2017-04-11 19:49:49 +03:00
|
|
|
/*
|
|
|
|
* Copyright(c) 2017 Intel Corporation. All rights reserved.
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of version 2 of the GNU General Public License as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* This program is distributed in the hope that it will be useful, but
|
|
|
|
* WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
* General Public License for more details.
|
|
|
|
*/
|
|
|
|
#include <linux/pagemap.h>
|
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/mount.h>
|
|
|
|
#include <linux/magic.h>
|
2017-05-08 20:55:27 +03:00
|
|
|
#include <linux/genhd.h>
|
2017-04-11 19:49:49 +03:00
|
|
|
#include <linux/cdev.h>
|
|
|
|
#include <linux/hash.h>
|
|
|
|
#include <linux/slab.h>
|
2017-05-29 22:57:56 +03:00
|
|
|
#include <linux/uio.h>
|
2017-01-25 05:44:18 +03:00
|
|
|
#include <linux/dax.h>
|
2017-04-11 19:49:49 +03:00
|
|
|
#include <linux/fs.h>
|
|
|
|
|
|
|
|
static dev_t dax_devt;
|
|
|
|
DEFINE_STATIC_SRCU(dax_srcu);
|
|
|
|
static struct vfsmount *dax_mnt;
|
|
|
|
static DEFINE_IDA(dax_minor_ida);
|
|
|
|
static struct kmem_cache *dax_cache __read_mostly;
|
|
|
|
static struct super_block *dax_superblock __read_mostly;
|
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
#define DAX_HASH_SIZE (PAGE_SIZE / sizeof(struct hlist_head))
|
|
|
|
static struct hlist_head dax_host_list[DAX_HASH_SIZE];
|
|
|
|
static DEFINE_SPINLOCK(dax_host_lock);
|
|
|
|
|
2017-04-11 19:49:49 +03:00
|
|
|
int dax_read_lock(void)
|
|
|
|
{
|
|
|
|
return srcu_read_lock(&dax_srcu);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_read_lock);
|
|
|
|
|
|
|
|
void dax_read_unlock(int id)
|
|
|
|
{
|
|
|
|
srcu_read_unlock(&dax_srcu, id);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_read_unlock);
|
|
|
|
|
2017-05-14 02:18:21 +03:00
|
|
|
#ifdef CONFIG_BLOCK
|
2017-08-30 19:16:38 +03:00
|
|
|
#include <linux/blkdev.h>
|
|
|
|
|
2017-05-08 20:55:27 +03:00
|
|
|
int bdev_dax_pgoff(struct block_device *bdev, sector_t sector, size_t size,
|
|
|
|
pgoff_t *pgoff)
|
|
|
|
{
|
|
|
|
phys_addr_t phys_off = (get_start_sect(bdev) + sector) * 512;
|
|
|
|
|
|
|
|
if (pgoff)
|
|
|
|
*pgoff = PHYS_PFN(phys_off);
|
|
|
|
if (phys_off % PAGE_SIZE || size % PAGE_SIZE)
|
|
|
|
return -EINVAL;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL(bdev_dax_pgoff);
|
|
|
|
|
2017-09-03 20:17:53 +03:00
|
|
|
#if IS_ENABLED(CONFIG_FS_DAX)
|
2017-08-30 19:16:38 +03:00
|
|
|
struct dax_device *fs_dax_get_by_bdev(struct block_device *bdev)
|
|
|
|
{
|
|
|
|
if (!blk_queue_dax(bdev->bd_queue))
|
|
|
|
return NULL;
|
|
|
|
return fs_dax_get_by_host(bdev->bd_disk->disk_name);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(fs_dax_get_by_bdev);
|
2017-09-03 20:17:53 +03:00
|
|
|
#endif
|
2017-08-30 19:16:38 +03:00
|
|
|
|
2017-05-08 20:55:27 +03:00
|
|
|
/**
|
|
|
|
* __bdev_dax_supported() - Check if the device supports dax for filesystem
|
|
|
|
* @sb: The superblock of the device
|
|
|
|
* @blocksize: The block size of the device
|
|
|
|
*
|
|
|
|
* This is a library function for filesystems to check if the block device
|
|
|
|
* can be mounted with dax option.
