diff mbox

[v9,1/5] configfs: Allow dynamic group creation

Message ID 1445603558-27693-2-git-send-email-daniel.baluta@intel.com (mailing list archive)
State New, archived
Headers show

Commit Message

Daniel Baluta Oct. 23, 2015, 12:32 p.m. UTC 
We don't want to hardcode default groups at subsystem
creation time. We export:
	* configfs_register_group
	* configfs_unregister_group
to allow drivers to programatically create/destroy groups
later, after module init time.

This is needed for IIO configfs support.

Suggested-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Daniel Baluta <daniel.baluta@intel.com>
---
 fs/configfs/dir.c        | 102 +++++++++++++++++++++++++++++++++++++++++++++++
 include/linux/configfs.h |   8 ++++
 2 files changed, 110 insertions(+)

Comments

Christoph Hellwig Oct. 23, 2015, 12:48 p.m. UTC | #1 
Hi Daniel,

all these functions look great, but can you also move your new
iio_sw_trigger_type_configfs_register/unregister functions to the
core code as those are the wrappers that everyone would have to write,
e.g. something like:

int configfs_register_default_group(struct config_group *parent_group,
		const char *name, struct config_item_type *item_type)
{
	struct config_group *group;

	group = kzalloc(sizeof(*group), GFP_KERNEL);
	if (!group)
		return -ENOMEM;
	config_group_init_type_name(group, name, item_type);

	ret = configfs_register_group(parent_group, group);
	if (ret)
		kfree(group);
	return ret;
}

and the same on the unregister side?

With that:

Reviewed-by: Christoph Hellwig <hch@lst.de>
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Daniel Baluta Oct. 23, 2015, 1:08 p.m. UTC | #2 
On Fri, Oct 23, 2015 at 3:48 PM, Christoph Hellwig <hch@lst.de> wrote:
> Hi Daniel,
>
> all these functions look great, but can you also move your new
> iio_sw_trigger_type_configfs_register/unregister functions to the
> core code as those are the wrappers that everyone would have to write,
> e.g. something like:
>
> int configfs_register_default_group(struct config_group *parent_group,
>                 const char *name, struct config_item_type *item_type)
> {
>         struct config_group *group;
>
>         group = kzalloc(sizeof(*group), GFP_KERNEL);
>         if (!group)
>                 return -ENOMEM;
>         config_group_init_type_name(group, name, item_type);
>
>         ret = configfs_register_group(parent_group, group);
>         if (ret)
>                 kfree(group);
>         return ret;
> }
>
> and the same on the unregister side?
Yes, this is a good idea. Will do!
>
> With that:
>
> Reviewed-by: Christoph Hellwig <hch@lst.de>

thanks,
Daniel.
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Daniel Baluta Oct. 23, 2015, 2:09 p.m. UTC | #3 
Christoph, one question. See below.

On Fri, Oct 23, 2015 at 3:48 PM, Christoph Hellwig <hch@lst.de> wrote:
> Hi Daniel,
>
> all these functions look great, but can you also move your new
> iio_sw_trigger_type_configfs_register/unregister functions to the
> core code as those are the wrappers that everyone would have to write,
> e.g. something like:
>
> int configfs_register_default_group(struct config_group *parent_group,
>                 const char *name, struct config_item_type *item_type)
> {
>         struct config_group *group;
>
>         group = kzalloc(sizeof(*group), GFP_KERNEL);
>         if (!group)
>                 return -ENOMEM;
>         config_group_init_type_name(group, name, item_type);

Until this point this is the implementation of configfs_alloc_group.

We have two options:

1) Remove configfs_alloc_group/configfs_free_group as people will directly
use our new configfs_register_default_group that does the allocation.

2) Keep configfs_alloc/group/configfs_gree_group and use it here.

