
|
|
psched_time_t now;
|
|
|
|
psched_tdiff_t delay = 0;
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int pmask;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2007-03-24 01:29:25 +07:00
|
|
|
now = psched_get_time();
|
2007-03-16 15:22:20 +07:00
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
pmask = q->pmask;
|
|
|
|
q->pmask = 0;
|
|
|
|
|
|
|
|
while (pmask) {
|
|
|
|
int prio = ffz(~pmask);
|
2007-03-16 15:22:20 +07:00
|
|
|
psched_tdiff_t tmp;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
pmask &= ~(1<<prio);
|
|
|
|
|
2007-03-16 15:22:20 +07:00
|
|
|
tmp = cbq_undelay_prio(q, prio, now);
|
2005-04-17 05:20:36 +07:00
|
|
|
if (tmp > 0) {
|
|
|
|
q->pmask |= 1<<prio;
|
|
|
|
if (tmp < delay || delay == 0)
|
|
|
|
delay = tmp;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (delay) {
|
2007-03-16 15:22:20 +07:00
|
|
|
ktime_t time;
|
|
|
|
|
2016-12-25 18:30:41 +07:00
|
|
|
time = 0;
|
2009-06-15 16:31:47 +07:00
|
|
|
time = ktime_add_ns(time, PSCHED_TICKS2NS(now + delay));
|
2014-09-21 08:01:30 +07:00
|
|
|
hrtimer_start(&q->delay_timer, time, HRTIMER_MODE_ABS_PINNED);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2008-08-19 10:51:18 +07:00
|
|
|
__netif_schedule(qdisc_root(sch));
|
2007-03-16 15:22:20 +07:00
|
|
|
return HRTIMER_NORESTART;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2007-02-09 21:25:16 +07:00
|
|
|
/*
|
2011-01-20 02:26:56 +07:00
|
|
|
* It is mission critical procedure.
|
|
|
|
*
|
|
|
|
* We "regenerate" toplevel cutoff, if transmitting class
|
|
|
|
* has backlog and it is not regulated. It is not part of
|
|
|
|
* original CBQ description, but looks more reasonable.
|
|
|
|
* Probably, it is wrong. This question needs further investigation.
|
|
|
|
*/
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static inline void
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_update_toplevel(struct cbq_sched_data *q, struct cbq_class *cl,
|
|
|
|
struct cbq_class *borrowed)
|
|
|
|
{
|
|
|
|
if (cl && q->toplevel >= borrowed->level) {
|
|
|
|
if (cl->q->q.qlen > 1) {
|
|
|
|
do {
|
2007-03-24 01:28:30 +07:00
|
|
|
if (borrowed->undertime == PSCHED_PASTPERFECT) {
|
2005-04-17 05:20:36 +07:00
|
|
|
q->toplevel = borrowed->level;
|
|
|
|
return;
|
|
|
|
}
|
2011-01-20 02:26:56 +07:00
|
|
|
} while ((borrowed = borrowed->borrow) != NULL);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
2007-02-09 21:25:16 +07:00
|
|
|
#if 0
|
2005-04-17 05:20:36 +07:00
|
|
|
/* It is not necessary now. Uncommenting it
|
|
|
|
will save CPU cycles, but decrease fairness.
|
|
|
|
*/
|
|
|
|
q->toplevel = TC_CBQ_MAXLEVEL;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
|
|
|
cbq_update(struct cbq_sched_data *q)
|
|
|
|
{
|
|
|
|
struct cbq_class *this = q->tx_class;
|
|
|
|
struct cbq_class *cl = this;
|
|
|
|
int len = q->tx_len;
|
2014-08-14 15:27:47 +07:00
|
|
|
psched_time_t now;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
q->tx_class = NULL;
|
2014-08-14 15:27:47 +07:00
|
|
|
/* Time integrator. We calculate EOS time
|
|
|
|
* by adding expected packet transmission time.
|
|
|
|
*/
|
|
|
|
now = q->now + L2T(&q->link, len);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
for ( ; cl; cl = cl->share) {
|
|
|
|
long avgidle = cl->avgidle;
|
|
|
|
long idle;
|
|
|
|
|
|
|
|
cl->bstats.packets++;
|
|
|
|
cl->bstats.bytes += len;
|
|
|
|
|
|
|
|
/*
|
2011-01-20 02:26:56 +07:00
|
|
|
* (now - last) is total time between packet right edges.
|
|
|
|
* (last_pktlen/rate) is "virtual" busy time, so that
|
|
|
|
*
|
|
|
|
* idle = (now - last) - last_pktlen/rate
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
|
2014-08-14 15:27:47 +07:00
|
|
|
idle = now - cl->last;
|
2005-04-17 05:20:36 +07:00
|
|
|
if ((unsigned long)idle > 128*1024*1024) {
|
|
|
|
avgidle = cl->maxidle;
|
|
|
|
} else {
|
|
|
|
idle -= L2T(cl, len);
|
|
|
|
|
|
|
|
/* true_avgidle := (1-W)*true_avgidle + W*idle,
|
2011-01-20 02:26:56 +07:00
|
|
|
* where W=2^{-ewma_log}. But cl->avgidle is scaled:
|
|
|
|
* cl->avgidle == true_avgidle/W,
|
|
|
|
* hence:
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
avgidle += idle - (avgidle>>cl->ewma_log);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (avgidle <= 0) {
|
|
|
|
/* Overlimit or at-limit */
|
|
|
|
|
|
|
|
if (avgidle < cl->minidle)
|
|
|
|
avgidle = cl->minidle;
|
|
|
|
|
|
|
|
cl->avgidle = avgidle;
|
|
|
|
|
|
|
|
/* Calculate expected time, when this class
|
2011-01-20 02:26:56 +07:00
|
|
|
* will be allowed to send.
|
|
|
|
* It will occur, when:
|
|
|
|
* (1-W)*true_avgidle + W*delay = 0, i.e.
|
|
|
|
* idle = (1/W - 1)*(-true_avgidle)
|
|
|
|
* or
|
|
|
|
* idle = (1 - W)*(-cl->avgidle);
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
idle = (-avgidle) - ((-avgidle) >> cl->ewma_log);
|
|
|
|
|
|
|
|
/*
|
2011-01-20 02:26:56 +07:00
|
|
|
* That is not all.
|
|
|
|
* To maintain the rate allocated to the class,
|
|
|
|
* we add to undertime virtual clock,
|
|
|
|
* necessary to complete transmitted packet.
|
|
|
|
* (len/phys_bandwidth has been already passed
|
|
|
|
* to the moment of cbq_update)
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
|
|
|
|
idle -= L2T(&q->link, len);
|
|
|
|
idle += L2T(cl, len);
|
|
|
|
|
2014-08-14 15:27:47 +07:00
|
|
|
cl->undertime = now + idle;
|
2005-04-17 05:20:36 +07:00
|
|
|
} else {
|
|
|
|
/* Underlimit */
|
|
|
|
|
2007-03-24 01:28:30 +07:00
|
|
|
cl->undertime = PSCHED_PASTPERFECT;
|
2005-04-17 05:20:36 +07:00
|
|
|
if (avgidle > cl->maxidle)
|
|
|
|
cl->avgidle = cl->maxidle;
|
|
|
|
else
|
|
|
|
cl->avgidle = avgidle;
|
|
|
|
}
|
2014-08-14 15:27:47 +07:00
|
|
|
if ((s64)(now - cl->last) > 0)
|
|
|
|
cl->last = now;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
cbq_update_toplevel(q, this, q->tx_borrowed);
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static inline struct cbq_class *
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_under_limit(struct cbq_class *cl)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(cl->qdisc);
|
|
|
|
struct cbq_class *this_cl = cl;
|
|
|
|
|
|
|
|
if (cl->tparent == NULL)
|
|
|
|
return cl;
|
|
|
|
|
2007-03-24 01:28:30 +07:00
|
|
|
if (cl->undertime == PSCHED_PASTPERFECT || q->now >= cl->undertime) {
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->delayed = 0;
|
|
|
|
return cl;
|
|
|
|
}
|
|
|
|
|
|
|
|
do {
|
|
|
|
/* It is very suspicious place. Now overlimit
|
2011-01-20 02:26:56 +07:00
|
|
|
* action is generated for not bounded classes
|
|
|
|
* only if link is completely congested.
|
|
|
|
* Though it is in agree with ancestor-only paradigm,
|
|
|
|
* it looks very stupid. Particularly,
|
|
|
|
* it means that this chunk of code will either
|
|
|
|
* never be called or result in strong amplification
|
|
|
|
* of burstiness. Dangerous, silly, and, however,
|
|
|
|
* no another solution exists.
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
2011-01-20 02:26:56 +07:00
|
|
|
cl = cl->borrow;
|
|
|
|
if (!cl) {
|
2005-04-17 05:20:36 +07:00
|
|
|
this_cl->qstats.overlimits++;
|
2016-06-09 05:27:39 +07:00
|
|
|
cbq_overlimit(this_cl);
|
2005-04-17 05:20:36 +07:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
if (cl->level > q->toplevel)
|
|
|
|
return NULL;
|
2007-03-24 01:28:30 +07:00
|
|
|
} while (cl->undertime != PSCHED_PASTPERFECT && q->now < cl->undertime);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
cl->delayed = 0;
|
|
|
|
return cl;
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static inline struct sk_buff *
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_dequeue_prio(struct Qdisc *sch, int prio)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
struct cbq_class *cl_tail, *cl_prev, *cl;
|
|
|
|
struct sk_buff *skb;
|
|
|
|
int deficit;
|
|
|
|
|
|
|
|
cl_tail = cl_prev = q->active[prio];
|
|
|
|
cl = cl_prev->next_alive;
|
|
|
|
|
|
|
|
do {
|
|
|
|
deficit = 0;
|
|
|
|
|
|
|
|
/* Start round */
|
|
|
|
do {
|
|
|
|
struct cbq_class *borrow = cl;
|
|
|
|
|
|
|
|
if (cl->q->q.qlen &&
|
|
|
|
(borrow = cbq_under_limit(cl)) == NULL)
|
|
|
|
goto skip_class;
|
|
|
|
|
|
|
|
if (cl->deficit <= 0) {
|
|
|
|
/* Class exhausted its allotment per
|
2011-01-20 02:26:56 +07:00
|
|
|
* this round. Switch to the next one.
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
deficit = 1;
|
|
|
|
cl->deficit += cl->quantum;
|
|
|
|
goto next_class;
|
|
|
|
}
|
|
|
|
|
|
|
|
skb = cl->q->dequeue(cl->q);
|
|
|
|
|
|
|
|
/* Class did not give us any skb :-(
|
2011-01-20 02:26:56 +07:00
|
|
|
* It could occur even if cl->q->q.qlen != 0
|
|
|
|
* f.e. if cl->q == "tbf"
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
if (skb == NULL)
|
|
|
|
goto skip_class;
|
|
|
|
|
2008-07-20 14:08:27 +07:00
|
|
|
cl->deficit -= qdisc_pkt_len(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
q->tx_class = cl;
|
|
|
|
q->tx_borrowed = borrow;
|
|
|
|
if (borrow != cl) {
|
|
|
|
#ifndef CBQ_XSTATS_BORROWS_BYTES
|
|
|
|
borrow->xstats.borrows++;
|
|
|
|
cl->xstats.borrows++;
|
|
|
|
#else
|
2008-07-20 14:08:27 +07:00
|
|
|
borrow->xstats.borrows += qdisc_pkt_len(skb);
|
|
|
|
cl->xstats.borrows += qdisc_pkt_len(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
#endif
|
|
|
|
}
|
2008-07-20 14:08:27 +07:00
|
|
|
q->tx_len = qdisc_pkt_len(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (cl->deficit <= 0) {
|
|
|
|
q->active[prio] = cl;
|
|
|
|
cl = cl->next_alive;
|
|
|
|
cl->deficit += cl->quantum;
|
|
|
|
}
|
|
|
|
return skb;
|
|
|
|
|
|
|
|
skip_class:
|
|
|
|
if (cl->q->q.qlen == 0 || prio != cl->cpriority) {
|
|
|
|
/* Class is empty or penalized.
|
2011-01-20 02:26:56 +07:00
|
|
|
* Unlink it from active chain.
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
cl_prev->next_alive = cl->next_alive;
|
|
|
|
cl->next_alive = NULL;
|
|
|
|
|
|
|
|
/* Did cl_tail point to it? */
|
|
|
|
if (cl == cl_tail) {
|
|
|
|
/* Repair it! */
|
|
|
|
cl_tail = cl_prev;
|
|
|
|
|
|
|
|
/* Was it the last class in this band? */
|
|
|
|
if (cl == cl_tail) {
|
|
|
|
/* Kill the band! */
|
|
|
|
q->active[prio] = NULL;
|
|
|
|
q->activemask &= ~(1<<prio);
|
|
|
|
if (cl->q->q.qlen)
|
|
|
|
cbq_activate_class(cl);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
q->active[prio] = cl_tail;
|
|
|
|
}
|
|
|
|
if (cl->q->q.qlen)
|
|
|
|
cbq_activate_class(cl);
|
|
|
|
|
|
|
|
cl = cl_prev;
|
|
|
|
}
|
|
|
|
|
|
|
|
next_class:
|
|
|
|
cl_prev = cl;
|
|
|
|
cl = cl->next_alive;
|
|
|
|
} while (cl_prev != cl_tail);
|
|
|
|
} while (deficit);
|
|
|
|
|
|
|
|
q->active[prio] = cl_prev;
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static inline struct sk_buff *
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_dequeue_1(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
struct sk_buff *skb;
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int activemask;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
activemask = q->activemask & 0xFF;
|
2005-04-17 05:20:36 +07:00
|
|
|
while (activemask) {
|
|
|
|
int prio = ffz(~activemask);
|
|
|
|
activemask &= ~(1<<prio);
|
|
|
|
skb = cbq_dequeue_prio(sch, prio);
|
|
|
|
if (skb)
|
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct sk_buff *
|
|
|
|
cbq_dequeue(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
struct sk_buff *skb;
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
psched_time_t now;
|
|
|
|
|
2007-03-24 01:29:25 +07:00
|
|
|
now = psched_get_time();
|
2014-08-14 15:27:47 +07:00
|
|
|
|
|
|
|
if (q->tx_class)
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_update(q);
|
2014-08-14 15:27:47 +07:00
|
|
|
|
|
|
|
q->now = now;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
for (;;) {
|
|
|
|
q->wd_expires = 0;
|
|
|
|
|
|
|
|
skb = cbq_dequeue_1(sch);
|
|
|
|
if (skb) {
|
2011-01-21 14:31:33 +07:00
|
|
|
qdisc_bstats_update(sch, skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
sch->q.qlen--;
|
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* All the classes are overlimit.
|
2011-01-20 02:26:56 +07:00
|
|
|
*
|
|
|
|
* It is possible, if:
|
|
|
|
*
|
|
|
|
* 1. Scheduler is empty.
|
|
|
|
* 2. Toplevel cutoff inhibited borrowing.
|
|
|
|
* 3. Root class is overlimit.
|
|
|
|
*
|
|
|
|
* Reset 2d and 3d conditions and retry.
|
|
|
|
*
|
|
|
|
* Note, that NS and cbq-2.0 are buggy, peeking
|
|
|
|
* an arbitrary class is appropriate for ancestor-only
|
|
|
|
* sharing, but not for toplevel algorithm.
|
|
|
|
*
|
|
|
|
* Our version is better, but slower, because it requires
|
|
|
|
* two passes, but it is unavoidable with top-level sharing.
|
|
|
|
*/
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (q->toplevel == TC_CBQ_MAXLEVEL &&
|
2007-03-24 01:28:30 +07:00
|
|
|
q->link.undertime == PSCHED_PASTPERFECT)
|
2005-04-17 05:20:36 +07:00
|
|
|
break;
|
|
|
|
|
|
|
|
q->toplevel = TC_CBQ_MAXLEVEL;
|
2007-03-24 01:28:30 +07:00
|
|
|
q->link.undertime = PSCHED_PASTPERFECT;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
/* No packets in scheduler or nobody wants to give them to us :-(
|
2011-01-20 02:26:56 +07:00
|
|
|
* Sigh... start watchdog timer in the last case.
