Re: [PATCH 4/6] sysfs: use sysfs_lock to protect the sysfs_dirent tree

[Cornelia Huck <cornelia.huck@de.ibm.com>]

On Tue, 29 May 2007 01:36:27 +0900,
Tejun Heo <htejun@gmail.com> wrote:

> @@ -795,6 +822,8 @@ static int sysfs_readdir(struct file * f
>  			i++;
>  			/* fallthrough */
>  		default:
> +			spin_lock(&sysfs_lock);
> +
>  			pos = &parent_sd->s_children;
>  			while (*pos != cursor)
>  				pos = &(*pos)->s_sibling;
> @@ -827,6 +856,8 @@ static int sysfs_readdir(struct file * f
>  			/* put cursor back in */
>  			cursor->s_sibling = *pos;
>  			*pos = cursor;
> +
> +			spin_unlock(&sysfs_lock);
>  	}
>  	return 0;
>  }

Here's the cause of the "sleeping function called" I saw. filldir() is
called under sysfs_lock, but calls copy_to_user()... This means you
can't use sysfs_lock for protection here.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/