|
|
|
|
*
|
|
|
|
* Return: negative errno if unsupported, 0 if supported.
|
|
|
|
*/
|
|
|
|
int __bdev_dax_supported(struct super_block *sb, int blocksize)
|
|
|
|
{
|
|
|
|
struct block_device *bdev = sb->s_bdev;
|
|
|
|
struct dax_device *dax_dev;
|
|
|
|
pgoff_t pgoff;
|
|
|
|
int err, id;
|
|
|
|
void *kaddr;
|
|
|
|
pfn_t pfn;
|
|
|
|
long len;
|
|
|
|
|
|
|
|
if (blocksize != PAGE_SIZE) {
|
2017-10-14 21:32:33 +03:00
|
|
|
pr_debug("VFS (%s): error: unsupported blocksize for dax\n",
|
2017-05-08 20:55:27 +03:00
|
|
|
sb->s_id);
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
err = bdev_dax_pgoff(bdev, 0, PAGE_SIZE, &pgoff);
|
|
|
|
if (err) {
|
2017-10-14 21:32:33 +03:00
|
|
|
pr_debug("VFS (%s): error: unaligned partition for dax\n",
|
2017-05-08 20:55:27 +03:00
|
|
|
sb->s_id);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
dax_dev = dax_get_by_host(bdev->bd_disk->disk_name);
|
|
|
|
if (!dax_dev) {
|
2017-10-14 21:32:33 +03:00
|
|
|
pr_debug("VFS (%s): error: device does not support dax\n",
|
2017-05-08 20:55:27 +03:00
|
|
|
sb->s_id);
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
}
|
|
|
|
|
|
|
|
id = dax_read_lock();
|
|
|
|
len = dax_direct_access(dax_dev, pgoff, 1, &kaddr, &pfn);
|
|
|
|
dax_read_unlock(id);
|
|
|
|
|
|
|
|
put_dax(dax_dev);
|
|
|
|
|
|
|
|
if (len < 1) {
|
2017-10-14 21:32:33 +03:00
|
|
|
pr_debug("VFS (%s): error: dax access failed (%ld)\n",
|
2017-05-08 20:55:27 +03:00
|
|
|
sb->s_id, len);
|
|
|
|
return len < 0 ? len : -EIO;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(__bdev_dax_supported);
|
2017-05-14 02:18:21 +03:00
|
|
|
#endif
|
2017-05-08 20:55:27 +03:00
|
|
|
|
2017-06-28 03:59:28 +03:00
|
|
|
enum dax_device_flags {
|
|
|
|
/* !alive + rcu grace period == no new operations / mappings */
|
|
|
|
DAXDEV_ALIVE,
|
2017-06-27 07:28:41 +03:00
|
|
|
/* gate whether dax_flush() calls the low level flush routine */
|
|
|
|
DAXDEV_WRITE_CACHE,
|
2017-06-28 03:59:28 +03:00
|
|
|
};
|
|
|
|
|
2017-04-11 19:49:49 +03:00
|
|
|
/**
|
|
|
|
* struct dax_device - anchor object for dax services
|
|
|
|
* @inode: core vfs
|
|
|
|
* @cdev: optional character interface for "device dax"
|
2017-04-20 01:14:31 +03:00
|
|
|
* @host: optional name for lookups where the device path is not available
|
2017-04-11 19:49:49 +03:00
|
|
|
* @private: dax driver private data
|
2017-06-28 03:59:28 +03:00
|
|
|
* @flags: state and boolean properties
|
2017-04-11 19:49:49 +03:00
|
|
|
*/
|
|
|
|
struct dax_device {
|
2017-04-20 01:14:31 +03:00
|
|
|
struct hlist_node list;
|
2017-04-11 19:49:49 +03:00
|
|
|
struct inode inode;
|
|
|
|
struct cdev cdev;
|
2017-04-20 01:14:31 +03:00
|
|
|
const char *host;
|
2017-04-11 19:49:49 +03:00
|
|
|
void *private;
|
2017-06-28 03:59:28 +03:00
|
|
|
unsigned long flags;
|
2017-01-25 05:44:18 +03:00
|
|
|
const struct dax_operations *ops;
|
2017-04-11 19:49:49 +03:00
|
|
|
};
|
|
|
|
|
2017-06-27 07:28:41 +03:00
|
|
|
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;
|
|
|
|
|
2017-09-01 04:47:43 +03:00
|
|
|
#ifndef CONFIG_ARCH_HAS_PMEM_API
|
|
|
|
if (a == &dev_attr_write_cache.attr)
|
2017-06-27 07:28:41 +03:00
|
|
|
return 0;
|
2017-09-01 04:47:43 +03:00
|
|
|
#endif
|
2017-06-27 07:28:41 +03:00
|
|
|
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);
|
|
|
|
|
2017-01-27 07:37:35 +03:00
|
|
|
/**
|
|
|
|
* dax_direct_access() - translate a device pgoff to an absolute pfn
|
|
|
|
* @dax_dev: a dax_device instance representing the logical memory range
|
|
|
|
* @pgoff: offset in pages from the start of the device to translate
|
|
|
|
* @nr_pages: number of consecutive pages caller can handle relative to @pfn
|
|
|
|
* @kaddr: output parameter that returns a virtual address mapping of pfn
|
|
|
|
* @pfn: output parameter that returns an absolute pfn translation of @pgoff
|
|
|
|
*
|
|
|
|
* Return: negative errno if an error occurs, otherwise the number of
|
|
|
|
* pages accessible at the device relative @pgoff.
|
|
|
|
*/
|
|
|
|
long dax_direct_access(struct dax_device *dax_dev, pgoff_t pgoff, long nr_pages,
|
|
|
|
void **kaddr, pfn_t *pfn)
|
|
|
|
{
|
|
|
|
long avail;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The device driver is allowed to sleep, in order to make the
|
|
|
|
* memory directly accessible.
|
|
|
|
*/
|
|
|
|
might_sleep();
|
|
|
|
|
|
|
|
if (!dax_dev)
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
|
|
|
if (!dax_alive(dax_dev))
|
|
|
|
return -ENXIO;
|
|
|
|
|
|
|
|
if (nr_pages < 0)
|
|
|
|
return nr_pages;
|
|
|
|
|
|
|
|
avail = dax_dev->ops->direct_access(dax_dev, pgoff, nr_pages,
|
|
|
|
kaddr, pfn);
|
|
|
|
if (!avail)
|
|
|
|
return -ERANGE;
|
|
|
|
return min(avail, nr_pages);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_direct_access);
|
|
|
|
|
2017-05-29 22:57:56 +03:00
|
|
|
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);
|
|
|
|
|
2017-09-01 04:47:43 +03:00
|
|
|
#ifdef CONFIG_ARCH_HAS_PMEM_API
|
|
|
|
void arch_wb_cache_pmem(void *addr, size_t size);
|
|
|
|
void dax_flush(struct dax_device *dax_dev, void *addr, size_t size)
|
2017-05-29 23:02:52 +03:00
|
|
|
{
|
2017-09-01 04:47:43 +03:00
|
|
|
if (unlikely(!test_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags)))
|
2017-06-27 07:28:41 +03:00
|
|
|
return;
|
|
|
|
|
2017-09-01 04:47:43 +03:00
|
|
|
arch_wb_cache_pmem(addr, size);
|
2017-05-29 23:02:52 +03:00
|
|
|
}
|
2017-09-01 04:47:43 +03:00
|
|
|
#else
|
|
|
|
void dax_flush(struct dax_device *dax_dev, void *addr, size_t size)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
#endif
|
2017-05-29 23:02:52 +03:00
|
|
|
EXPORT_SYMBOL_GPL(dax_flush);
|
|
|
|
|
2017-06-27 07:28:41 +03:00
|
|
|
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);
|
|
|
|
|
2017-07-26 16:35:09 +03:00
|
|
|
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);
|
|
|
|
|
2017-04-11 19:49:49 +03:00
|
|
|
bool dax_alive(struct dax_device *dax_dev)
|
|
|
|
{
|
|
|
|
lockdep_assert_held(&dax_srcu);
|
2017-06-28 03:59:28 +03:00
|
|
|
return test_bit(DAXDEV_ALIVE, &dax_dev->flags);
|
2017-04-11 19:49:49 +03:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_alive);
|
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
static int dax_host_hash(const char *host)
|
|
|
|
{
|
|
|
|
return hashlen_hash(hashlen_string("DAX", host)) % DAX_HASH_SIZE;
|
|
|
|
}
|
|
|
|
|
2017-04-11 19:49:49 +03:00
|
|
|
/*
|
|
|
|
* Note, rcu is not protecting the liveness of dax_dev, rcu is ensuring
|
|
|
|
* that any fault handlers or operations that might have seen
|
|
|
|
* dax_alive(), have completed. Any operations that start after
|
|
|
|
* synchronize_srcu() has run will abort upon seeing !dax_alive().
|
|
|
|
*/
|
|
|
|
void kill_dax(struct dax_device *dax_dev)
|
|
|
|
{
|
|
|
|
if (!dax_dev)
|
|
|
|
return;
|
|
|
|
|
2017-06-28 03:59:28 +03:00
|
|
|
clear_bit(DAXDEV_ALIVE, &dax_dev->flags);
|
2017-04-20 01:14:31 +03:00
|
|
|
|
2017-04-11 19:49:49 +03:00
|
|
|
synchronize_srcu(&dax_srcu);
|
2017-04-20 01:14:31 +03:00
|
|
|
|
|
|
|
spin_lock(&dax_host_lock);
|
|
|
|
hlist_del_init(&dax_dev->list);
|
|
|
|
spin_unlock(&dax_host_lock);
|
|
|
|
|
2017-04-11 19:49:49 +03:00
|
|
|
dax_dev->private = NULL;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(kill_dax);
|
|
|
|
|
|
|
|
static struct inode *dax_alloc_inode(struct super_block *sb)
|
|
|
|
{
|
|
|
|
struct dax_device *dax_dev;
|
2017-06-09 18:50:49 +03:00
|
|
|
struct inode *inode;
|
2017-04-11 19:49:49 +03:00
|
|
|
|
|
|
|
dax_dev = kmem_cache_alloc(dax_cache, GFP_KERNEL);
|
2017-11-14 17:59:54 +03:00
|
|
|
if (!dax_dev)
|
|
|
|
return NULL;
|
|
|
|
|
2017-06-09 18:50:49 +03:00
|
|
|
inode = &dax_dev->inode;
|
|
|
|
inode->i_rdev = 0;
|
|
|
|
return inode;
|
2017-04-11 19:49:49 +03:00
|
|
|
}
|
|
|
|
|
|
|
|
static struct dax_device *to_dax_dev(struct inode *inode)
|
|
|
|
{
|
|
|
|
return container_of(inode, struct dax_device, inode);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void dax_i_callback(struct rcu_head *head)
|
|
|
|
{
|
|
|
|
struct inode *inode = container_of(head, struct inode, i_rcu);
|
|
|
|
struct dax_device *dax_dev = to_dax_dev(inode);
|
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
kfree(dax_dev->host);
|
|
|
|
dax_dev->host = NULL;
|
2017-06-09 18:50:49 +03:00
|
|
|
if (inode->i_rdev)
|
|
|
|
ida_simple_remove(&dax_minor_ida, MINOR(inode->i_rdev));
|
2017-04-11 19:49:49 +03:00
|
|
|
kmem_cache_free(dax_cache, dax_dev);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void dax_destroy_inode(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct dax_device *dax_dev = to_dax_dev(inode);
|
|
|
|
|
2017-06-28 03:59:28 +03:00
|
|
|
WARN_ONCE(test_bit(DAXDEV_ALIVE, &dax_dev->flags),
|
2017-04-11 19:49:49 +03:00
|
|
|
"kill_dax() must be called before final iput()\n");
|
|
|
|
call_rcu(&inode->i_rcu, dax_i_callback);
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct super_operations dax_sops = {
|
|
|
|
.statfs = simple_statfs,
|
|
|
|
.alloc_inode = dax_alloc_inode,
|
|
|
|
.destroy_inode = dax_destroy_inode,
|
|
|
|
.drop_inode = generic_delete_inode,
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct dentry *dax_mount(struct file_system_type *fs_type,
|
|
|
|
int flags, const char *dev_name, void *data)
|
|
|
|
{
|
|
|
|
return mount_pseudo(fs_type, "dax:", &dax_sops, NULL, DAXFS_MAGIC);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct file_system_type dax_fs_type = {
|
|
|
|
.name = "dax",
|
|
|
|
.mount = dax_mount,
|
|
|
|
.kill_sb = kill_anon_super,
|
|
|
|
};
|
|
|
|
|
|
|
|
static int dax_test(struct inode *inode, void *data)
|
|
|
|
{
|
|
|
|
dev_t devt = *(dev_t *) data;
|
|
|
|
|
|
|
|
return inode->i_rdev == devt;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int dax_set(struct inode *inode, void *data)
|
|
|
|
{
|
|
|
|
dev_t devt = *(dev_t *) data;
|
|
|
|
|
|
|
|
inode->i_rdev = devt;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct dax_device *dax_dev_get(dev_t devt)
|
|
|
|
{
|
|
|
|
struct dax_device *dax_dev;
|
|
|
|
struct inode *inode;
|
|
|
|
|
|
|
|
inode = iget5_locked(dax_superblock, hash_32(devt + DAXFS_MAGIC, 31),
|
|
|
|
dax_test, dax_set, &devt);
|
|
|
|
|
|
|
|
if (!inode)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
dax_dev = to_dax_dev(inode);
|
|
|
|
if (inode->i_state & I_NEW) {
|
2017-06-28 03:59:28 +03:00
|
|
|
set_bit(DAXDEV_ALIVE, &dax_dev->flags);
|
2017-04-11 19:49:49 +03:00
|
|
|
inode->i_cdev = &dax_dev->cdev;
|
|
|
|
inode->i_mode = S_IFCHR;
|
|
|
|
inode->i_flags = S_DAX;
|
|
|
|
mapping_set_gfp_mask(&inode->i_data, GFP_USER);
|
|
|
|
unlock_new_inode(inode);
|
|
|
|
}
|
|
|
|
|
|
|
|
return dax_dev;
|
|
|
|
}
|
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
static void dax_add_host(struct dax_device *dax_dev, const char *host)
|
|
|
|
{
|
|
|
|
int hash;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Unconditionally init dax_dev since it's coming from a
|
|
|
|
* non-zeroed slab cache
|
|
|
|
*/
|
|
|
|
INIT_HLIST_NODE(&dax_dev->list);
|
|
|
|
dax_dev->host = host;
|
|
|
|
if (!host)
|
|
|
|
return;
|
|
|
|
|
|
|
|
hash = dax_host_hash(host);
|
|
|
|
spin_lock(&dax_host_lock);
|
|
|
|
hlist_add_head(&dax_dev->list, &dax_host_list[hash]);
|
|
|
|
spin_unlock(&dax_host_lock);
|
|
|
|
}
|
|
|
|
|
2017-01-25 05:44:18 +03:00
|
|
|
struct dax_device *alloc_dax(void *private, const char *__host,
|
|
|
|
const struct dax_operations *ops)
|
2017-04-11 19:49:49 +03:00
|
|
|
{
|
|
|
|
struct dax_device *dax_dev;
|
2017-04-20 01:14:31 +03:00
|
|
|
const char *host;
|
2017-04-11 19:49:49 +03:00
|
|
|
dev_t devt;
|
|
|
|
int minor;
|
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
host = kstrdup(__host, GFP_KERNEL);
|
|
|
|
if (__host && !host)
|
|
|
|
return NULL;
|
|
|
|
|
2017-05-08 22:33:53 +03:00
|
|
|
minor = ida_simple_get(&dax_minor_ida, 0, MINORMASK+1, GFP_KERNEL);
|
2017-04-11 19:49:49 +03:00
|
|
|
if (minor < 0)
|
2017-04-20 01:14:31 +03:00
|
|
|
goto err_minor;
|
2017-04-11 19:49:49 +03:00
|
|
|
|
|
|
|
devt = MKDEV(MAJOR(dax_devt), minor);
|
|
|
|
dax_dev = dax_dev_get(devt);
|
|
|
|
if (!dax_dev)
|
2017-04-20 01:14:31 +03:00
|
|
|
goto err_dev;
|
2017-04-11 19:49:49 +03:00
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
dax_add_host(dax_dev, host);
|
2017-01-25 05:44:18 +03:00
|
|
|
dax_dev->ops = ops;
|
2017-04-11 19:49:49 +03:00
|
|
|
dax_dev->private = private;
|
|
|
|
return dax_dev;
|
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
err_dev:
|
2017-04-11 19:49:49 +03:00
|
|
|
ida_simple_remove(&dax_minor_ida, minor);
|
2017-04-20 01:14:31 +03:00
|
|
|
err_minor:
|
|
|
|
kfree(host);
|
2017-04-11 19:49:49 +03:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(alloc_dax);
|
|
|
|
|
|
|
|
void put_dax(struct dax_device *dax_dev)
|
|
|
|
{
|
|
|
|
if (!dax_dev)
|
|
|
|
return;
|
|
|
|
iput(&dax_dev->inode);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(put_dax);
|
|
|
|
|
2017-04-20 01:14:31 +03:00
|
|
|
/**
|
|
|
|
* dax_get_by_host() - temporary lookup mechanism for filesystem-dax
|
|
|
|
* @host: alternate name for the device registered by a dax driver
|
|
|
|
*/
|
|
|
|
struct dax_device *dax_get_by_host(const char *host)
|
|
|
|
{
|
|
|
|
struct dax_device *dax_dev, *found = NULL;
|
|
|
|
int hash, id;
|
|
|
|
|
|
|
|
if (!host)
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
hash = dax_host_hash(host);
|
|
|
|
|
|
|
|
id = dax_read_lock();
|
|
|
|
spin_lock(&dax_host_lock);
|
|
|
|
hlist_for_each_entry(dax_dev, &dax_host_list[hash], list) {
|
|
|
|
if (!dax_alive(dax_dev)
|
|
|
|
|| strcmp(host, dax_dev->host) != 0)
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (igrab(&dax_dev->inode))
|
|
|
|
found = dax_dev;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
spin_unlock(&dax_host_lock);
|
|
|
|
dax_read_unlock(id);
|
|
|
|
|
|
|
|
return found;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_get_by_host);
|
|
|
|
|
2017-04-11 19:49:49 +03:00
|
|
|
/**
|
|
|
|
* inode_dax: convert a public inode into its dax_dev
|
|
|
|
* @inode: An inode with i_cdev pointing to a dax_dev
|
|
|
|
*
|
|
|
|
* Note this is not equivalent to to_dax_dev() which is for private
|
|
|
|
* internal use where we know the inode filesystem type == dax_fs_type.
|
|
|
|
*/
|
|
|
|
struct dax_device *inode_dax(struct inode *inode)
|
|
|
|
{
|
|
|
|
struct cdev *cdev = inode->i_cdev;
|
|
|
|
|
|
|
|
return container_of(cdev, struct dax_device, cdev);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(inode_dax);
|
|
|
|
|
|
|
|
struct inode *dax_inode(struct dax_device *dax_dev)
|
|
|
|
{
|
|
|
|
return &dax_dev->inode;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_inode);
|
|
|
|
|
|
|
|
void *dax_get_private(struct dax_device *dax_dev)
|
|
|
|
{
|
|
|
|
return dax_dev->private;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(dax_get_private);
|
|
|
|
|
|
|
|
static void init_once(void *_dax_dev)
|
|
|
|
{
|
|
|
|
struct dax_device *dax_dev = _dax_dev;
|
|
|
|
struct inode *inode = &dax_dev->inode;
|
|
|
|
|
2017-06-09 18:50:49 +03:00
|
|
|
memset(dax_dev, 0, sizeof(*dax_dev));
|
2017-04-11 19:49:49 +03:00
|
|
|
inode_init_once(inode);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __dax_fs_init(void)
|
|
|
|
{
|
|
|
|
int rc;
|
|
|
|
|
|
|
|
dax_cache = kmem_cache_create("dax_cache", sizeof(struct dax_device), 0,
|
|
|
|
(SLAB_HWCACHE_ALIGN|SLAB_RECLAIM_ACCOUNT|
|
|
|
|
SLAB_MEM_SPREAD|SLAB_ACCOUNT),
|
|
|
|
init_once);
|
|
|
|
if (!dax_cache)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
|
|
|
rc = register_filesystem(&dax_fs_type);
|
|
|
|
if (rc)
|
|
|
|
goto err_register_fs;
|
|
|
|
|
|
|
|
dax_mnt = kern_mount(&dax_fs_type);
|
|
|
|
if (IS_ERR(dax_mnt)) {
|
|
|
|
rc = PTR_ERR(dax_mnt);
|
|
|
|
goto err_mount;
|
|
|
|
}
|
|
|
|
dax_superblock = dax_mnt->mnt_sb;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
err_mount:
|
|
|
|
unregister_filesystem(&dax_fs_type);
|
|
|
|
err_register_fs:
|
|
|
|
kmem_cache_destroy(dax_cache);
|
|
|
|
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __dax_fs_exit(void)
|
|
|
|
{
|
|
|
|
kern_unmount(dax_mnt);
|
|
|
|
unregister_filesystem(&dax_fs_type);
|
|
|
|
kmem_cache_destroy(dax_cache);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int __init dax_fs_init(void)
|
|
|
|
{
|
|
|
|
int rc;
|
|
|
|
|
|
|
|
rc = __dax_fs_init();
|
|
|
|
if (rc)
|
|
|
|
return rc;
|
|
|
|
|
2017-05-08 22:33:53 +03:00
|
|
|
rc = alloc_chrdev_region(&dax_devt, 0, MINORMASK+1, "dax");
|
2017-04-11 19:49:49 +03:00
|
|
|
if (rc)
|
|
|
|
__dax_fs_exit();
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __exit dax_fs_exit(void)
|
|
|
|
{
|
2017-05-08 22:33:53 +03:00
|
|
|
unregister_chrdev_region(dax_devt, MINORMASK+1);
|
2017-04-11 19:49:49 +03:00
|
|
|
ida_destroy(&dax_minor_ida);
|
|
|
|
__dax_fs_exit();
|
|
|
|
}
|
|
|
|
|
|
|
|
MODULE_AUTHOR("Intel Corporation");
|
|
|
|
MODULE_LICENSE("GPL v2");
|
|
|
|
subsys_initcall(dax_fs_init);
|
|
|
|
module_exit(dax_fs_exit);
|