lkml.org 
[lkml]   [2021]   [Jul]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v12 15/17] counter: Implement events_queue_size sysfs attribute
On Tue, Jul 06, 2021 at 02:40:13PM +0300, Jarkko Nikula wrote:
> Hi
>
> On 7/5/21 11:19 AM, William Breathitt Gray wrote:
> > The events_queue_size sysfs attribute provides a way for users to
> > dynamically configure the Counter events queue size for the Counter
> > character device interface. The size is in number of struct
> > counter_event data structures. The number of elements will be rounded-up
> > to a power of 2 due to a requirement of the kfifo_alloc function called
> > during reallocation of the queue.
> >
> ...
> > diff --git a/drivers/counter/counter-chrdev.c b/drivers/counter/counter-chrdev.c
> > index 92805b1f65b8..13644c87d02a 100644
> > --- a/drivers/counter/counter-chrdev.c
> > +++ b/drivers/counter/counter-chrdev.c
> > @@ -323,6 +323,9 @@ static int counter_chrdev_open(struct inode *inode, struct file *filp)
> > typeof(*counter),
> > chrdev);
> >
> > + if (!mutex_trylock(&counter->chrdev_lock))
> > + return -EBUSY;
> > +
> > get_device(&counter->dev);
> > filp->private_data = counter;
> >
> > @@ -339,6 +342,7 @@ static int counter_chrdev_release(struct inode *inode, struct file *filp)
> > return err;
> >
> > put_device(&counter->dev);
> > + mutex_unlock(&counter->chrdev_lock);
> >
> > return 0;
> > }
>
> I got two separate mutex warnings from counter_chrdev_open() by doing
> blind "cat /dev/counter0". First one due mutex being uninitialized:
>
> [ 441.057342] DEBUG_LOCKS_WARN_ON(lock->magic != lock)
> [ 441.057355] WARNING: CPU: 2 PID: 366 at kernel/locking/mutex.c:1416
> mutex_trylock+0xf2/0x130
> ...
> [ 441.217331] Call Trace:
> [ 441.220062] counter_chrdev_open+0x21/0x60 [counter]
> ...
>
> which I fixed trivially by (please be free to use it)
>
> --- a/drivers/counter/counter-chrdev.c
> +++ b/drivers/counter/counter-chrdev.c
> @@ -364,6 +364,7 @@ int counter_chrdev_add(struct counter_device *const
> counter)
> spin_lock_init(&counter->events_list_lock);
> init_waitqueue_head(&counter->events_wait);
> mutex_init(&counter->events_lock);
> + mutex_init(&counter->chrdev_lock);
>
> /* Initialize character device */
> cdev_init(&counter->chrdev, &counter_fops);

Thanks, I'll add this line in.

> and after that
>
> [ 16.564403] ================================================
> [ 16.570725] WARNING: lock held when returning to user space!
> [ 16.577044] 5.13.0-next-20210706+ #4 Not tainted
> [ 16.582198] ------------------------------------------------
> [ 16.588507] cat/331 is leaving the kernel with locks still held!
> [ 16.595214] 1 lock held by cat/331:
> [ 16.599103] #0: ffff888102bb3630
> (&counter->chrdev_lock){+.+.}-{3:3}, at: counter_chrdev_open+0x21/0x60
> [counter]
>
> Jarkko

I'm not sure how to resolve this warning. The purpose of this lock is to
limit chrdev to a single open at a time. To accomplish this I grab this
lock in counter_chrdev_open() and hold it until counter_chrdev_release()
is called. Is there a better way to accomplish this?

William Breathitt Gray
[unhandled content-type:application/pgp-signature]
\
 
 \ /
  Last update: 2021-07-10 12:25    [W:0.145 / U:0.252 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site