2005-09-11 19:15:07 -07:00
|
|
|
/*
|
|
|
|
* connector.c
|
connector: create connector workqueue only while needed once
The netlink connector uses its own workqueue to relay the datas sent
from userspace to the appropriate callback. If you launch the test
from Documentation/connector and change it a bit to send a high flow
of data, you will see thousands of events coming to the "cqueue"
workqueue by looking at the workqueue tracer.
This flow of events can be sent very quickly. So, to not encumber the
kevent workqueue and delay other jobs, the "cqueue" workqueue should
remain.
But this workqueue is pointless most of the time, it will always be
created (assuming you have built it of course) although only
developpers with specific needs will use it.
So avoid this "most of the time useless task", this patch proposes to
create this workqueue only when needed once. The first jobs to be
sent to connector callbacks will be sent to kevent while the "cqueue"
thread creation will be scheduled to kevent too.
The following jobs will continue to be scheduled to keventd until the
cqueue workqueue is created, and then the rest of the jobs will
continue to perform as usual, through this dedicated workqueue.
Each time I tested this patch, only the first event was sent to
keventd, the rest has been sent to cqueue which have been created
quickly.
Also, this patch fixes some trailing whitespaces on the connector files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 00:22:04 -07:00
|
|
|
*
|
2009-07-21 12:43:51 -07:00
|
|
|
* 2004+ Copyright (c) Evgeniy Polyakov <zbr@ioremap.net>
|
2005-09-11 19:15:07 -07:00
|
|
|
* All rights reserved.
|
connector: create connector workqueue only while needed once
The netlink connector uses its own workqueue to relay the datas sent
from userspace to the appropriate callback. If you launch the test
from Documentation/connector and change it a bit to send a high flow
of data, you will see thousands of events coming to the "cqueue"
workqueue by looking at the workqueue tracer.
This flow of events can be sent very quickly. So, to not encumber the
kevent workqueue and delay other jobs, the "cqueue" workqueue should
remain.
But this workqueue is pointless most of the time, it will always be
created (assuming you have built it of course) although only
developpers with specific needs will use it.
So avoid this "most of the time useless task", this patch proposes to
create this workqueue only when needed once. The first jobs to be
sent to connector callbacks will be sent to kevent while the "cqueue"
thread creation will be scheduled to kevent too.
The following jobs will continue to be scheduled to keventd until the
cqueue workqueue is created, and then the rest of the jobs will
continue to perform as usual, through this dedicated workqueue.
Each time I tested this patch, only the first event was sent to
keventd, the rest has been sent to cqueue which have been created
quickly.
Also, this patch fixes some trailing whitespaces on the connector files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 00:22:04 -07:00
|
|
|
*
|
2005-09-11 19:15:07 -07:00
|
|
|
* 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.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License
|
|
|
|
* along with this program; if not, write to the Free Software
|
|
|
|
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/list.h>
|
|
|
|
#include <linux/skbuff.h>
|
|
|
|
#include <linux/netlink.h>
|
|
|
|
#include <linux/moduleparam.h>
|
|
|
|
#include <linux/connector.h>
|
2006-03-23 04:00:21 -07:00
|
|
|
#include <linux/mutex.h>
|
2008-06-27 20:03:24 -07:00
|
|
|
#include <linux/proc_fs.h>
|
|
|
|
#include <linux/spinlock.h>
|
2005-09-11 19:15:07 -07:00
|
|
|
|
|
|
|
#include <net/sock.h>
|
|
|
|
|
|
|
|
MODULE_LICENSE("GPL");
|
2009-07-21 12:43:51 -07:00
|
|
|
MODULE_AUTHOR("Evgeniy Polyakov <zbr@ioremap.net>");
|
2005-09-11 19:15:07 -07:00
|
|
|
MODULE_DESCRIPTION("Generic userspace <-> kernelspace connector.");
|
|
|
|
|
|
|
|
static struct cn_dev cdev;
|
|
|
|
|
2008-02-26 19:25:53 -07:00
|
|
|
static int cn_already_initialized;
|
2005-09-11 19:15:07 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* msg->seq and msg->ack are used to determine message genealogy.
