License cleanup: add SPDX GPL-2.0 license identifier to files with no license
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 21:07:57 +07:00
|
|
|
/* SPDX-License-Identifier: GPL-2.0 */
|
2005-04-17 05:20:36 +07:00
|
|
|
#ifndef __NET_SCHED_GENERIC_H
|
|
|
|
#define __NET_SCHED_GENERIC_H
|
|
|
|
|
|
|
|
#include <linux/netdevice.h>
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/rcupdate.h>
|
|
|
|
#include <linux/pkt_sched.h>
|
|
|
|
#include <linux/pkt_cls.h>
|
2014-09-29 01:52:56 +07:00
|
|
|
#include <linux/percpu.h>
|
qdisc: bulk dequeue support for qdiscs with TCQ_F_ONETXQUEUE
Based on DaveM's recent API work on dev_hard_start_xmit(), that allows
sending/processing an entire skb list.
This patch implements qdisc bulk dequeue, by allowing multiple packets
to be dequeued in dequeue_skb().
The optimization principle for this is two fold, (1) to amortize
locking cost and (2) avoid expensive tailptr update for notifying HW.
(1) Several packets are dequeued while holding the qdisc root_lock,
amortizing locking cost over several packet. The dequeued SKB list is
processed under the TXQ lock in dev_hard_start_xmit(), thus also
amortizing the cost of the TXQ lock.
(2) Further more, dev_hard_start_xmit() will utilize the skb->xmit_more
API to delay HW tailptr update, which also reduces the cost per
packet.
One restriction of the new API is that every SKB must belong to the
same TXQ. This patch takes the easy way out, by restricting bulk
dequeue to qdisc's with the TCQ_F_ONETXQUEUE flag, that specifies the
qdisc only have attached a single TXQ.
Some detail about the flow; dev_hard_start_xmit() will process the skb
list, and transmit packets individually towards the driver (see
xmit_one()). In case the driver stops midway in the list, the
remaining skb list is returned by dev_hard_start_xmit(). In
sch_direct_xmit() this returned list is requeued by dev_requeue_skb().
To avoid overshooting the HW limits, which results in requeuing, the
patch limits the amount of bytes dequeued, based on the drivers BQL
limits. In-effect bulking will only happen for BQL enabled drivers.
Small amounts for extra HoL blocking (2x MTU/0.24ms) were
measured at 100Mbit/s, with bulking 8 packets, but the
oscillating nature of the measurement indicate something, like
sched latency might be causing this effect. More comparisons
show, that this oscillation goes away occationally. Thus, we
disregard this artifact completely and remove any "magic" bulking
limit.
For now, as a conservative approach, stop bulking when seeing TSO and
segmented GSO packets. They already benefit from bulking on their own.
A followup patch add this, to allow easier bisect-ability for finding
regressions.
Jointed work with Hannes, Daniel and Florian.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-02 03:35:59 +07:00
|
|
|
#include <linux/dynamic_queue_limits.h>
|
2017-05-17 16:08:01 +07:00
|
|
|
#include <linux/list.h>
|
2017-07-04 19:53:07 +07:00
|
|
|
#include <linux/refcount.h>
|
2017-10-27 08:24:28 +07:00
|
|
|
#include <linux/workqueue.h>
|
2019-02-11 15:55:32 +07:00
|
|
|
#include <linux/mutex.h>
|
2019-08-26 20:44:57 +07:00
|
|
|
#include <linux/rwsem.h>
|
2019-08-26 20:44:58 +07:00
|
|
|
#include <linux/atomic.h>
|
2019-11-02 21:17:47 +07:00
|
|
|
#include <linux/hashtable.h>
|
2005-04-17 05:20:36 +07:00
|
|
|
#include <net/gen_stats.h>
|
2007-03-23 01:55:50 +07:00
|
|
|
#include <net/rtnetlink.h>
|
2019-07-19 23:20:15 +07:00
|
|
|
#include <net/flow_offload.h>
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
struct Qdisc_ops;
|
|
|
|
struct qdisc_walker;
|
|
|
|
struct tcf_walker;
|
|
|
|
struct module;
|
2018-09-14 21:46:18 +07:00
|
|
|
struct bpf_flow_keys;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct qdisc_rate_table {
|
2005-04-17 05:20:36 +07:00
|
|
|
struct tc_ratespec rate;
|
|
|
|
u32 data[256];
|
|
|
|
struct qdisc_rate_table *next;
|
|
|
|
int refcnt;
|
|
|
|
};
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
enum qdisc_state_t {
|
2008-07-16 16:15:04 +07:00
|
|
|
__QDISC_STATE_SCHED,
|
2008-08-18 11:51:03 +07:00
|
|
|
__QDISC_STATE_DEACTIVATED,
|
2008-07-16 14:56:32 +07:00
|
|
|
};
|
|
|
|
|
2008-07-20 14:08:47 +07:00
|
|
|
struct qdisc_size_table {
|
2011-01-20 10:48:19 +07:00
|
|
|
struct rcu_head rcu;
|
2008-07-20 14:08:47 +07:00
|
|
|
struct list_head list;
|
|
|
|
struct tc_sizespec szopts;
|
|
|
|
int refcnt;
|
|
|
|
u16 data[];
|
|
|
|
};
|
|
|
|
|
2016-09-18 05:57:34 +07:00
|
|
|
/* similar to sk_buff_head, but skb->prev pointer is undefined. */
|
|
|
|
struct qdisc_skb_head {
|
|
|
|
struct sk_buff *head;
|
|
|
|
struct sk_buff *tail;
|
2019-04-10 19:32:41 +07:00
|
|
|
__u32 qlen;
|
2016-09-18 05:57:34 +07:00
|
|
|
spinlock_t lock;
|
|
|
|
};
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct Qdisc {
|
2016-06-22 13:16:49 +07:00
|
|
|
int (*enqueue)(struct sk_buff *skb,
|
|
|
|
struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free);
|
|
|
|
struct sk_buff * (*dequeue)(struct Qdisc *sch);
|
2011-10-21 04:45:43 +07:00
|
|
|
unsigned int flags;
|
2009-02-01 16:12:42 +07:00
|
|
|
#define TCQ_F_BUILTIN 1
|
2011-01-20 12:27:16 +07:00
|
|
|
#define TCQ_F_INGRESS 2
|
|
|
|
#define TCQ_F_CAN_BYPASS 4
|
|
|
|
#define TCQ_F_MQROOT 8
|
2012-12-11 22:54:33 +07:00
|
|
|
#define TCQ_F_ONETXQUEUE 0x10 /* dequeue_skb() can assume all skbs are for
|
|
|
|
* q->dev_queue : It can test
|
|
|
|
* netif_xmit_frozen_or_stopped() before
|
|
|
|
* dequeueing next packet.
|
|
|
|
* Its true for MQ/MQPRIO slaves, or non
|
|
|
|
* multiqueue device.
|
|
|
|
*/
|
2009-02-01 16:12:42 +07:00
|
|
|
#define TCQ_F_WARN_NONWC (1 << 16)
|
2014-09-29 01:52:56 +07:00
|
|
|
#define TCQ_F_CPUSTATS 0x20 /* run using percpu statistics */
|
2015-12-02 11:08:51 +07:00
|
|
|
#define TCQ_F_NOPARENT 0x40 /* root of its hierarchy :
|
|
|
|
* qdisc_tree_decrease_qlen() should stop.
|
|
|
|
*/
|
2017-03-08 22:03:32 +07:00
|
|
|
#define TCQ_F_INVISIBLE 0x80 /* invisible by default in dump */
|
2017-12-08 00:54:25 +07:00
|
|
|
#define TCQ_F_NOLOCK 0x100 /* qdisc does not require locking */
|
2017-12-14 20:54:29 +07:00
|
|
|
#define TCQ_F_OFFLOADED 0x200 /* qdisc is offloaded to HW */
|
2013-06-06 22:43:22 +07:00
|
|
|
u32 limit;
|
2011-10-21 04:45:43 +07:00
|
|
|
const struct Qdisc_ops *ops;
|
2011-01-20 10:48:19 +07:00
|
|
|
struct qdisc_size_table __rcu *stab;
|
2016-08-10 16:05:15 +07:00
|
|
|
struct hlist_node hash;
|
2005-04-17 05:20:36 +07:00
|
|
|
u32 handle;
|
|
|
|
u32 parent;
|
2008-07-19 10:54:17 +07:00
|
|
|
|
net: reorder struct Qdisc for better SMP performance
dev_queue_xmit() needs to dirty fields "state", "q", "bstats" and "qstats"
On x86_64 arch, they currently span three cache lines, involving more
cache line ping pongs than necessary, making longer holding of queue spinlock.
We can reduce this to one cache line, by grouping all read-mostly fields
at the beginning of structure. (Or should I say, all highly modified fields
at the end :) )
Before patch :
offsetof(struct Qdisc, state)=0x38
offsetof(struct Qdisc, q)=0x48
offsetof(struct Qdisc, bstats)=0x80
offsetof(struct Qdisc, qstats)=0x90
sizeof(struct Qdisc)=0xc8
After patch :
offsetof(struct Qdisc, state)=0x80
offsetof(struct Qdisc, q)=0x88
offsetof(struct Qdisc, bstats)=0xa0
offsetof(struct Qdisc, qstats)=0xac
sizeof(struct Qdisc)=0xc0
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-20 15:33:32 +07:00
|
|
|
struct netdev_queue *dev_queue;
|
|
|
|
|
2016-12-05 00:48:16 +07:00
|
|
|
struct net_rate_estimator __rcu *rate_est;
|
2015-01-30 08:30:12 +07:00
|
|
|
struct gnet_stats_basic_cpu __percpu *cpu_bstats;
|
|
|
|
struct gnet_stats_queue __percpu *cpu_qstats;
|
net: sched: shrink struct Qdisc
The struct Qdisc has a lot of holes, especially after commit
a53851e2c321 ("net: sched: explicit locking in gso_cpu fallback"),
which as a side effect, moved the fields just after 'busylock'
on a new cacheline.
Since both 'padded' and 'refcnt' are not updated frequently, and
there is a hole before 'gso_skb', we can move such fields there,
saving a cacheline without any performance side effect.
Before this commit:
pahole -C Qdisc net/sche/sch_generic.o
# ...
/* size: 384, cachelines: 6, members: 25 */
/* sum members: 236, holes: 3, sum holes: 92 */
/* padding: 56 */
After this commit:
pahole -C Qdisc net/sche/sch_generic.o
# ...
/* size: 320, cachelines: 5, members: 25 */
/* sum members: 236, holes: 2, sum holes: 28 */
/* padding: 56 */
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-05-25 21:28:44 +07:00
|
|
|
int padded;
|
|
|
|
refcount_t refcnt;
|
2015-01-30 08:30:12 +07:00
|
|
|
|
net: reorder struct Qdisc for better SMP performance
dev_queue_xmit() needs to dirty fields "state", "q", "bstats" and "qstats"
On x86_64 arch, they currently span three cache lines, involving more
cache line ping pongs than necessary, making longer holding of queue spinlock.
We can reduce this to one cache line, by grouping all read-mostly fields
at the beginning of structure. (Or should I say, all highly modified fields
at the end :) )
Before patch :
offsetof(struct Qdisc, state)=0x38
offsetof(struct Qdisc, q)=0x48
offsetof(struct Qdisc, bstats)=0x80
offsetof(struct Qdisc, qstats)=0x90
sizeof(struct Qdisc)=0xc8
After patch :
offsetof(struct Qdisc, state)=0x80
offsetof(struct Qdisc, q)=0x88
offsetof(struct Qdisc, bstats)=0xa0
offsetof(struct Qdisc, qstats)=0xac
sizeof(struct Qdisc)=0xc0
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-03-20 15:33:32 +07:00
|
|
|
/*
|
|
|
|
* For performance sake on SMP, we put highly modified fields at the end
|
|
|
|
*/
|
2017-12-08 00:55:45 +07:00
|
|
|
struct sk_buff_head gso_skb ____cacheline_aligned_in_smp;
|
2016-09-18 05:57:34 +07:00
|
|
|
struct qdisc_skb_head q;
|
2015-01-30 08:30:12 +07:00
|
|
|
struct gnet_stats_basic_packed bstats;
|
2016-06-06 23:37:15 +07:00
|
|
|
seqcount_t running;
|
2015-01-30 08:30:12 +07:00
|
|
|
struct gnet_stats_queue qstats;
|
net_sched: generalize bulk dequeue
When qdisc bulk dequeue was added in linux-3.18 (commit
5772e9a3463b "qdisc: bulk dequeue support for qdiscs
with TCQ_F_ONETXQUEUE"), it was constrained to some
specific qdiscs.