|
|
|
|
*/
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (sch->q.qlen) {
|
2014-09-29 01:53:29 +07:00
|
|
|
qdisc_qstats_overlimit(sch);
|
2007-03-16 15:21:11 +07:00
|
|
|
if (q->wd_expires)
|
|
|
|
qdisc_watchdog_schedule(&q->watchdog,
|
2007-03-17 02:31:28 +07:00
|
|
|
now + q->wd_expires);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* CBQ class maintanance routines */
|
|
|
|
|
|
|
|
static void cbq_adjust_levels(struct cbq_class *this)
|
|
|
|
{
|
|
|
|
if (this == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
do {
|
|
|
|
int level = 0;
|
|
|
|
struct cbq_class *cl;
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
cl = this->children;
|
|
|
|
if (cl) {
|
2005-04-17 05:20:36 +07:00
|
|
|
do {
|
|
|
|
if (cl->level > level)
|
|
|
|
level = cl->level;
|
|
|
|
} while ((cl = cl->sibling) != this->children);
|
|
|
|
}
|
2011-01-20 02:26:56 +07:00
|
|
|
this->level = level + 1;
|
2005-04-17 05:20:36 +07:00
|
|
|
} while ((this = this->tparent) != NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_normalize_quanta(struct cbq_sched_data *q, int prio)
|
|
|
|
{
|
|
|
|
struct cbq_class *cl;
|
2008-07-06 13:22:05 +07:00
|
|
|
unsigned int h;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (q->quanta[prio] == 0)
|
|
|
|
return;
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
for (h = 0; h < q->clhash.hashsize; h++) {
|
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 08:06:00 +07:00
|
|
|
hlist_for_each_entry(cl, &q->clhash.hash[h], common.hnode) {
|
2005-04-17 05:20:36 +07:00
|
|
|
/* BUGGGG... Beware! This expression suffer of
|
2011-01-20 02:26:56 +07:00
|
|
|
* arithmetic overflows!
|
2005-04-17 05:20:36 +07:00
|
|
|
*/
|
|
|
|
if (cl->priority == prio) {
|
|
|
|
cl->quantum = (cl->weight*cl->allot*q->nclasses[prio])/
|
|
|
|
q->quanta[prio];
|
|
|
|
}
|
2013-12-10 19:55:32 +07:00
|
|
|
if (cl->quantum <= 0 ||
|
|
|
|
cl->quantum > 32*qdisc_dev(cl->qdisc)->mtu) {
|
2013-12-23 16:38:58 +07:00
|
|
|
pr_warn("CBQ: class %08x has bad quantum==%ld, repaired.\n",
|
|
|
|
cl->common.classid, cl->quantum);
|
2008-07-09 07:06:30 +07:00
|
|
|
cl->quantum = qdisc_dev(cl->qdisc)->mtu/2 + 1;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_sync_defmap(struct cbq_class *cl)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(cl->qdisc);
|
|
|
|
struct cbq_class *split = cl->split;
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int h;
|
2005-04-17 05:20:36 +07:00
|
|
|
int i;
|
|
|
|
|
|
|
|
if (split == NULL)
|
|
|
|
return;
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
for (i = 0; i <= TC_PRIO_MAX; i++) {
|
|
|
|
if (split->defaults[i] == cl && !(cl->defmap & (1<<i)))
|
2005-04-17 05:20:36 +07:00
|
|
|
split->defaults[i] = NULL;
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
for (i = 0; i <= TC_PRIO_MAX; i++) {
|
2005-04-17 05:20:36 +07:00
|
|
|
int level = split->level;
|
|
|
|
|
|
|
|
if (split->defaults[i])
|
|
|
|
continue;
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
for (h = 0; h < q->clhash.hashsize; h++) {
|
2005-04-17 05:20:36 +07:00
|
|
|
struct cbq_class *c;
|
|
|
|
|
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 08:06:00 +07:00
|
|
|
hlist_for_each_entry(c, &q->clhash.hash[h],
|
2008-07-06 13:22:05 +07:00
|
|
|
common.hnode) {
|
2005-04-17 05:20:36 +07:00
|
|
|
if (c->split == split && c->level < level &&
|
2011-01-20 02:26:56 +07:00
|
|
|
c->defmap & (1<<i)) {
|
2005-04-17 05:20:36 +07:00
|
|
|
split->defaults[i] = c;
|
|
|
|
level = c->level;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_change_defmap(struct cbq_class *cl, u32 splitid, u32 def, u32 mask)
|
|
|
|
{
|
|
|
|
struct cbq_class *split = NULL;
|
|
|
|
|
|
|
|
if (splitid == 0) {
|
2011-01-20 02:26:56 +07:00
|
|
|
split = cl->split;
|
|
|
|
if (!split)
|
2005-04-17 05:20:36 +07:00
|
|
|
return;
|
2008-07-06 13:22:05 +07:00
|
|
|
splitid = split->common.classid;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
if (split == NULL || split->common.classid != splitid) {
|
2005-04-17 05:20:36 +07:00
|
|
|
for (split = cl->tparent; split; split = split->tparent)
|
2008-07-06 13:22:05 +07:00
|
|
|
if (split->common.classid == splitid)
|
2005-04-17 05:20:36 +07:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (split == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (cl->split != split) {
|
|
|
|
cl->defmap = 0;
|
|
|
|
cbq_sync_defmap(cl);
|
|
|
|
cl->split = split;
|
2011-01-20 02:26:56 +07:00
|
|
|
cl->defmap = def & mask;
|
2005-04-17 05:20:36 +07:00
|
|
|
} else
|
2011-01-20 02:26:56 +07:00
|
|
|
cl->defmap = (cl->defmap & ~mask) | (def & mask);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
cbq_sync_defmap(cl);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_unlink_class(struct cbq_class *this)
|
|
|
|
{
|
|
|
|
struct cbq_class *cl, **clp;
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(this->qdisc);
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
qdisc_class_hash_remove(&q->clhash, &this->common);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (this->tparent) {
|
2011-01-20 02:26:56 +07:00
|
|
|
clp = &this->sibling;
|
2005-04-17 05:20:36 +07:00
|
|
|
cl = *clp;
|
|
|
|
do {
|
|
|
|
if (cl == this) {
|
|
|
|
*clp = cl->sibling;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
clp = &cl->sibling;
|
|
|
|
} while ((cl = *clp) != this->sibling);
|
|
|
|
|
|
|
|
if (this->tparent->children == this) {
|
|
|
|
this->tparent->children = this->sibling;
|
|
|
|
if (this->sibling == this)
|
|
|
|
this->tparent->children = NULL;
|
|
|
|
}
|
|
|
|
} else {
|
2008-07-26 11:43:18 +07:00
|
|
|
WARN_ON(this->sibling != this);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_link_class(struct cbq_class *this)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(this->qdisc);
|
|
|
|
struct cbq_class *parent = this->tparent;
|
|
|
|
|
|
|
|
this->sibling = this;
|
2008-07-06 13:22:05 +07:00
|
|
|
qdisc_class_hash_insert(&q->clhash, &this->common);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (parent == NULL)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (parent->children == NULL) {
|
|
|
|
parent->children = this;
|
|
|
|
} else {
|
|
|
|
this->sibling = parent->children->sibling;
|
|
|
|
parent->children->sibling = this;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2011-01-20 02:26:56 +07:00
|
|
|
cbq_reset(struct Qdisc *sch)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
struct cbq_class *cl;
|
|
|
|
int prio;
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int h;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
q->activemask = 0;
|
|
|
|
q->pmask = 0;
|
|
|
|
q->tx_class = NULL;
|
|
|
|
q->tx_borrowed = NULL;
|
2007-03-16 15:21:11 +07:00
|
|
|
qdisc_watchdog_cancel(&q->watchdog);
|
2009-09-02 07:59:25 +07:00
|
|
|
hrtimer_cancel(&q->delay_timer);
|
2005-04-17 05:20:36 +07:00
|
|
|
q->toplevel = TC_CBQ_MAXLEVEL;
|
2007-03-24 01:29:25 +07:00
|
|
|
q->now = psched_get_time();
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
for (prio = 0; prio <= TC_CBQ_MAXPRIO; prio++)
|
|
|
|
q->active[prio] = NULL;
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
for (h = 0; h < q->clhash.hashsize; h++) {
|
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 08:06:00 +07:00
|
|
|
hlist_for_each_entry(cl, &q->clhash.hash[h], common.hnode) {
|
2005-04-17 05:20:36 +07:00
|
|
|
qdisc_reset(cl->q);
|
|
|
|
|
|
|
|
cl->next_alive = NULL;
|
2007-03-24 01:28:30 +07:00
|
|
|
cl->undertime = PSCHED_PASTPERFECT;
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->avgidle = cl->maxidle;
|
|
|
|