|
|
|
|
* When someone sends message it puts there locally unique sequence
|
|
|
|
* and random acknowledge numbers. Sequence number may be copied into
|
|
|
|
* nlmsghdr->nlmsg_seq too.
|
|
|
|
*
|
|
|
|
* Sequence number is incremented with each message to be sent.
|
|
|
|
*
|
|
|
|
* If we expect reply to our message then the sequence number in
|
|
|
|
* received message MUST be the same as in original message, and
|
|
|
|
* acknowledge number MUST be the same + 1.
|
|
|
|
*
|
|
|
|
* If we receive a message and its sequence number is not equal to the
|
|
|
|
* one we are expecting then it is a new message.
|
|
|
|
*
|
|
|
|
* If we receive a message and its sequence number is the same as one
|
|
|
|
* we are expecting but it's acknowledgement number is not equal to
|
|
|
|
* the acknowledgement number in the original message + 1, then it is
|
|
|
|
* a new message.
|
|
|
|
*
|
|
|
|
*/
|
2005-10-06 23:46:04 -07:00
|
|
|
int cn_netlink_send(struct cn_msg *msg, u32 __group, gfp_t gfp_mask)
|
2005-09-11 19:15:07 -07:00
|
|
|
{
|
|
|
|
struct cn_callback_entry *__cbq;
|
|
|
|
unsigned int size;
|
|
|
|
struct sk_buff *skb;
|
|
|
|
struct nlmsghdr *nlh;
|
|
|
|
struct cn_msg *data;
|
|
|
|
struct cn_dev *dev = &cdev;
|
|
|
|
u32 group = 0;
|
|
|
|
int found = 0;
|
|
|
|
|
|
|
|
if (!__group) {
|
|
|
|
spin_lock_bh(&dev->cbdev->queue_lock);
|
|
|
|
list_for_each_entry(__cbq, &dev->cbdev->queue_list,
|
|
|
|
callback_entry) {
|
2005-09-26 15:06:50 -07:00
|
|
|
if (cn_cb_equal(&__cbq->id.id, &msg->id)) {
|
2005-09-11 19:15:07 -07:00
|
|
|
found = 1;
|
|
|
|
group = __cbq->group;
|
2008-01-04 02:54:38 -07:00
|
|
|
break;
|
2005-09-11 19:15:07 -07:00
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_bh(&dev->cbdev->queue_lock);
|
|
|
|
|
|
|
|
if (!found)
|
|
|
|
return -ENODEV;
|
|
|
|
} else {
|
|
|
|
group = __group;
|
|
|
|
}
|
|
|
|
|
2006-03-20 23:21:40 -07:00
|
|
|
if (!netlink_has_listeners(dev->nls, group))
|
|
|
|
return -ESRCH;
|
|
|
|
|
2005-09-11 19:15:07 -07:00
|
|
|
size = NLMSG_SPACE(sizeof(*msg) + msg->len);
|
|
|
|
|
|
|
|
skb = alloc_skb(size, gfp_mask);
|
|
|
|
if (!skb)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
nlh = NLMSG_PUT(skb, 0, msg->seq, NLMSG_DONE, size - sizeof(*nlh));
|
|
|
|
|
|
|
|
data = NLMSG_DATA(nlh);
|
|
|
|
|
|
|
|
memcpy(data, msg, sizeof(*data) + msg->len);
|
|
|
|
|
|
|
|
NETLINK_CB(skb).dst_group = group;
|
|
|
|
|
2006-03-20 23:21:40 -07:00
|
|
|
return netlink_broadcast(dev->nls, skb, 0, group, gfp_mask);
|
2005-09-11 19:15:07 -07:00
|
|
|
|
|
|
|
nlmsg_failure:
|
|
|
|
kfree_skb(skb);
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
2006-06-23 02:05:46 -07:00
|
|
|
EXPORT_SYMBOL_GPL(cn_netlink_send);
|
2005-09-11 19:15:07 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Callback helper - queues work and setup destructor for given data.