With some extra care, we can extend this to all qdiscs,
so that typical traffic shaping solutions can benefit from
small batches (8 packets in this patch).
For example, HTB is often used on some multi queue device.
And bonding/team are multi queue devices...
Idea is to bulk-dequeue packets mapping to the same transmit queue.
This brings between 35 and 80 % performance increase in HTB setup
under pressure on a bonding setup :
1) NUMA node contention : 610,000 pps -> 1,110,000 pps
2) No node contention : 1,380,000 pps -> 1,930,000 pps
Now we should work to add batches on the enqueue() side ;)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: John Fastabend <john.r.fastabend@intel.com>
Cc: Jesper Dangaard Brouer <brouer@redhat.com>
Cc: Hannes Frederic Sowa <hannes@stressinduktion.org>
Cc: Florian Westphal <fw@strlen.de>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2016-06-22 13:16:52 +07:00
|
|
|
unsigned long state;
|
|
|
|
struct Qdisc *next_sched;
|
2017-12-08 00:56:23 +07:00
|
|
|
struct sk_buff_head skb_bad_txq;
|
2013-06-06 22:43:22 +07:00
|
|
|
|
|
|
|
spinlock_t busylock ____cacheline_aligned_in_smp;
|
2018-05-15 21:24:36 +07:00
|
|
|
spinlock_t seqlock;
|
2019-03-22 22:01:55 +07:00
|
|
|
|
|
|
|
/* for NOLOCK qdisc, true if there are no enqueued skbs */
|
|
|
|
bool empty;
|
2018-09-24 23:22:51 +07:00
|
|
|
struct rcu_head rcu;
|
2005-04-17 05:20:36 +07:00
|
|
|
};
|
|
|
|
|
2017-08-25 11:12:28 +07:00
|
|
|
static inline void qdisc_refcount_inc(struct Qdisc *qdisc)
|
|
|
|
{
|
|
|
|
if (qdisc->flags & TCQ_F_BUILTIN)
|
|
|
|
return;
|
|
|
|
refcount_inc(&qdisc->refcnt);
|
|
|
|
}
|
|
|
|
|
2018-09-24 23:22:52 +07:00
|
|
|
/* Intended to be used by unlocked users, when concurrent qdisc release is
|
|
|
|
* possible.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static inline struct Qdisc *qdisc_refcount_inc_nz(struct Qdisc *qdisc)
|
|
|
|
{
|
|
|
|
if (qdisc->flags & TCQ_F_BUILTIN)
|
|
|
|
return qdisc;
|
|
|
|
if (refcount_inc_not_zero(&qdisc->refcnt))
|
|
|
|
return qdisc;
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2018-05-15 21:24:36 +07:00
|
|
|
static inline bool qdisc_is_running(struct Qdisc *qdisc)
|
2010-06-02 17:23:51 +07:00
|
|
|
{
|
2018-05-15 15:50:31 +07:00
|
|
|
if (qdisc->flags & TCQ_F_NOLOCK)
|
2018-05-15 21:24:36 +07:00
|
|
|
return spin_is_locked(&qdisc->seqlock);
|
2016-06-06 23:37:15 +07:00
|
|
|
return (raw_read_seqcount(&qdisc->running) & 1) ? true : false;
|
2010-06-02 17:23:51 +07:00
|
|
|
}
|
|
|
|
|
2019-04-10 19:32:39 +07:00
|
|
|
static inline bool qdisc_is_percpu_stats(const struct Qdisc *q)
|
|
|
|
{
|
|
|
|
return q->flags & TCQ_F_CPUSTATS;
|
|
|
|
}
|
|
|
|
|
2019-03-22 22:01:55 +07:00
|
|
|
static inline bool qdisc_is_empty(const struct Qdisc *qdisc)
|
|
|
|
{
|
2019-04-10 19:32:39 +07:00
|
|
|
if (qdisc_is_percpu_stats(qdisc))
|
2019-11-08 23:45:23 +07:00
|
|
|
return READ_ONCE(qdisc->empty);
|
|
|
|
return !READ_ONCE(qdisc->q.qlen);
|
2019-03-22 22:01:55 +07:00
|
|
|
}
|
|
|
|
|
2010-06-02 17:23:51 +07:00
|
|
|
static inline bool qdisc_run_begin(struct Qdisc *qdisc)
|
|
|
|
{
|
2018-05-15 15:50:31 +07:00
|
|
|
if (qdisc->flags & TCQ_F_NOLOCK) {
|
2018-05-15 21:24:36 +07:00
|
|
|
if (!spin_trylock(&qdisc->seqlock))
|
2018-05-15 15:50:31 +07:00
|
|
|
return false;
|
2019-11-08 23:45:23 +07:00
|
|
|
WRITE_ONCE(qdisc->empty, false);
|
2018-05-15 15:50:31 +07:00
|
|
|
} else if (qdisc_is_running(qdisc)) {
|
2011-01-20 12:27:16 +07:00
|
|
|
return false;
|
2018-05-15 15:50:31 +07:00
|
|
|
}
|
2016-06-09 21:45:11 +07:00
|
|
|
/* Variant of write_seqcount_begin() telling lockdep a trylock
|
|
|
|
* was attempted.
|
|
|
|
*/
|
|
|
|
raw_write_seqcount_begin(&qdisc->running);
|
|
|
|
seqcount_acquire(&qdisc->running.dep_map, 0, 1, _RET_IP_);
|
2011-01-20 12:27:16 +07:00
|
|
|
return true;
|
2010-06-02 17:23:51 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void qdisc_run_end(struct Qdisc *qdisc)
|
|
|
|
{
|
2016-06-06 23:37:15 +07:00
|
|
|
write_seqcount_end(&qdisc->running);
|
2018-05-15 15:50:31 +07:00
|
|
|
if (qdisc->flags & TCQ_F_NOLOCK)
|
2018-05-15 21:24:36 +07:00
|
|
|
spin_unlock(&qdisc->seqlock);
|
2011-01-20 12:27:16 +07:00
|
|
|
}
|
|
|
|
|
qdisc: bulk dequeue support for qdiscs with TCQ_F_ONETXQUEUE
Based on DaveM's recent API work on dev_hard_start_xmit(), that allows
sending/processing an entire skb list.
This patch implements qdisc bulk dequeue, by allowing multiple packets
to be dequeued in dequeue_skb().
The optimization principle for this is two fold, (1) to amortize
locking cost and (2) avoid expensive tailptr update for notifying HW.
(1) Several packets are dequeued while holding the qdisc root_lock,
amortizing locking cost over several packet. The dequeued SKB list is
processed under the TXQ lock in dev_hard_start_xmit(), thus also
amortizing the cost of the TXQ lock.
(2) Further more, dev_hard_start_xmit() will utilize the skb->xmit_more
API to delay HW tailptr update, which also reduces the cost per
packet.
One restriction of the new API is that every SKB must belong to the
same TXQ. This patch takes the easy way out, by restricting bulk
dequeue to qdisc's with the TCQ_F_ONETXQUEUE flag, that specifies the
qdisc only have attached a single TXQ.
Some detail about the flow; dev_hard_start_xmit() will process the skb
list, and transmit packets individually towards the driver (see
xmit_one()). In case the driver stops midway in the list, the
remaining skb list is returned by dev_hard_start_xmit(). In
sch_direct_xmit() this returned list is requeued by dev_requeue_skb().
To avoid overshooting the HW limits, which results in requeuing, the
patch limits the amount of bytes dequeued, based on the drivers BQL
limits. In-effect bulking will only happen for BQL enabled drivers.
Small amounts for extra HoL blocking (2x MTU/0.24ms) were
measured at 100Mbit/s, with bulking 8 packets, but the
oscillating nature of the measurement indicate something, like
sched latency might be causing this effect. More comparisons
show, that this oscillation goes away occationally. Thus, we
disregard this artifact completely and remove any "magic" bulking
limit.
For now, as a conservative approach, stop bulking when seeing TSO and
segmented GSO packets. They already benefit from bulking on their own.
A followup patch add this, to allow easier bisect-ability for finding
regressions.
Jointed work with Hannes, Daniel and Florian.
Signed-off-by: Jesper Dangaard Brouer <brouer@redhat.com>
Signed-off-by: Hannes Frederic Sowa <hannes@stressinduktion.org>
Signed-off-by: Daniel Borkmann <dborkman@redhat.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-02 03:35:59 +07:00
|
|
|
static inline bool qdisc_may_bulk(const struct Qdisc *qdisc)
|
|
|
|
{
|
|
|
|
return qdisc->flags & TCQ_F_ONETXQUEUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline int qdisc_avail_bulklimit(const struct netdev_queue *txq)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_BQL
|
|
|
|
/* Non-BQL migrated drivers will return 0, too. */
|
|
|
|
return dql_avail(&txq->dql);
|
|
|
|
#else
|
|
|
|
return 0;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct Qdisc_class_ops {
|
2019-02-11 15:55:46 +07:00
|
|
|
unsigned int flags;
|
2005-04-17 05:20:36 +07:00
|
|
|
/* Child qdisc manipulation */
|
2009-09-15 16:53:07 +07:00
|
|
|
struct netdev_queue * (*select_queue)(struct Qdisc *, struct tcmsg *);
|
2005-04-17 05:20:36 +07:00
|
|
|
int (*graft)(struct Qdisc *, unsigned long cl,
|
2017-12-21 00:35:17 +07:00
|
|
|
struct Qdisc *, struct Qdisc **,
|
|
|
|
struct netlink_ext_ack *extack);
|
2005-04-17 05:20:36 +07:00
|
|
|
struct Qdisc * (*leaf)(struct Qdisc *, unsigned long cl);
|
2006-11-30 08:35:48 +07:00
|
|
|
void (*qlen_notify)(struct Qdisc *, unsigned long);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
/* Class manipulation routines */
|
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
|
|
|
unsigned long (*find)(struct Qdisc *, u32 classid);
|
2005-04-17 05:20:36 +07:00
|
|
|
int (*change)(struct Qdisc *, u32, u32,
|
2017-12-21 00:35:15 +07:00
|
|
|
struct nlattr **, unsigned long *,
|
|
|
|
struct netlink_ext_ack *);
|
2005-04-17 05:20:36 +07:00
|
|
|
int (*delete)(struct Qdisc *, unsigned long);
|
|
|
|
void (*walk)(struct Qdisc *, struct qdisc_walker * arg);
|
|
|
|
|
|
|
|
/* Filter manipulation */
|
2017-12-05 06:39:59 +07:00
|
|
|
struct tcf_block * (*tcf_block)(struct Qdisc *sch,
|
2017-12-21 00:35:16 +07:00
|
|
|
unsigned long arg,
|
|
|
|
struct netlink_ext_ack *extack);
|
2005-04-17 05:20:36 +07:00
|
|
|
unsigned long (*bind_tcf)(struct Qdisc *, unsigned long,
|
|
|
|
u32 classid);
|
|
|
|
void (*unbind_tcf)(struct Qdisc *, unsigned long);
|
|
|
|
|
|
|
|
/* rtnetlink specific */
|
|
|
|
int (*dump)(struct Qdisc *, unsigned long,
|
|
|
|
struct sk_buff *skb, struct tcmsg*);
|
|
|
|
int (*dump_stats)(struct Qdisc *, unsigned long,
|
|
|
|
struct gnet_dump *);
|
|
|
|
};
|
|
|
|
|
2019-02-11 15:55:46 +07:00
|
|
|
/* Qdisc_class_ops flag values */
|
|
|
|
|
|
|
|
/* Implements API that doesn't require rtnl lock */
|
|
|
|
enum qdisc_class_ops_flags {
|
|
|
|
QDISC_CLASS_OPS_DOIT_UNLOCKED = 1,
|
|
|
|
};
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct Qdisc_ops {
|
2005-04-17 05:20:36 +07:00
|
|
|
struct Qdisc_ops *next;
|
2007-11-14 16:44:41 +07:00
|
|
|
const struct Qdisc_class_ops *cl_ops;
|
2005-04-17 05:20:36 +07:00
|
|
|
char id[IFNAMSIZ];
|
|
|
|
int priv_size;
|
2017-12-08 00:55:26 +07:00
|
|
|
unsigned int static_flags;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2016-06-22 13:16:49 +07:00
|
|
|
int (*enqueue)(struct sk_buff *skb,
|
|
|
|
struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free);
|
2005-04-17 05:20:36 +07:00
|
|
|
struct sk_buff * (*dequeue)(struct Qdisc *);
|
2008-10-31 14:43:45 +07:00
|
|
|