cl->deficit = cl->quantum;
|
|
|
|
cl->cpriority = cl->priority;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
sch->q.qlen = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static int cbq_set_lss(struct cbq_class *cl, struct tc_cbq_lssopt *lss)
|
|
|
|
{
|
2011-01-20 02:26:56 +07:00
|
|
|
if (lss->change & TCF_CBQ_LSS_FLAGS) {
|
|
|
|
cl->share = (lss->flags & TCF_CBQ_LSS_ISOLATED) ? NULL : cl->tparent;
|
|
|
|
cl->borrow = (lss->flags & TCF_CBQ_LSS_BOUNDED) ? NULL : cl->tparent;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
2011-01-20 02:26:56 +07:00
|
|
|
if (lss->change & TCF_CBQ_LSS_EWMA)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->ewma_log = lss->ewma_log;
|
2011-01-20 02:26:56 +07:00
|
|
|
if (lss->change & TCF_CBQ_LSS_AVPKT)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->avpkt = lss->avpkt;
|
2011-01-20 02:26:56 +07:00
|
|
|
if (lss->change & TCF_CBQ_LSS_MINIDLE)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->minidle = -(long)lss->minidle;
|
2011-01-20 02:26:56 +07:00
|
|
|
if (lss->change & TCF_CBQ_LSS_MAXIDLE) {
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->maxidle = lss->maxidle;
|
|
|
|
cl->avgidle = lss->maxidle;
|
|
|
|
}
|
2011-01-20 02:26:56 +07:00
|
|
|
if (lss->change & TCF_CBQ_LSS_OFFTIME)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->offtime = lss->offtime;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_rmprio(struct cbq_sched_data *q, struct cbq_class *cl)
|
|
|
|
{
|
|
|
|
q->nclasses[cl->priority]--;
|
|
|
|
q->quanta[cl->priority] -= cl->weight;
|
|
|
|
cbq_normalize_quanta(q, cl->priority);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_addprio(struct cbq_sched_data *q, struct cbq_class *cl)
|
|
|
|
{
|
|
|
|
q->nclasses[cl->priority]++;
|
|
|
|
q->quanta[cl->priority] += cl->weight;
|
|
|
|
cbq_normalize_quanta(q, cl->priority);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int cbq_set_wrr(struct cbq_class *cl, struct tc_cbq_wrropt *wrr)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(cl->qdisc);
|
|
|
|
|
|
|
|
if (wrr->allot)
|
|
|
|
cl->allot = wrr->allot;
|
|
|
|
if (wrr->weight)
|
|
|
|
cl->weight = wrr->weight;
|
|
|
|
if (wrr->priority) {
|
2011-01-20 02:26:56 +07:00
|
|
|
cl->priority = wrr->priority - 1;
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->cpriority = cl->priority;
|
|
|
|
if (cl->priority >= cl->priority2)
|
2011-01-20 02:26:56 +07:00
|
|
|
cl->priority2 = TC_CBQ_MAXPRIO - 1;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
cbq_addprio(q, cl);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int cbq_set_fopt(struct cbq_class *cl, struct tc_cbq_fopt *fopt)
|
|
|
|
{
|
|
|
|
cbq_change_defmap(cl, fopt->split, fopt->defmap, fopt->defchange);
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-01-24 11:35:39 +07:00
|
|
|
static const struct nla_policy cbq_policy[TCA_CBQ_MAX + 1] = {
|
|
|
|
[TCA_CBQ_LSSOPT] = { .len = sizeof(struct tc_cbq_lssopt) },
|
|
|
|
[TCA_CBQ_WRROPT] = { .len = sizeof(struct tc_cbq_wrropt) },
|
|
|
|
[TCA_CBQ_FOPT] = { .len = sizeof(struct tc_cbq_fopt) },
|
|
|
|
[TCA_CBQ_OVL_STRATEGY] = { .len = sizeof(struct tc_cbq_ovl) },
|
|
|
|
[TCA_CBQ_RATE] = { .len = sizeof(struct tc_ratespec) },
|
|
|
|
[TCA_CBQ_RTAB] = { .type = NLA_BINARY, .len = TC_RTAB_SIZE },
|
|
|
|
[TCA_CBQ_POLICE] = { .len = sizeof(struct tc_cbq_police) },
|
|
|
|
};
|
|
|
|
|
2017-12-21 00:35:13 +07:00
|
|
|
static int cbq_init(struct Qdisc *sch, struct nlattr *opt,
|
|
|
|
struct netlink_ext_ack *extack)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2008-01-23 13:11:17 +07:00
|
|
|
struct nlattr *tb[TCA_CBQ_MAX + 1];
|
2005-04-17 05:20:36 +07:00
|
|
|
struct tc_ratespec *r;
|
2008-01-24 11:33:32 +07:00
|
|
|
int err;
|
|
|
|
|
2017-08-30 16:49:01 +07:00
|
|
|
qdisc_watchdog_init(&q->watchdog, sch);
|
|
|
|
hrtimer_init(&q->delay_timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS_PINNED);
|
|
|
|
q->delay_timer.function = cbq_undelay;
|
|
|
|
|
|
|
|
if (!opt)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2017-04-12 19:34:07 +07:00
|
|
|
err = nla_parse_nested(tb, TCA_CBQ_MAX, opt, cbq_policy, NULL);
|
2008-01-24 11:33:32 +07:00
|
|
|
if (err < 0)
|
|
|
|
return err;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2017-12-21 00:35:11 +07:00
|
|
|
if (!tb[TCA_CBQ_RTAB] || !tb[TCA_CBQ_RATE])
|
2005-04-17 05:20:36 +07:00
|
|
|
return -EINVAL;
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
r = nla_data(tb[TCA_CBQ_RATE]);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2017-12-21 00:35:11 +07:00
|
|
|
q->link.R_tab = qdisc_get_rtab(r, tb[TCA_CBQ_RTAB]);
|
|
|
|
if (!q->link.R_tab)
|
2005-04-17 05:20:36 +07:00
|
|
|
return -EINVAL;
|
|
|
|
|
2017-11-28 00:37:21 +07:00
|
|
|
err = tcf_block_get(&q->link.block, &q->link.filter_list, sch);
|
|
|
|
if (err)
|
|
|
|
goto put_rtab;
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
err = qdisc_class_hash_init(&q->clhash);
|
|
|
|
if (err < 0)
|
2017-11-28 00:37:21 +07:00
|
|
|
goto put_block;
|
2008-07-06 13:22:05 +07:00
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
q->link.sibling = &q->link;
|
2008-07-06 13:22:05 +07:00
|
|
|
q->link.common.classid = sch->handle;
|
2005-04-17 05:20:36 +07:00
|
|
|
q->link.qdisc = sch;
|
2010-10-16 20:04:08 +07:00
|
|
|
q->link.q = qdisc_create_dflt(sch->dev_queue, &pfifo_qdisc_ops,
|
|
|
|
sch->handle);
|
|
|
|
if (!q->link.q)
|
2005-04-17 05:20:36 +07:00
|
|
|
q->link.q = &noop_qdisc;
|
2017-03-08 22:03:32 +07:00
|
|
|
else
|
|
|
|
qdisc_hash_add(q->link.q, true);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
q->link.priority = TC_CBQ_MAXPRIO - 1;
|
|
|
|
q->link.priority2 = TC_CBQ_MAXPRIO - 1;
|
|
|
|
q->link.cpriority = TC_CBQ_MAXPRIO - 1;
|
2008-07-09 07:06:30 +07:00
|
|
|
q->link.allot = psched_mtu(qdisc_dev(sch));
|
2005-04-17 05:20:36 +07:00
|
|
|
q->link.quantum = q->link.allot;
|
|
|
|
q->link.weight = q->link.R_tab->rate.rate;
|
|
|
|
|
|
|
|
q->link.ewma_log = TC_CBQ_DEF_EWMA;
|
|
|
|
q->link.avpkt = q->link.allot/2;
|
|
|
|
q->link.minidle = -0x7FFFFFFF;
|
|
|
|
|
|
|
|
q->toplevel = TC_CBQ_MAXLEVEL;
|
2007-03-24 01:29:25 +07:00
|
|
|
q->now = psched_get_time();
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
cbq_link_class(&q->link);
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
if (tb[TCA_CBQ_LSSOPT])
|
|
|
|
cbq_set_lss(&q->link, nla_data(tb[TCA_CBQ_LSSOPT]));
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
cbq_addprio(q, &q->link);
|
|
|
|
return 0;
|
2008-07-06 13:22:05 +07:00
|
|
|
|
2017-11-28 00:37:21 +07:00
|
|
|
put_block:
|
|
|
|
tcf_block_put(q->link.block);
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
put_rtab:
|
|
|
|
qdisc_put_rtab(q->link.R_tab);
|
|
|
|
return err;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static int cbq_dump_rate(struct sk_buff *skb, struct cbq_class *cl)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
2007-04-20 10:29:13 +07:00
|
|
|
unsigned char *b = skb_tail_pointer(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2012-03-29 16:11:39 +07:00
|
|
|
if (nla_put(skb, TCA_CBQ_RATE, sizeof(cl->R_tab->rate), &cl->R_tab->rate))
|
|
|
|
goto nla_put_failure;
|
2005-04-17 05:20:36 +07:00
|
|
|
return skb->len;
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
nla_put_failure:
|
2007-03-26 13:06:12 +07:00
|
|
|
nlmsg_trim(skb, b);
|
2005-04-17 05:20:36 +07:00
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static int cbq_dump_lss(struct sk_buff *skb, struct cbq_class *cl)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
2007-04-20 10:29:13 +07:00
|
|
|