|
|
|
|
*/
|
2009-10-01 19:40:07 -07:00
|
|
|
static int cn_call_callback(struct sk_buff *skb)
|
2005-09-11 19:15:07 -07:00
|
|
|
{
|
2007-03-07 13:55:39 -07:00
|
|
|
struct cn_callback_entry *__cbq, *__new_cbq;
|
2005-09-11 19:15:07 -07:00
|
|
|
struct cn_dev *dev = &cdev;
|
2009-10-01 19:40:04 -07:00
|
|
|
struct cn_msg *msg = NLMSG_DATA(nlmsg_hdr(skb));
|
2005-09-26 15:06:50 -07:00
|
|
|
int err = -ENODEV;
|
2005-09-11 19:15:07 -07:00
|
|
|
|
|
|
|
spin_lock_bh(&dev->cbdev->queue_lock);
|
|
|
|
list_for_each_entry(__cbq, &dev->cbdev->queue_list, callback_entry) {
|
2005-09-26 15:06:50 -07:00
|
|
|
if (cn_cb_equal(&__cbq->id.id, &msg->id)) {
|
2006-12-18 02:53:58 -07:00
|
|
|
if (likely(!work_pending(&__cbq->work) &&
|
2009-10-01 19:40:07 -07:00
|
|
|
__cbq->data.skb == NULL)) {
|
2009-10-01 19:40:04 -07:00
|
|
|
__cbq->data.skb = skb;
|
2005-09-11 19:15:07 -07:00
|
|
|
|
connector: create connector workqueue only while needed once
The netlink connector uses its own workqueue to relay the datas sent
from userspace to the appropriate callback. If you launch the test
from Documentation/connector and change it a bit to send a high flow
of data, you will see thousands of events coming to the "cqueue"
workqueue by looking at the workqueue tracer.
This flow of events can be sent very quickly. So, to not encumber the
kevent workqueue and delay other jobs, the "cqueue" workqueue should
remain.
But this workqueue is pointless most of the time, it will always be
created (assuming you have built it of course) although only
developpers with specific needs will use it.
So avoid this "most of the time useless task", this patch proposes to
create this workqueue only when needed once. The first jobs to be
sent to connector callbacks will be sent to kevent while the "cqueue"
thread creation will be scheduled to kevent too.
The following jobs will continue to be scheduled to keventd until the
cqueue workqueue is created, and then the rest of the jobs will
continue to perform as usual, through this dedicated workqueue.
Each time I tested this patch, only the first event was sent to
keventd, the rest has been sent to cqueue which have been created
quickly.
Also, this patch fixes some trailing whitespaces on the connector files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 00:22:04 -07:00
|
|
|
if (queue_cn_work(__cbq, &__cbq->work))
|
2005-09-26 15:06:50 -07:00
|
|
|
err = 0;
|
2008-01-04 02:52:02 -07:00
|
|
|
else
|
|
|
|
err = -EINVAL;
|
2005-09-11 19:15:07 -07:00
|
|
|
} else {
|
2005-09-26 15:06:50 -07:00
|
|
|
struct cn_callback_data *d;
|
connector: create connector workqueue only while needed once
The netlink connector uses its own workqueue to relay the datas sent
from userspace to the appropriate callback. If you launch the test
from Documentation/connector and change it a bit to send a high flow
of data, you will see thousands of events coming to the "cqueue"
workqueue by looking at the workqueue tracer.
This flow of events can be sent very quickly. So, to not encumber the
kevent workqueue and delay other jobs, the "cqueue" workqueue should
remain.
But this workqueue is pointless most of the time, it will always be
created (assuming you have built it of course) although only
developpers with specific needs will use it.
So avoid this "most of the time useless task", this patch proposes to
create this workqueue only when needed once. The first jobs to be
sent to connector callbacks will be sent to kevent while the "cqueue"
thread creation will be scheduled to kevent too.
The following jobs will continue to be scheduled to keventd until the
cqueue workqueue is created, and then the rest of the jobs will
continue to perform as usual, through this dedicated workqueue.
Each time I tested this patch, only the first event was sent to
keventd, the rest has been sent to cqueue which have been created
quickly.
