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

/*
 * Persistent Memory Driver
 *
 * Copyright (c) 2014-2015, Intel Corporation.
 * Copyright (c) 2015, Christoph Hellwig <hch@lst.de>.
 * Copyright (c) 2015, Boaz Harrosh <boaz@plexistor.com>.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope 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 <asm/cacheflush.h>
#include <linux/blkdev.h>
#include <linux/hdreg.h>
#include <linux/init.h>
#include <linux/platform_device.h>
#include <linux/module.h>
#include <linux/memory_hotplug.h>
#include <linux/moduleparam.h>
#include <linux/vmalloc.h>
#include <linux/slab.h>
#include <linux/pmem.h>
#include <linux/nd.h>
#include "pfn.h"
#include "nd.h"

struct pmem_device {
	struct request_queue	*pmem_queue;
	struct gendisk		*pmem_disk;
	struct nd_namespace_common *ndns;

	/* One contiguous memory region per device */
	phys_addr_t		phys_addr;
	/* when non-zero this device is hosting a 'pfn' instance */
	phys_addr_t		data_offset;
	void __pmem		*virt_addr;
	size_t			size;
};

static int pmem_major;

static void pmem_do_bvec(struct pmem_device *pmem, struct page *page,
			unsigned int len, unsigned int off, int rw,
			sector_t sector)
{
	void *mem = kmap_atomic(page);
	phys_addr_t pmem_off = sector * 512 + pmem->data_offset;
	void __pmem *pmem_addr = pmem->virt_addr + pmem_off;

	if (rw == READ) {
		memcpy_from_pmem(mem + off, pmem_addr, len);
		flush_dcache_page(page);
	} else {
		flush_dcache_page(page);
		memcpy_to_pmem(pmem_addr, mem + off, len);
	}

	kunmap_atomic(mem);
}

static blk_qc_t pmem_make_request(struct request_queue *q, struct bio *bio)
{
	bool do_acct;
	unsigned long start;
	struct bio_vec bvec;
	struct bvec_iter iter;
	struct block_device *bdev = bio->bi_bdev;
	struct pmem_device *pmem = bdev->bd_disk->private_data;

	do_acct = nd_iostat_start(bio, &start);
	bio_for_each_segment(bvec, bio, iter)
		pmem_do_bvec(pmem, bvec.bv_page, bvec.bv_len, bvec.bv_offset,
				bio_data_dir(bio), iter.bi_sector);
	if (do_acct)
		nd_iostat_end(bio, start);

	if (bio_data_dir(bio))
		wmb_pmem();

	bio_endio(bio);
	return BLK_QC_T_NONE;
}

static int pmem_rw_page(struct block_device *bdev, sector_t sector,
		       struct page *page, int rw)
{
	struct pmem_device *pmem = bdev->bd_disk->private_data;

	pmem_do_bvec(pmem, page, PAGE_CACHE_SIZE, 0, rw, sector);
	if (rw & WRITE)
		wmb_pmem();
	page_endio(page, rw & WRITE, 0);

	return 0;
}

static long pmem_direct_access(struct block_device *bdev, sector_t sector,
		      void __pmem **kaddr, unsigned long *pfn)
{
	struct pmem_device *pmem = bdev->bd_disk->private_data;
	resource_size_t offset = sector * 512 + pmem->data_offset;

	*kaddr = pmem->virt_addr + offset;
	*pfn = (pmem->phys_addr + offset) >> PAGE_SHIFT;

	return pmem->size - offset;
}

static const struct block_device_operations pmem_fops = {
	.owner =		THIS_MODULE,
	.rw_page =		pmem_rw_page,
	.direct_access =	pmem_direct_access,
	.revalidate_disk =	nvdimm_revalidate_disk,
};

static struct pmem_device *pmem_alloc(struct device *dev,
		struct resource *res, int id)
{
	struct pmem_device *pmem;