unsigned char *b = skb_tail_pointer(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
struct tc_cbq_lssopt opt;
|
|
|
|
|
|
|
|
opt.flags = 0;
|
|
|
|
if (cl->borrow == NULL)
|
|
|
|
opt.flags |= TCF_CBQ_LSS_BOUNDED;
|
|
|
|
if (cl->share == NULL)
|
|
|
|
opt.flags |= TCF_CBQ_LSS_ISOLATED;
|
|
|
|
opt.ewma_log = cl->ewma_log;
|
|
|
|
opt.level = cl->level;
|
|
|
|
opt.avpkt = cl->avpkt;
|
|
|
|
opt.maxidle = cl->maxidle;
|
|
|
|
opt.minidle = (u32)(-cl->minidle);
|
|
|
|
opt.offtime = cl->offtime;
|
|
|
|
opt.change = ~0;
|
2012-03-29 16:11:39 +07:00
|
|
|
if (nla_put(skb, TCA_CBQ_LSSOPT, sizeof(opt), &opt))
|
|
|
|
goto nla_put_failure;
|
2005-04-17 05:20:36 +07:00
|
|
|
return skb->len;
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
nla_put_failure:
|
2007-03-26 13:06:12 +07:00
|
|
|
nlmsg_trim(skb, b);
|
2005-04-17 05:20:36 +07:00
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static int cbq_dump_wrr(struct sk_buff *skb, struct cbq_class *cl)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
2007-04-20 10:29:13 +07:00
|
|
|
unsigned char *b = skb_tail_pointer(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
struct tc_cbq_wrropt opt;
|
|
|
|
|
2013-07-30 14:16:21 +07:00
|
|
|
memset(&opt, 0, sizeof(opt));
|
2005-04-17 05:20:36 +07:00
|
|
|
opt.flags = 0;
|
|
|
|
opt.allot = cl->allot;
|
2011-01-20 02:26:56 +07:00
|
|
|
opt.priority = cl->priority + 1;
|
|
|
|
opt.cpriority = cl->cpriority + 1;
|
2005-04-17 05:20:36 +07:00
|
|
|
opt.weight = cl->weight;
|
2012-03-29 16:11:39 +07:00
|
|
|
if (nla_put(skb, TCA_CBQ_WRROPT, sizeof(opt), &opt))
|
|
|
|
goto nla_put_failure;
|
2005-04-17 05:20:36 +07:00
|
|
|
return skb->len;
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
nla_put_failure:
|
2007-03-26 13:06:12 +07:00
|
|
|
nlmsg_trim(skb, b);
|
2005-04-17 05:20:36 +07:00
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static int cbq_dump_fopt(struct sk_buff *skb, struct cbq_class *cl)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
2007-04-20 10:29:13 +07:00
|
|
|
unsigned char *b = skb_tail_pointer(skb);
|
2005-04-17 05:20:36 +07:00
|
|
|
struct tc_cbq_fopt opt;
|
|
|
|
|
|
|
|
if (cl->split || cl->defmap) {
|
2008-07-06 13:22:05 +07:00
|
|
|
opt.split = cl->split ? cl->split->common.classid : 0;
|
2005-04-17 05:20:36 +07:00
|
|
|
opt.defmap = cl->defmap;
|
|
|
|
opt.defchange = ~0;
|
2012-03-29 16:11:39 +07:00
|
|
|
if (nla_put(skb, TCA_CBQ_FOPT, sizeof(opt), &opt))
|
|
|
|
goto nla_put_failure;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
return skb->len;
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
nla_put_failure:
|
2007-03-26 13:06:12 +07:00
|
|
|
nlmsg_trim(skb, b);
|
2005-04-17 05:20:36 +07:00
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int cbq_dump_attr(struct sk_buff *skb, struct cbq_class *cl)
|
|
|
|
{
|
|
|
|
if (cbq_dump_lss(skb, cl) < 0 ||
|
|
|
|
cbq_dump_rate(skb, cl) < 0 ||
|
|
|
|
cbq_dump_wrr(skb, cl) < 0 ||
|
|
|
|
cbq_dump_fopt(skb, cl) < 0)
|
|
|
|
return -1;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int cbq_dump(struct Qdisc *sch, struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2008-01-24 11:34:11 +07:00
|
|
|
struct nlattr *nest;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2008-01-24 11:34:11 +07:00
|
|
|
nest = nla_nest_start(skb, TCA_OPTIONS);
|
|
|
|
if (nest == NULL)
|
|
|
|
goto nla_put_failure;
|
2005-04-17 05:20:36 +07:00
|
|
|
if (cbq_dump_attr(skb, &q->link) < 0)
|
2008-01-23 13:11:17 +07:00
|
|
|
goto nla_put_failure;
|
2014-03-12 09:20:32 +07:00
|
|
|
return nla_nest_end(skb, nest);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
nla_put_failure:
|
2008-01-24 11:34:11 +07:00
|
|
|
nla_nest_cancel(skb, nest);
|
2005-04-17 05:20:36 +07:00
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
cbq_dump_stats(struct Qdisc *sch, struct gnet_dump *d)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
|
|
|
|
q->link.xstats.avgidle = q->link.avgidle;
|
|
|
|
return gnet_stats_copy_app(d, &q->link.xstats, sizeof(q->link.xstats));
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
cbq_dump_class(struct Qdisc *sch, unsigned long arg,
|
|
|
|
struct sk_buff *skb, struct tcmsg *tcm)
|
|
|
|
{
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
2008-01-24 11:34:11 +07:00
|
|
|
struct nlattr *nest;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (cl->tparent)
|
2008-07-06 13:22:05 +07:00
|
|
|
tcm->tcm_parent = cl->tparent->common.classid;
|
2005-04-17 05:20:36 +07:00
|
|
|
else
|
|
|
|
tcm->tcm_parent = TC_H_ROOT;
|
2008-07-06 13:22:05 +07:00
|
|
|
tcm->tcm_handle = cl->common.classid;
|
2005-04-17 05:20:36 +07:00
|
|
|
tcm->tcm_info = cl->q->handle;
|
|
|
|
|
2008-01-24 11:34:11 +07:00
|
|
|
nest = nla_nest_start(skb, TCA_OPTIONS);
|
|
|
|
if (nest == NULL)
|
|
|
|
goto nla_put_failure;
|
2005-04-17 05:20:36 +07:00
|
|
|
if (cbq_dump_attr(skb, cl) < 0)
|
2008-01-23 13:11:17 +07:00
|
|
|
goto nla_put_failure;
|
2014-03-12 09:20:32 +07:00
|
|
|
return nla_nest_end(skb, nest);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
nla_put_failure:
|
2008-01-24 11:34:11 +07:00
|
|
|
nla_nest_cancel(skb, nest);
|
2005-04-17 05:20:36 +07:00
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
cbq_dump_class_stats(struct Qdisc *sch, unsigned long arg,
|
|
|
|
struct gnet_dump *d)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
cl->xstats.avgidle = cl->avgidle;
|
|
|
|
cl->xstats.undertime = 0;
|
|
|
|
|
2007-03-24 01:28:30 +07:00
|
|
|
if (cl->undertime != PSCHED_PASTPERFECT)
|
2007-03-24 01:28:55 +07:00
|
|
|
cl->xstats.undertime = cl->undertime - q->now;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2016-06-06 23:37:16 +07:00
|
|
|
if (gnet_stats_copy_basic(qdisc_root_sleeping_running(sch),
|
|
|
|
d, NULL, &cl->bstats) < 0 ||
|
2016-12-05 00:48:16 +07:00
|
|
|
gnet_stats_copy_rate_est(d, &cl->rate_est) < 0 ||
|
2014-09-29 01:54:24 +07:00
|
|
|
gnet_stats_copy_queue(d, NULL, &cl->qstats, cl->q->q.qlen) < 0)
|
2005-04-17 05:20:36 +07:00
|
|
|
return -1;
|
|
|
|
|
|
|
|
return gnet_stats_copy_app(d, &cl->xstats, sizeof(cl->xstats));
|
|
|
|
}
|
|
|
|
|
|
|
|
static int cbq_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new,
|
|
|
|
struct Qdisc **old)
|
|
|
|
{
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2009-09-04 13:41:17 +07:00
|
|
|
if (new == NULL) {
|
2010-10-16 20:04:08 +07:00
|
|
|
new = qdisc_create_dflt(sch->dev_queue,
|
2009-09-04 13:41:17 +07:00
|
|
|
&pfifo_qdisc_ops, cl->common.classid);
|
|
|
|
if (new == NULL)
|
|
|
|
return -ENOBUFS;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
2009-09-04 13:41:17 +07:00
|
|
|
|
2016-02-26 05:55:00 +07:00
|
|
|
*old = qdisc_replace(sch, new, &cl->q);
|
2009-09-04 13:41:17 +07:00
|
|
|
return 0;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static struct Qdisc *cbq_leaf(struct Qdisc *sch, unsigned long arg)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2009-09-04 13:41:17 +07:00
|
|
|
return cl->q;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2006-12-08 15:25:55 +07:00
|
|
|
static void cbq_qlen_notify(struct Qdisc *sch, unsigned long arg)
|
|
|
|
{
|
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
|
|
|
|
2017-08-15 20:39:59 +07:00
|
|
|
cbq_deactivate_class(cl);
|
2006-12-08 15:25:55 +07:00
|
|
|
}
|
|
|
|
|
net_sched: remove tc class reference counting
For TC classes, their ->get() and ->put() are always paired, and the
reference counting is completely useless, because:
1) For class modification and dumping paths, we already hold RTNL lock,
so all of these ->get(),->change(),->put() are atomic.