struct sk_buff * (*peek)(struct Qdisc *);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2017-12-21 00:35:13 +07:00
|
|
|
int (*init)(struct Qdisc *sch, struct nlattr *arg,
|
|
|
|
struct netlink_ext_ack *extack);
|
2005-04-17 05:20:36 +07:00
|
|
|
void (*reset)(struct Qdisc *);
|
|
|
|
void (*destroy)(struct Qdisc *);
|
2017-12-05 06:39:59 +07:00
|
|
|
int (*change)(struct Qdisc *sch,
|
2017-12-21 00:35:14 +07:00
|
|
|
struct nlattr *arg,
|
|
|
|
struct netlink_ext_ack *extack);
|
2017-12-05 06:39:59 +07:00
|
|
|
void (*attach)(struct Qdisc *sch);
|
2018-01-26 09:26:23 +07:00
|
|
|
int (*change_tx_queue_len)(struct Qdisc *, unsigned int);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
int (*dump)(struct Qdisc *, struct sk_buff *);
|
|
|
|
int (*dump_stats)(struct Qdisc *, struct gnet_dump *);
|
|
|
|
|
2018-01-17 17:46:52 +07:00
|
|
|
void (*ingress_block_set)(struct Qdisc *sch,
|
|
|
|
u32 block_index);
|
|
|
|
void (*egress_block_set)(struct Qdisc *sch,
|
|
|
|
u32 block_index);
|
|
|
|
u32 (*ingress_block_get)(struct Qdisc *sch);
|
|
|
|
u32 (*egress_block_get)(struct Qdisc *sch);
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
struct module *owner;
|
|
|
|
};
|
|
|
|
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct tcf_result {
|
2017-05-17 16:08:03 +07:00
|
|
|
union {
|
|
|
|
struct {
|
|
|
|
unsigned long class;
|
|
|
|
u32 classid;
|
|
|
|
};
|
|
|
|
const struct tcf_proto *goto_tp;
|
2018-07-30 19:30:44 +07:00
|
|
|
|
2019-06-25 05:13:35 +07:00
|
|
|
/* used in the skb_tc_reinsert function */
|
2018-07-30 19:30:44 +07:00
|
|
|
struct {
|
|
|
|
bool ingress;
|
|
|
|
struct gnet_stats_queue *qstats;
|
|
|
|
};
|
2017-05-17 16:08:03 +07:00
|
|
|
};
|
2005-04-17 05:20:36 +07:00
|
|
|
};
|
|
|
|
|
2018-07-23 14:23:07 +07:00
|
|
|
struct tcf_chain;
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct tcf_proto_ops {
|
2013-12-16 11:15:11 +07:00
|
|
|
struct list_head head;
|
2005-04-17 05:20:36 +07:00
|
|
|
char kind[IFNAMSIZ];
|
|
|
|
|
2011-07-06 06:25:42 +07:00
|
|
|
int (*classify)(struct sk_buff *,
|
|
|
|
const struct tcf_proto *,
|
|
|
|
struct tcf_result *);
|
2005-04-17 05:20:36 +07:00
|
|
|
int (*init)(struct tcf_proto*);
|
2019-02-11 15:55:45 +07:00
|
|
|
void (*destroy)(struct tcf_proto *tp, bool rtnl_held,
|
2018-01-25 03:54:13 +07:00
|
|
|
struct netlink_ext_ack *extack);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2017-08-05 11:31:43 +07:00
|
|
|
void* (*get)(struct tcf_proto*, u32 handle);
|
2019-02-11 15:55:44 +07:00
|
|
|
void (*put)(struct tcf_proto *tp, void *f);
|
2013-01-14 12:15:39 +07:00
|
|
|
int (*change)(struct net *net, struct sk_buff *,
|
2012-05-26 02:42:45 +07:00
|
|
|
struct tcf_proto*, unsigned long,
|
2008-01-23 13:11:33 +07:00
|
|
|
u32 handle, struct nlattr **,
|
2019-02-11 15:55:45 +07:00
|
|
|
void **, bool, bool,
|
2018-01-18 23:20:51 +07:00
|
|
|
struct netlink_ext_ack *);
|
2018-01-18 23:20:49 +07:00
|
|
|
int (*delete)(struct tcf_proto *tp, void *arg,
|
2019-02-11 15:55:45 +07:00
|
|
|
bool *last, bool rtnl_held,
|
2018-01-18 23:20:53 +07:00
|
|
|
struct netlink_ext_ack *);
|
net/sched: add delete_empty() to filters and use it in cls_flower
Revert "net/sched: cls_u32: fix refcount leak in the error path of
u32_change()", and fix the u32 refcount leak in a more generic way that
preserves the semantic of rule dumping.
On tc filters that don't support lockless insertion/removal, there is no
need to guard against concurrent insertion when a removal is in progress.
Therefore, for most of them we can avoid a full walk() when deleting, and
just decrease the refcount, like it was done on older Linux kernels.
This fixes situations where walk() was wrongly detecting a non-empty
filter, like it happened with cls_u32 in the error path of change(), thus
leading to failures in the following tdc selftests:
6aa7: (filter, u32) Add/Replace u32 with source match and invalid indev
6658: (filter, u32) Add/Replace u32 with custom hash table and invalid handle
74c2: (filter, u32) Add/Replace u32 filter with invalid hash table id
On cls_flower, and on (future) lockless filters, this check is necessary:
move all the check_empty() logic in a callback so that each filter
can have its own implementation. For cls_flower, it's sufficient to check
if no IDRs have been allocated.
This reverts commit 275c44aa194b7159d1191817b20e076f55f0e620.
Changes since v1:
- document the need for delete_empty() when TCF_PROTO_OPS_DOIT_UNLOCKED
is used, thanks to Vlad Buslov
- implement delete_empty() without doing fl_walk(), thanks to Vlad Buslov
- squash revert and new fix in a single patch, to be nice with bisect
tests that run tdc on u32 filter, thanks to Dave Miller
Fixes: 275c44aa194b ("net/sched: cls_u32: fix refcount leak in the error path of u32_change()")
Fixes: 6676d5e416ee ("net: sched: set dedicated tcf_walker flag when tp is empty")
Suggested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Suggested-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Reviewed-by: Vlad Buslov <vladbu@mellanox.com>
Tested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-12-28 22:36:58 +07:00
|
|
|
bool (*delete_empty)(struct tcf_proto *tp);
|
2019-02-11 15:55:45 +07:00
|
|
|
void (*walk)(struct tcf_proto *tp,
|
|
|
|
struct tcf_walker *arg, bool rtnl_held);
|
2018-06-26 04:30:05 +07:00
|
|
|
int (*reoffload)(struct tcf_proto *tp, bool add,
|
2019-07-19 23:20:15 +07:00
|
|
|
flow_setup_cb_t *cb, void *cb_priv,
|
2018-06-26 04:30:05 +07:00
|
|
|
struct netlink_ext_ack *extack);
|
2019-08-26 20:45:00 +07:00
|
|
|
void (*hw_add)(struct tcf_proto *tp,
|
|
|
|
void *type_data);
|
|
|
|
void (*hw_del)(struct tcf_proto *tp,
|
|
|
|
void *type_data);
|
2020-01-24 07:26:18 +07:00
|
|
|
void (*bind_class)(void *, u32, unsigned long,
|
|
|
|
void *, unsigned long);
|
2018-07-23 14:23:07 +07:00
|
|
|
void * (*tmplt_create)(struct net *net,
|
|
|
|
struct tcf_chain *chain,
|
|
|
|
struct nlattr **tca,
|
|
|
|
struct netlink_ext_ack *extack);
|
|
|
|
void (*tmplt_destroy)(void *tmplt_priv);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
/* rtnetlink specific */
|
2017-08-05 11:31:43 +07:00
|
|
|
int (*dump)(struct net*, struct tcf_proto*, void *,
|
2019-02-11 15:55:45 +07:00
|
|
|
struct sk_buff *skb, struct tcmsg*,
|
|
|
|
bool);
|
2018-07-23 14:23:07 +07:00
|
|
|
int (*tmplt_dump)(struct sk_buff *skb,
|
|
|
|
struct net *net,
|
|
|
|
void *tmplt_priv);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
struct module *owner;
|
2019-02-11 15:55:45 +07:00
|
|
|
int flags;
|
|
|
|
};
|
|
|
|
|
net/sched: add delete_empty() to filters and use it in cls_flower
Revert "net/sched: cls_u32: fix refcount leak in the error path of
u32_change()", and fix the u32 refcount leak in a more generic way that
preserves the semantic of rule dumping.
On tc filters that don't support lockless insertion/removal, there is no
need to guard against concurrent insertion when a removal is in progress.
Therefore, for most of them we can avoid a full walk() when deleting, and
just decrease the refcount, like it was done on older Linux kernels.
This fixes situations where walk() was wrongly detecting a non-empty
filter, like it happened with cls_u32 in the error path of change(), thus
leading to failures in the following tdc selftests:
6aa7: (filter, u32) Add/Replace u32 with source match and invalid indev
6658: (filter, u32) Add/Replace u32 with custom hash table and invalid handle
74c2: (filter, u32) Add/Replace u32 filter with invalid hash table id
On cls_flower, and on (future) lockless filters, this check is necessary:
move all the check_empty() logic in a callback so that each filter
can have its own implementation. For cls_flower, it's sufficient to check
if no IDRs have been allocated.
This reverts commit 275c44aa194b7159d1191817b20e076f55f0e620.
Changes since v1:
- document the need for delete_empty() when TCF_PROTO_OPS_DOIT_UNLOCKED
is used, thanks to Vlad Buslov
- implement delete_empty() without doing fl_walk(), thanks to Vlad Buslov
- squash revert and new fix in a single patch, to be nice with bisect
tests that run tdc on u32 filter, thanks to Dave Miller
Fixes: 275c44aa194b ("net/sched: cls_u32: fix refcount leak in the error path of u32_change()")
Fixes: 6676d5e416ee ("net: sched: set dedicated tcf_walker flag when tp is empty")
Suggested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Suggested-by: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Reviewed-by: Vlad Buslov <vladbu@mellanox.com>
Tested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2019-12-28 22:36:58 +07:00
|
|
|
/* Classifiers setting TCF_PROTO_OPS_DOIT_UNLOCKED in tcf_proto_ops->flags
|
|
|
|
* are expected to implement tcf_proto_ops->delete_empty(), otherwise race
|
|
|
|
* conditions can occur when filters are inserted/deleted simultaneously.
|
|
|
|
*/
|
2019-02-11 15:55:45 +07:00
|
|
|
enum tcf_proto_ops_flags {
|
|
|
|
TCF_PROTO_OPS_DOIT_UNLOCKED = 1,
|
2005-04-17 05:20:36 +07:00
|
|
|
};
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct tcf_proto {
|
2005-04-17 05:20:36 +07:00
|
|
|
/* Fast access part */
|
2014-09-13 10:05:27 +07:00
|
|
|
struct tcf_proto __rcu *next;
|
|
|
|
void __rcu *root;
|
2018-07-30 19:30:43 +07:00
|
|
|
|
|
|
|
/* called under RCU BH lock*/
|
2011-07-06 06:25:42 +07:00
|
|
|
int (*classify)(struct sk_buff *,
|
|
|
|
const struct tcf_proto *,
|
|
|
|
struct tcf_result *);
|
2006-11-21 09:07:51 +07:00
|
|
|
__be16 protocol;
|
2005-04-17 05:20:36 +07:00
|
|
|
|
|
|
|
/* All the rest */
|
|
|
|
u32 prio;
|
|
|
|
void *data;
|
2011-07-06 06:25:42 +07:00
|
|
|
const struct tcf_proto_ops *ops;
|
2017-05-17 16:08:01 +07:00
|
|
|
struct tcf_chain *chain;
|
2019-02-11 15:55:41 +07:00
|
|
|
/* Lock protects tcf_proto shared state and can be used by unlocked
|
|
|
|
* classifiers to protect their private data.
