4 files changed, 150 insertions, 22 deletions

diff --git a/drivers/dax/device-dax.h b/drivers/dax/device-dax.h
index fdcd9769ffde..688b051750bd 100644
--- a/drivers/dax/device-dax.h
+++ b/drivers/dax/device-dax.h
@@ -21,5 +21,5 @@ struct dax_region *alloc_dax_region(struct device *parent,
 		int region_id, struct resource *res, unsigned int align,
 		void *addr, unsigned long flags);
 struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,
-		struct resource *res, int count);
+		int id, struct resource *res, int count);
 #endif /* __DEVICE_DAX_H__ */
diff --git a/drivers/dax/device.c b/drivers/dax/device.c
index 006e657dfcb9..e9f3b3e4bbf4 100644
--- a/drivers/dax/device.c
+++ b/drivers/dax/device.c
@@ -499,6 +499,7 @@ static int dax_open(struct inode *inode, struct file *filp)
 	inode->i_mapping = __dax_inode->i_mapping;
 	inode->i_mapping->host = __dax_inode;
 	filp->f_mapping = inode->i_mapping;
+	filp->f_wb_err = filemap_sample_wb_err(filp->f_mapping);
 	filp->private_data = dev_dax;
 	inode->i_flags = S_DAX;
 
@@ -528,7 +529,8 @@ static void dev_dax_release(struct device *dev)
 	struct dax_region *dax_region = dev_dax->region;
 	struct dax_device *dax_dev = dev_dax->dax_dev;
 
-	ida_simple_remove(&dax_region->ida, dev_dax->id);
+	if (dev_dax->id >= 0)
+		ida_simple_remove(&dax_region->ida, dev_dax->id);
 	dax_region_put(dax_region);
 	put_dax(dax_dev);
 	kfree(dev_dax);
@@ -558,7 +560,7 @@ static void unregister_dev_dax(void *dev)
 }
 
 struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,
-		struct resource *res, int count)
+		int id, struct resource *res, int count)
 {
 	struct device *parent = dax_region->dev;
 	struct dax_device *dax_dev;
@@ -566,7 +568,10 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,
 	struct inode *inode;
 	struct device *dev;
 	struct cdev *cdev;
-	int rc = 0, i;
+	int rc, i;
+
+	if (!count)
+		return ERR_PTR(-EINVAL);
 
 	dev_dax = kzalloc(sizeof(*dev_dax) + sizeof(*res) * count, GFP_KERNEL);
 	if (!dev_dax)
@@ -586,10 +591,16 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,
 	if (i < count)
 		goto err_id;
 
-	dev_dax->id = ida_simple_get(&dax_region->ida, 0, 0, GFP_KERNEL);
-	if (dev_dax->id < 0) {
-		rc = dev_dax->id;
-		goto err_id;
+	if (id < 0) {
+		id = ida_simple_get(&dax_region->ida, 0, 0, GFP_KERNEL);
+		dev_dax->id = id;
+		if (id < 0) {
+			rc = id;
+			goto err_id;
+		}
+	} else {
+		/* region provider owns @id lifetime */
+		dev_dax->id = -1;
 	}
 
 	/*
@@ -597,8 +608,10 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,
 	 * device outside of mmap of the resulting character device.
 	 */
 	dax_dev = alloc_dax(dev_dax, NULL, NULL);
-	if (!dax_dev)
+	if (!dax_dev) {
+		rc = -ENOMEM;
 		goto err_dax;
+	}
 
 	/* from here on we're committed to teardown via dax_dev_release() */
 	dev = &dev_dax->dev;
@@ -619,7 +632,7 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,
 	dev->parent = parent;
 	dev->groups = dax_attribute_groups;
 	dev->release = dev_dax_release;
-	dev_set_name(dev, "dax%d.%d", dax_region->id, dev_dax->id);
+	dev_set_name(dev, "dax%d.%d", dax_region->id, id);
 
 	rc = cdev_device_add(cdev, dev);
 	if (rc) {
@@ -635,7 +648,8 @@ struct dev_dax *devm_create_dev_dax(struct dax_region *dax_region,
 	return dev_dax;
 