2) For filter bindiing/unbinding, we use other reference counter than
this one, and they should have RTNL lock too.
3) For ->qlen_notify(), it is special because it is called on ->enqueue()
path, but we already hold qdisc tree lock there, and we hold this
tree lock when graft or delete the class too, so it should not be gone
or changed until we release the tree lock.
Therefore, this patch removes ->get() and ->put(), but:
1) Adds a new ->find() to find the pointer to a class by classid, no
refcnt.
2) Move the original class destroy upon the last refcnt into ->delete(),
right after releasing tree lock. This is fine because the class is
already removed from hash when holding the lock.
For those who also use ->put() as ->unbind(), just rename them to reflect
this change.
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-25 06:51:29 +07:00
|
|
|
static unsigned long cbq_find(struct Qdisc *sch, u32 classid)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
|
net_sched: remove tc class reference counting
For TC classes, their ->get() and ->put() are always paired, and the
reference counting is completely useless, because:
1) For class modification and dumping paths, we already hold RTNL lock,
so all of these ->get(),->change(),->put() are atomic.
2) For filter bindiing/unbinding, we use other reference counter than
this one, and they should have RTNL lock too.
3) For ->qlen_notify(), it is special because it is called on ->enqueue()
path, but we already hold qdisc tree lock there, and we hold this
tree lock when graft or delete the class too, so it should not be gone
or changed until we release the tree lock.
Therefore, this patch removes ->get() and ->put(), but:
1) Adds a new ->find() to find the pointer to a class by classid, no
refcnt.
2) Move the original class destroy upon the last refcnt into ->delete(),
right after releasing tree lock. This is fine because the class is
already removed from hash when holding the lock.
For those who also use ->put() as ->unbind(), just rename them to reflect
this change.
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-25 06:51:29 +07:00
|
|
|
return (unsigned long)cbq_class_lookup(q, classid);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_destroy_class(struct Qdisc *sch, struct cbq_class *cl)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
|
2008-07-26 11:43:18 +07:00
|
|
|
WARN_ON(cl->filters);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2017-05-17 16:07:55 +07:00
|
|
|
tcf_block_put(cl->block);
|
2005-04-17 05:20:36 +07:00
|
|
|
qdisc_destroy(cl->q);
|
|
|
|
qdisc_put_rtab(cl->R_tab);
|
2016-12-05 00:48:16 +07:00
|
|
|
gen_kill_estimator(&cl->rate_est);
|
2005-04-17 05:20:36 +07:00
|
|
|
if (cl != &q->link)
|
|
|
|
kfree(cl);
|
|
|
|
}
|
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
static void cbq_destroy(struct Qdisc *sch)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 08:06:00 +07:00
|
|
|
struct hlist_node *next;
|
2005-04-17 05:20:36 +07:00
|
|
|
struct cbq_class *cl;
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int h;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2007-07-15 14:03:05 +07:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
2005-04-17 05:20:36 +07:00
|
|
|
q->rx_class = NULL;
|
|
|
|
#endif
|
|
|
|
/*
|
|
|
|
* Filters must be destroyed first because we don't destroy the
|
|
|
|
* classes from root to leafs which means that filters can still
|
|
|
|
* be bound to classes which have been destroyed already. --TGR '04
|
|
|
|
*/
|
2008-07-06 13:22:05 +07:00
|
|
|
for (h = 0; h < q->clhash.hashsize; h++) {
|
2017-08-15 20:35:21 +07:00
|
|
|
hlist_for_each_entry(cl, &q->clhash.hash[h], common.hnode) {
|
2017-05-17 16:07:55 +07:00
|
|
|
tcf_block_put(cl->block);
|
2017-08-15 20:35:21 +07:00
|
|
|
cl->block = NULL;
|
|
|
|
}
|
2007-06-06 06:06:59 +07:00
|
|
|
}
|
2008-07-06 13:22:05 +07:00
|
|
|
for (h = 0; h < q->clhash.hashsize; h++) {
|
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 08:06:00 +07:00
|
|
|
hlist_for_each_entry_safe(cl, next, &q->clhash.hash[h],
|
2008-07-06 13:22:05 +07:00
|
|
|
common.hnode)
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_destroy_class(sch, cl);
|
|
|
|
}
|
2008-07-06 13:22:05 +07:00
|
|
|
qdisc_class_hash_destroy(&q->clhash);
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
2008-01-23 13:11:17 +07:00
|
|
|
cbq_change_class(struct Qdisc *sch, u32 classid, u32 parentid, struct nlattr **tca,
|
2017-12-21 00:35:15 +07:00
|
|
|
unsigned long *arg, struct netlink_ext_ack *extack)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
int err;
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl = (struct cbq_class *)*arg;
|
2008-01-23 13:11:17 +07:00
|
|
|
struct nlattr *opt = tca[TCA_OPTIONS];
|
|
|
|
struct nlattr *tb[TCA_CBQ_MAX + 1];
|
2005-04-17 05:20:36 +07:00
|
|
|
struct cbq_class *parent;
|
|
|
|
struct qdisc_rate_table *rtab = NULL;
|
|
|
|
|
2017-12-21 00:35:11 +07:00
|
|
|
if (!opt)
|
2005-04-17 05:20:36 +07:00
|
|
|
return -EINVAL;
|
|
|
|
|
2017-04-12 19:34:07 +07:00
|
|
|
err = nla_parse_nested(tb, TCA_CBQ_MAX, opt, cbq_policy, NULL);
|
2008-01-24 11:33:32 +07:00
|
|
|
if (err < 0)
|
|
|
|
return err;
|
|
|
|
|
cbq: remove TCA_CBQ_POLICE support
iproute2 doesn't implement any cbq option that results in this attribute
being sent to kernel.
To make use of it, user would have to
- patch iproute2
- add a class
- attach a qdisc to the class (default pfifo doesn't work as
q->handle is 0 and cbq_set_police() is a no-op in this case)
- re-'add' the same class (tc class change ...) again
- user must also specifiy a defmap (e.g. 'split 1:0 defmap 3f'), since
this 'police' feature relies on its presence
- the added qdisc must be one of bfifo, pfifo or netem
If all of these conditions are met and _some_ leaf qdiscs, namely
p/bfifo, netem, plug or tbf would drop a packet, kernel calls back into
cbq, which will attempt to re-queue the skb into a different class
as indicated by the parents' defmap entry for TC_PRIO_BESTEFFORT.
[ i.e. we behave as if tc_classify returned TC_ACT_RECLASSIFY ].
This feature, which isn't documented or implemented in iproute2,
and isn't implemented consistently (most qdiscs like sfq, codel, etc
drop right away instead of attempting this reclassification) is the
sole reason for the reshape_fail and __parent member in Qdisc struct.