|
|
|
|
*/
|
|
|
|
spinlock_t lock;
|
|
|
|
bool deleting;
|
2019-02-11 15:55:39 +07:00
|
|
|
refcount_t refcnt;
|
2014-09-13 10:05:27 +07:00
|
|
|
struct rcu_head rcu;
|
2019-11-02 21:17:47 +07:00
|
|
|
struct hlist_node destroy_ht_node;
|
2005-04-17 05:20:36 +07:00
|
|
|
};
|
|
|
|
|
2008-07-20 14:08:47 +07:00
|
|
|
struct qdisc_skb_cb {
|
2019-04-22 22:55:44 +07:00
|
|
|
struct {
|
|
|
|
unsigned int pkt_len;
|
|
|
|
u16 slave_dev_queue_mapping;
|
|
|
|
u16 tc_classid;
|
2018-09-14 21:46:18 +07:00
|
|
|
};
|
2014-09-18 22:02:05 +07:00
|
|
|
#define QDISC_CB_PRIV_LEN 20
|
|
|
|
unsigned char data[QDISC_CB_PRIV_LEN];
|
2008-07-20 14:08:47 +07:00
|
|
|
};
|
|
|
|
|
2017-11-03 17:46:24 +07:00
|
|
|
typedef void tcf_chain_head_change_t(struct tcf_proto *tp_head, void *priv);
|
|
|
|
|
2017-05-17 16:07:59 +07:00
|
|
|
struct tcf_chain {
|
2019-02-11 15:55:38 +07:00
|
|
|
/* Protects filter_chain. */
|
|
|
|
struct mutex filter_chain_lock;
|
2017-05-17 16:07:59 +07:00
|
|
|
struct tcf_proto __rcu *filter_chain;
|
2017-05-17 16:08:01 +07:00
|
|
|
struct list_head list;
|
|
|
|
struct tcf_block *block;
|
|
|
|
u32 index; /* chain index */
|
|
|
|
unsigned int refcnt;
|
2018-07-27 14:45:05 +07:00
|
|
|
unsigned int action_refcnt;
|
2018-07-23 14:23:06 +07:00
|
|
|
bool explicitly_created;
|
2019-02-11 15:55:42 +07:00
|
|
|
bool flushing;
|
2018-07-23 14:23:07 +07:00
|
|
|
const struct tcf_proto_ops *tmplt_ops;
|
|
|
|
void *tmplt_priv;
|
2019-03-20 21:00:16 +07:00
|
|
|
struct rcu_head rcu;
|
2017-05-17 16:07:55 +07:00
|
|
|
};
|
|
|
|
|
2017-05-17 16:07:59 +07:00
|
|
|
struct tcf_block {
|
2019-02-11 15:55:32 +07:00
|
|
|
/* Lock protects tcf_block and lifetime-management data of chains
|
|
|
|
* attached to the block (refcnt, action_refcnt, explicitly_created).
|
|
|
|
*/
|
|
|
|
struct mutex lock;
|
2017-05-17 16:08:01 +07:00
|
|
|
struct list_head chain_list;
|
2018-01-17 17:46:46 +07:00
|
|
|
u32 index; /* block index for shared blocks */
|
net_sched: fix tcm_parent in tc filter dump
When we tell kernel to dump filters from root (ffff:ffff),
those filters on ingress (ffff:0000) are matched, but their
true parents must be dumped as they are. However, kernel
dumps just whatever we tell it, that is either ffff:ffff
or ffff:0000:
$ nl-cls-list --dev=dummy0 --parent=root
cls basic dev dummy0 id none parent root prio 49152 protocol ip match-all
cls basic dev dummy0 id :1 parent root prio 49152 protocol ip match-all
$ nl-cls-list --dev=dummy0 --parent=ffff:
cls basic dev dummy0 id none parent ffff: prio 49152 protocol ip match-all
cls basic dev dummy0 id :1 parent ffff: prio 49152 protocol ip match-all
This is confusing and misleading, more importantly this is
a regression since 4.15, so the old behavior must be restored.
And, when tc filters are installed on a tc class, the parent
should be the classid, rather than the qdisc handle. Commit
edf6711c9840 ("net: sched: remove classid and q fields from tcf_proto")
removed the classid we save for filters, we can just restore
this classid in tcf_block.
Steps to reproduce this:
ip li set dev dummy0 up
tc qd add dev dummy0 ingress
tc filter add dev dummy0 parent ffff: protocol arp basic action pass
tc filter show dev dummy0 root
Before this patch:
filter protocol arp pref 49152 basic
filter protocol arp pref 49152 basic handle 0x1
action order 1: gact action pass
random type none pass val 0
index 1 ref 1 bind 1
After this patch:
filter parent ffff: protocol arp pref 49152 basic
filter parent ffff: protocol arp pref 49152 basic handle 0x1
action order 1: gact action pass
random type none pass val 0
index 1 ref 1 bind 1
Fixes: a10fa20101ae ("net: sched: propagate q and parent from caller down to tcf_fill_node")
Fixes: edf6711c9840 ("net: sched: remove classid and q fields from tcf_proto")
Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2020-05-01 10:53:49 +07:00
|
|
|
u32 classid; /* which class this block belongs to */
|
2018-09-24 23:22:54 +07:00
|
|
|
refcount_t refcnt;
|
2017-10-13 19:00:58 +07:00
|
|
|
struct net *net;
|
2017-10-13 19:00:57 +07:00
|
|
|
struct Qdisc *q;
|
2019-08-26 20:44:57 +07:00
|
|
|
struct rw_semaphore cb_lock; /* protects cb_list and offload counters */
|
2019-07-19 23:20:16 +07:00
|
|
|
struct flow_block flow_block;
|
2018-01-17 17:46:48 +07:00
|
|
|
struct list_head owner_list;
|
|
|
|
bool keep_dst;
|
2019-08-26 20:44:58 +07:00
|
|
|
atomic_t offloadcnt; /* Number of oddloaded filters */
|
2018-01-17 17:46:50 +07:00
|
|
|
unsigned int nooffloaddevcnt; /* Number of devs unable to do offload */
|
2019-08-26 20:45:01 +07:00
|
|
|
unsigned int lockeddevcnt; /* Number of devs that require rtnl lock. */
|
2018-07-23 14:23:05 +07:00
|
|
|
struct {
|
|
|
|
struct tcf_chain *chain;
|
|
|
|
struct list_head filter_chain_list;
|
|
|
|
} chain0;
|
2018-09-24 23:22:57 +07:00
|
|
|
struct rcu_head rcu;
|
2019-11-02 21:17:47 +07:00
|
|
|
DECLARE_HASHTABLE(proto_destroy_ht, 7);
|
|
|
|
struct mutex proto_destroy_lock; /* Lock for proto_destroy hashtable. */
|
2017-05-17 16:07:59 +07:00
|
|
|
};
|
|
|
|
|
2019-02-11 15:55:38 +07:00
|
|
|
#ifdef CONFIG_PROVE_LOCKING
|
|
|
|
static inline bool lockdep_tcf_chain_is_locked(struct tcf_chain *chain)
|
|
|
|
{
|
|
|
|
return lockdep_is_held(&chain->filter_chain_lock);
|
|
|
|
}
|
2019-02-11 15:55:41 +07:00
|
|
|
|
|
|
|
static inline bool lockdep_tcf_proto_is_locked(struct tcf_proto *tp)
|
|
|
|
{
|
|
|
|
return lockdep_is_held(&tp->lock);
|
|
|
|
}
|
2019-02-11 15:55:38 +07:00
|
|
|
#else
|
|
|
|
static inline bool lockdep_tcf_chain_is_locked(struct tcf_block *chain)
|
|
|
|
{
|
|
|
|
return true;
|
|
|
|
}
|
2019-02-11 15:55:41 +07:00
|
|
|
|
|
|
|
static inline bool lockdep_tcf_proto_is_locked(struct tcf_proto *tp)
|
|
|
|
{
|
|
|
|
return true;
|
|
|
|
}
|
2019-02-11 15:55:38 +07:00
|
|
|
#endif /* #ifdef CONFIG_PROVE_LOCKING */
|
|
|
|
|
|
|
|
#define tcf_chain_dereference(p, chain) \
|
|
|
|
rcu_dereference_protected(p, lockdep_tcf_chain_is_locked(chain))
|
|
|
|
|
2019-02-11 15:55:41 +07:00
|
|
|
#define tcf_proto_dereference(p, tp) \
|
|
|
|
rcu_dereference_protected(p, lockdep_tcf_proto_is_locked(tp))
|
|
|
|
|
2012-02-07 03:14:37 +07:00
|
|
|
static inline void qdisc_cb_private_validate(const struct sk_buff *skb, int sz)
|
|
|
|
{
|
|
|
|
struct qdisc_skb_cb *qcb;
|
2012-06-12 13:03:51 +07:00
|
|
|
|
|
|
|
BUILD_BUG_ON(sizeof(skb->cb) < offsetof(struct qdisc_skb_cb, data) + sz);
|
2012-02-07 03:14:37 +07:00
|
|
|
BUILD_BUG_ON(sizeof(qcb->data) < sz);
|
|
|
|
}
|
|
|
|
|
2019-04-10 19:32:41 +07:00
|
|
|
static inline int qdisc_qlen_cpu(const struct Qdisc *q)
|
|
|
|
{
|
|
|
|
return this_cpu_ptr(q->cpu_qstats)->qlen;
|
|
|
|
}
|
|
|
|
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline int qdisc_qlen(const struct Qdisc *q)
|
2009-08-06 08:44:21 +07:00
|
|
|
{
|
|
|
|
return q->q.qlen;
|
|
|
|
}
|
|
|
|
|
2019-04-10 19:32:41 +07:00
|
|
|
static inline int qdisc_qlen_sum(const struct Qdisc *q)
|
2017-12-08 00:57:00 +07:00
|
|
|
{
|
2019-04-10 19:32:41 +07:00
|
|
|
__u32 qlen = q->qstats.qlen;
|
|
|
|
int i;
|
2017-12-08 00:57:00 +07:00
|
|
|
|
2019-04-10 19:32:41 +07:00
|
|
|
if (qdisc_is_percpu_stats(q)) {
|
|
|
|
for_each_possible_cpu(i)
|
|
|
|
qlen += per_cpu_ptr(q->cpu_qstats, i)->qlen;
|
|
|
|
} else {
|
2018-05-26 11:53:30 +07:00
|
|
|
qlen += q->q.qlen;
|
2019-04-10 19:32:41 +07:00
|
|
|
}
|
2017-12-08 00:57:00 +07:00
|
|
|
|
|
|
|
return qlen;
|
|
|
|
}
|
|
|
|
|
2011-01-09 15:30:54 +07:00
|
|
|
static inline struct qdisc_skb_cb *qdisc_skb_cb(const struct sk_buff *skb)
|
2008-07-20 14:08:47 +07:00
|
|
|
{
|
|
|
|
return (struct qdisc_skb_cb *)skb->cb;
|
|
|
|
}
|
|
|
|
|
2008-07-17 14:53:03 +07:00
|
|
|
static inline spinlock_t *qdisc_lock(struct Qdisc *qdisc)
|
|
|
|
{
|
|
|
|
return &qdisc->q.lock;
|
|
|
|
}
|
|
|
|
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline struct Qdisc *qdisc_root(const struct Qdisc *qdisc)
|
2008-07-16 15:42:40 +07:00
|
|
|
{
|
2014-09-13 10:04:52 +07:00
|
|
|
struct Qdisc *q = rcu_dereference_rtnl(qdisc->dev_queue->qdisc);
|
|
|
|
|
|
|
|
return q;
|
2008-07-16 15:42:40 +07:00
|
|
|
}
|
|
|
|
|
2019-09-25 03:11:26 +07:00
|
|
|
static inline struct Qdisc *qdisc_root_bh(const struct Qdisc *qdisc)
|
|
|
|
{
|
|
|
|
return rcu_dereference_bh(qdisc->dev_queue->qdisc);
|
|
|
|
}
|
|
|
|
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline struct Qdisc *qdisc_root_sleeping(const struct Qdisc *qdisc)
|
2008-08-21 19:11:14 +07:00
|
|
|
{
|
|
|
|
return qdisc->dev_queue->qdisc_sleeping;
|
|
|
|
}
|
|
|
|
|
2008-08-03 13:27:37 +07:00
|
|
|
/* The qdisc root lock is a mechanism by which to top level
|
|
|
|
* of a qdisc tree can be locked from any qdisc node in the
|
|
|
|
* forest. This allows changing the configuration of some
|
|
|
|
* aspect of the qdisc tree while blocking out asynchronous
|
|
|
|
* qdisc access in the packet processing paths.
|
|
|
|
*
|
|
|
|
* It is only legal to do this when the root will not change
|
|
|
|
* on us. Otherwise we'll potentially lock the wrong qdisc
|
|
|
|
* root. This is enforced by holding the RTNL semaphore, which
|
|
|
|
* all users of this lock accessor must do.