Also, this patch fixes some trailing whitespaces on the connector files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 00:22:04 -07:00
|
|
|
|
2007-03-07 13:55:39 -07:00
|
|
|
err = -ENOMEM;
|
|
|
|
__new_cbq = kzalloc(sizeof(struct cn_callback_entry), GFP_ATOMIC);
|
|
|
|
if (__new_cbq) {
|
|
|
|
d = &__new_cbq->data;
|
2009-10-01 19:40:04 -07:00
|
|
|
d->skb = skb;
|
2005-09-26 15:06:50 -07:00
|
|
|
d->callback = __cbq->data.callback;
|
2007-03-07 13:55:39 -07:00
|
|
|
d->free = __new_cbq;
|
2005-09-26 15:06:50 -07:00
|
|
|
|
connector: create connector workqueue only while needed once
The netlink connector uses its own workqueue to relay the datas sent
from userspace to the appropriate callback. If you launch the test
from Documentation/connector and change it a bit to send a high flow
of data, you will see thousands of events coming to the "cqueue"
workqueue by looking at the workqueue tracer.
This flow of events can be sent very quickly. So, to not encumber the
kevent workqueue and delay other jobs, the "cqueue" workqueue should
remain.
But this workqueue is pointless most of the time, it will always be
created (assuming you have built it of course) although only
developpers with specific needs will use it.
So avoid this "most of the time useless task", this patch proposes to
create this workqueue only when needed once. The first jobs to be
sent to connector callbacks will be sent to kevent while the "cqueue"
thread creation will be scheduled to kevent too.
The following jobs will continue to be scheduled to keventd until the
cqueue workqueue is created, and then the rest of the jobs will
continue to perform as usual, through this dedicated workqueue.
Each time I tested this patch, only the first event was sent to
keventd, the rest has been sent to cqueue which have been created
quickly.
Also, this patch fixes some trailing whitespaces on the connector files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 00:22:04 -07:00
|
|
|
__new_cbq->pdev = __cbq->pdev;
|
|
|
|
|
2007-03-07 13:55:39 -07:00
|
|
|
INIT_WORK(&__new_cbq->work,
|
2006-12-18 02:53:58 -07:00
|
|
|
&cn_queue_wrapper);
|
2007-03-07 13:55:39 -07:00
|
|
|
|
connector: create connector workqueue only while needed once
The netlink connector uses its own workqueue to relay the datas sent
from userspace to the appropriate callback. If you launch the test
from Documentation/connector and change it a bit to send a high flow
of data, you will see thousands of events coming to the "cqueue"
workqueue by looking at the workqueue tracer.
This flow of events can be sent very quickly. So, to not encumber the
kevent workqueue and delay other jobs, the "cqueue" workqueue should
remain.
But this workqueue is pointless most of the time, it will always be
created (assuming you have built it of course) although only
developpers with specific needs will use it.
So avoid this "most of the time useless task", this patch proposes to
create this workqueue only when needed once. The first jobs to be
sent to connector callbacks will be sent to kevent while the "cqueue"
thread creation will be scheduled to kevent too.
The following jobs will continue to be scheduled to keventd until the
cqueue workqueue is created, and then the rest of the jobs will
continue to perform as usual, through this dedicated workqueue.
Each time I tested this patch, only the first event was sent to
keventd, the rest has been sent to cqueue which have been created
quickly.
Also, this patch fixes some trailing whitespaces on the connector files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 00:22:04 -07:00
|
|
|
if (queue_cn_work(__new_cbq,
|
2007-03-07 13:55:39 -07:00
|
|
|
&__new_cbq->work))
|
2005-09-26 15:06:50 -07:00
|
|
|
err = 0;
|
|
|
|
else {
|
2007-03-07 13:55:39 -07:00
|
|
|
kfree(__new_cbq);
|
2005-09-26 15:06:50 -07:00
|
|
|
err = -EINVAL;
|
|
|
|
}
|
2007-03-07 13:55:39 -07:00
|
|
|
}
|
2005-09-11 19:15:07 -07:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
spin_unlock_bh(&dev->cbdev->queue_lock);
|
|
|
|
|
2005-09-26 15:06:50 -07:00
|
|
|
return err;
|
2005-09-11 19:15:07 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Main netlink receiving function.
|
|
|
|
*
|
2008-01-04 02:55:01 -07:00
|
|
|
* It checks skb, netlink header and msg sizes, and calls callback helper.