So remove TCA_CBQ_POLICE support from the kernel, reject it via EOPNOTSUPP
so userspace knows we don't support it, and then remove no-longer needed
infrastructure in followup commit.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-06-09 05:27:40 +07:00
|
|
|
if (tb[TCA_CBQ_OVL_STRATEGY] || tb[TCA_CBQ_POLICE])
|
2016-06-09 05:27:39 +07:00
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
if (cl) {
|
|
|
|
/* Check parent */
|
|
|
|
if (parentid) {
|
2008-07-06 13:22:05 +07:00
|
|
|
if (cl->tparent &&
|
|
|
|
cl->tparent->common.classid != parentid)
|
2005-04-17 05:20:36 +07:00
|
|
|
return -EINVAL;
|
|
|
|
if (!cl->tparent && parentid != TC_H_ROOT)
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
if (tb[TCA_CBQ_RATE]) {
|
2008-11-26 12:13:31 +07:00
|
|
|
rtab = qdisc_get_rtab(nla_data(tb[TCA_CBQ_RATE]),
|
|
|
|
tb[TCA_CBQ_RTAB]);
|
2005-04-17 05:20:36 +07:00
|
|
|
if (rtab == NULL)
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2008-11-26 12:13:31 +07:00
|
|
|
if (tca[TCA_RATE]) {
|
2014-09-29 01:52:56 +07:00
|
|
|
err = gen_replace_estimator(&cl->bstats, NULL,
|
|
|
|
&cl->rate_est,
|
2016-06-06 23:37:16 +07:00
|
|
|
NULL,
|
|
|
|
qdisc_root_sleeping_running(sch),
|
2008-11-26 12:13:31 +07:00
|
|
|
tca[TCA_RATE]);
|
|
|
|
if (err) {
|
2013-12-17 14:29:17 +07:00
|
|
|
qdisc_put_rtab(rtab);
|
2008-11-26 12:13:31 +07:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
/* Change class parameters */
|
|
|
|
sch_tree_lock(sch);
|
|
|
|
|
|
|
|
if (cl->next_alive != NULL)
|
|
|
|
cbq_deactivate_class(cl);
|
|
|
|
|
|
|
|
if (rtab) {
|
2008-11-20 19:11:36 +07:00
|
|
|
qdisc_put_rtab(cl->R_tab);
|
|
|
|
cl->R_tab = rtab;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
if (tb[TCA_CBQ_LSSOPT])
|
|
|
|
cbq_set_lss(cl, nla_data(tb[TCA_CBQ_LSSOPT]));
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
if (tb[TCA_CBQ_WRROPT]) {
|
2005-04-17 05:20:36 +07:00
|
|
|
cbq_rmprio(q, cl);
|
2008-01-23 13:11:17 +07:00
|
|
|
cbq_set_wrr(cl, nla_data(tb[TCA_CBQ_WRROPT]));
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
if (tb[TCA_CBQ_FOPT])
|
|
|
|
cbq_set_fopt(cl, nla_data(tb[TCA_CBQ_FOPT]));
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (cl->q->q.qlen)
|
|
|
|
cbq_activate_class(cl);
|
|
|
|
|
|
|
|
sch_tree_unlock(sch);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (parentid == TC_H_ROOT)
|
|
|
|
return -EINVAL;
|
|
|
|
|
2017-12-21 00:35:11 +07:00
|
|
|
if (!tb[TCA_CBQ_WRROPT] || !tb[TCA_CBQ_RATE] || !tb[TCA_CBQ_LSSOPT])
|
2005-04-17 05:20:36 +07:00
|
|
|
return -EINVAL;
|
|
|
|
|
2008-01-23 13:11:17 +07:00
|
|
|
rtab = qdisc_get_rtab(nla_data(tb[TCA_CBQ_RATE]), tb[TCA_CBQ_RTAB]);
|
2005-04-17 05:20:36 +07:00
|
|
|
if (rtab == NULL)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
if (classid) {
|
|
|
|
err = -EINVAL;
|
2011-01-20 02:26:56 +07:00
|
|
|
if (TC_H_MAJ(classid ^ sch->handle) ||
|
|
|
|
cbq_class_lookup(q, classid))
|
2005-04-17 05:20:36 +07:00
|
|
|
goto failure;
|
|
|
|
} else {
|
|
|
|
int i;
|
2011-01-20 02:26:56 +07:00
|
|
|
classid = TC_H_MAKE(sch->handle, 0x8000);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-01-20 02:26:56 +07:00
|
|
|
for (i = 0; i < 0x8000; i++) {
|
2005-04-17 05:20:36 +07:00
|
|
|
if (++q->hgenerator >= 0x8000)
|
|
|
|
q->hgenerator = 1;
|
|
|
|
if (cbq_class_lookup(q, classid|q->hgenerator) == NULL)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
err = -ENOSR;
|
|
|
|
if (i >= 0x8000)
|
|
|
|
goto failure;
|
|
|
|
classid = classid|q->hgenerator;
|
|
|
|
}
|
|
|
|
|
|
|
|
parent = &q->link;
|
|
|
|
if (parentid) {
|
|
|
|
parent = cbq_class_lookup(q, parentid);
|
|
|
|
err = -EINVAL;
|
2017-12-21 00:35:11 +07:00
|
|
|
if (!parent)
|
2005-04-17 05:20:36 +07:00
|
|
|
goto failure;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = -ENOBUFS;
|
2006-07-22 04:51:30 +07:00
|
|
|
cl = kzalloc(sizeof(*cl), GFP_KERNEL);
|
2005-04-17 05:20:36 +07:00
|
|
|
if (cl == NULL)
|
|
|
|
goto failure;
|
2008-11-26 12:13:31 +07:00
|
|
|
|
2017-10-13 19:00:57 +07:00
|
|
|
err = tcf_block_get(&cl->block, &cl->filter_list, sch);
|
2017-05-17 16:07:55 +07:00
|
|
|
if (err) {
|
|
|
|
kfree(cl);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2008-11-26 12:13:31 +07:00
|
|
|
if (tca[TCA_RATE]) {
|
2014-09-29 01:52:56 +07:00
|
|
|
err = gen_new_estimator(&cl->bstats, NULL, &cl->rate_est,
|
2016-06-06 23:37:16 +07:00
|
|
|
NULL,
|
|
|
|
qdisc_root_sleeping_running(sch),
|
2008-11-26 12:13:31 +07:00
|
|
|
tca[TCA_RATE]);
|
|
|
|
if (err) {
|
2017-05-17 16:07:55 +07:00
|
|
|
tcf_block_put(cl->block);
|
2008-11-26 12:13:31 +07:00
|
|
|
kfree(cl);
|
|
|
|
goto failure;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->R_tab = rtab;
|
|
|
|
rtab = NULL;
|
2010-10-16 20:04:08 +07:00
|
|
|
cl->q = qdisc_create_dflt(sch->dev_queue, &pfifo_qdisc_ops, classid);
|
|
|
|
if (!cl->q)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->q = &noop_qdisc;
|
2017-03-08 22:03:32 +07:00
|
|
|
else
|
|
|
|
qdisc_hash_add(cl->q, true);
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
cl->common.classid = classid;
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->tparent = parent;
|
|
|
|
cl->qdisc = sch;
|
|
|
|
cl->allot = parent->allot;
|
|
|
|
cl->quantum = cl->allot;
|
|
|
|
cl->weight = cl->R_tab->rate.rate;
|
|
|
|
|
|
|
|
sch_tree_lock(sch);
|
|
|
|
cbq_link_class(cl);
|
|
|
|
cl->borrow = cl->tparent;
|
|
|
|
if (cl->tparent != &q->link)
|
|
|
|
cl->share = cl->tparent;
|
|
|
|
cbq_adjust_levels(parent);
|
|
|
|
cl->minidle = -0x7FFFFFFF;
|
2008-01-23 13:11:17 +07:00
|
|
|
cbq_set_lss(cl, nla_data(tb[TCA_CBQ_LSSOPT]));
|
|
|
|
cbq_set_wrr(cl, nla_data(tb[TCA_CBQ_WRROPT]));
|
2011-01-20 02:26:56 +07:00
|
|
|
if (cl->ewma_log == 0)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->ewma_log = q->link.ewma_log;
|
2011-01-20 02:26:56 +07:00
|
|
|
if (cl->maxidle == 0)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->maxidle = q->link.maxidle;
|
2011-01-20 02:26:56 +07:00
|
|
|
if (cl->avpkt == 0)
|
2005-04-17 05:20:36 +07:00
|
|
|
cl->avpkt = q->link.avpkt;
|
2008-01-23 13:11:17 +07:00
|
|
|
if (tb[TCA_CBQ_FOPT])
|
|
|
|
cbq_set_fopt(cl, nla_data(tb[TCA_CBQ_FOPT]));
|
2005-04-17 05:20:36 +07:00
|
|
|
sch_tree_unlock(sch);
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
qdisc_class_hash_grow(sch, &q->clhash);
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
*arg = (unsigned long)cl;
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
failure:
|
|
|
|
qdisc_put_rtab(rtab);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int cbq_delete(struct Qdisc *sch, unsigned long arg)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
2016-02-26 05:55:01 +07:00
|
|
|
unsigned int qlen, backlog;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (cl->filters || cl->children || cl == &q->link)
|
|
|
|
return -EBUSY;
|
|
|
|
|
|
|
|
sch_tree_lock(sch);
|
|
|
|
|
2006-12-08 15:25:55 +07:00
|
|
|
qlen = cl->q->q.qlen;
|
2016-02-26 05:55:01 +07:00
|
|
|
backlog = cl->q->qstats.backlog;
|
2006-12-08 15:25:55 +07:00
|
|
|
qdisc_reset(cl->q);
|
2016-02-26 05:55:01 +07:00
|
|
|
qdisc_tree_reduce_backlog(cl->q, qlen, backlog);
|
2006-12-08 15:25:55 +07:00
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
if (cl->next_alive)
|
|
|
|
cbq_deactivate_class(cl);
|
|
|
|
|
|
|
|
if (q->tx_borrowed == cl)
|
|
|
|
q->tx_borrowed = q->tx_class;
|
|
|
|
if (q->tx_class == cl) {
|
|
|
|
q->tx_class = NULL;
|
|
|
|
q->tx_borrowed = NULL;
|
|
|
|
}
|
2007-07-15 14:03:05 +07:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
2005-04-17 05:20:36 +07:00
|
|
|
if (q->rx_class == cl)
|
|
|
|
q->rx_class = NULL;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
cbq_unlink_class(cl);
|
|
|
|
cbq_adjust_levels(cl->tparent);
|
|
|
|
cl->defmap = 0;
|
|
|
|
cbq_sync_defmap(cl);
|
|
|
|
|
|
|
|
cbq_rmprio(q, cl);
|
|
|
|
sch_tree_unlock(sch);
|
|
|
|
|
net_sched: remove tc class reference counting
For TC classes, their ->get() and ->put() are always paired, and the
reference counting is completely useless, because:
1) For class modification and dumping paths, we already hold RTNL lock,
so all of these ->get(),->change(),->put() are atomic.