|
|
|
|
*/
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline spinlock_t *qdisc_root_lock(const struct Qdisc *qdisc)
|
2008-07-16 15:42:40 +07:00
|
|
|
{
|
|
|
|
struct Qdisc *root = qdisc_root(qdisc);
|
|
|
|
|
2008-08-03 13:27:37 +07:00
|
|
|
ASSERT_RTNL();
|
2008-07-17 14:53:03 +07:00
|
|
|
return qdisc_lock(root);
|
2008-07-16 15:42:40 +07:00
|
|
|
}
|
|
|
|
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline spinlock_t *qdisc_root_sleeping_lock(const struct Qdisc *qdisc)
|
2008-08-27 16:25:17 +07:00
|
|
|
{
|
|
|
|
struct Qdisc *root = qdisc_root_sleeping(qdisc);
|
|
|
|
|
|
|
|
ASSERT_RTNL();
|
|
|
|
return qdisc_lock(root);
|
|
|
|
}
|
|
|
|
|
2016-06-06 23:37:16 +07:00
|
|
|
static inline seqcount_t *qdisc_root_sleeping_running(const struct Qdisc *qdisc)
|
|
|
|
{
|
|
|
|
struct Qdisc *root = qdisc_root_sleeping(qdisc);
|
|
|
|
|
|
|
|
ASSERT_RTNL();
|
|
|
|
return &root->running;
|
|
|
|
}
|
|
|
|
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline struct net_device *qdisc_dev(const struct Qdisc *qdisc)
|
2008-07-09 07:06:30 +07:00
|
|
|
{
|
|
|
|
return qdisc->dev_queue->dev;
|
|
|
|
}
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline void sch_tree_lock(const struct Qdisc *q)
|
2008-07-16 17:12:24 +07:00
|
|
|
{
|
2008-08-27 16:27:10 +07:00
|
|
|
spin_lock_bh(qdisc_root_sleeping_lock(q));
|
2008-07-16 17:12:24 +07:00
|
|
|
}
|
|
|
|
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline void sch_tree_unlock(const struct Qdisc *q)
|
2008-07-16 17:12:24 +07:00
|
|
|
{
|
2008-08-27 16:27:10 +07:00
|
|
|
spin_unlock_bh(qdisc_root_sleeping_lock(q));
|
2008-07-16 17:12:24 +07:00
|
|
|
}
|
|
|
|
|
2005-07-06 04:14:30 +07:00
|
|
|
extern struct Qdisc noop_qdisc;
|
|
|
|
extern struct Qdisc_ops noop_qdisc_ops;
|
2009-09-06 15:58:51 +07:00
|
|
|
extern struct Qdisc_ops pfifo_fast_ops;
|
|
|
|
extern struct Qdisc_ops mq_qdisc_ops;
|
2015-08-28 02:21:38 +07:00
|
|
|
extern struct Qdisc_ops noqueue_qdisc_ops;
|
2013-08-28 06:19:08 +07:00
|
|
|
extern const struct Qdisc_ops *default_qdisc_ops;
|
2016-03-02 23:21:43 +07:00
|
|
|
static inline const struct Qdisc_ops *
|
|
|
|
get_default_qdisc_ops(const struct net_device *dev, int ntx)
|
|
|
|
{
|
|
|
|
return ntx < dev->real_num_tx_queues ?
|
|
|
|
default_qdisc_ops : &pfifo_fast_ops;
|
|
|
|
}
|
2005-07-06 04:14:30 +07:00
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct Qdisc_class_common {
|
2008-07-06 13:21:31 +07:00
|
|
|
u32 classid;
|
|
|
|
struct hlist_node hnode;
|
|
|
|
};
|
|
|
|
|
2009-11-03 10:26:03 +07:00
|
|
|
struct Qdisc_class_hash {
|
2008-07-06 13:21:31 +07:00
|
|
|
struct hlist_head *hash;
|
|
|
|
unsigned int hashsize;
|
|
|
|
unsigned int hashmask;
|
|
|
|
unsigned int hashelems;
|
|
|
|
};
|
|
|
|
|
|
|
|
static inline unsigned int qdisc_class_hash(u32 id, u32 mask)
|
|
|
|
{
|
|
|
|
id ^= id >> 8;
|
|
|
|
id ^= id >> 4;
|
|
|
|
return id & mask;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline struct Qdisc_class_common *
|
2011-10-21 04:45:43 +07:00
|
|
|
qdisc_class_find(const struct Qdisc_class_hash *hash, u32 id)
|
2008-07-06 13:21:31 +07:00
|
|
|
{
|
|
|
|
struct Qdisc_class_common *cl;
|
|
|
|
unsigned int h;
|
|
|
|
|
2017-08-18 14:23:24 +07:00
|
|
|
if (!id)
|
|
|
|
return NULL;
|
|
|
|
|
2008-07-06 13:21:31 +07:00
|
|
|
h = qdisc_class_hash(id, hash->hashmask);
|
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, &hash->hash[h], hnode) {
|
2008-07-06 13:21:31 +07:00
|
|
|
if (cl->classid == id)
|
|
|
|
return cl;
|
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2017-10-27 16:35:34 +07:00
|
|
|
static inline int tc_classid_to_hwtc(struct net_device *dev, u32 classid)
|
|
|
|
{
|
|
|
|
u32 hwtc = TC_H_MIN(classid) - TC_H_MIN_PRIORITY;
|
|
|
|
|
|
|
|
return (hwtc < netdev_get_num_tc(dev)) ? hwtc : -EINVAL;
|
|
|
|
}
|
|
|
|
|
2013-07-31 12:47:13 +07:00
|
|
|
int qdisc_class_hash_init(struct Qdisc_class_hash *);
|
|
|
|
void qdisc_class_hash_insert(struct Qdisc_class_hash *,
|
|
|
|
struct Qdisc_class_common *);
|
|
|
|
void qdisc_class_hash_remove(struct Qdisc_class_hash *,
|
|
|
|
struct Qdisc_class_common *);
|
|
|
|
void qdisc_class_hash_grow(struct Qdisc *, struct Qdisc_class_hash *);
|
|
|
|
void qdisc_class_hash_destroy(struct Qdisc_class_hash *);
|
|
|
|
|
2018-01-26 09:26:23 +07:00
|
|
|
int dev_qdisc_change_tx_queue_len(struct net_device *dev);
|
2013-07-31 12:47:13 +07:00
|
|
|
void dev_init_scheduler(struct net_device *dev);
|
|
|
|
void dev_shutdown(struct net_device *dev);
|
|
|
|
void dev_activate(struct net_device *dev);
|
|
|
|
void dev_deactivate(struct net_device *dev);
|
|
|
|
void dev_deactivate_many(struct list_head *head);
|
|
|
|
struct Qdisc *dev_graft_qdisc(struct netdev_queue *dev_queue,
|
|
|
|
struct Qdisc *qdisc);
|
|
|
|
void qdisc_reset(struct Qdisc *qdisc);
|
2018-09-24 23:22:50 +07:00
|
|
|
void qdisc_put(struct Qdisc *qdisc);
|
2018-09-24 23:22:51 +07:00
|
|
|
void qdisc_put_unlocked(struct Qdisc *qdisc);
|
2019-01-09 23:10:57 +07:00
|
|
|
void qdisc_tree_reduce_backlog(struct Qdisc *qdisc, int n, int len);
|
2018-11-08 08:33:34 +07:00
|
|
|
#ifdef CONFIG_NET_SCHED
|
|
|
|
int qdisc_offload_dump_helper(struct Qdisc *q, enum tc_setup_type type,
|
|
|
|
void *type_data);
|
2018-11-08 08:33:37 +07:00
|
|
|
void qdisc_offload_graft_helper(struct net_device *dev, struct Qdisc *sch,
|
|
|
|
struct Qdisc *new, struct Qdisc *old,
|
|
|
|
enum tc_setup_type type, void *type_data,
|
|
|
|
struct netlink_ext_ack *extack);
|
2018-11-08 08:33:34 +07:00
|
|
|
#else
|
|
|
|
static inline int
|
|
|
|
qdisc_offload_dump_helper(struct Qdisc *q, enum tc_setup_type type,
|
|
|
|
void *type_data)
|
|
|
|
{
|
|
|
|
q->flags &= ~TCQ_F_OFFLOADED;
|
|
|
|
return 0;
|
|
|
|
}
|
2018-11-08 08:33:37 +07:00
|
|
|
|
|
|
|
static inline void
|
|
|
|
qdisc_offload_graft_helper(struct net_device *dev, struct Qdisc *sch,
|
|
|
|
struct Qdisc *new, struct Qdisc *old,
|
|
|
|
enum tc_setup_type type, void *type_data,
|
|
|
|
struct netlink_ext_ack *extack)
|
|
|
|
{
|
|
|
|
}
|
2018-11-08 08:33:34 +07:00
|
|
|
#endif
|
2013-07-31 12:47:13 +07:00
|
|
|
struct Qdisc *qdisc_alloc(struct netdev_queue *dev_queue,
|
2017-12-21 00:35:20 +07:00
|
|
|
const struct Qdisc_ops *ops,
|
|
|
|
struct netlink_ext_ack *extack);
|
net, sched: fix panic when updating miniq {b,q}stats
While working on fixing another bug, I ran into the following panic
on arm64 by simply attaching clsact qdisc, adding a filter and running
traffic on ingress to it:
[...]
[ 178.188591] Unable to handle kernel read from unreadable memory at virtual address 810fb501f000
[ 178.197314] Mem abort info:
[ 178.200121] ESR = 0x96000004
[ 178.203168] Exception class = DABT (current EL), IL = 32 bits
[ 178.209095] SET = 0, FnV = 0
[ 178.212157] EA = 0, S1PTW = 0
[ 178.215288] Data abort info:
[ 178.218175] ISV = 0, ISS = 0x00000004
[ 178.222019] CM = 0, WnR = 0
[ 178.224997] user pgtable: 4k pages, 48-bit VAs, pgd = 0000000023cb3f33
[ 178.231531] [0000810fb501f000] *pgd=0000000000000000
[ 178.236508] Internal error: Oops: 96000004 [#1] SMP
[...]
[ 178.311855] CPU: 73 PID: 2497 Comm: ping Tainted: G W 4.15.0-rc7+ #5
[ 178.319413] Hardware name: FOXCONN R2-1221R-A4/C2U4N_MB, BIOS G31FB18A 03/31/2017
[ 178.326887] pstate: 60400005 (nZCv daif +PAN -UAO)
[ 178.331685] pc : __netif_receive_skb_core+0x49c/0xac8
[ 178.336728] lr : __netif_receive_skb+0x28/0x78
[ 178.341161] sp : ffff00002344b750
[ 178.344465] x29: ffff00002344b750 x28: ffff810fbdfd0580
[ 178.349769] x27: 0000000000000000 x26: ffff000009378000
[...]
[ 178.418715] x1 : 0000000000000054 x0 : 0000000000000000
[ 178.424020] Process ping (pid: 2497, stack limit = 0x000000009f0a3ff4)
[ 178.430537] Call trace:
[ 178.432976] __netif_receive_skb_core+0x49c/0xac8
[ 178.437670] __netif_receive_skb+0x28/0x78
[ 178.441757] process_backlog+0x9c/0x160
[ 178.445584] net_rx_action+0x2f8/0x3f0
[...]
Reason is that sch_ingress and sch_clsact are doing mini_qdisc_pair_init()
which sets up miniq pointers to cpu_{b,q}stats from the underlying qdisc.
Problem is that this cannot work since they are actually set up right after
the qdisc ->init() callback in qdisc_create(), so first packet going into
sch_handle_ingress() tries to call mini_qdisc_bstats_cpu_update() and we
therefore panic.
In order to fix this, allocation of {b,q}stats needs to happen before we
call into ->init(). In net-next, there's already such option through commit
d59f5ffa59d8 ("net: sched: a dflt qdisc may be used with per cpu stats").
However, the bug needs to be fixed in net still for 4.15. Thus, include
these bits to reduce any merge churn and reuse the static_flags field to
set TCQ_F_CPUSTATS, and remove the allocation from qdisc_create() since
there is no other user left. Prashant Bhole ran into the same issue but
for net-next, thus adding him below as well as co-author. Same issue was
also reported by Sandipan Das when using bcc.
