2008-06-24 01:32:33

by Li Zefan

Subject: [PATCH] CONNECTOR: add a proc entry to list connectors

I got a problem when I wanted to check if the kernel supports process
event connector, and It seems there's no way to do this check.

At best I can check if the kernel supports connector or not, by looking
into /proc/net/netlink, or maybe checking the return value of bind() to
see if it's ENOENT.

So how about add /proc/net/connector to list all supported connectors?
# cat /proc/net/connector
Name ID
connector 4294967295:4294967295
cn_proc 1:1
w1 3:1

Signed-off-by: Li Zefan <[email protected]>
---
drivers/connector/connector.c | 41 +++++++++++++++++++++++++++++++++++++++++
1 files changed, 41 insertions(+), 0 deletions(-)

diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c
index 85e2ba7..08f9983 100644
--- a/drivers/connector/connector.c
+++ b/drivers/connector/connector.c
@@ -27,6 +27,8 @@
#include <linux/moduleparam.h>
#include <linux/connector.h>
#include <linux/mutex.h>
+#include <linux/proc_fs.h>
+#include <linux/spinlock.h>

#include <net/sock.h>

@@ -403,6 +405,41 @@ static void cn_callback(void *data)
mutex_unlock(&notify_lock);
}

+static int cn_proc_show(struct seq_file *m, void *v)
+{
+ struct cn_queue_dev *dev = cdev.cbdev;
+ struct cn_callback_entry *cbq;
+ unsigned long flags;
+
+ seq_printf(m, "Name ID\n");
+
+ spin_lock_irqsave(&dev->queue_lock, flags);
+
+ list_for_each_entry(cbq, &dev->queue_list, callback_entry) {
+ seq_printf(m, "%-9s %u:%u\n",
+ cbq->id.name,
+ cbq->id.id.idx,
+ cbq->id.id.val);
+ }
+
+ spin_unlock_irqrestore(&dev->queue_lock, flags);
+
+ 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 = seq_release
+};
+
static int __devinit cn_init(void)
{
struct cn_dev *dev = &cdev;
@@ -434,6 +471,8 @@ static int __devinit cn_init(void)
return -EINVAL;
}

+ proc_net_fops_create(&init_net, "connector", S_IRUGO, &cn_file_ops);
+
return 0;
}

@@ -443,6 +482,8 @@ static void __devexit cn_fini(void)

cn_already_initialized = 0;

+ proc_net_remove(&init_net, "connector");
+
cn_del_callback(&dev->id);
cn_queue_free_dev(dev->cbdev);
netlink_kernel_release(dev->nls);
--
1.5.4.rc3

2008-06-24 01:46:21

by David Miller

Subject: Re: [PATCH] CONNECTOR: add a proc entry to list connectors

From: Li Zefan <[email protected]>
Date: Tue, 24 Jun 2008 09:30:44 +0800

> I got a problem when I wanted to check if the kernel supports process
> event connector, and It seems there's no way to do this check.

Try to create a connector socket.

2008-06-24 01:59:14

by Li Zefan

Subject: Re: [PATCH] CONNECTOR: add a proc entry to list connectors

David Miller wrote:
> From: Li Zefan <[email protected]>
> Date: Tue, 24 Jun 2008 09:30:44 +0800
>
>> I got a problem when I wanted to check if the kernel supports process
>> event connector, and It seems there's no way to do this check.
>
> Try to create a connector socket.
>

Do you mean socket(PF_NETLINK, SOCK_DGRAM, NETLINK_CONNECTOR); ?

But this has nothing to do with process event connector. If it's not
supported, both socket() and bind() return successfully, just I can
recv() nothing.

2008-06-24 06:50:20

by Evgeniy Polyakov

Subject: Re: [PATCH] CONNECTOR: add a proc entry to list connectors

Hi.

On Tue, Jun 24, 2008 at 09:30:44AM +0800, Li Zefan ([email protected]) wrote:
> I got a problem when I wanted to check if the kernel supports process
> event connector, and It seems there's no way to do this check.
>
> At best I can check if the kernel supports connector or not, by looking
> into /proc/net/netlink, or maybe checking the return value of bind() to
> see if it's ENOENT.
>
> So how about add /proc/net/connector to list all supported connectors?
> # cat /proc/net/connector
> Name ID
> connector 4294967295:4294967295
> cn_proc 1:1
> w1 3:1

I like the idea, but if it is not appropriate way, I can extend
connector core itself so it will support not only add/remove
notification of added/removed users, but will also allow to query if
given user exists.

> Signed-off-by: Li Zefan <[email protected]>

Ack.

> +static int cn_proc_show(struct seq_file *m, void *v)
> +{
> + struct cn_queue_dev *dev = cdev.cbdev;
> + struct cn_callback_entry *cbq;
> + unsigned long flags;
> +
> + seq_printf(m, "Name ID\n");
> +
> + spin_lock_irqsave(&dev->queue_lock, flags);

It can be BH lock only.

--
Evgeniy Polyakov