  err_dax:
-	ida_simple_remove(&dax_region->ida, dev_dax->id);
+	if (dev_dax->id >= 0)
+		ida_simple_remove(&dax_region->ida, dev_dax->id);
  err_id:
 	kfree(dev_dax);
 
diff --git a/drivers/dax/pmem.c b/drivers/dax/pmem.c
index 9f2a0b4fd801..8d8c852ba8f2 100644
--- a/drivers/dax/pmem.c
+++ b/drivers/dax/pmem.c
@@ -58,13 +58,12 @@ static void dax_pmem_percpu_kill(void *data)
 
 static int dax_pmem_probe(struct device *dev)
 {
-	int rc;
 	void *addr;
 	struct resource res;
+	int rc, id, region_id;
 	struct nd_pfn_sb *pfn_sb;
 	struct dev_dax *dev_dax;
 	struct dax_pmem *dax_pmem;
-	struct nd_region *nd_region;
 	struct nd_namespace_io *nsio;
 	struct dax_region *dax_region;
 	struct nd_namespace_common *ndns;
@@ -123,14 +122,17 @@ static int dax_pmem_probe(struct device *dev)
 	/* adjust the dax_region resource to the start of data */
 	res.start += le64_to_cpu(pfn_sb->dataoff);
 
-	nd_region = to_nd_region(dev->parent);
-	dax_region = alloc_dax_region(dev, nd_region->id, &res,
+	rc = sscanf(dev_name(&ndns->dev), "namespace%d.%d", &region_id, &id);
+	if (rc != 2)
+		return -EINVAL;
+
+	dax_region = alloc_dax_region(dev, region_id, &res,
 			le32_to_cpu(pfn_sb->align), addr, PFN_DEV|PFN_MAP);
 	if (!dax_region)
 		return -ENOMEM;
 
 	/* TODO: support for subdividing a dax region... */
-	dev_dax = devm_create_dev_dax(dax_region, &res, 1);
+	dev_dax = devm_create_dev_dax(dax_region, id, &res, 1);
 
 	/* child dev_dax instances now own the lifetime of the dax_region */
 	dax_region_put(dax_region);
diff --git a/drivers/dax/super.c b/drivers/dax/super.c
index 922d0823f8ec..938eb4868f7f 100644
--- a/drivers/dax/super.c
+++ b/drivers/dax/super.c
@@ -18,6 +18,7 @@
 #include <linux/cdev.h>
 #include <linux/hash.h>
 #include <linux/slab.h>
+#include <linux/uio.h>
 #include <linux/dax.h>
 #include <linux/fs.h>
 
@@ -115,13 +116,20 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
 EXPORT_SYMBOL_GPL(__bdev_dax_supported);
 #endif
 
+enum dax_device_flags {
+	/* !alive + rcu grace period == no new operations / mappings */
+	DAXDEV_ALIVE,
+	/* gate whether dax_flush() calls the low level flush routine */
+	DAXDEV_WRITE_CACHE,
+};
+
 /**
  * struct dax_device - anchor object for dax services
  * @inode: core vfs
  * @cdev: optional character interface for "device dax"
  * @host: optional name for lookups where the device path is not available
  * @private: dax driver private data
- * @alive: !alive + rcu grace period == no new operations / mappings
+ * @flags: state and boolean properties
  */
 struct dax_device {
 	struct hlist_node list;
@@ -129,10 +137,75 @@ struct dax_device {
 	struct cdev cdev;
 	const char *host;
 	void *private;
-	bool alive;
+	unsigned long flags;
 	const struct dax_operations *ops;
 };
 
