kernfs: implement kernfs_ops->atomic_write_len

A write to a kernfs_node is buffered through a kernel buffer.  Writes
<= PAGE_SIZE are performed atomically, while larger ones are executed
in PAGE_SIZE chunks.  While this is enough for sysfs, cgroup which is
scheduled to be converted to use kernfs needs a bit more control over
it.

This patch adds kernfs_ops->atomic_write_len.  If not set (zero), the
behavior stays the same.  If set, writes upto the size are executed
atomically and larger writes are rejected with -E2BIG.

A different implementation strategy would be allowing configuring
chunking size while making the original write size available to the
write method; however, such strategy, while being more complicated,
doesn't really buy anything.  If the write implementation has to
handle chunking, the specific chunk size shouldn't matter all that
much.

Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

2 files changed, 37 insertions, 20 deletions

diff --git a/fs/kernfs/file.c b/fs/kernfs/file.c
index 10a8c91c49d..ddcb471b9cc 100644
--- a/fs/kernfs/file.c
+++ b/fs/kernfs/file.c
@@ -252,19 +252,9 @@ static ssize_t kernfs_fop_write(struct file *file, const char __user *user_buf,
 				size_t count, loff_t *ppos)
 {
 	struct kernfs_open_file *of = kernfs_of(file);
-	ssize_t len = min_t(size_t, count, PAGE_SIZE);
 	const struct kernfs_ops *ops;
-	char *buf;
-
-	buf = kmalloc(len + 1, GFP_KERNEL);
-	if (!buf)
-		return -ENOMEM;
-
-	if (copy_from_user(buf, user_buf, len)) {
-		len = -EFAULT;
-		goto out_free;
-	}
-	buf[len] = '\0';	/* guarantee string termination */
+	char *buf = NULL;
+	ssize_t len;
 
 	/*
 	 * @of->mutex nests outside active ref and is just to ensure that
@@ -273,22 +263,45 @@ static ssize_t kernfs_fop_write(struct file *file, const char __user *user_buf,
 	mutex_lock(&of->mutex);
 	if (!kernfs_get_active(of->kn)) {
 		mutex_unlock(&of->mutex);
-		len = -ENODEV;
-		goto out_free;
+		return -ENODEV;
 	}
 
 	ops = kernfs_ops(of->kn);
-	if (ops->write)
-		len = ops->write(of, buf, len, *ppos);
-	else
+	if (!ops->write) {
 		len = -EINVAL;
+		goto out_unlock;
+	}
+
+	if (ops->atomic_write_len) {
+		len = count;
+		if (len > ops->atomic_write_len) {
+			len = -E2BIG;
+			goto out_unlock;
+		}
+	} else {
+		len = min_t(size_t, count, PAGE_SIZE);
+	}
+
+	buf = kmalloc(len + 1, GFP_KERNEL);
+	if (!buf) {
+		len = -ENOMEM;
+		goto out_unlock;
+	}
 
+	if (copy_from_user(buf, user_buf, len)) {
+		len = -EFAULT;
+		goto out_unlock;
+	}
+	buf[len] = '\0';	/* guarantee string termination */
+
+	len = ops->write(of, buf, len, *ppos);
+out_unlock:
 	kernfs_put_active(of->kn);
 	mutex_unlock(&of->mutex);
 
 	if (len > 0)
 		*ppos += len;
-out_free:
+
 	kfree(buf);
 	return len;
 }
diff --git a/include/linux/kernfs.h b/include/linux/kernfs.h
index 4520c86f5cb..47f5235a097 100644
--- a/include/linux/kernfs.h
+++ b/include/linux/kernfs.h
@@ -178,9 +178,13 @@ struct kernfs_ops {
 			loff_t off);
 
 	/*
-	 * write() is bounced through kernel buffer and a write larger than
-	 * PAGE_SIZE results in partial operation of PAGE_SIZE.
+	 * write() is bounced through kernel buffer.  If atomic_write_len
+	 * is not set, a write larger than PAGE_SIZE results in partial
+	 * operations of PAGE_SIZE chunks.  If atomic_write_len is set,
+	 * writes upto the specified size are executed atomically but
+	 * larger ones are rejected with -E2BIG.
 	 */
+	size_t atomic_write_len;
 	ssize_t (*write)(struct kernfs_open_file *of, char *buf, size_t bytes,
 			 loff_t off);