|
2005-09-11 19:15:07 -07:00
|
|
|
*/
|
|
|
|
static void cn_rx_skb(struct sk_buff *__skb)
|
|
|
|
{
|
|
|
|
struct nlmsghdr *nlh;
|
|
|
|
int err;
|
|
|
|
struct sk_buff *skb;
|
|
|
|
|
|
|
|
skb = skb_get(__skb);
|
|
|
|
|
|
|
|
if (skb->len >= NLMSG_SPACE(0)) {
|
2007-04-25 19:08:35 -07:00
|
|
|
nlh = nlmsg_hdr(skb);
|
2005-09-11 19:15:07 -07:00
|
|
|
|
|
|
|
if (nlh->nlmsg_len < sizeof(struct cn_msg) ||
|
|
|
|
skb->len < nlh->nlmsg_len ||
|
|
|
|
nlh->nlmsg_len > CONNECTOR_MAX_MSG_SIZE) {
|
|
|
|
kfree_skb(skb);
|
2007-10-30 20:41:49 -07:00
|
|
|
return;
|
2005-09-11 19:15:07 -07:00
|
|
|
}
|
|
|
|
|
2009-10-01 19:40:07 -07:00
|
|
|
err = cn_call_callback(skb);
|
2005-09-11 19:15:07 -07:00
|
|
|
if (err < 0)
|
|
|
|
kfree_skb(skb);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Callback add routing - adds callback with given ID and name.
|
|
|
|
* If there is registered callback with the same ID it will not be added.
|
|
|
|
*
|
|
|
|
* May sleep.
|
|
|
|
*/
|
2009-07-17 10:13:21 -07:00
|
|
|
int cn_add_callback(struct cb_id *id, char *name,
|
2009-10-01 19:40:05 -07:00
|
|
|
void (*callback)(struct cn_msg *, struct netlink_skb_parms *))
|
2005-09-11 19:15:07 -07:00
|
|
|
{
|
|
|
|
int err;
|
|
|
|
struct cn_dev *dev = &cdev;
|
|
|
|
|
2006-06-19 23:42:53 -07:00
|
|
|
if (!cn_already_initialized)
|
|
|
|
return -EAGAIN;
|
|
|
|
|
2005-09-26 15:06:50 -07:00
|
|
|
err = cn_queue_add_callback(dev->cbdev, name, id, callback);
|
|
|
|
if (err)
|
2005-09-11 19:15:07 -07:00
|
|
|
return err;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
2006-06-23 02:05:46 -07:00
|
|
|
EXPORT_SYMBOL_GPL(cn_add_callback);
|
2005-09-11 19:15:07 -07:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Callback remove routing - removes callback
|
|
|
|
* with given ID.
|
|
|
|
* If there is no registered callback with given
|
|
|
|
* ID nothing happens.
|
|
|
|
*
|
|
|
|
* May sleep while waiting for reference counter to become zero.
|
|
|
|
*/
|
|
|
|
void cn_del_callback(struct cb_id *id)
|
|
|
|
{
|
|
|
|
struct cn_dev *dev = &cdev;
|
|
|
|
|
|
|
|
cn_queue_del_callback(dev->cbdev, id);
|
|
|
|
}
|
2006-06-23 02:05:46 -07:00
|
|
|
EXPORT_SYMBOL_GPL(cn_del_callback);
|
2005-09-11 19:15:07 -07:00
|
|
|
|
2008-06-27 20:03:24 -07:00
|
|
|
static int cn_proc_show(struct seq_file *m, void *v)
|
|
|
|
{
|
|
|
|
struct cn_queue_dev *dev = cdev.cbdev;
|
|
|
|
struct cn_callback_entry *cbq;
|
|
|
|
|
|
|
|
seq_printf(m, "Name ID\n");
|
|
|
|
|
|
|
|
spin_lock_bh(&dev->queue_lock);
|
|
|
|
|
|
|
|
list_for_each_entry(cbq, &dev->queue_list, callback_entry) {
|
|
|
|
seq_printf(m, "%-15s %u:%u\n",
|
|
|
|
cbq->id.name,
|
|
|
|
cbq->id.id.idx,
|
|
|
|
cbq->id.id.val);
|
|
|
|
}
|
|
|
|
|
|
|
|
spin_unlock_bh(&dev->queue_lock);
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int cn_proc_open(struct inode *inode, struct file *file)
|
|
|
|
{
|
|
|
|
return single_open(file, cn_proc_show, NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct file_operations cn_file_ops = {
|
|
|
|
.owner = THIS_MODULE,
|
|
|
|
.open = cn_proc_open,
|
|
|
|
.read = seq_read,
|
|
|
|
.llseek = seq_lseek,
|
|