Fixes: 46209401f8f6 ("net: core: introduce mini_Qdisc and eliminate usage of tp->q for clsact fastpath")
Reference: https://lists.iovisor.org/pipermail/iovisor-dev/2018-January/001190.html
Reported-by: Sandipan Das <sandipan@linux.vnet.ibm.com>
Co-authored-by: Prashant Bhole <bhole_prashant_q7@lab.ntt.co.jp>
Co-authored-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Pirko <jiri@resnulli.us>
Signed-off-by: David S. Miller <davem@davemloft.net>
2018-01-16 05:12:09 +07:00
|
|
|
void qdisc_free(struct Qdisc *qdisc);
|
2013-07-31 12:47:13 +07:00
|
|
|
struct Qdisc *qdisc_create_dflt(struct netdev_queue *dev_queue,
|
2017-12-21 00:35:21 +07:00
|
|
|
const struct Qdisc_ops *ops, u32 parentid,
|
|
|
|
struct netlink_ext_ack *extack);
|
2013-07-31 12:47:13 +07:00
|
|
|
void __qdisc_calculate_pkt_len(struct sk_buff *skb,
|
|
|
|
const struct qdisc_size_table *stab);
|
2015-09-16 13:05:43 +07:00
|
|
|
int skb_do_redirect(struct sk_buff *);
|
2005-04-17 05:20:36 +07:00
|
|
|
|
2016-01-07 21:50:22 +07:00
|
|
|
static inline bool skb_at_tc_ingress(const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
2017-01-08 05:06:37 +07:00
|
|
|
return skb->tc_at_ingress;
|
2016-01-07 21:50:22 +07:00
|
|
|
#else
|
|
|
|
return false;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2017-01-08 05:06:35 +07:00
|
|
|
static inline bool skb_skip_tc_classify(struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
|
|
|
if (skb->tc_skip_classify) {
|
|
|
|
skb->tc_skip_classify = 0;
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
2018-02-28 20:59:15 +07:00
|
|
|
/* Reset all TX qdiscs greater than index of a device. */
|
2010-07-01 20:21:57 +07:00
|
|
|
static inline void qdisc_reset_all_tx_gt(struct net_device *dev, unsigned int i)
|
2008-07-09 12:59:10 +07:00
|
|
|
{
|
2010-07-01 20:21:35 +07:00
|
|
|
struct Qdisc *qdisc;
|
|
|
|
|
2010-07-01 20:21:57 +07:00
|
|
|
for (; i < dev->num_tx_queues; i++) {
|
2014-09-13 10:04:52 +07:00
|
|
|
qdisc = rtnl_dereference(netdev_get_tx_queue(dev, i)->qdisc);
|
2010-07-01 20:21:35 +07:00
|
|
|
if (qdisc) {
|
|
|
|
spin_lock_bh(qdisc_lock(qdisc));
|
|
|
|
qdisc_reset(qdisc);
|
|
|
|
spin_unlock_bh(qdisc_lock(qdisc));
|
|
|
|
}
|
|
|
|
}
|
2008-07-09 12:59:10 +07:00
|
|
|
}
|
|
|
|
|
2008-07-09 13:00:25 +07:00
|
|
|
/* Are all TX queues of the device empty? */
|
|
|
|
static inline bool qdisc_all_tx_empty(const struct net_device *dev)
|
|
|
|
{
|
2008-07-17 14:34:19 +07:00
|
|
|
unsigned int i;
|
2014-09-13 10:04:52 +07:00
|
|
|
|
|
|
|
rcu_read_lock();
|
2008-07-17 14:34:19 +07:00
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
2014-09-13 10:04:52 +07:00
|
|
|
const struct Qdisc *q = rcu_dereference(txq->qdisc);
|
2008-07-09 13:00:25 +07:00
|
|
|
|
2019-04-10 19:32:38 +07:00
|
|
|
if (!qdisc_is_empty(q)) {
|
2014-09-13 10:04:52 +07:00
|
|
|
rcu_read_unlock();
|
2008-07-17 14:34:19 +07:00
|
|
|
return false;
|
2014-09-13 10:04:52 +07:00
|
|
|
}
|
2008-07-17 14:34:19 +07:00
|
|
|
}
|
2014-09-13 10:04:52 +07:00
|
|
|
rcu_read_unlock();
|
2008-07-17 14:34:19 +07:00
|
|
|
return true;
|
2008-07-09 13:00:25 +07:00
|
|
|
}
|
|
|
|
|
2008-07-09 13:01:06 +07:00
|
|
|
/* Are any of the TX qdiscs changing? */
|
2011-10-21 04:45:43 +07:00
|
|
|
static inline bool qdisc_tx_changing(const struct net_device *dev)
|
2008-07-09 13:01:06 +07:00
|
|
|
{
|
2008-07-17 14:34:19 +07:00
|
|
|
unsigned int i;
|
2014-09-13 10:04:52 +07:00
|
|
|
|
2008-07-17 14:34:19 +07:00
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
2014-09-13 10:04:52 +07:00
|
|
|
if (rcu_access_pointer(txq->qdisc) != txq->qdisc_sleeping)
|
2008-07-17 14:34:19 +07:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
return false;
|
2008-07-09 13:01:06 +07:00
|
|
|
}
|
|
|
|
|
2008-07-17 14:34:19 +07:00
|
|
|
/* Is the device using the noop qdisc on all queues? */
|
2008-07-09 13:01:27 +07:00
|
|
|
static inline bool qdisc_tx_is_noop(const struct net_device *dev)
|
|
|
|
{
|
2008-07-17 14:34:19 +07:00
|
|
|
unsigned int i;
|
2014-09-13 10:04:52 +07:00
|
|
|
|
2008-07-17 14:34:19 +07:00
|
|
|
for (i = 0; i < dev->num_tx_queues; i++) {
|
|
|
|
struct netdev_queue *txq = netdev_get_tx_queue(dev, i);
|
2014-09-13 10:04:52 +07:00
|
|
|
if (rcu_access_pointer(txq->qdisc) != &noop_qdisc)
|
2008-07-17 14:34:19 +07:00
|
|
|
return false;
|
|
|
|
}
|
|
|
|
return true;
|
2008-07-09 13:01:27 +07:00
|
|
|
}
|
|
|
|
|
2011-01-09 15:30:54 +07:00
|
|
|
static inline unsigned int qdisc_pkt_len(const struct sk_buff *skb)
|
2008-07-20 14:08:27 +07:00
|
|
|
{
|
2008-07-20 14:08:47 +07:00
|
|
|
return qdisc_skb_cb(skb)->pkt_len;
|
2008-07-20 14:08:27 +07:00
|
|
|
}
|
|
|
|
|
2008-08-05 12:39:11 +07:00
|
|
|
/* additional qdisc xmit flags (NET_XMIT_MASK in linux/netdevice.h) */
|
2008-08-05 12:31:03 +07:00
|
|
|
enum net_xmit_qdisc_t {
|
|
|
|
__NET_XMIT_STOLEN = 0x00010000,
|
2008-08-05 12:39:11 +07:00
|
|
|
__NET_XMIT_BYPASS = 0x00020000,
|
2008-08-05 12:31:03 +07:00
|
|
|
};
|
|
|
|
|
2008-08-05 12:39:11 +07:00
|
|
|
#ifdef CONFIG_NET_CLS_ACT
|
2008-08-05 12:31:03 +07:00
|
|
|
#define net_xmit_drop_count(e) ((e) & __NET_XMIT_STOLEN ? 0 : 1)
|
|
|
|
#else
|
|
|
|
#define net_xmit_drop_count(e) (1)
|
|
|
|
#endif
|
|
|
|
|
2011-01-20 10:48:19 +07:00
|
|
|
static inline void qdisc_calculate_pkt_len(struct sk_buff *skb,
|
|
|
|
const struct Qdisc *sch)
|
2008-07-20 14:08:04 +07:00
|
|
|
{
|
2008-07-21 08:13:01 +07:00
|
|
|
#ifdef CONFIG_NET_SCHED
|
2011-01-20 10:48:19 +07:00
|
|
|
struct qdisc_size_table *stab = rcu_dereference_bh(sch->stab);
|
|
|
|
|
|
|
|
if (stab)
|
|
|
|
__qdisc_calculate_pkt_len(skb, stab);
|
2008-07-21 08:13:01 +07:00
|
|
|
#endif
|
2011-01-20 10:48:19 +07:00
|
|
|
}
|
|
|
|
|
2016-06-22 13:16:49 +07:00
|
|
|
static inline int qdisc_enqueue(struct sk_buff *skb, struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free)
|
2011-01-20 10:48:19 +07:00
|
|
|
{
|
|
|
|
qdisc_calculate_pkt_len(skb, sch);
|
2016-06-22 13:16:49 +07:00
|
|
|
return sch->enqueue(skb, sch, to_free);
|
2008-07-20 14:08:04 +07:00
|
|
|
}
|
|
|
|
|
2016-05-13 19:55:35 +07:00
|
|
|
static inline void _bstats_update(struct gnet_stats_basic_packed *bstats,
|
|
|
|
__u64 bytes, __u32 packets)
|
|
|
|
{
|
|
|
|
bstats->bytes += bytes;
|
|
|
|
bstats->packets += packets;
|
|
|
|
}
|
|
|
|
|
2011-01-09 15:30:54 +07:00
|
|
|
static inline void bstats_update(struct gnet_stats_basic_packed *bstats,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
2016-05-13 19:55:35 +07:00
|
|
|
_bstats_update(bstats,
|
|
|
|
qdisc_pkt_len(skb),
|
|
|
|
skb_is_gso(skb) ? skb_shinfo(skb)->gso_segs : 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void _bstats_cpu_update(struct gnet_stats_basic_cpu *bstats,
|
|
|
|
__u64 bytes, __u32 packets)
|
|
|
|
{
|
|
|
|
u64_stats_update_begin(&bstats->syncp);
|
|
|
|
_bstats_update(&bstats->bstats, bytes, packets);
|
|
|
|
u64_stats_update_end(&bstats->syncp);
|
2011-01-09 15:30:54 +07:00
|
|
|
}
|
|
|
|
|
2015-07-06 19:18:03 +07:00
|
|
|
static inline void bstats_cpu_update(struct gnet_stats_basic_cpu *bstats,
|
|
|
|
const struct sk_buff *skb)
|
2014-09-29 01:52:56 +07:00
|
|
|
{
|
|
|
|
u64_stats_update_begin(&bstats->syncp);
|
|
|
|
bstats_update(&bstats->bstats, skb);
|
|
|
|
u64_stats_update_end(&bstats->syncp);
|
|
|
|
}
|
|
|
|
|
2015-07-06 19:18:03 +07:00
|
|
|
static inline void qdisc_bstats_cpu_update(struct Qdisc *sch,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
bstats_cpu_update(this_cpu_ptr(sch->cpu_bstats), skb);
|
|
|
|
}
|
|
|
|
|
2011-01-09 15:30:54 +07:00
|
|
|
static inline void qdisc_bstats_update(struct Qdisc *sch,
|
|
|
|
const struct sk_buff *skb)
|
2009-08-06 08:44:21 +07:00
|
|
|
{
|
2011-01-09 15:30:54 +07:00
|
|
|
bstats_update(&sch->bstats, skb);
|
2009-08-06 08:44:21 +07:00
|
|
|
}
|
|
|
|
|
2014-09-29 01:53:29 +07:00
|
|
|
static inline void qdisc_qstats_backlog_dec(struct Qdisc *sch,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
sch->qstats.backlog -= qdisc_pkt_len(skb);
|
|
|
|
}
|
|
|
|
|
2017-12-08 00:55:07 +07:00
|
|
|
static inline void qdisc_qstats_cpu_backlog_dec(struct Qdisc *sch,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
this_cpu_sub(sch->cpu_qstats->backlog, qdisc_pkt_len(skb));
|
|
|
|
}
|
|
|
|
|
2014-09-29 01:53:29 +07:00
|
|
|
static inline void qdisc_qstats_backlog_inc(struct Qdisc *sch,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
sch->qstats.backlog += qdisc_pkt_len(skb);
|
|
|
|
}
|
|
|
|
|
2017-12-08 00:55:07 +07:00
|
|
|