	pmem = devm_kzalloc(dev, sizeof(*pmem), GFP_KERNEL);
	if (!pmem)
		return ERR_PTR(-ENOMEM);

	pmem->phys_addr = res->start;
	pmem->size = resource_size(res);
	if (!arch_has_wmb_pmem())
		dev_warn(dev, "unable to guarantee persistence of writes\n");

	if (!devm_request_mem_region(dev, pmem->phys_addr, pmem->size,
			dev_name(dev))) {
		dev_warn(dev, "could not reserve region [0x%pa:0x%zx]\n",
				&pmem->phys_addr, pmem->size);
		return ERR_PTR(-EBUSY);
	}

	if (pmem_should_map_pages(dev))
		pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, res);
	else
		pmem->virt_addr = (void __pmem *) devm_memremap(dev,
				pmem->phys_addr, pmem->size,
				ARCH_MEMREMAP_PMEM);

	if (IS_ERR(pmem->virt_addr))
		return (void __force *) pmem->virt_addr;

	return pmem;
}

static void pmem_detach_disk(struct pmem_device *pmem)
{
	if (!pmem->pmem_disk)
		return;

	del_gendisk(pmem->pmem_disk);
	put_disk(pmem->pmem_disk);
	blk_cleanup_queue(pmem->pmem_queue);
}

static int pmem_attach_disk(struct device *dev,
		struct nd_namespace_common *ndns, struct pmem_device *pmem)
{
	int nid = dev_to_node(dev);
	struct gendisk *disk;

	pmem->pmem_queue = blk_alloc_queue_node(GFP_KERNEL, nid);
	if (!pmem->pmem_queue)
		return -ENOMEM;

	blk_queue_make_request(pmem->pmem_queue, pmem_make_request);
	blk_queue_physical_block_size(pmem->pmem_queue, PAGE_SIZE);
	blk_queue_max_hw_sectors(pmem->pmem_queue, UINT_MAX);
	blk_queue_bounce_limit(pmem->pmem_queue, BLK_BOUNCE_ANY);
	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, pmem->pmem_queue);

	disk = alloc_disk_node(0, nid);
	if (!disk) {
		blk_cleanup_queue(pmem->pmem_queue);
		return -ENOMEM;
	}

	disk->major		= pmem_major;
	disk->first_minor	= 0;
	disk->fops		= &pmem_fops;
	disk->private_data	= pmem;
	disk->queue		= pmem->pmem_queue;
	disk->flags		= GENHD_FL_EXT_DEVT;
	nvdimm_namespace_disk_name(ndns, disk->disk_name);
	disk->driverfs_dev = dev;
	set_capacity(disk, (pmem->size - pmem->data_offset) / 512);
	pmem->pmem_disk = disk;

	add_disk(disk);
	revalidate_disk(disk);

	return 0;
}

static int pmem_rw_bytes(struct nd_namespace_common *ndns,
		resource_size_t offset, void *buf, size_t size, int rw)
{
	struct pmem_device *pmem = dev_get_drvdata(ndns->claim);

	if (unlikely(offset + size > pmem->size)) {
		dev_WARN_ONCE(&ndns->dev, 1, "request out of range\n");
		return -EFAULT;
	}

	if (rw == READ)
		memcpy_from_pmem(buf, pmem->virt_addr + offset, size);
	else {
		memcpy_to_pmem(pmem->virt_addr + offset, buf, size);
		wmb_pmem();
	}

	return 0;
}

static int nd_pfn_init(struct nd_pfn *nd_pfn)
{
	struct nd_pfn_sb *pfn_sb = kzalloc(sizeof(*pfn_sb), GFP_KERNEL);
	struct pmem_device *pmem = dev_get_drvdata(&nd_pfn->dev);
	struct nd_namespace_common *ndns = nd_pfn->ndns;
	struct nd_region *nd_region;
	unsigned long npfns;
	phys_addr_t offset;
	u64 checksum;
	int rc;

