kernfs: switch global kernfs_rwsem lock to per-fs lock

The kernfs implementation has big lock granularity(kernfs_rwsem) so
every kernfs-based(e.g., sysfs, cgroup) fs are able to compete the
lock. It makes trouble for some cases to wait the global lock
for a long time even though they are totally independent contexts
each other.

A general example is process A goes under direct reclaim with holding
the lock when it accessed the file in sysfs and process B is waiting
the lock with exclusive mode and then process C is waiting the lock
until process B could finish the job after it gets the lock from
process A.

This patch switches the global kernfs_rwsem to per-fs lock, which
put the rwsem into kernfs_root.

Suggested-by: Tejun Heo <tj@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Minchan Kim <minchan@kernel.org>
Link: https://lore.kernel.org/r/20211118230008.2679780-1-minchan@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

1 files changed, 14 insertions, 8 deletions

diff --git a/fs/kernfs/inode.c b/fs/kernfs/inode.c
index c0eae1725435..3d783d80f5da 100644
--- a/fs/kernfs/inode.c
+++ b/fs/kernfs/inode.c
@@ -99,10 +99,11 @@ int __kernfs_setattr(struct kernfs_node *kn, const struct iattr *iattr)
 int kernfs_setattr(struct kernfs_node *kn, const struct iattr *iattr)
 {
 	int ret;
+	struct kernfs_root *root = kernfs_root(kn);
 
-	down_write(&kernfs_rwsem);
+	down_write(&root->kernfs_rwsem);
 	ret = __kernfs_setattr(kn, iattr);
-	up_write(&kernfs_rwsem);
+	up_write(&root->kernfs_rwsem);
 	return ret;
 }
 
@@ -111,12 +112,14 @@ int kernfs_iop_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
 {
 	struct inode *inode = d_inode(dentry);
 	struct kernfs_node *kn = inode->i_private;
+	struct kernfs_root *root;
 	int error;
 
 	if (!kn)
 		return -EINVAL;
 
-	down_write(&kernfs_rwsem);
+	root = kernfs_root(kn);
+	down_write(&root->kernfs_rwsem);
 	error = setattr_prepare(&init_user_ns, dentry, iattr);
 	if (error)
 		goto out;
@@ -129,7 +132,7 @@ int kernfs_iop_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
 	setattr_copy(&init_user_ns, inode, iattr);
 
 out:
-	up_write(&kernfs_rwsem);
+	up_write(&root->kernfs_rwsem);
 	return error;
 }
 
@@ -184,13 +187,14 @@ int kernfs_iop_getattr(struct user_namespace *mnt_userns,
 {
 	struct inode *inode = d_inode(path->dentry);
 	struct kernfs_node *kn = inode->i_private;
+	struct kernfs_root *root = kernfs_root(kn);
 
-	down_read(&kernfs_rwsem);
+	down_read(&root->kernfs_rwsem);
 	spin_lock(&inode->i_lock);
 	kernfs_refresh_inode(kn, inode);
 	generic_fillattr(&init_user_ns, inode, stat);
 	spin_unlock(&inode->i_lock);
-	up_read(&kernfs_rwsem);
+	up_read(&root->kernfs_rwsem);
 
 	return 0;
 }
@@ -274,19 +278,21 @@ int kernfs_iop_permission(struct user_namespace *mnt_userns,
 			  struct inode *inode, int mask)
 {
 	struct kernfs_node *kn;
+	struct kernfs_root *root;
 	int ret;
 
 	if (mask & MAY_NOT_BLOCK)
 		return -ECHILD;
 
 	kn = inode->i_private;
+	root = kernfs_root(kn);
 
-	down_read(&kernfs_rwsem);
+	down_read(&root->kernfs_rwsem);
 	spin_lock(&inode->i_lock);
 	kernfs_refresh_inode(kn, inode);
 	ret = generic_permission(&init_user_ns, inode, mask);
 	spin_unlock(&inode->i_lock);
-	up_read(&kernfs_rwsem);
+	up_read(&root->kernfs_rwsem);
 
 	return ret;
 }