I'm in favour of 1) variant, as it makes little sense to have a wrapper over
kmalloc + config_group_init_type_name.

thanks,
Daniel.
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c
index c81ce7f..bf76795 100644
--- a/fs/configfs/dir.c
+++ b/fs/configfs/dir.c
@@ -1636,6 +1636,108 @@  const struct file_operations configfs_dir_operations = {
 	.iterate	= configfs_readdir,
 };
 
+/**
+ * configfs_alloc_group() - allocate a configfs group
+ * @name:	group name
+ * @item_type:	config_item associated with the allocated group
+ *
+ * Dynamically allocates and initialize a group specified by name and
+ * config_item_type. We need kzalloc'ed memory because default_group
+ * is empty.
+ *
+ * Return: allocated config group or ERR_PTR() on error
+ */
+struct config_group *
+configfs_alloc_group(const char *name, struct config_item_type *item_type)
+{
+	struct config_group *group;
+
+	group = kzalloc(sizeof(*group), GFP_KERNEL);
+	if (!group)
+		return ERR_PTR(-ENOMEM);
+	config_group_init_type_name(group, name, item_type);
+
+	return group;
+}
+EXPORT_SYMBOL(configfs_alloc_group);
+
+/**
+ * configfs_free_group() - frees memory for a configfs group
+ * @group:	group to be 'eliberated'
+ */
+void configfs_free_group(struct config_group *group)
+{
+	kfree(group);
+}
+EXPORT_SYMBOL(configfs_free_group);
+
+/**
+ * configfs_register_group - creates a parent-child relation between two groups
+ * @parent_group:	parent group
+ * @group:		child group
+ *
+ * link groups, creates dentry for the child and attaches it to the
+ * parent dentry.
+ *
+ * Return: 0 on success, negative errno code on error
+ */
+int configfs_register_group(struct config_group *parent_group,
+			    struct config_group *group)
+{
+	struct configfs_subsystem *subsys = parent_group->cg_subsys;
+	struct dentry *parent;
+	int ret;
+
+	mutex_lock(&subsys->su_mutex);
+	link_group(parent_group, group);
+	mutex_unlock(&subsys->su_mutex);
+
+	parent = parent_group->cg_item.ci_dentry;
+
+	mutex_lock_nested(&d_inode(parent)->i_mutex, I_MUTEX_PARENT);
+	ret = create_default_group(parent_group, group);
+	if (!ret) {
+		spin_lock(&configfs_dirent_lock);
+		configfs_dir_set_ready(group->cg_item.ci_dentry->d_fsdata);
+		spin_unlock(&configfs_dirent_lock);
+	}
+	mutex_unlock(&d_inode(parent)->i_mutex);
+	return ret;
+}
+EXPORT_SYMBOL(configfs_register_group);
+
+/**
+ * configfs_unregister_group() - unregisters a child group from its parent
+ *
+ * @group: parent group to be unregistered
+ *
+ * Undoes condigfs_register_group()
+ */
+void configfs_unregister_group(struct config_group *group)
+{
+	struct configfs_subsystem *subsys = group->cg_subsys;
+	struct dentry *dentry = group->cg_item.ci_dentry;
+	struct dentry *parent = group->cg_item.ci_parent->ci_dentry;
+
+	mutex_lock_nested(&d_inode(parent)->i_mutex, I_MUTEX_PARENT);
+	spin_lock(&configfs_dirent_lock);
+	configfs_detach_prep(dentry, NULL);
+	spin_unlock(&configfs_dirent_lock);
+
+	configfs_detach_group(&group->cg_item);
+	d_inode(dentry)->i_flags |= S_DEAD;
+	dont_mount(dentry);
+	d_delete(dentry);
+	mutex_unlock(&d_inode(parent)->i_mutex);
+
+	dput(dentry);
+
+	mutex_lock(&subsys->su_mutex);
+	unlink_group(group);
+	mutex_unlock(&subsys->su_mutex);
+}
+EXPORT_SYMBOL(configfs_unregister_group);
+
 int configfs_register_subsystem(struct configfs_subsystem *subsys)
 {
 	int err;
diff --git a/include/linux/configfs.h b/include/linux/configfs.h
index 63a36e8..c522a5e 100644
--- a/include/linux/configfs.h
+++ b/include/linux/configfs.h
@@ -252,6 +252,14 @@  static inline struct configfs_subsystem *to_configfs_subsystem(struct config_gro
 int configfs_register_subsystem(struct configfs_subsystem *subsys);
 void configfs_unregister_subsystem(struct configfs_subsystem *subsys);
 
+struct config_group *
+configfs_alloc_group(const char *name, struct config_item_type *item_type);
+void configfs_free_group(struct config_group *group);
+
+int configfs_register_group(struct config_group *parent_group,
+			    struct config_group *group);
+void configfs_unregister_group(struct config_group *group);
+
 /* These functions can sleep and can alloc with GFP_KERNEL */
 /* WARNING: These cannot be called underneath configfs callbacks!! */
 int configfs_depend_item(struct configfs_subsystem *subsys, struct config_item *target);