	if (!pfn_sb)
		return -ENOMEM;

	nd_pfn->pfn_sb = pfn_sb;
	rc = nd_pfn_validate(nd_pfn);
	if (rc == 0 || rc == -EBUSY)
		return rc;

	nd_region = to_nd_region(nd_pfn->dev.parent);
	if (nd_region->ro) {
		dev_info(&nd_pfn->dev,
				"%s is read-only, unable to init metadata\n",
				dev_name(&nd_region->dev));
		goto err;
	}

	memset(pfn_sb, 0, sizeof(*pfn_sb));
	npfns = (pmem->size - SZ_8K) / SZ_4K;
	/*
	 * Note, we use 64 here for the standard size of struct page,
	 * debugging options may cause it to be larger in which case the
	 * implementation will limit the pfns advertised through
	 * ->direct_access() to those that are included in the memmap.
	 */
	if (nd_pfn->mode == PFN_MODE_PMEM)
		offset = ALIGN(SZ_8K + 64 * npfns, PMD_SIZE);
	else if (nd_pfn->mode == PFN_MODE_RAM)
		offset = SZ_8K;
	else
		goto err;

	npfns = (pmem->size - offset) / SZ_4K;
	pfn_sb->mode = cpu_to_le32(nd_pfn->mode);
	pfn_sb->dataoff = cpu_to_le64(offset);
	pfn_sb->npfns = cpu_to_le64(npfns);
	memcpy(pfn_sb->signature, PFN_SIG, PFN_SIG_LEN);
	memcpy(pfn_sb->uuid, nd_pfn->uuid, 16);
	pfn_sb->version_major = cpu_to_le16(1);
	checksum = nd_sb_checksum((struct nd_gen_sb *) pfn_sb);
	pfn_sb->checksum = cpu_to_le64(checksum);

	rc = nvdimm_write_bytes(ndns, SZ_4K, pfn_sb, sizeof(*pfn_sb));
	if (rc)
		goto err;

	return 0;
 err:
	nd_pfn->pfn_sb = NULL;
	kfree(pfn_sb);
	return -ENXIO;
}

static int nvdimm_namespace_detach_pfn(struct nd_namespace_common *ndns)
{
	struct nd_pfn *nd_pfn = to_nd_pfn(ndns->claim);
	struct pmem_device *pmem;

	/* free pmem disk */
	pmem = dev_get_drvdata(&nd_pfn->dev);
	pmem_detach_disk(pmem);

	/* release nd_pfn resources */
	kfree(nd_pfn->pfn_sb);
	nd_pfn->pfn_sb = NULL;

	return 0;
}

static int nvdimm_namespace_attach_pfn(struct nd_namespace_common *ndns)
{
	struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
	struct nd_pfn *nd_pfn = to_nd_pfn(ndns->claim);
	struct device *dev = &nd_pfn->dev;
	struct vmem_altmap *altmap;
	struct nd_region *nd_region;
	struct nd_pfn_sb *pfn_sb;
	struct pmem_device *pmem;
	phys_addr_t offset;
	int rc;

	if (!nd_pfn->uuid || !nd_pfn->ndns)
		return -ENODEV;

	nd_region = to_nd_region(dev->parent);
	rc = nd_pfn_init(nd_pfn);
	if (rc)
		return rc;

	pfn_sb = nd_pfn->pfn_sb;
	offset = le64_to_cpu(pfn_sb->dataoff);
	nd_pfn->mode = le32_to_cpu(nd_pfn->pfn_sb->mode);
	if (nd_pfn->mode == PFN_MODE_RAM) {
		if (offset != SZ_8K)
			return -EINVAL;
		nd_pfn->npfns = le64_to_cpu(pfn_sb->npfns);
		altmap = NULL;
	} else {
		rc = -ENXIO;
		goto err;
	}

	/* establish pfn range for lookup, and switch to direct map */
	pmem = dev_get_drvdata(dev);
	devm_memunmap(dev, (void __force *) pmem->virt_addr);
	pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, &nsio->res);
	if (IS_ERR(pmem->virt_addr)) {
		rc = PTR_ERR(pmem->virt_addr);
		goto err;
	}

	/* attach pmem disk in "pfn-mode" */
	pmem->data_offset = offset;
	rc = pmem_attach_disk(dev, ndns, pmem);
	if (rc)
		goto err;

	return rc;
 err:
	nvdimm_namespace_detach_pfn(ndns);
	return rc;
}

static int nd_pmem_probe(struct device *dev)
{
	struct nd_region *nd_region = to_nd_region(dev->parent);
	struct nd_namespace_common *ndns;
	struct nd_namespace_io *nsio;
	struct pmem_device *pmem;

	ndns = nvdimm_namespace_common_probe(dev);
	if (IS_ERR(ndns))
		return PTR_ERR(ndns);

	nsio = to_nd_namespace_io(&ndns->dev);
	pmem = pmem_alloc(dev, &nsio->res, nd_region->id);
	if (IS_ERR(pmem))
		return PTR_ERR(pmem);

	pmem->ndns = ndns;
	dev_set_drvdata(dev, pmem);
	ndns->rw_bytes = pmem_rw_bytes;

	if (is_nd_btt(dev))
		return nvdimm_namespace_attach_btt(ndns);

	if (is_nd_pfn(dev))
		return nvdimm_namespace_attach_pfn(ndns);

	if (nd_btt_probe(ndns, pmem) == 0) {
		/* we'll come back as btt-pmem */
		return -ENXIO;
	}

	if (nd_pfn_probe(ndns, pmem) == 0) {
		/* we'll come back as pfn-pmem */
		return -ENXIO;
	}

	return pmem_attach_disk(dev, ndns, pmem);
}

static int nd_pmem_remove(struct device *dev)
{
	struct pmem_device *pmem = dev_get_drvdata(dev);

	if (is_nd_btt(dev))
		nvdimm_namespace_detach_btt(pmem->ndns);
	else if (is_nd_pfn(dev))
		nvdimm_namespace_detach_pfn(pmem->ndns);
	else
		pmem_detach_disk(pmem);

	return 0;
}

MODULE_ALIAS("pmem");
MODULE_ALIAS_ND_DEVICE(ND_DEVICE_NAMESPACE_IO);
MODULE_ALIAS_ND_DEVICE(ND_DEVICE_NAMESPACE_PMEM);
static struct nd_device_driver nd_pmem_driver = {
	.probe = nd_pmem_probe,
	.remove = nd_pmem_remove,
	.drv = {
		.name = "nd_pmem",
	},
	.type = ND_DRIVER_NAMESPACE_IO | ND_DRIVER_NAMESPACE_PMEM,
};

static int __init pmem_init(void)
{
	int error;

	pmem_major = register_blkdev(0, "pmem");
	if (pmem_major < 0)
		return pmem_major;

	error = nd_driver_register(&nd_pmem_driver);
	if (error) {
		unregister_blkdev(pmem_major, "pmem");
		return error;
	}

	return 0;
}
module_init(pmem_init);

static void pmem_exit(void)
{
	driver_unregister(&nd_pmem_driver.drv);
	unregister_blkdev(pmem_major, "pmem");
}
module_exit(pmem_exit);

MODULE_AUTHOR("Ross Zwisler <ross.zwisler@linux.intel.com>");
MODULE_LICENSE("GPL v2");
Commit	Line	Data
9e853f23 RZ	1	/*
	2	* Persistent Memory Driver
	3	*
9f53f9fa	4	* Copyright (c) 2014-2015, Intel Corporation.
9e853f23 RZ	5	* Copyright (c) 2015, Christoph Hellwig <hch@lst.de>.
	6	* Copyright (c) 2015, Boaz Harrosh <boaz@plexistor.com>.
	7	*
	8	* This program is free software; you can redistribute it and/or modify it
	9	* under the terms and conditions of the GNU General Public License,
	10	* version 2, as published by the Free Software Foundation.
	11	*
	12	* This program is distributed in the hope it will be useful, but WITHOUT
	13	* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
	14	* FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
	15	* more details.
	16	*/
	17
	18	#include <asm/cacheflush.h>
	19	#include <linux/blkdev.h>
	20	#include <linux/hdreg.h>
	21	#include <linux/init.h>
	22	#include <linux/platform_device.h>
	23	#include <linux/module.h>
32ab0a3f	24	#include <linux/memory_hotplug.h>
9e853f23	25	#include <linux/moduleparam.h>
32ab0a3f	26	#include <linux/vmalloc.h>
9e853f23	27	#include <linux/slab.h>
61031952	28	#include <linux/pmem.h>
9f53f9fa	29	#include <linux/nd.h>
32ab0a3f	30	#include "pfn.h"
9f53f9fa	31	#include "nd.h"
9e853f23 RZ	32
	33	struct pmem_device {
	34	struct request_queue *pmem_queue;
	35	struct gendisk *pmem_disk;
32ab0a3f	36	struct nd_namespace_common *ndns;
9e853f23 RZ	37
	38	/* One contiguous memory region per device */
	39	phys_addr_t phys_addr;
32ab0a3f DW	40	/* when non-zero this device is hosting a 'pfn' instance */
32ab0a3f DW	41	phys_addr_t data_offset;
61031952	42	void __pmem *virt_addr;
9e853f23 RZ	43	size_t size;
	44	};
	45
	46	static int pmem_major;
9e853f23 RZ	47
	48	static void pmem_do_bvec(struct pmem_device pmem, struct page page,
	49	unsigned int len, unsigned int off, int rw,
	50	sector_t sector)
	51	{
	52	void *mem = kmap_atomic(page);
32ab0a3f	53	phys_addr_t pmem_off = sector * 512 + pmem->data_offset;
61031952	54	void __pmem *pmem_addr = pmem->virt_addr + pmem_off;
9e853f23 RZ	55
9e853f23 RZ	56	if (rw == READ) {
61031952	57	memcpy_from_pmem(mem + off, pmem_addr, len);
9e853f23 RZ	58	flush_dcache_page(page);
	59	} else {
	60	flush_dcache_page(page);
61031952	61	memcpy_to_pmem(pmem_addr, mem + off, len);
9e853f23 RZ	62	}
	63
	64	kunmap_atomic(mem);
	65	}
	66
dece1635	67	static blk_qc_t pmem_make_request(struct request_queue q, struct bio bio)
9e853f23	68	{
f0dc089c DW	69	bool do_acct;
f0dc089c DW	70	unsigned long start;
9e853f23	71	struct bio_vec bvec;
9e853f23	72	struct bvec_iter iter;
edc870e5 DW	73	struct block_device *bdev = bio->bi_bdev;
edc870e5 DW	74	struct pmem_device *pmem = bdev->bd_disk->private_data;
9e853f23	75
f0dc089c	76	do_acct = nd_iostat_start(bio, &start);
edc870e5	77	bio_for_each_segment(bvec, bio, iter)
9e853f23	78	pmem_do_bvec(pmem, bvec.bv_page, bvec.bv_len, bvec.bv_offset,
edc870e5	79	bio_data_dir(bio), iter.bi_sector);
f0dc089c DW	80	if (do_acct)
f0dc089c DW	81	nd_iostat_end(bio, start);
61031952 RZ	82
	83	if (bio_data_dir(bio))
	84	wmb_pmem();
	85
4246a0b6	86	bio_endio(bio);
dece1635	87	return BLK_QC_T_NONE;
9e853f23 RZ	88	}
	89
	90	static int pmem_rw_page(struct block_device *bdev, sector_t sector,
	91	struct page *page, int rw)
	92	{
	93	struct pmem_device *pmem = bdev->bd_disk->private_data;
	94
	95	pmem_do_bvec(pmem, page, PAGE_CACHE_SIZE, 0, rw, sector);
ba8fe0f8 RZ	96	if (rw & WRITE)
ba8fe0f8 RZ	97	wmb_pmem();
9e853f23 RZ	98	page_endio(page, rw & WRITE, 0);
	99
	100	return 0;
	101	}
	102
	103	static long pmem_direct_access(struct block_device *bdev, sector_t sector,
cb389b9c	104	void __pmem *kaddr, unsigned long pfn)
9e853f23 RZ	105	{
9e853f23 RZ	106	struct pmem_device *pmem = bdev->bd_disk->private_data;
32ab0a3f	107	resource_size_t offset = sector * 512 + pmem->data_offset;
589e75d1	108
e2e05394	109	*kaddr = pmem->virt_addr + offset;
9e853f23 RZ	110	*pfn = (pmem->phys_addr + offset) >> PAGE_SHIFT;
9e853f23 RZ	111
589e75d1	112	return pmem->size - offset;
9e853f23 RZ	113	}
	114
	115	static const struct block_device_operations pmem_fops = {
	116	.owner = THIS_MODULE,
	117	.rw_page = pmem_rw_page,
	118	.direct_access = pmem_direct_access,
58138820	119	.revalidate_disk = nvdimm_revalidate_disk,
9e853f23 RZ	120	};
9e853f23 RZ	121
9f53f9fa DW	122	static struct pmem_device pmem_alloc(struct device dev,
9f53f9fa DW	123	struct resource *res, int id)
9e853f23 RZ	124	{
9e853f23 RZ	125	struct pmem_device *pmem;
9e853f23	126
708ab62b	127	pmem = devm_kzalloc(dev, sizeof(*pmem), GFP_KERNEL);
9e853f23	128	if (!pmem)
8c2f7e86	129	return ERR_PTR(-ENOMEM);
9e853f23 RZ	130
	131	pmem->phys_addr = res->start;
	132	pmem->size = resource_size(res);
96601adb	133	if (!arch_has_wmb_pmem())
61031952	134	dev_warn(dev, "unable to guarantee persistence of writes\n");
9e853f23	135
708ab62b CH	136	if (!devm_request_mem_region(dev, pmem->phys_addr, pmem->size,
708ab62b CH	137	dev_name(dev))) {
9f53f9fa DW	138	dev_warn(dev, "could not reserve region [0x%pa:0x%zx]\n",
9f53f9fa DW	139	&pmem->phys_addr, pmem->size);
8c2f7e86	140	return ERR_PTR(-EBUSY);
9e853f23 RZ	141	}
9e853f23 RZ	142
b36f4761 DW	143	if (pmem_should_map_pages(dev))
	144	pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, res);
	145	else
a639315d DW	146	pmem->virt_addr = (void __pmem *) devm_memremap(dev,
	147	pmem->phys_addr, pmem->size,
	148	ARCH_MEMREMAP_PMEM);
b36f4761 DW	149
	150	if (IS_ERR(pmem->virt_addr))
	151	return (void __force *) pmem->virt_addr;
8c2f7e86 DW	152
	153	return pmem;
	154	}
	155
	156	static void pmem_detach_disk(struct pmem_device *pmem)
	157	{
32ab0a3f DW	158	if (!pmem->pmem_disk)
	159	return;
	160
8c2f7e86 DW	161	del_gendisk(pmem->pmem_disk);
	162	put_disk(pmem->pmem_disk);