static inline void qdisc_qstats_cpu_backlog_inc(struct Qdisc *sch,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
this_cpu_add(sch->cpu_qstats->backlog, qdisc_pkt_len(skb));
|
|
|
|
}
|
|
|
|
|
2019-04-10 19:32:41 +07:00
|
|
|
static inline void qdisc_qstats_cpu_qlen_inc(struct Qdisc *sch)
|
2017-12-08 00:55:07 +07:00
|
|
|
{
|
2019-04-10 19:32:41 +07:00
|
|
|
this_cpu_inc(sch->cpu_qstats->qlen);
|
2017-12-08 00:55:07 +07:00
|
|
|
}
|
|
|
|
|
2019-04-10 19:32:41 +07:00
|
|
|
static inline void qdisc_qstats_cpu_qlen_dec(struct Qdisc *sch)
|
2017-12-08 00:55:07 +07:00
|
|
|
{
|
2019-04-10 19:32:41 +07:00
|
|
|
this_cpu_dec(sch->cpu_qstats->qlen);
|
2017-12-08 00:55:07 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void qdisc_qstats_cpu_requeues_inc(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
this_cpu_inc(sch->cpu_qstats->requeues);
|
|
|
|
}
|
|
|
|
|
2014-09-29 01:53:29 +07:00
|
|
|
static inline void __qdisc_qstats_drop(struct Qdisc *sch, int count)
|
|
|
|
{
|
|
|
|
sch->qstats.drops += count;
|
|
|
|
}
|
|
|
|
|
2015-07-06 19:18:03 +07:00
|
|
|
static inline void qstats_drop_inc(struct gnet_stats_queue *qstats)
|
2014-09-29 01:53:29 +07:00
|
|
|
{
|
2015-07-06 19:18:03 +07:00
|
|
|
qstats->drops++;
|
2014-09-29 01:53:29 +07:00
|
|
|
}
|
|
|
|
|
2015-07-06 19:18:03 +07:00
|
|
|
static inline void qstats_overlimit_inc(struct gnet_stats_queue *qstats)
|
2014-09-29 01:54:24 +07:00
|
|
|
{
|
2015-07-06 19:18:03 +07:00
|
|
|
qstats->overlimits++;
|
|
|
|
}
|
2014-09-29 01:54:24 +07:00
|
|
|
|
2015-07-06 19:18:03 +07:00
|
|
|
static inline void qdisc_qstats_drop(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
qstats_drop_inc(&sch->qstats);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void qdisc_qstats_cpu_drop(struct Qdisc *sch)
|
|
|
|
{
|
2016-08-25 00:23:34 +07:00
|
|
|
this_cpu_inc(sch->cpu_qstats->drops);
|
2014-09-29 01:54:24 +07:00
|
|
|
}
|
|
|
|
|
2014-09-29 01:53:29 +07:00
|
|
|
static inline void qdisc_qstats_overlimit(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
sch->qstats.overlimits++;
|
|
|
|
}
|
|
|
|
|
2019-03-28 22:53:12 +07:00
|
|
|
static inline int qdisc_qstats_copy(struct gnet_dump *d, struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
__u32 qlen = qdisc_qlen_sum(sch);
|
|
|
|
|
|
|
|
return gnet_stats_copy_queue(d, sch->cpu_qstats, &sch->qstats, qlen);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void qdisc_qstats_qlen_backlog(struct Qdisc *sch, __u32 *qlen,
|
|
|
|
__u32 *backlog)
|
|
|
|
{
|
|
|
|
struct gnet_stats_queue qstats = { 0 };
|
|
|
|
__u32 len = qdisc_qlen_sum(sch);
|
|
|
|
|
|
|
|
__gnet_stats_copy_queue(&qstats, sch->cpu_qstats, &sch->qstats, len);
|
|
|
|
*qlen = qstats.qlen;
|
|
|
|
*backlog = qstats.backlog;
|
|
|
|
}
|
|
|
|
|
2019-03-28 22:53:13 +07:00
|
|
|
static inline void qdisc_tree_flush_backlog(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
__u32 qlen, backlog;
|
|
|
|
|
|
|
|
qdisc_qstats_qlen_backlog(sch, &qlen, &backlog);
|
|
|
|
qdisc_tree_reduce_backlog(sch, qlen, backlog);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void qdisc_purge_queue(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
__u32 qlen, backlog;
|
|
|
|
|
|
|
|
qdisc_qstats_qlen_backlog(sch, &qlen, &backlog);
|
|
|
|
qdisc_reset(sch);
|
|
|
|
qdisc_tree_reduce_backlog(sch, qlen, backlog);
|
|
|
|
}
|
|
|
|
|
2016-09-18 05:57:34 +07:00
|
|
|
static inline void qdisc_skb_head_init(struct qdisc_skb_head *qh)
|
|
|
|
{
|
|
|
|
qh->head = NULL;
|
|
|
|
qh->tail = NULL;
|
|
|
|
qh->qlen = 0;
|
|
|
|
}
|
|
|
|
|
2018-07-30 06:22:13 +07:00
|
|
|
static inline void __qdisc_enqueue_tail(struct sk_buff *skb,
|
|
|
|
struct qdisc_skb_head *qh)
|
2005-06-19 12:57:26 +07:00
|
|
|
{
|
2016-09-18 05:57:34 +07:00
|
|
|
struct sk_buff *last = qh->tail;
|
|
|
|
|
|
|
|
if (last) {
|
|
|
|
skb->next = NULL;
|
|
|
|
last->next = skb;
|
|
|
|
qh->tail = skb;
|
|
|
|
} else {
|
|
|
|
qh->tail = skb;
|
|
|
|
qh->head = skb;
|
|
|
|
}
|
|
|
|
qh->qlen++;
|
2005-06-19 12:57:26 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline int qdisc_enqueue_tail(struct sk_buff *skb, struct Qdisc *sch)
|
|
|
|
{
|
2018-07-30 06:22:13 +07:00
|
|
|
__qdisc_enqueue_tail(skb, &sch->q);
|
|
|
|
qdisc_qstats_backlog_inc(sch, skb);
|
|
|
|
return NET_XMIT_SUCCESS;
|
2005-06-19 12:57:26 +07:00
|
|
|
}
|
|
|
|
|
2018-07-30 06:33:28 +07:00
|
|
|
static inline void __qdisc_enqueue_head(struct sk_buff *skb,
|
|
|
|
struct qdisc_skb_head *qh)
|
|
|
|
{
|
|
|
|
skb->next = qh->head;
|
|
|
|
|
|
|
|
if (!qh->head)
|
|
|
|
qh->tail = skb;
|
|
|
|
qh->head = skb;
|
|
|
|
qh->qlen++;
|
|
|
|
}
|
|
|
|
|
2016-09-18 05:57:34 +07:00
|
|
|
static inline struct sk_buff *__qdisc_dequeue_head(struct qdisc_skb_head *qh)
|
2005-06-19 12:57:26 +07:00
|
|
|
{
|
2016-09-18 05:57:34 +07:00
|
|
|
struct sk_buff *skb = qh->head;
|
|
|
|
|
|
|
|
if (likely(skb != NULL)) {
|
|
|
|
qh->head = skb->next;
|
|
|
|
qh->qlen--;
|
|
|
|
if (qh->head == NULL)
|
|
|
|
qh->tail = NULL;
|
|
|
|
skb->next = NULL;
|
|
|
|
}
|
2005-06-19 12:57:26 +07:00
|
|
|
|
2016-09-18 05:57:32 +07:00
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline struct sk_buff *qdisc_dequeue_head(struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
struct sk_buff *skb = __qdisc_dequeue_head(&sch->q);
|
|
|
|
|
2011-01-21 14:31:33 +07:00
|
|
|
if (likely(skb != NULL)) {
|
2014-09-29 01:53:29 +07:00
|
|
|
qdisc_qstats_backlog_dec(sch, skb);
|
2011-01-21 14:31:33 +07:00
|
|
|
qdisc_bstats_update(sch, skb);
|
|
|
|
}
|
2005-06-19 12:57:26 +07:00
|
|
|
|
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
|
2016-06-22 13:16:49 +07:00
|
|
|
/* Instead of calling kfree_skb() while root qdisc lock is held,
|
|
|
|
* queue the skb for future freeing at end of __dev_xmit_skb()
|
|
|
|
*/
|
|
|
|
static inline void __qdisc_drop(struct sk_buff *skb, struct sk_buff **to_free)
|
|
|
|
{
|
|
|
|
skb->next = *to_free;
|
|
|
|
*to_free = skb;
|
|
|
|
}
|
|
|
|
|
2018-03-06 00:52:54 +07:00
|
|
|
static inline void __qdisc_drop_all(struct sk_buff *skb,
|
|
|
|
struct sk_buff **to_free)
|
|
|
|
{
|
|
|
|
if (skb->prev)
|
|
|
|
skb->prev->next = *to_free;
|
|
|
|
else
|
|
|
|
skb->next = *to_free;
|
|
|
|
*to_free = skb;
|
|
|
|
}
|
|
|
|
|
2010-01-24 19:30:59 +07:00
|
|
|
static inline unsigned int __qdisc_queue_drop_head(struct Qdisc *sch,
|
2016-09-18 05:57:34 +07:00
|
|
|
struct qdisc_skb_head *qh,
|
2016-06-22 13:16:49 +07:00
|
|
|
struct sk_buff **to_free)
|
2010-01-24 19:30:59 +07:00
|
|
|
{
|
2016-09-18 05:57:34 +07:00
|
|
|
struct sk_buff *skb = __qdisc_dequeue_head(qh);
|
2010-01-24 19:30:59 +07:00
|
|
|
|
|
|
|
if (likely(skb != NULL)) {
|
|
|
|
unsigned int len = qdisc_pkt_len(skb);
|
2016-06-22 13:16:49 +07:00
|
|
|
|
2014-09-29 01:53:29 +07:00
|
|
|
qdisc_qstats_backlog_dec(sch, skb);
|
2016-06-22 13:16:49 +07:00
|
|
|
__qdisc_drop(skb, to_free);
|
2010-01-24 19:30:59 +07:00
|
|
|
return len;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-06-22 13:16:49 +07:00
|
|
|
static inline unsigned int qdisc_queue_drop_head(struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free)
|
2010-01-24 19:30:59 +07:00
|
|
|
{
|
2016-06-22 13:16:49 +07:00
|
|
|
return __qdisc_queue_drop_head(sch, &sch->q, to_free);
|
2005-06-19 12:57:26 +07:00
|
|
|
}
|
|
|
|
|
2008-10-31 14:44:18 +07:00
|
|
|
static inline struct sk_buff *qdisc_peek_head(struct Qdisc *sch)
|
|
|
|
{
|
2016-09-18 05:57:34 +07:00
|
|
|
const struct qdisc_skb_head *qh = &sch->q;
|
|
|
|
|
|
|
|
return qh->head;
|
2008-10-31 14:44:18 +07:00
|
|
|
}
|
|
|
|
|
2008-10-31 14:47:01 +07:00
|
|
|
/* generic pseudo peek method for non-work-conserving qdisc */
|
|
|
|
static inline struct sk_buff *qdisc_peek_dequeued(struct Qdisc *sch)
|
|
|
|
{
|
2017-12-08 00:55:45 +07:00
|
|
|
struct sk_buff *skb = skb_peek(&sch->gso_skb);
|
|
|
|
|
2008-10-31 14:47:01 +07:00
|
|
|
/* we can reuse ->gso_skb because peek isn't called for root qdiscs */
|
2017-12-08 00:55:45 +07:00
|
|
|
if (!skb) {
|
|
|
|
skb = sch->dequeue(sch);
|
|
|
|
|
|
|
|
if (skb) {
|
|
|
|
__skb_queue_head(&sch->gso_skb, skb);
|
2008-11-06 07:02:34 +07:00
|
|
|
/* it's still part of the queue */
|
2017-12-08 00:55:45 +07:00
|
|
|
qdisc_qstats_backlog_inc(sch, skb);
|
2008-11-06 07:02:34 +07:00
|
|
|
sch->q.qlen++;
|
2016-06-04 05:05:57 +07:00
|
|
|
}
|
2008-11-06 07:02:34 +07:00
|
|
|
}
|
2008-10-31 14:47:01 +07:00
|
|
|
|
2017-12-08 00:55:45 +07:00
|
|
|
return skb;
|
2008-10-31 14:47:01 +07:00
|
|
|
}
|
|
|