2) For filter bindiing/unbinding, we use other reference counter than
this one, and they should have RTNL lock too.
3) For ->qlen_notify(), it is special because it is called on ->enqueue()
path, but we already hold qdisc tree lock there, and we hold this
tree lock when graft or delete the class too, so it should not be gone
or changed until we release the tree lock.
Therefore, this patch removes ->get() and ->put(), but:
1) Adds a new ->find() to find the pointer to a class by classid, no
refcnt.
2) Move the original class destroy upon the last refcnt into ->delete(),
right after releasing tree lock. This is fine because the class is
already removed from hash when holding the lock.
For those who also use ->put() as ->unbind(), just rename them to reflect
this change.
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-25 06:51:29 +07:00
|
|
|
cbq_destroy_class(sch, cl);
|
2005-04-17 05:20:36 +07:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2017-05-17 16:07:55 +07:00
|
|
|
static struct tcf_block *cbq_tcf_block(struct Qdisc *sch, unsigned long arg)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
|
|
|
|
|
|
|
if (cl == NULL)
|
|
|
|
cl = &q->link;
|
|
|
|
|
2017-05-17 16:07:55 +07:00
|
|
|
return cl->block;
|
2005-04-17 05:20:36 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static unsigned long cbq_bind_filter(struct Qdisc *sch, unsigned long parent,
|
|
|
|
u32 classid)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *p = (struct cbq_class *)parent;
|
2005-04-17 05:20:36 +07:00
|
|
|
struct cbq_class *cl = cbq_class_lookup(q, classid);
|
|
|
|
|
|
|
|
if (cl) {
|
|
|
|
if (p && p->level <= cl->level)
|
|
|
|
return 0;
|
|
|
|
cl->filters++;
|
|
|
|
return (unsigned long)cl;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_unbind_filter(struct Qdisc *sch, unsigned long arg)
|
|
|
|
{
|
2011-01-20 02:26:56 +07:00
|
|
|
struct cbq_class *cl = (struct cbq_class *)arg;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
cl->filters--;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void cbq_walk(struct Qdisc *sch, struct qdisc_walker *arg)
|
|
|
|
{
|
|
|
|
struct cbq_sched_data *q = qdisc_priv(sch);
|
2008-07-06 13:22:05 +07:00
|
|
|
struct cbq_class *cl;
|
2011-01-20 02:26:56 +07:00
|
|
|
unsigned int h;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
if (arg->stop)
|
|
|
|
return;
|
|
|
|
|
2008-07-06 13:22:05 +07:00
|
|
|
for (h = 0; h < q->clhash.hashsize; h++) {
|
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 08:06:00 +07:00
|
|
|
hlist_for_each_entry(cl, &q->clhash.hash[h], common.hnode) {
|
2005-04-17 05:20:36 +07:00
|
|
|
if (arg->count < arg->skip) {
|
|
|
|
arg->count++;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
if (arg->fn(sch, (unsigned long)cl, arg) < 0) {
|
|
|
|
arg->stop = 1;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
arg->count++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-11-14 16:44:41 +07:00
|
|
|
static const struct Qdisc_class_ops cbq_class_ops = {
|
2005-04-17 05:20:36 +07:00
|
|
|
.graft = cbq_graft,
|
|
|
|
.leaf = cbq_leaf,
|
2006-12-08 15:25:55 +07:00
|
|
|
.qlen_notify = cbq_qlen_notify,
|
net_sched: remove tc class reference counting
For TC classes, their ->get() and ->put() are always paired, and the
reference counting is completely useless, because:
1) For class modification and dumping paths, we already hold RTNL lock,
so all of these ->get(),->change(),->put() are atomic.
2) For filter bindiing/unbinding, we use other reference counter than
this one, and they should have RTNL lock too.
3) For ->qlen_notify(), it is special because it is called on ->enqueue()
path, but we already hold qdisc tree lock there, and we hold this
tree lock when graft or delete the class too, so it should not be gone
or changed until we release the tree lock.
Therefore, this patch removes ->get() and ->put(), but:
1) Adds a new ->find() to find the pointer to a class by classid, no
refcnt.
2) Move the original class destroy upon the last refcnt into ->delete(),
right after releasing tree lock. This is fine because the class is
already removed from hash when holding the lock.
For those who also use ->put() as ->unbind(), just rename them to reflect
this change.
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2017-08-25 06:51:29 +07:00
|
|
|
.find = cbq_find,
|
2005-04-17 05:20:36 +07:00
|
|
|
.change = cbq_change_class,
|
|
|
|
.delete = cbq_delete,
|
|
|
|
.walk = cbq_walk,
|
2017-05-17 16:07:55 +07:00
|
|
|
.tcf_block = cbq_tcf_block,
|
2005-04-17 05:20:36 +07:00
|
|
|
.bind_tcf = cbq_bind_filter,
|
|
|
|
.unbind_tcf = cbq_unbind_filter,
|
|
|
|
.dump = cbq_dump_class,
|
|
|
|
.dump_stats = cbq_dump_class_stats,
|
|
|
|
};
|
|
|
|
|
2007-11-14 16:44:41 +07:00
|
|
|
static struct Qdisc_ops cbq_qdisc_ops __read_mostly = {
|
2005-04-17 05:20:36 +07:00
|
|
|
.next = NULL,
|
|
|
|
.cl_ops = &cbq_class_ops,
|
|
|
|
.id = "cbq",
|
|
|
|
.priv_size = sizeof(struct cbq_sched_data),
|
|
|
|
.enqueue = cbq_enqueue,
|
|
|
|
.dequeue = cbq_dequeue,
|
2008-10-31 14:47:01 +07:00
|
|
|
.peek = qdisc_peek_dequeued,
|
2005-04-17 05:20:36 +07:00
|
|
|
.init = cbq_init,
|
|
|
|
.reset = cbq_reset,
|
|
|
|
.destroy = cbq_destroy,
|
|
|
|
.change = NULL,
|
|
|
|
.dump = cbq_dump,
|
|
|
|
.dump_stats = cbq_dump_stats,
|
|
|
|
.owner = THIS_MODULE,
|
|
|
|
};
|
|
|
|
|
|
|
|
static int __init cbq_module_init(void)
|
|
|
|
{
|
|
|
|
return register_qdisc(&cbq_qdisc_ops);
|
|
|
|
}
|
2007-02-09 21:25:16 +07:00
|
|
|
static void __exit cbq_module_exit(void)
|
2005-04-17 05:20:36 +07:00
|
|
|
{
|
|
|
|
unregister_qdisc(&cbq_qdisc_ops);
|
|
|
|
}
|
|
|
|
module_init(cbq_module_init)
|
|
|
|
module_exit(cbq_module_exit)
|
|
|
|
MODULE_LICENSE("GPL");
|