public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH] hiddev: Switch to unlocked_ioctl
@ 2008-05-22 21:22 Alan Cox
  2008-05-26  8:55 ` Jiri Kosina
  0 siblings, 1 reply; 3+ messages in thread
From: Alan Cox @ 2008-05-22 21:22 UTC (permalink / raw)
  To: linux-kernel, dmitry.torokhov, linux-usb

Push down the BKL. In some cases compat_ioctl already doesn't take the
BKL so we don't either. Some of the locking here seems already dubious
and object lifetimes want documenting

Signed-off-by: Alan Cox <alan@redhat.com>

diff --git a/drivers/hid/usbhid/hiddev.c b/drivers/hid/usbhid/hiddev.c
index 95cc192..fe4f364 100644
--- a/drivers/hid/usbhid/hiddev.c
+++ b/drivers/hid/usbhid/hiddev.c
@@ -406,6 +406,7 @@ static noinline int hiddev_ioctl_usage(struct hiddev *hiddev, unsigned int cmd,
 	uref_multi = kmalloc(sizeof(struct hiddev_usage_ref_multi), GFP_KERNEL);
 	if (!uref_multi)
 		return -ENOMEM;
+	lock_kernel();
 	uref = &uref_multi->uref;
 	if (cmd == HIDIOCGUSAGES || cmd == HIDIOCSUSAGES) {
 		if (copy_from_user(uref_multi, user_arg,
@@ -501,12 +502,15 @@ static noinline int hiddev_ioctl_usage(struct hiddev *hiddev, unsigned int cmd,
 		}
 
 goodreturn:
+		unlock_kernel();
 		kfree(uref_multi);
 		return 0;
 fault:
+		unlock_kernel();
 		kfree(uref_multi);
 		return -EFAULT;
 inval:
+		unlock_kernel();
 		kfree(uref_multi);
 		return -EINVAL;
 	}
@@ -540,7 +544,7 @@ static noinline int hiddev_ioctl_string(struct hiddev *hiddev, unsigned int cmd,
 	return len;
 }
 
-static int hiddev_ioctl(struct inode *inode, struct file *file, unsigned int cmd, unsigned long arg)
+static long hiddev_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 {
 	struct hiddev_list *list = file->private_data;
 	struct hiddev *hiddev = list->hiddev;
@@ -555,7 +559,10 @@ static int hiddev_ioctl(struct inode *inode, struct file *file, unsigned int cmd
 	struct usbhid_device *usbhid = hid->driver_data;
 	void __user *user_arg = (void __user *)arg;
 	int i;
+	
+	/* Called without BKL by compat methods so no BKL taken */
 
+	/* FIXME: Who or what stop this racing with a disconnect ?? */
 	if (!hiddev->exist)
 		return -EIO;
 
@@ -768,7 +775,7 @@ static const struct file_operations hiddev_fops = {
 	.poll =		hiddev_poll,
 	.open =		hiddev_open,
 	.release =	hiddev_release,
-	.ioctl =	hiddev_ioctl,
+	.unlocked_ioctl =	hiddev_ioctl,
 	.fasync =	hiddev_fasync,
 #ifdef CONFIG_COMPAT
 	.compat_ioctl	= hiddev_compat_ioctl,

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] hiddev: Switch to unlocked_ioctl
  2008-05-22 21:22 [PATCH] hiddev: Switch to unlocked_ioctl Alan Cox
@ 2008-05-26  8:55 ` Jiri Kosina
  2008-05-26 10:07   ` Oliver Neukum
  0 siblings, 1 reply; 3+ messages in thread
From: Jiri Kosina @ 2008-05-26  8:55 UTC (permalink / raw)
  To: Alan Cox; +Cc: linux-kernel, Dmitry Torokhov, linux-usb, Oliver Neukum

On Thu, 22 May 2008, Alan Cox wrote:

> Push down the BKL. In some cases compat_ioctl already doesn't take the
> BKL so we don't either. 

Thanks.

> @@ -555,7 +559,10 @@ static int hiddev_ioctl(struct inode *inode, struct file *file, unsigned int cmd
>  	struct usbhid_device *usbhid = hid->driver_data;
>  	void __user *user_arg = (void __user *)arg;
>  	int i;
> +	
> +	/* Called without BKL by compat methods so no BKL taken */
>  
> +	/* FIXME: Who or what stop this racing with a disconnect ?? */
>  	if (!hiddev->exist)
>  		return -EIO;

Oliver (added to CC) -- I vaguely remember you already fixing this race, 
but I don't seem to be able to find the patch anywhere. Do I just remember 
incorrectly?

Thanks,

-- 
Jiri Kosina

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] hiddev: Switch to unlocked_ioctl
  2008-05-26  8:55 ` Jiri Kosina
@ 2008-05-26 10:07   ` Oliver Neukum
  0 siblings, 0 replies; 3+ messages in thread
From: Oliver Neukum @ 2008-05-26 10:07 UTC (permalink / raw)
  To: Jiri Kosina; +Cc: Alan Cox, linux-kernel, Dmitry Torokhov, linux-usb

Am Montag 26 Mai 2008 10:55:09 schrieb Jiri Kosina:
> On Thu, 22 May 2008, Alan Cox wrote:
> 
> > Push down the BKL. In some cases compat_ioctl already doesn't take the
> > BKL so we don't either. 
> 
> Thanks.
> 
> > @@ -555,7 +559,10 @@ static int hiddev_ioctl(struct inode *inode, struct file *file, unsigned int cmd
> >       struct usbhid_device *usbhid = hid->driver_data;
> >       void __user *user_arg = (void __user *)arg;
> >       int i;
> > +     
> > +     /* Called without BKL by compat methods so no BKL taken */
> >  
> > +     /* FIXME: Who or what stop this racing with a disconnect ?? */
> >       if (!hiddev->exist)
> >               return -EIO;
> 
> Oliver (added to CC) -- I vaguely remember you already fixing this race, 
> but I don't seem to be able to find the patch anywhere. Do I just remember 
> incorrectly?

I fixed a race between disconnect and open. Ioctl was not involved.

	Regards
		Oliver

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2008-05-26 10:08 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-05-22 21:22 [PATCH] hiddev: Switch to unlocked_ioctl Alan Cox
2008-05-26  8:55 ` Jiri Kosina
2008-05-26 10:07   ` Oliver Neukum

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox