[deliverable/linux.git] / drivers / nvdimm / core.c

/*
 * Copyright(c) 2013-2015 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/libnvdimm.h>
#include <linux/badblocks.h>
#include <linux/export.h>
#include <linux/module.h>
#include <linux/blkdev.h>
#include <linux/device.h>
#include <linux/ctype.h>
#include <linux/ndctl.h>
#include <linux/mutex.h>
#include <linux/slab.h>
#include <linux/io.h>
#include "nd-core.h"
#include "nd.h"

LIST_HEAD(nvdimm_bus_list);
DEFINE_MUTEX(nvdimm_bus_list_mutex);

void nvdimm_bus_lock(struct device *dev)
{
	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);

	if (!nvdimm_bus)
		return;
	mutex_lock(&nvdimm_bus->reconfig_mutex);
}
EXPORT_SYMBOL(nvdimm_bus_lock);

void nvdimm_bus_unlock(struct device *dev)
{
	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);

	if (!nvdimm_bus)
		return;
	mutex_unlock(&nvdimm_bus->reconfig_mutex);
}
EXPORT_SYMBOL(nvdimm_bus_unlock);

bool is_nvdimm_bus_locked(struct device *dev)
{
	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);

	if (!nvdimm_bus)
		return false;
	return mutex_is_locked(&nvdimm_bus->reconfig_mutex);
}
EXPORT_SYMBOL(is_nvdimm_bus_locked);

struct nvdimm_map {
	struct nvdimm_bus *nvdimm_bus;
	struct list_head list;
	resource_size_t offset;
	unsigned long flags;
	size_t size;
	union {
		void *mem;
		void __iomem *iomem;
	};
	struct kref kref;
};

static struct nvdimm_map *find_nvdimm_map(struct device *dev,
		resource_size_t offset)
{
	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
	struct nvdimm_map *nvdimm_map;

	list_for_each_entry(nvdimm_map, &nvdimm_bus->mapping_list, list)
		if (nvdimm_map->offset == offset)
			return nvdimm_map;
	return NULL;
}

static struct nvdimm_map *alloc_nvdimm_map(struct device *dev,
		resource_size_t offset, size_t size, unsigned long flags)
{
	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
	struct nvdimm_map *nvdimm_map;

	nvdimm_map = kzalloc(sizeof(*nvdimm_map), GFP_KERNEL);
	if (!nvdimm_map)
		return NULL;

	INIT_LIST_HEAD(&nvdimm_map->list);
	nvdimm_map->nvdimm_bus = nvdimm_bus;
	nvdimm_map->offset = offset;
	nvdimm_map->flags = flags;
	nvdimm_map->size = size;
	kref_init(&nvdimm_map->kref);

	if (!request_mem_region(offset, size, dev_name(&nvdimm_bus->dev)))
		goto err_request_region;

	if (flags)
		nvdimm_map->mem = memremap(offset, size, flags);
	else
		nvdimm_map->iomem = ioremap(offset, size);

	if (!nvdimm_map->mem)
		goto err_map;

	dev_WARN_ONCE(dev, !is_nvdimm_bus_locked(dev), "%s: bus unlocked!",
			__func__);
	list_add(&nvdimm_map->list, &nvdimm_bus->mapping_list);

	return nvdimm_map;

 err_map:
	release_mem_region(offset, size);
 err_request_region:
	kfree(nvdimm_map);
	return NULL;
}

static void nvdimm_map_release(struct kref *kref)
{
	struct nvdimm_bus *nvdimm_bus;
	struct nvdimm_map *nvdimm_map;

	nvdimm_map = container_of(kref, struct nvdimm_map, kref);
	nvdimm_bus = nvdimm_map->nvdimm_bus;

	dev_dbg(&nvdimm_bus->dev, "%s: %pa\n", __func__, &nvdimm_map->offset);
	list_del(&nvdimm_map->list);
	if (nvdimm_map->flags)
		memunmap(nvdimm_map->mem);
	else
		iounmap(nvdimm_map->iomem);
	release_mem_region(nvdimm_map->offset, nvdimm_map->size);
	kfree(nvdimm_map);
}

static void nvdimm_map_put(void *data)
{
	struct nvdimm_map *nvdimm_map = data;
	struct nvdimm_bus *nvdimm_bus = nvdimm_map->nvdimm_bus;

	nvdimm_bus_lock(&nvdimm_bus->dev);
	kref_put(&nvdimm_map->kref, nvdimm_map_release);
	nvdimm_bus_unlock(&nvdimm_bus->dev);
}

/**
 * devm_nvdimm_memremap - map a resource that is shared across regions
 * @dev: device that will own a reference to the shared mapping
 * @offset: physical base address of the mapping
 * @size: mapping size
 * @flags: memremap flags, or, if zero, perform an ioremap instead
 */
void *devm_nvdimm_memremap(struct device *dev, resource_size_t offset,
		size_t size, unsigned long flags)
{
	struct nvdimm_map *nvdimm_map;

	nvdimm_bus_lock(dev);
	nvdimm_map = find_nvdimm_map(dev, offset);
	if (!nvdimm_map)
		nvdimm_map = alloc_nvdimm_map(dev, offset, size, flags);
	else
		kref_get(&nvdimm_map->kref);
	nvdimm_bus_unlock(dev);

	if (devm_add_action_or_reset(dev, nvdimm_map_put, nvdimm_map))
		return NULL;

	return nvdimm_map->mem;
}
EXPORT_SYMBOL_GPL(devm_nvdimm_memremap);

u64 nd_fletcher64(void *addr, size_t len, bool le)
{
	u32 *buf = addr;
	u32 lo32 = 0;
	u64 hi32 = 0;
	int i;

	for (i = 0; i < len / sizeof(u32); i++) {
		lo32 += le ? le32_to_cpu((__le32) buf[i]) : buf[i];
		hi32 += lo32;
	}

	return hi32 << 32 | lo32;
}
EXPORT_SYMBOL_GPL(nd_fletcher64);

struct nvdimm_bus_descriptor *to_nd_desc(struct nvdimm_bus *nvdimm_bus)
{
	/* struct nvdimm_bus definition is private to libnvdimm */
	return nvdimm_bus->nd_desc;
}
EXPORT_SYMBOL_GPL(to_nd_desc);

struct device *to_nvdimm_bus_dev(struct nvdimm_bus *nvdimm_bus)
{
	/* struct nvdimm_bus definition is private to libnvdimm */
	return &nvdimm_bus->dev;
}
EXPORT_SYMBOL_GPL(to_nvdimm_bus_dev);

static bool is_uuid_sep(char sep)
{
	if (sep == '\n' || sep == '-' || sep == ':' || sep == '\0')
		return true;
	return false;
}

static int nd_uuid_parse(struct device *dev, u8 *uuid_out, const char *buf,
		size_t len)
{
	const char *str = buf;
	u8 uuid[16];
	int i;

	for (i = 0; i < 16; i++) {
		if (!isxdigit(str[0]) || !isxdigit(str[1])) {
			dev_dbg(dev, "%s: pos: %d buf[%zd]: %c buf[%zd]: %c\n",
					__func__, i, str - buf, str[0],
					str + 1 - buf, str[1]);
			return -EINVAL;
		}

		uuid[i] = (hex_to_bin(str[0]) << 4) | hex_to_bin(str[1]);
		str += 2;
		if (is_uuid_sep(*str))
			str++;
	}

	memcpy(uuid_out, uuid, sizeof(uuid));
	return 0;
}

/**
 * nd_uuid_store: common implementation for writing 'uuid' sysfs attributes
 * @dev: container device for the uuid property
 * @uuid_out: uuid buffer to replace
 * @buf: raw sysfs buffer to parse
 *
 * Enforce that uuids can only be changed while the device is disabled
 * (driver detached)
 * LOCKING: expects device_lock() is held on entry
 */
int nd_uuid_store(struct device *dev, u8 **uuid_out, const char *buf,
		size_t len)
{
	u8 uuid[16];
	int rc;

	if (dev->driver)
		return -EBUSY;

	rc = nd_uuid_parse(dev, uuid, buf, len);
	if (rc)
		return rc;

	kfree(*uuid_out);
	*uuid_out = kmemdup(uuid, sizeof(uuid), GFP_KERNEL);
	if (!(*uuid_out))
		return -ENOMEM;

	return 0;
}

ssize_t nd_sector_size_show(unsigned long current_lbasize,
		const unsigned long *supported, char *buf)
{
	ssize_t len = 0;
	int i;

	for (i = 0; supported[i]; i++)
		if (current_lbasize == supported[i])
			len += sprintf(buf + len, "[%ld] ", supported[i]);
		else
			len += sprintf(buf + len, "%ld ", supported[i]);
	len += sprintf(buf + len, "\n");
	return len;
}

ssize_t nd_sector_size_store(struct device *dev, const char *buf,
		unsigned long *current_lbasize, const unsigned long *supported)
{
	unsigned long lbasize;
	int rc, i;

	if (dev->driver)
		return -EBUSY;

	rc = kstrtoul(buf, 0, &lbasize);
	if (rc)
		return rc;

	for (i = 0; supported[i]; i++)
		if (lbasize == supported[i])
			break;

	if (supported[i]) {
		*current_lbasize = lbasize;
		return 0;
	} else {
		return -EINVAL;
	}
}

void __nd_iostat_start(struct bio *bio, unsigned long *start)
{
	struct gendisk *disk = bio->bi_bdev->bd_disk;
	const int rw = bio_data_dir(bio);
	int cpu = part_stat_lock();

	*start = jiffies;
	part_round_stats(cpu, &disk->part0);
	part_stat_inc(cpu, &disk->part0, ios[rw]);
	part_stat_add(cpu, &disk->part0, sectors[rw], bio_sectors(bio));
	part_inc_in_flight(&disk->part0, rw);
	part_stat_unlock();
}
EXPORT_SYMBOL(__nd_iostat_start);

void nd_iostat_end(struct bio *bio, unsigned long start)
{
	struct gendisk *disk = bio->bi_bdev->bd_disk;
	unsigned long duration = jiffies - start;
	const int rw = bio_data_dir(bio);
	int cpu = part_stat_lock();

	part_stat_add(cpu, &disk->part0, ticks[rw], duration);
	part_round_stats(cpu, &disk->part0);
	part_dec_in_flight(&disk->part0, rw);
	part_stat_unlock();
}
EXPORT_SYMBOL(nd_iostat_end);

static ssize_t commands_show(struct device *dev,
		struct device_attribute *attr, char *buf)
{
	int cmd, len = 0;
	struct nvdimm_bus *nvdimm_bus = to_nvdimm_bus(dev);
	struct nvdimm_bus_descriptor *nd_desc = nvdimm_bus->nd_desc;

	for_each_set_bit(cmd, &nd_desc->cmd_mask, BITS_PER_LONG)
		len += sprintf(buf + len, "%s ", nvdimm_bus_cmd_name(cmd));
	len += sprintf(buf + len, "\n");
	return len;
}
static DEVICE_ATTR_RO(commands);

static const char *nvdimm_bus_provider(struct nvdimm_bus *nvdimm_bus)
{
	struct nvdimm_bus_descriptor *nd_desc = nvdimm_bus->nd_desc;
	struct device *parent = nvdimm_bus->dev.parent;

	if (nd_desc->provider_name)
		return nd_desc->provider_name;
	else if (parent)
		return dev_name(parent);
	else
		return "unknown";
}

static ssize_t provider_show(struct device *dev,
		struct device_attribute *attr, char *buf)
{
	struct nvdimm_bus *nvdimm_bus = to_nvdimm_bus(dev);