+static ssize_t write_cache_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct dax_device *dax_dev = dax_get_by_host(dev_name(dev));
+	ssize_t rc;
+
+	WARN_ON_ONCE(!dax_dev);
+	if (!dax_dev)
+		return -ENXIO;
+
+	rc = sprintf(buf, "%d\n", !!test_bit(DAXDEV_WRITE_CACHE,
+				&dax_dev->flags));
+	put_dax(dax_dev);
+	return rc;
+}
+
+static ssize_t write_cache_store(struct device *dev,
+		struct device_attribute *attr, const char *buf, size_t len)
+{
+	bool write_cache;
+	int rc = strtobool(buf, &write_cache);
+	struct dax_device *dax_dev = dax_get_by_host(dev_name(dev));
+
+	WARN_ON_ONCE(!dax_dev);
+	if (!dax_dev)
+		return -ENXIO;
+
+	if (rc)
+		len = rc;
+	else if (write_cache)
+		set_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+	else
+		clear_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+
+	put_dax(dax_dev);
+	return len;
+}
+static DEVICE_ATTR_RW(write_cache);
+
+static umode_t dax_visible(struct kobject *kobj, struct attribute *a, int n)
+{
+	struct device *dev = container_of(kobj, typeof(*dev), kobj);
+	struct dax_device *dax_dev = dax_get_by_host(dev_name(dev));
+
+	WARN_ON_ONCE(!dax_dev);
+	if (!dax_dev)
+		return 0;
+
+	if (a == &dev_attr_write_cache.attr && !dax_dev->ops->flush)
+		return 0;
+	return a->mode;
+}
+
+static struct attribute *dax_attributes[] = {
+	&dev_attr_write_cache.attr,
+	NULL,
+};
+
+struct attribute_group dax_attribute_group = {
+	.name = "dax",
+	.attrs = dax_attributes,
+	.is_visible = dax_visible,
+};
+EXPORT_SYMBOL_GPL(dax_attribute_group);
+
 /**
  * dax_direct_access() - translate a device pgoff to an absolute pfn
  * @dax_dev: a dax_device instance representing the logical memory range
@@ -172,10 +245,49 @@ long dax_direct_access(struct dax_device *dax_dev, pgoff_t pgoff, long nr_pages,
 }
 EXPORT_SYMBOL_GPL(dax_direct_access);
 
+size_t dax_copy_from_iter(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
+		size_t bytes, struct iov_iter *i)
+{
+	if (!dax_alive(dax_dev))
+		return 0;
+
+	return dax_dev->ops->copy_from_iter(dax_dev, pgoff, addr, bytes, i);
+}
+EXPORT_SYMBOL_GPL(dax_copy_from_iter);
+
+void dax_flush(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
+		size_t size)
+{
+	if (!dax_alive(dax_dev))
+		return;
+
+	if (!test_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags))
+		return;
+
+	if (dax_dev->ops->flush)
+		dax_dev->ops->flush(dax_dev, pgoff, addr, size);
+}
+EXPORT_SYMBOL_GPL(dax_flush);
+
+void dax_write_cache(struct dax_device *dax_dev, bool wc)
+{
+	if (wc)
+		set_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+	else
+		clear_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+}
+EXPORT_SYMBOL_GPL(dax_write_cache);
+
+bool dax_write_cache_enabled(struct dax_device *dax_dev)
+{
+	return test_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+}
+EXPORT_SYMBOL_GPL(dax_write_cache_enabled);
+
 bool dax_alive(struct dax_device *dax_dev)
 {
 	lockdep_assert_held(&dax_srcu);
-	return dax_dev->alive;
+	return test_bit(DAXDEV_ALIVE, &dax_dev->flags);
 }
 EXPORT_SYMBOL_GPL(dax_alive);
 
@@ -195,7 +307,7 @@ void kill_dax(struct dax_device *dax_dev)
 	if (!dax_dev)
 		return;
 
-	dax_dev->alive = false;
+	clear_bit(DAXDEV_ALIVE, &dax_dev->flags);
 
 	synchronize_srcu(&dax_srcu);
 
@@ -239,7 +351,7 @@ static void dax_destroy_inode(struct inode *inode)
 {
 	struct dax_device *dax_dev = to_dax_dev(inode);
 
-	WARN_ONCE(dax_dev->alive,
+	WARN_ONCE(test_bit(DAXDEV_ALIVE, &dax_dev->flags),
 			"kill_dax() must be called before final iput()\n");
 	call_rcu(&inode->i_rcu, dax_i_callback);
 }
@@ -291,7 +403,7 @@ static struct dax_device *dax_dev_get(dev_t devt)
 
 	dax_dev = to_dax_dev(inode);
 	if (inode->i_state & I_NEW) {
-		dax_dev->alive = true;
+		set_bit(DAXDEV_ALIVE, &dax_dev->flags);
 		inode->i_cdev = &dax_dev->cdev;
 		inode->i_mode = S_IFCHR;
 		inode->i_flags = S_DAX;