|
|
.release = single_release
|
|
|
|
};
|
|
|
|
|
2006-06-19 23:42:53 -07:00
|
|
|
static int __devinit cn_init(void)
|
2005-09-11 19:15:07 -07:00
|
|
|
{
|
|
|
|
struct cn_dev *dev = &cdev;
|
|
|
|
|
2007-10-10 21:15:29 -07:00
|
|
|
dev->input = cn_rx_skb;
|
2005-09-11 19:15:07 -07:00
|
|
|
|
2007-09-12 04:05:38 -07:00
|
|
|
dev->nls = netlink_kernel_create(&init_net, NETLINK_CONNECTOR,
|
2005-09-11 19:15:07 -07:00
|
|
|
CN_NETLINK_USERS + 0xf,
|
2007-04-20 14:14:21 -07:00
|
|
|
dev->input, NULL, THIS_MODULE);
|
2005-09-11 19:15:07 -07:00
|
|
|
if (!dev->nls)
|
|
|
|
return -EIO;
|
|
|
|
|
|
|
|
dev->cbdev = cn_queue_alloc_dev("cqueue", dev->nls);
|
|
|
|
if (!dev->cbdev) {
|
2008-01-28 15:41:19 -07:00
|
|
|
netlink_kernel_release(dev->nls);
|
2005-09-11 19:15:07 -07:00
|
|
|
return -EINVAL;
|
|
|
|
}
|
connector: create connector workqueue only while needed once
The netlink connector uses its own workqueue to relay the datas sent
from userspace to the appropriate callback. If you launch the test
from Documentation/connector and change it a bit to send a high flow
of data, you will see thousands of events coming to the "cqueue"
workqueue by looking at the workqueue tracer.
This flow of events can be sent very quickly. So, to not encumber the
kevent workqueue and delay other jobs, the "cqueue" workqueue should
remain.
But this workqueue is pointless most of the time, it will always be
created (assuming you have built it of course) although only
developpers with specific needs will use it.
So avoid this "most of the time useless task", this patch proposes to
create this workqueue only when needed once. The first jobs to be
sent to connector callbacks will be sent to kevent while the "cqueue"
thread creation will be scheduled to kevent too.
The following jobs will continue to be scheduled to keventd until the
cqueue workqueue is created, and then the rest of the jobs will
continue to perform as usual, through this dedicated workqueue.
Each time I tested this patch, only the first event was sent to
keventd, the rest has been sent to cqueue which have been created
quickly.
Also, this patch fixes some trailing whitespaces on the connector files.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
2009-02-03 00:22:04 -07:00
|
|
|
|
2006-06-19 23:42:53 -07:00
|
|
|
cn_already_initialized = 1;
|
2005-09-11 19:15:07 -07:00
|
|
|
|
2008-06-27 20:03:24 -07:00
|
|
|
proc_net_fops_create(&init_net, "connector", S_IRUGO, &cn_file_ops);
|
|
|
|
|
2005-09-11 19:15:07 -07:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2006-06-19 23:42:53 -07:00
|
|
|
static void __devexit cn_fini(void)
|
2005-09-11 19:15:07 -07:00
|
|
|
{
|
|
|
|
struct cn_dev *dev = &cdev;
|
|
|
|
|
|
|
|
cn_already_initialized = 0;
|
|
|
|
|
2008-06-27 20:03:24 -07:00
|
|
|
proc_net_remove(&init_net, "connector");
|
|
|
|
|
2005-09-11 19:15:07 -07:00
|
|
|
cn_queue_free_dev(dev->cbdev);
|
2008-01-28 15:41:19 -07:00
|
|
|
netlink_kernel_release(dev->nls);
|
2005-09-11 19:15:07 -07:00
|
|
|
}
|
|
|
|
|
2006-06-19 23:42:53 -07:00
|
|
|
subsys_initcall(cn_init);
|
2005-09-11 19:15:07 -07:00
|
|
|
module_exit(cn_fini);
|