	return sprintf(buf, "%s\n", nvdimm_bus_provider(nvdimm_bus));
}
static DEVICE_ATTR_RO(provider);

static int flush_namespaces(struct device *dev, void *data)
{
	device_lock(dev);
	device_unlock(dev);
	return 0;
}

static int flush_regions_dimms(struct device *dev, void *data)
{
	device_lock(dev);
	device_unlock(dev);
	device_for_each_child(dev, NULL, flush_namespaces);
	return 0;
}

static ssize_t wait_probe_show(struct device *dev,
		struct device_attribute *attr, char *buf)
{
	struct nvdimm_bus *nvdimm_bus = to_nvdimm_bus(dev);
	struct nvdimm_bus_descriptor *nd_desc = nvdimm_bus->nd_desc;
	int rc;

	if (nd_desc->flush_probe) {
		rc = nd_desc->flush_probe(nd_desc);
		if (rc)
			return rc;
	}
	nd_synchronize();
	device_for_each_child(dev, NULL, flush_regions_dimms);
	return sprintf(buf, "1\n");
}
static DEVICE_ATTR_RO(wait_probe);

static struct attribute *nvdimm_bus_attributes[] = {
	&dev_attr_commands.attr,
	&dev_attr_wait_probe.attr,
	&dev_attr_provider.attr,
	NULL,
};

struct attribute_group nvdimm_bus_attribute_group = {
	.attrs = nvdimm_bus_attributes,
};
EXPORT_SYMBOL_GPL(nvdimm_bus_attribute_group);

static void set_badblock(struct badblocks *bb, sector_t s, int num)
{
	dev_dbg(bb->dev, "Found a poison range (0x%llx, 0x%llx)\n",
			(u64) s * 512, (u64) num * 512);
	/* this isn't an error as the hardware will still throw an exception */
	if (badblocks_set(bb, s, num, 1))
		dev_info_once(bb->dev, "%s: failed for sector %llx\n",
				__func__, (u64) s);
}

/**
 * __add_badblock_range() - Convert a physical address range to bad sectors
 * @bb:		badblocks instance to populate
 * @ns_offset:	namespace offset where the error range begins (in bytes)
 * @len:	number of bytes of poison to be added
 *
 * This assumes that the range provided with (ns_offset, len) is within
 * the bounds of physical addresses for this namespace, i.e. lies in the
 * interval [ns_start, ns_start + ns_size)
 */
static void __add_badblock_range(struct badblocks *bb, u64 ns_offset, u64 len)
{
	const unsigned int sector_size = 512;
	sector_t start_sector;
	u64 num_sectors;
	u32 rem;

	start_sector = div_u64(ns_offset, sector_size);
	num_sectors = div_u64_rem(len, sector_size, &rem);
	if (rem)
		num_sectors++;

	if (unlikely(num_sectors > (u64)INT_MAX)) {
		u64 remaining = num_sectors;
		sector_t s = start_sector;

		while (remaining) {
			int done = min_t(u64, remaining, INT_MAX);

			set_badblock(bb, s, done);
			remaining -= done;
			s += done;
		}
	} else
		set_badblock(bb, start_sector, num_sectors);
}

static void badblocks_populate(struct list_head *poison_list,
		struct badblocks *bb, const struct resource *res)
{
	struct nd_poison *pl;

	if (list_empty(poison_list))
		return;

	list_for_each_entry(pl, poison_list, list) {
		u64 pl_end = pl->start + pl->length - 1;

		/* Discard intervals with no intersection */
		if (pl_end < res->start)
			continue;
		if (pl->start >  res->end)
			continue;
		/* Deal with any overlap after start of the namespace */
		if (pl->start >= res->start) {
			u64 start = pl->start;
			u64 len;

			if (pl_end <= res->end)
				len = pl->length;
			else
				len = res->start + resource_size(res)
					- pl->start;
			__add_badblock_range(bb, start - res->start, len);
			continue;
		}
		/* Deal with overlap for poison starting before the namespace */
		if (pl->start < res->start) {
			u64 len;

			if (pl_end < res->end)
				len = pl->start + pl->length - res->start;
			else
				len = resource_size(res);
			__add_badblock_range(bb, 0, len);
		}
	}
}

/**
 * nvdimm_badblocks_populate() - Convert a list of poison ranges to badblocks
 * @region: parent region of the range to interrogate
 * @bb: badblocks instance to populate
 * @res: resource range to consider
 *
 * The poison list generated during bus initialization may contain
 * multiple, possibly overlapping physical address ranges.  Compare each
 * of these ranges to the resource range currently being initialized,
 * and add badblocks entries for all matching sub-ranges
 */
void nvdimm_badblocks_populate(struct nd_region *nd_region,
		struct badblocks *bb, const struct resource *res)
{
	struct nvdimm_bus *nvdimm_bus;
	struct list_head *poison_list;

	if (!is_nd_pmem(&nd_region->dev)) {
		dev_WARN_ONCE(&nd_region->dev, 1,
				"%s only valid for pmem regions\n", __func__);
		return;
	}
	nvdimm_bus = walk_to_nvdimm_bus(&nd_region->dev);
	poison_list = &nvdimm_bus->poison_list;

	nvdimm_bus_lock(&nvdimm_bus->dev);
	badblocks_populate(poison_list, bb, res);
	nvdimm_bus_unlock(&nvdimm_bus->dev);
}
EXPORT_SYMBOL_GPL(nvdimm_badblocks_populate);

static int add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length)
{
	struct nd_poison *pl;

	pl = kzalloc(sizeof(*pl), GFP_KERNEL);
	if (!pl)
		return -ENOMEM;

	pl->start = addr;
	pl->length = length;
	list_add_tail(&pl->list, &nvdimm_bus->poison_list);

	return 0;
}

static int bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length)
{
	struct nd_poison *pl;

	if (list_empty(&nvdimm_bus->poison_list))
		return add_poison(nvdimm_bus, addr, length);

	/*
	 * There is a chance this is a duplicate, check for those first.
	 * This will be the common case as ARS_STATUS returns all known
	 * errors in the SPA space, and we can't query it per region
	 */
	list_for_each_entry(pl, &nvdimm_bus->poison_list, list)
		if (pl->start == addr) {
			/* If length has changed, update this list entry */
			if (pl->length != length)
				pl->length = length;
			return 0;
		}

	/*
	 * If not a duplicate or a simple length update, add the entry as is,
	 * as any overlapping ranges will get resolved when the list is consumed
	 * and converted to badblocks
	 */
	return add_poison(nvdimm_bus, addr, length);
}

int nvdimm_bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length)
{
	int rc;

	nvdimm_bus_lock(&nvdimm_bus->dev);
	rc = bus_add_poison(nvdimm_bus, addr, length);
	nvdimm_bus_unlock(&nvdimm_bus->dev);

	return rc;
}
EXPORT_SYMBOL_GPL(nvdimm_bus_add_poison);

#ifdef CONFIG_BLK_DEV_INTEGRITY
int nd_integrity_init(struct gendisk *disk, unsigned long meta_size)
{
	struct blk_integrity bi;

	if (meta_size == 0)
		return 0;

	memset(&bi, 0, sizeof(bi));

	bi.tuple_size = meta_size;
	bi.tag_size = meta_size;

	blk_integrity_register(disk, &bi);
	blk_queue_max_integrity_segments(disk->queue, 1);

	return 0;
}
EXPORT_SYMBOL(nd_integrity_init);

#else /* CONFIG_BLK_DEV_INTEGRITY */
int nd_integrity_init(struct gendisk *disk, unsigned long meta_size)
{
	return 0;
}
EXPORT_SYMBOL(nd_integrity_init);

#endif

static __init int libnvdimm_init(void)
{
	int rc;

	rc = nvdimm_bus_init();
	if (rc)
		return rc;
	rc = nvdimm_init();
	if (rc)
		goto err_dimm;
	rc = nd_region_init();
	if (rc)
		goto err_region;
	return 0;
 err_region:
	nvdimm_exit();
 err_dimm:
	nvdimm_bus_exit();
	return rc;
}

static __exit void libnvdimm_exit(void)
{
	WARN_ON(!list_empty(&nvdimm_bus_list));
	nd_region_exit();
	nvdimm_exit();
	nvdimm_bus_exit();
	nd_region_devs_exit();
	nvdimm_devs_exit();
}

MODULE_LICENSE("GPL v2");
MODULE_AUTHOR("Intel Corporation");
subsys_initcall(libnvdimm_init);
module_exit(libnvdimm_exit);
Commit	Line	Data
b94d5230 DW	1	/*
	2	* Copyright(c) 2013-2015 Intel Corporation. All rights reserved.
	3	*
	4	* This program is free software; you can redistribute it and/or modify
	5	* it under the terms of version 2 of the GNU General Public License as
	6	* published by the Free Software Foundation.
	7	*
	8	* This program is distributed in the hope that it will be useful, but
	9	* WITHOUT ANY WARRANTY; without even the implied warranty of
	10	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
	11	* General Public License for more details.
	12	*/
	13	#include <linux/libnvdimm.h>
b95f5f43	14	#include <linux/badblocks.h>
b94d5230 DW	15	#include <linux/export.h>
b94d5230 DW	16	#include <linux/module.h>
41cd8b70	17	#include <linux/blkdev.h>
b94d5230	18	#include <linux/device.h>
bf9bccc1	19	#include <linux/ctype.h>
62232e45	20	#include <linux/ndctl.h>
45def22c	21	#include <linux/mutex.h>
b94d5230	22	#include <linux/slab.h>
29b9aa0a	23	#include <linux/io.h>
b94d5230	24	#include "nd-core.h"
4d88a97a	25	#include "nd.h"
b94d5230	26
e6dfb2de DW	27	LIST_HEAD(nvdimm_bus_list);
e6dfb2de DW	28	DEFINE_MUTEX(nvdimm_bus_list_mutex);
b94d5230	29
3d88002e DW	30	void nvdimm_bus_lock(struct device *dev)
	31	{
	32	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
	33
	34	if (!nvdimm_bus)
	35	return;
	36	mutex_lock(&nvdimm_bus->reconfig_mutex);
	37	}
	38	EXPORT_SYMBOL(nvdimm_bus_lock);
	39
	40	void nvdimm_bus_unlock(struct device *dev)
	41	{
	42	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
	43
	44	if (!nvdimm_bus)
	45	return;
	46	mutex_unlock(&nvdimm_bus->reconfig_mutex);
	47	}
	48	EXPORT_SYMBOL(nvdimm_bus_unlock);
	49
	50	bool is_nvdimm_bus_locked(struct device *dev)
	51	{
	52	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
	53
	54	if (!nvdimm_bus)
	55	return false;
	56	return mutex_is_locked(&nvdimm_bus->reconfig_mutex);
	57	}
	58	EXPORT_SYMBOL(is_nvdimm_bus_locked);
	59
29b9aa0a DW	60	struct nvdimm_map {
	61	struct nvdimm_bus *nvdimm_bus;
	62	struct list_head list;
	63	resource_size_t offset;
	64	unsigned long flags;
	65	size_t size;
	66	union {
	67	void *mem;
	68	void __iomem *iomem;
	69	};
	70	struct kref kref;
	71	};
	72
	73	static struct nvdimm_map find_nvdimm_map(struct device dev,
	74	resource_size_t offset)
	75	{
	76	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
	77	struct nvdimm_map *nvdimm_map;
	78
	79	list_for_each_entry(nvdimm_map, &nvdimm_bus->mapping_list, list)
	80	if (nvdimm_map->offset == offset)
	81	return nvdimm_map;
	82	return NULL;
	83	}
	84
	85	static struct nvdimm_map alloc_nvdimm_map(struct device dev,
	86	resource_size_t offset, size_t size, unsigned long flags)
	87	{
	88	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
	89	struct nvdimm_map *nvdimm_map;
	90
	91	nvdimm_map = kzalloc(sizeof(*nvdimm_map), GFP_KERNEL);
	92	if (!nvdimm_map)
	93	return NULL;
	94
	95	INIT_LIST_HEAD(&nvdimm_map->list);
	96	nvdimm_map->nvdimm_bus = nvdimm_bus;
	97	nvdimm_map->offset = offset;
	98	nvdimm_map->flags = flags;
	99	nvdimm_map->size = size;
	100	kref_init(&nvdimm_map->kref);
	101
	102	if (!request_mem_region(offset, size, dev_name(&nvdimm_bus->dev)))
	103	goto err_request_region;
	104
	105	if (flags)
	106	nvdimm_map->mem = memremap(offset, size, flags);
	107	else
	108	nvdimm_map->iomem = ioremap(offset, size);
	109
	110	if (!nvdimm_map->mem)
	111	goto err_map;
	112
	113	dev_WARN_ONCE(dev, !is_nvdimm_bus_locked(dev), "%s: bus unlocked!",
	114	__func__);
	115	list_add(&nvdimm_map->list, &nvdimm_bus->mapping_list);
	116
	117	return nvdimm_map;
	118
	119	err_map:
	120	release_mem_region(offset, size);
	121	err_request_region:
	122	kfree(nvdimm_map);
	123	return NULL;
124	}
125
126	static void nvdimm_map_release(struct kref *kref)
127	{
128	struct nvdimm_bus *nvdimm_bus;
129	struct nvdimm_map *nvdimm_map;
130
131	nvdimm_map = container_of(kref, struct nvdimm_map, kref);
132	nvdimm_bus = nvdimm_map->nvdimm_bus;
133
134	dev_dbg(&nvdimm_bus->dev, "%s: %pa\n", __func__, &nvdimm_map->offset);
135	list_del(&nvdimm_map->list);
136	if (nvdimm_map->flags)
137	memunmap(nvdimm_map->mem);
138	else
139	iounmap(nvdimm_map->iomem);
140	release_mem_region(nvdimm_map->offset, nvdimm_map->size);
141	kfree(nvdimm_map);
142	}
143
144	static void nvdimm_map_put(void *data)
145	{
146	struct nvdimm_map *nvdimm_map = data;
147	struct nvdimm_bus *nvdimm_bus = nvdimm_map->nvdimm_bus;
148
149	nvdimm_bus_lock(&nvdimm_bus->dev);
150	kref_put(&nvdimm_map->kref, nvdimm_map_release);
151	nvdimm_bus_unlock(&nvdimm_bus->dev);
152	}
153
154	/**
155	* devm_nvdimm_memremap - map a resource that is shared across regions
156	* @dev: device that will own a reference to the shared mapping
157	* @offset: physical base address of the mapping
158	* @size: mapping size
159	* @flags: memremap flags, or, if zero, perform an ioremap instead
160	*/
161	void devm_nvdimm_memremap(struct device dev, resource_size_t offset,
162	size_t size, unsigned long flags)
163	{
164	struct nvdimm_map *nvdimm_map;
165
166	nvdimm_bus_lock(dev);
167	nvdimm_map = find_nvdimm_map(dev, offset);
168	if (!nvdimm_map)
169	nvdimm_map = alloc_nvdimm_map(dev, offset, size, flags);
170	else
171	kref_get(&nvdimm_map->kref);
172	nvdimm_bus_unlock(dev);
173
174	if (devm_add_action_or_reset(dev, nvdimm_map_put, nvdimm_map))
175	return NULL;
176
177	return nvdimm_map->mem;
178	}
179	EXPORT_SYMBOL_GPL(devm_nvdimm_memremap);
180
eaf96153 DW	181	u64 nd_fletcher64(void *addr, size_t len, bool le)
	182	{
	183	u32 *buf = addr;
	184	u32 lo32 = 0;
	185	u64 hi32 = 0;
	186	int i;
	187
	188	for (i = 0; i < len / sizeof(u32); i++) {
	189	lo32 += le ? le32_to_cpu((__le32) buf[i]) : buf[i];
	190	hi32 += lo32;
	191	}
	192
	193	return hi32 << 32 \| lo32;
	194	}
	195	EXPORT_SYMBOL_GPL(nd_fletcher64);
	196
45def22c DW	197	struct nvdimm_bus_descriptor to_nd_desc(struct nvdimm_bus nvdimm_bus)
	198	{
	199	/* struct nvdimm_bus definition is private to libnvdimm */
	200	return nvdimm_bus->nd_desc;
	201	}
	202	EXPORT_SYMBOL_GPL(to_nd_desc);
	203
37b137ff VV	204	struct device to_nvdimm_bus_dev(struct nvdimm_bus nvdimm_bus)
	205	{
	206	/* struct nvdimm_bus definition is private to libnvdimm */
	207	return &nvdimm_bus->dev;
	208	}
	209	EXPORT_SYMBOL_GPL(to_nvdimm_bus_dev);
	210
bf9bccc1 DW	211	static bool is_uuid_sep(char sep)
	212	{
	213	if (sep == '\n' \|\| sep == '-' \|\| sep == ':' \|\| sep == '\0')
	214	return true;
	215	return false;
	216	}
	217
	218	static int nd_uuid_parse(struct device dev, u8 uuid_out, const char *buf,
	219	size_t len)
	220	{
	221	const char *str = buf;
	222	u8 uuid[16];
	223	int i;
	224
	225	for (i = 0; i < 16; i++) {
	226	if (!isxdigit(str[0]) \|\| !isxdigit(str[1])) {
	227	dev_dbg(dev, "%s: pos: %d buf[%zd]: %c buf[%zd]: %c\n",
	228	__func__, i, str - buf, str[0],
	229	str + 1 - buf, str[1]);
	230	return -EINVAL;
	231	}
	232
	233	uuid[i] = (hex_to_bin(str[0]) << 4) \| hex_to_bin(str[1]);
	234	str += 2;
	235	if (is_uuid_sep(*str))
	236	str++;
	237	}
	238
	239	memcpy(uuid_out, uuid, sizeof(uuid));
	240	return 0;
	241	}
	242
	243	/**
	244	* nd_uuid_store: common implementation for writing 'uuid' sysfs attributes
	245	* @dev: container device for the uuid property
	246	* @uuid_out: uuid buffer to replace
	247	* @buf: raw sysfs buffer to parse
	248	*
	249	* Enforce that uuids can only be changed while the device is disabled
	250	* (driver detached)
	251	* LOCKING: expects device_lock() is held on entry
	252	*/
	253	int nd_uuid_store(struct device dev, u8 uuid_out, const char buf,
	254	size_t len)
	255	{
	256	u8 uuid[16];
	257	int rc;
	258
	259	if (dev->driver)
	260	return -EBUSY;
	261
	262	rc = nd_uuid_parse(dev, uuid, buf, len);
	263	if (rc)
	264	return rc;
	265
	266	kfree(*uuid_out);
	267	*uuid_out = kmemdup(uuid, sizeof(uuid), GFP_KERNEL);
	268	if (!(*uuid_out))
	269	return -ENOMEM;
	270
	271	return 0;
	272	}
	273
1b40e09a DW	274	ssize_t nd_sector_size_show(unsigned long current_lbasize,
	275	const unsigned long supported, char buf)
	276	{
	277	ssize_t len = 0;
	278	int i;
	279
	280	for (i = 0; supported[i]; i++)
	281	if (current_lbasize == supported[i])
	282	len += sprintf(buf + len, "[%ld] ", supported[i]);
	283	else
	284	len += sprintf(buf + len, "%ld ", supported[i]);
	285	len += sprintf(buf + len, "\n");
	286	return len;
	287	}
	288
	289	ssize_t nd_sector_size_store(struct device dev, const char buf,
	290	unsigned long current_lbasize, const unsigned long supported)
	291	{
	292	unsigned long lbasize;
	293	int rc, i;
	294
	295	if (dev->driver)
	296	return -EBUSY;
	297
	298	rc = kstrtoul(buf, 0, &lbasize);
	299	if (rc)
	300	return rc;
	301
	302	for (i = 0; supported[i]; i++)
	303	if (lbasize == supported[i])
	304	break;
	305
	306	if (supported[i]) {
	307	*current_lbasize = lbasize;
	308	return 0;
	309	} else {
	310	return -EINVAL;
	311	}
	312	}
	313
f0dc089c DW	314	void __nd_iostat_start(struct bio bio, unsigned long start)
	315	{
	316	struct gendisk *disk = bio->bi_bdev->bd_disk;
	317	const int rw = bio_data_dir(bio);
	318	int cpu = part_stat_lock();
	319
	320	*start = jiffies;
	321	part_round_stats(cpu, &disk->part0);
	322	part_stat_inc(cpu, &disk->part0, ios[rw]);
	323	part_stat_add(cpu, &disk->part0, sectors[rw], bio_sectors(bio));
	324	part_inc_in_flight(&disk->part0, rw);
	325	part_stat_unlock();
	326	}
	327	EXPORT_SYMBOL(__nd_iostat_start);
	328
	329	void nd_iostat_end(struct bio *bio, unsigned long start)
	330	{
	331	struct gendisk *disk = bio->bi_bdev->bd_disk;
	332	unsigned long duration = jiffies - start;
	333	const int rw = bio_data_dir(bio);
	334	int cpu = part_stat_lock();
	335
	336	part_stat_add(cpu, &disk->part0, ticks[rw], duration);
	337	part_round_stats(cpu, &disk->part0);
	338	part_dec_in_flight(&disk->part0, rw);
	339	part_stat_unlock();
	340	}
	341	EXPORT_SYMBOL(nd_iostat_end);
	342
62232e45 DW	343	static ssize_t commands_show(struct device *dev,
	344	struct device_attribute attr, char buf)
	345	{
	346	int cmd, len = 0;
	347	struct nvdimm_bus *nvdimm_bus = to_nvdimm_bus(dev);
	348	struct nvdimm_bus_descriptor *nd_desc = nvdimm_bus->nd_desc;
	349
e3654eca	350	for_each_set_bit(cmd, &nd_desc->cmd_mask, BITS_PER_LONG)
62232e45 DW	351	len += sprintf(buf + len, "%s ", nvdimm_bus_cmd_name(cmd));
	352	len += sprintf(buf + len, "\n");
	353	return len;
	354	}
	355	static DEVICE_ATTR_RO(commands);
	356
45def22c DW	357	static const char nvdimm_bus_provider(struct nvdimm_bus nvdimm_bus)
	358	{
	359	struct nvdimm_bus_descriptor *nd_desc = nvdimm_bus->nd_desc;
	360	struct device *parent = nvdimm_bus->dev.parent;
	361
	362	if (nd_desc->provider_name)
	363	return nd_desc->provider_name;
	364	else if (parent)
	365	return dev_name(parent);
	366	else
	367	return "unknown";
	368	}
	369
	370	static ssize_t provider_show(struct device *dev,
	371	struct device_attribute attr, char buf)
	372	{
	373	struct nvdimm_bus *nvdimm_bus = to_nvdimm_bus(dev);
	374
	375	return sprintf(buf, "%s\n", nvdimm_bus_provider(nvdimm_bus));
	376	}
	377	static DEVICE_ATTR_RO(provider);
	378
4d88a97a DW	379	static int flush_namespaces(struct device dev, void data)
	380	{
	381	device_lock(dev);
	382	device_unlock(dev);
	383	return 0;
	384	}
	385
	386	static int flush_regions_dimms(struct device dev, void data)
	387	{
	388	device_lock(dev);
	389	device_unlock(dev);
	390	device_for_each_child(dev, NULL, flush_namespaces);
	391	return 0;
	392	}
	393
	394	static ssize_t wait_probe_show(struct device *dev,
	395	struct device_attribute attr, char buf)
	396	{
7ae0fa43 DW	397	struct nvdimm_bus *nvdimm_bus = to_nvdimm_bus(dev);
	398	struct nvdimm_bus_descriptor *nd_desc = nvdimm_bus->nd_desc;
	399	int rc;
	400
	401	if (nd_desc->flush_probe) {
	402	rc = nd_desc->flush_probe(nd_desc);
	403	if (rc)
	404	return rc;
	405	}
4d88a97a DW	406	nd_synchronize();
	407	device_for_each_child(dev, NULL, flush_regions_dimms);
	408	return sprintf(buf, "1\n");
	409	}
	410	static DEVICE_ATTR_RO(wait_probe);
	411
45def22c	412	static struct attribute *nvdimm_bus_attributes[] = {
62232e45	413	&dev_attr_commands.attr,
4d88a97a	414	&dev_attr_wait_probe.attr,
45def22c DW	415	&dev_attr_provider.attr,
	416	NULL,
	417	};
	418
	419	struct attribute_group nvdimm_bus_attribute_group = {
	420	.attrs = nvdimm_bus_attributes,
	421	};
	422	EXPORT_SYMBOL_GPL(nvdimm_bus_attribute_group);
	423
b95f5f43	424	static void set_badblock(struct badblocks *bb, sector_t s, int num)
87ba05df	425	{
b95f5f43	426	dev_dbg(bb->dev, "Found a poison range (0x%llx, 0x%llx)\n",
87ba05df DW	427	(u64) s * 512, (u64) num * 512);
87ba05df DW	428	/* this isn't an error as the hardware will still throw an exception */
b95f5f43 DW	429	if (badblocks_set(bb, s, num, 1))
b95f5f43 DW	430	dev_info_once(bb->dev, "%s: failed for sector %llx\n",
87ba05df DW	431	__func__, (u64) s);
	432	}
	433
0caeef63 VV	434	/**
0caeef63 VV	435	* __add_badblock_range() - Convert a physical address range to bad sectors
b95f5f43	436	* @bb: badblocks instance to populate
0caeef63 VV	437	* @ns_offset: namespace offset where the error range begins (in bytes)
	438	* @len: number of bytes of poison to be added
	439	*
	440	* This assumes that the range provided with (ns_offset, len) is within
	441	* the bounds of physical addresses for this namespace, i.e. lies in the
	442	* interval [ns_start, ns_start + ns_size)
	443	*/
b95f5f43	444	static void __add_badblock_range(struct badblocks *bb, u64 ns_offset, u64 len)
0caeef63	445	{
b95f5f43	446	const unsigned int sector_size = 512;
0caeef63 VV	447	sector_t start_sector;
	448	u64 num_sectors;
	449	u32 rem;
0caeef63 VV	450
	451	start_sector = div_u64(ns_offset, sector_size);
	452	num_sectors = div_u64_rem(len, sector_size, &rem);
	453	if (rem)
	454	num_sectors++;
	455
0caeef63 VV	456	if (unlikely(num_sectors > (u64)INT_MAX)) {
	457	u64 remaining = num_sectors;
	458	sector_t s = start_sector;
	459
	460	while (remaining) {
	461	int done = min_t(u64, remaining, INT_MAX);
	462
b95f5f43	463	set_badblock(bb, s, done);
0caeef63 VV	464	remaining -= done;
	465	s += done;
	466	}
0caeef63	467	} else
b95f5f43	468	set_badblock(bb, start_sector, num_sectors);
0caeef63 VV	469	}
0caeef63 VV	470
a3901802 DW	471	static void badblocks_populate(struct list_head *poison_list,
a3901802 DW	472	struct badblocks bb, const struct resource res)
0caeef63	473	{
0caeef63	474	struct nd_poison *pl;
0caeef63	475
0caeef63	476	if (list_empty(poison_list))
b95f5f43	477	return;
0caeef63 VV	478
	479	list_for_each_entry(pl, poison_list, list) {
	480	u64 pl_end = pl->start + pl->length - 1;
	481
	482	/* Discard intervals with no intersection */
5faecf4e	483	if (pl_end < res->start)
0caeef63	484	continue;
5faecf4e	485	if (pl->start > res->end)
0caeef63 VV	486	continue;
0caeef63 VV	487	/* Deal with any overlap after start of the namespace */
5faecf4e	488	if (pl->start >= res->start) {
0caeef63 VV	489	u64 start = pl->start;
	490	u64 len;
	491
5faecf4e	492	if (pl_end <= res->end)
0caeef63 VV	493	len = pl->length;
0caeef63 VV	494	else
5faecf4e DW	495	len = res->start + resource_size(res)
	496	- pl->start;
	497	__add_badblock_range(bb, start - res->start, len);
0caeef63 VV	498	continue;
	499	}
	500	/* Deal with overlap for poison starting before the namespace */
5faecf4e	501	if (pl->start < res->start) {
0caeef63 VV	502	u64 len;
0caeef63 VV	503
5faecf4e DW	504	if (pl_end < res->end)
5faecf4e DW	505	len = pl->start + pl->length - res->start;
0caeef63	506	else
5faecf4e	507	len = resource_size(res);
b95f5f43	508	__add_badblock_range(bb, 0, len);
0caeef63 VV	509	}
0caeef63 VV	510	}
0caeef63	511	}
5faecf4e DW	512
5faecf4e DW	513	/**
a3901802 DW	514	* nvdimm_badblocks_populate() - Convert a list of poison ranges to badblocks
	515	* @region: parent region of the range to interrogate
	516	* @bb: badblocks instance to populate
	517	* @res: resource range to consider
5faecf4e	518	*
a3901802 DW	519	* The poison list generated during bus initialization may contain
	520	* multiple, possibly overlapping physical address ranges. Compare each
	521	* of these ranges to the resource range currently being initialized,
	522	* and add badblocks entries for all matching sub-ranges
5faecf4e	523	*/
a3901802 DW	524	void nvdimm_badblocks_populate(struct nd_region *nd_region,
a3901802 DW	525	struct badblocks bb, const struct resource res)
5faecf4e	526	{
5faecf4e DW	527	struct nvdimm_bus *nvdimm_bus;
5faecf4e DW	528	struct list_head *poison_list;
5faecf4e	529
a3901802 DW	530	if (!is_nd_pmem(&nd_region->dev)) {
	531	dev_WARN_ONCE(&nd_region->dev, 1,
	532	"%s only valid for pmem regions\n", __func__);
	533	return;
	534	}
	535	nvdimm_bus = walk_to_nvdimm_bus(&nd_region->dev);
5faecf4e DW	536	poison_list = &nvdimm_bus->poison_list;
	537
	538	nvdimm_bus_lock(&nvdimm_bus->dev);
a3901802	539	badblocks_populate(poison_list, bb, res);
5faecf4e DW	540	nvdimm_bus_unlock(&nvdimm_bus->dev);
5faecf4e DW	541	}
a3901802	542	EXPORT_SYMBOL_GPL(nvdimm_badblocks_populate);
0caeef63	543
5faecf4e	544	static int add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length)
0caeef63 VV	545	{
	546	struct nd_poison *pl;
	547
	548	pl = kzalloc(sizeof(*pl), GFP_KERNEL);
	549	if (!pl)
	550	return -ENOMEM;
	551
	552	pl->start = addr;
	553	pl->length = length;
	554	list_add_tail(&pl->list, &nvdimm_bus->poison_list);
	555
	556	return 0;
	557	}
	558
5faecf4e	559	static int bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length)
0caeef63 VV	560	{
	561	struct nd_poison *pl;
	562
	563	if (list_empty(&nvdimm_bus->poison_list))
5faecf4e	564	return add_poison(nvdimm_bus, addr, length);
0caeef63 VV	565
	566	/*
	567	* There is a chance this is a duplicate, check for those first.
	568	* This will be the common case as ARS_STATUS returns all known
	569	* errors in the SPA space, and we can't query it per region
	570	*/
	571	list_for_each_entry(pl, &nvdimm_bus->poison_list, list)
	572	if (pl->start == addr) {
	573	/* If length has changed, update this list entry */
	574	if (pl->length != length)
	575	pl->length = length;
	576	return 0;
	577	}
	578
	579	/*
	580	* If not a duplicate or a simple length update, add the entry as is,
	581	* as any overlapping ranges will get resolved when the list is consumed
	582	* and converted to badblocks
	583	*/
5faecf4e DW	584	return add_poison(nvdimm_bus, addr, length);
	585	}
	586
	587	int nvdimm_bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr, u64 length)
	588	{
	589	int rc;
	590
	591	nvdimm_bus_lock(&nvdimm_bus->dev);
	592	rc = bus_add_poison(nvdimm_bus, addr, length);
	593	nvdimm_bus_unlock(&nvdimm_bus->dev);
	594
	595	return rc;
0caeef63 VV	596	}
	597	EXPORT_SYMBOL_GPL(nvdimm_bus_add_poison);
	598
41cd8b70	599	#ifdef CONFIG_BLK_DEV_INTEGRITY
41cd8b70 VV	600	int nd_integrity_init(struct gendisk *disk, unsigned long meta_size)
41cd8b70 VV	601	{
0f8087ec	602	struct blk_integrity bi;
41cd8b70	603
fcae6957 VV	604	if (meta_size == 0)
	605	return 0;
	606
8729bdea JT	607	memset(&bi, 0, sizeof(bi));
8729bdea JT	608
0f8087ec MP	609	bi.tuple_size = meta_size;
	610	bi.tag_size = meta_size;
	611
25520d55	612	blk_integrity_register(disk, &bi);
41cd8b70 VV	613	blk_queue_max_integrity_segments(disk->queue, 1);
	614
	615	return 0;
	616	}
	617	EXPORT_SYMBOL(nd_integrity_init);
	618
	619	#else /* CONFIG_BLK_DEV_INTEGRITY */
	620	int nd_integrity_init(struct gendisk *disk, unsigned long meta_size)
	621	{
	622	return 0;
	623	}
	624	EXPORT_SYMBOL(nd_integrity_init);
	625
	626	#endif
	627
45def22c DW	628	static __init int libnvdimm_init(void)
45def22c DW	629	{
4d88a97a DW	630	int rc;
	631
	632	rc = nvdimm_bus_init();
	633	if (rc)
	634	return rc;
	635	rc = nvdimm_init();
	636	if (rc)
	637	goto err_dimm;
3d88002e DW	638	rc = nd_region_init();
	639	if (rc)
	640	goto err_region;
4d88a97a	641	return 0;
3d88002e DW	642	err_region:
3d88002e DW	643	nvdimm_exit();
4d88a97a DW	644	err_dimm:
	645	nvdimm_bus_exit();
	646	return rc;
45def22c DW	647	}
	648
	649	static __exit void libnvdimm_exit(void)
	650	{
	651	WARN_ON(!list_empty(&nvdimm_bus_list));
3d88002e	652	nd_region_exit();
4d88a97a	653	nvdimm_exit();
45def22c	654	nvdimm_bus_exit();
b354aba0 DW	655	nd_region_devs_exit();
b354aba0 DW	656	nvdimm_devs_exit();
45def22c DW	657	}
45def22c DW	658
b94d5230 DW	659	MODULE_LICENSE("GPL v2");
b94d5230 DW	660	MODULE_AUTHOR("Intel Corporation");
45def22c DW	661	subsys_initcall(libnvdimm_init);
45def22c DW	662	module_exit(libnvdimm_exit);