|
|
2019-04-10 19:32:40 +07:00
|
|
|
static inline void qdisc_update_stats_at_dequeue(struct Qdisc *sch,
|
|
|
|
struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
if (qdisc_is_percpu_stats(sch)) {
|
|
|
|
qdisc_qstats_cpu_backlog_dec(sch, skb);
|
|
|
|
qdisc_bstats_cpu_update(sch, skb);
|
2019-04-10 19:32:41 +07:00
|
|
|
qdisc_qstats_cpu_qlen_dec(sch);
|
2019-04-10 19:32:40 +07:00
|
|
|
} else {
|
|
|
|
qdisc_qstats_backlog_dec(sch, skb);
|
|
|
|
qdisc_bstats_update(sch, skb);
|
|
|
|
sch->q.qlen--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void qdisc_update_stats_at_enqueue(struct Qdisc *sch,
|
|
|
|
unsigned int pkt_len)
|
|
|
|
{
|
|
|
|
if (qdisc_is_percpu_stats(sch)) {
|
2019-04-10 19:32:41 +07:00
|
|
|
qdisc_qstats_cpu_qlen_inc(sch);
|
2019-04-10 19:32:40 +07:00
|
|
|
this_cpu_add(sch->cpu_qstats->backlog, pkt_len);
|
|
|
|
} else {
|
|
|
|
sch->qstats.backlog += pkt_len;
|
|
|
|
sch->q.qlen++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-10-31 14:47:01 +07:00
|
|
|
/* use instead of qdisc->dequeue() for all qdiscs queried with ->peek() */
|
|
|
|
static inline struct sk_buff *qdisc_dequeue_peeked(struct Qdisc *sch)
|
|
|
|
{
|
2017-12-08 00:55:45 +07:00
|
|
|
struct sk_buff *skb = skb_peek(&sch->gso_skb);
|
2008-10-31 14:47:01 +07:00
|
|
|
|
2008-11-06 07:02:34 +07:00
|
|
|
if (skb) {
|
2017-12-08 00:55:45 +07:00
|
|
|
skb = __skb_dequeue(&sch->gso_skb);
|
2019-04-10 19:32:39 +07:00
|
|
|
if (qdisc_is_percpu_stats(sch)) {
|
|
|
|
qdisc_qstats_cpu_backlog_dec(sch, skb);
|
2019-04-10 19:32:41 +07:00
|
|
|
qdisc_qstats_cpu_qlen_dec(sch);
|
2019-04-10 19:32:39 +07:00
|
|
|
} else {
|
|
|
|
qdisc_qstats_backlog_dec(sch, skb);
|
|
|
|
sch->q.qlen--;
|
|
|
|
}
|
2008-11-06 07:02:34 +07:00
|
|
|
} else {
|
2008-10-31 14:47:01 +07:00
|
|
|
skb = sch->dequeue(sch);
|
2008-11-06 07:02:34 +07:00
|
|
|
}
|
2008-10-31 14:47:01 +07:00
|
|
|
|
|
|
|
return skb;
|
|
|
|
}
|
|
|
|
|
2016-09-18 05:57:34 +07:00
|
|
|
static inline void __qdisc_reset_queue(struct qdisc_skb_head *qh)
|
2005-06-19 12:57:26 +07:00
|
|
|
{
|
|
|
|
/*
|
|
|
|
* We do not know the backlog in bytes of this list, it
|
|
|
|
* is up to the caller to correct it
|
|
|
|
*/
|
2016-09-18 05:57:34 +07:00
|
|
|
ASSERT_RTNL();
|
|
|
|
if (qh->qlen) {
|
|
|
|
rtnl_kfree_skbs(qh->head, qh->tail);
|
|
|
|
|
|
|
|
qh->head = NULL;
|
|
|
|
qh->tail = NULL;
|
|
|
|
qh->qlen = 0;
|
2016-06-14 10:21:50 +07:00
|
|
|
}
|
2005-06-19 12:57:26 +07:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void qdisc_reset_queue(struct Qdisc *sch)
|
|
|
|
{
|
2016-06-14 10:21:50 +07:00
|
|
|
__qdisc_reset_queue(&sch->q);
|
2005-06-19 12:57:26 +07:00
|
|
|
sch->qstats.backlog = 0;
|
|
|
|
}
|
|
|
|
|
2016-02-26 05:55:00 +07:00
|
|
|
static inline struct Qdisc *qdisc_replace(struct Qdisc *sch, struct Qdisc *new,
|
|
|
|
struct Qdisc **pold)
|
|
|
|
{
|
|
|
|
struct Qdisc *old;
|
|
|
|
|
|
|
|
sch_tree_lock(sch);
|
|
|
|
old = *pold;
|
|
|
|
*pold = new;
|
2019-03-28 22:53:13 +07:00
|
|
|
if (old != NULL)
|
|
|
|
qdisc_tree_flush_backlog(old);
|
2016-02-26 05:55:00 +07:00
|
|
|
sch_tree_unlock(sch);
|
|
|
|
|
|
|
|
return old;
|
|
|
|
}
|
|
|
|
|
2016-06-14 10:21:50 +07:00
|
|
|
static inline void rtnl_qdisc_drop(struct sk_buff *skb, struct Qdisc *sch)
|
|
|
|
{
|
|
|
|
rtnl_kfree_skbs(skb, skb);
|
|
|
|
qdisc_qstats_drop(sch);
|
|
|
|
}
|
|
|
|
|
2017-12-08 00:55:07 +07:00
|
|
|
static inline int qdisc_drop_cpu(struct sk_buff *skb, struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free)
|
|
|
|
{
|
|
|
|
__qdisc_drop(skb, to_free);
|
|
|
|
qdisc_qstats_cpu_drop(sch);
|
|
|
|
|
|
|
|
return NET_XMIT_DROP;
|
|
|
|
}
|
2016-06-22 13:16:49 +07:00
|
|
|
|
|
|
|
static inline int qdisc_drop(struct sk_buff *skb, struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free)
|
2005-06-19 12:57:26 +07:00
|
|
|
{
|
2016-06-22 13:16:49 +07:00
|
|
|
__qdisc_drop(skb, to_free);
|
2014-09-29 01:53:29 +07:00
|
|
|
qdisc_qstats_drop(sch);
|
2005-06-19 12:57:26 +07:00
|
|
|
|
|
|
|
return NET_XMIT_DROP;
|
|
|
|
}
|
|
|
|
|
2018-03-06 00:52:54 +07:00
|
|
|
static inline int qdisc_drop_all(struct sk_buff *skb, struct Qdisc *sch,
|
|
|
|
struct sk_buff **to_free)
|
|
|
|
{
|
|
|
|
__qdisc_drop_all(skb, to_free);
|
|
|
|
qdisc_qstats_drop(sch);
|
|
|
|
|
|
|
|
return NET_XMIT_DROP;
|
|
|
|
}
|
|
|
|
|
2007-09-12 21:35:24 +07:00
|
|
|
/* Length to Time (L2T) lookup in a qdisc_rate_table, to determine how
|
|
|
|
long it will take to send a packet given its size.
|
|
|
|
*/
|
|
|
|
static inline u32 qdisc_l2t(struct qdisc_rate_table* rtab, unsigned int pktlen)
|
|
|
|
{
|
2007-09-12 21:36:28 +07:00
|
|
|
int slot = pktlen + rtab->rate.cell_align + rtab->rate.overhead;
|
|
|
|
if (slot < 0)
|
|
|
|
slot = 0;
|
2007-09-12 21:35:24 +07:00
|
|
|
slot >>= rtab->rate.cell_log;
|
|
|
|
if (slot > 255)
|
2010-09-23 03:43:57 +07:00
|
|
|
return rtab->data[255]*(slot >> 8) + rtab->data[slot & 0xFF];
|
2007-09-12 21:35:24 +07:00
|
|
|
return rtab->data[slot];
|
|
|
|
}
|
|
|
|
|
2013-02-12 07:12:03 +07:00
|
|
|
struct psched_ratecfg {
|
2013-06-07 03:56:19 +07:00
|
|
|
u64 rate_bytes_ps; /* bytes per second */
|
2013-06-02 20:55:05 +07:00
|
|
|
u32 mult;
|
|
|
|
u16 overhead;
|
2013-08-15 04:47:11 +07:00
|
|
|
u8 linklayer;
|
2013-06-02 20:55:05 +07:00
|
|
|
u8 shift;
|
2013-02-12 07:12:03 +07:00
|
|
|
};
|
|
|
|
|
|
|
|
static inline u64 psched_l2t_ns(const struct psched_ratecfg *r,
|
|
|
|
unsigned int len)
|
|
|
|
{
|
2013-08-15 04:47:11 +07:00
|
|
|
len += r->overhead;
|
|
|
|
|
|
|
|
if (unlikely(r->linklayer == TC_LINKLAYER_ATM))
|
|
|
|
return ((u64)(DIV_ROUND_UP(len,48)*53) * r->mult) >> r->shift;
|
|
|
|
|
|
|
|
return ((u64)len * r->mult) >> r->shift;
|
2013-02-12 07:12:03 +07:00
|
|
|
}
|
|
|
|
|
2013-07-31 12:47:13 +07:00
|
|
|
void psched_ratecfg_precompute(struct psched_ratecfg *r,
|
2013-09-19 23:10:03 +07:00
|
|
|
const struct tc_ratespec *conf,
|
|
|
|
u64 rate64);
|
2013-02-12 07:12:03 +07:00
|
|
|
|
2013-06-02 20:55:05 +07:00
|
|
|
static inline void psched_ratecfg_getrate(struct tc_ratespec *res,
|
|
|
|
const struct psched_ratecfg *r)
|
2013-02-12 07:12:03 +07:00
|
|
|
{
|
2013-06-02 20:55:05 +07:00
|
|
|
memset(res, 0, sizeof(*res));
|
2013-09-19 23:10:03 +07:00
|
|
|
|
|
|
|
/* legacy struct tc_ratespec has a 32bit @rate field
|
|
|
|
* Qdisc using 64bit rate should add new attributes
|
|
|
|
* in order to maintain compatibility.
|
|
|
|
*/
|
|
|
|
res->rate = min_t(u64, r->rate_bytes_ps, ~0U);
|
|
|
|
|
2013-06-02 20:55:05 +07:00
|
|
|
res->overhead = r->overhead;
|
2013-08-15 04:47:11 +07:00
|
|
|
res->linklayer = (r->linklayer & TC_LINKLAYER_MASK);
|
2013-02-12 07:12:03 +07:00
|
|
|
}
|
|
|
|
|
2017-11-03 17:46:25 +07:00
|
|
|
/* Mini Qdisc serves for specific needs of ingress/clsact Qdisc.
|
|
|
|
* The fast path only needs to access filter list and to update stats
|
|
|
|
*/
|
|
|
|
struct mini_Qdisc {
|
|
|
|
struct tcf_proto *filter_list;
|
2020-02-16 17:01:22 +07:00
|
|
|
struct tcf_block *block;
|
2017-11-03 17:46:25 +07:00
|
|
|
struct gnet_stats_basic_cpu __percpu *cpu_bstats;
|
|
|
|
struct gnet_stats_queue __percpu *cpu_qstats;
|
|
|
|
struct rcu_head rcu;
|
|
|
|
};
|
|
|
|
|
|
|
|
static inline void mini_qdisc_bstats_cpu_update(struct mini_Qdisc *miniq,
|
|
|
|
const struct sk_buff *skb)
|
|
|
|
{
|
|
|
|
bstats_cpu_update(this_cpu_ptr(miniq->cpu_bstats), skb);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void mini_qdisc_qstats_cpu_drop(struct mini_Qdisc *miniq)
|
|
|
|
{
|
|
|
|
this_cpu_inc(miniq->cpu_qstats->drops);
|
|
|
|
}
|
|
|
|
|
|
|
|
struct mini_Qdisc_pair {
|
|
|
|
struct mini_Qdisc miniq1;
|
|
|
|
struct mini_Qdisc miniq2;
|
|
|
|
struct mini_Qdisc __rcu **p_miniq;
|
|
|
|
};
|
|
|
|
|
|
|
|
void mini_qdisc_pair_swap(struct mini_Qdisc_pair *miniqp,
|
|
|
|
struct tcf_proto *tp_head);
|
|
|
|
void mini_qdisc_pair_init(struct mini_Qdisc_pair *miniqp, struct Qdisc *qdisc,
|
|
|
|
struct mini_Qdisc __rcu **p_miniq);
|
2020-02-16 17:01:22 +07:00
|
|
|
void mini_qdisc_pair_block_init(struct mini_Qdisc_pair *miniqp,
|
|
|
|
struct tcf_block *block);
|
2017-11-03 17:46:25 +07:00
|
|
|
|
2019-10-30 21:09:03 +07:00
|
|
|
static inline int skb_tc_reinsert(struct sk_buff *skb, struct tcf_result *res)
|
2018-07-30 19:30:44 +07:00
|
|
|
{
|
2019-10-30 21:09:03 +07:00
|
|
|
return res->ingress ? netif_receive_skb(skb) : dev_queue_xmit(skb);
|
2018-07-30 19:30:44 +07:00
|
|
|
}
|
|
|
|
|
2005-04-17 05:20:36 +07:00
|
|
|
#endif
|