	163	blk_cleanup_queue(pmem->pmem_queue);
	164	}
	165
32ab0a3f DW	166	static int pmem_attach_disk(struct device *dev,
32ab0a3f DW	167	struct nd_namespace_common ndns, struct pmem_device pmem)
8c2f7e86	168	{
538ea4aa	169	int nid = dev_to_node(dev);
8c2f7e86	170	struct gendisk *disk;
9e853f23	171
538ea4aa	172	pmem->pmem_queue = blk_alloc_queue_node(GFP_KERNEL, nid);
9e853f23	173	if (!pmem->pmem_queue)
8c2f7e86	174	return -ENOMEM;
9e853f23 RZ	175
9e853f23 RZ	176	blk_queue_make_request(pmem->pmem_queue, pmem_make_request);
6b47496a	177	blk_queue_physical_block_size(pmem->pmem_queue, PAGE_SIZE);
43d3fa3a	178	blk_queue_max_hw_sectors(pmem->pmem_queue, UINT_MAX);
9e853f23	179	blk_queue_bounce_limit(pmem->pmem_queue, BLK_BOUNCE_ANY);
0f51c4fa	180	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, pmem->pmem_queue);
9e853f23	181
538ea4aa	182	disk = alloc_disk_node(0, nid);
8c2f7e86 DW	183	if (!disk) {
	184	blk_cleanup_queue(pmem->pmem_queue);
	185	return -ENOMEM;
	186	}
9e853f23	187
9e853f23	188	disk->major = pmem_major;
9f53f9fa	189	disk->first_minor = 0;
9e853f23 RZ	190	disk->fops = &pmem_fops;
	191	disk->private_data = pmem;
	192	disk->queue = pmem->pmem_queue;
	193	disk->flags = GENHD_FL_EXT_DEVT;
5212e11f	194	nvdimm_namespace_disk_name(ndns, disk->disk_name);
32ab0a3f DW	195	disk->driverfs_dev = dev;
32ab0a3f DW	196	set_capacity(disk, (pmem->size - pmem->data_offset) / 512);
9e853f23 RZ	197	pmem->pmem_disk = disk;
	198
	199	add_disk(disk);
58138820	200	revalidate_disk(disk);
9e853f23	201
8c2f7e86 DW	202	return 0;
8c2f7e86 DW	203	}
9e853f23	204
8c2f7e86 DW	205	static int pmem_rw_bytes(struct nd_namespace_common *ndns,
	206	resource_size_t offset, void *buf, size_t size, int rw)
	207	{
	208	struct pmem_device *pmem = dev_get_drvdata(ndns->claim);
	209
	210	if (unlikely(offset + size > pmem->size)) {
	211	dev_WARN_ONCE(&ndns->dev, 1, "request out of range\n");
	212	return -EFAULT;
	213	}
	214
	215	if (rw == READ)
61031952 RZ	216	memcpy_from_pmem(buf, pmem->virt_addr + offset, size);
	217	else {
	218	memcpy_to_pmem(pmem->virt_addr + offset, buf, size);
	219	wmb_pmem();
	220	}
8c2f7e86 DW	221
	222	return 0;
	223	}
	224
32ab0a3f DW	225	static int nd_pfn_init(struct nd_pfn *nd_pfn)
	226	{
	227	struct nd_pfn_sb pfn_sb = kzalloc(sizeof(pfn_sb), GFP_KERNEL);
	228	struct pmem_device *pmem = dev_get_drvdata(&nd_pfn->dev);
	229	struct nd_namespace_common *ndns = nd_pfn->ndns;
	230	struct nd_region *nd_region;
	231	unsigned long npfns;
	232	phys_addr_t offset;
	233	u64 checksum;
	234	int rc;
	235
	236	if (!pfn_sb)
	237	return -ENOMEM;
	238
	239	nd_pfn->pfn_sb = pfn_sb;
	240	rc = nd_pfn_validate(nd_pfn);
	241	if (rc == 0 \|\| rc == -EBUSY)
	242	return rc;
	243
32ab0a3f DW	244	nd_region = to_nd_region(nd_pfn->dev.parent);
	245	if (nd_region->ro) {
	246	dev_info(&nd_pfn->dev,
	247	"%s is read-only, unable to init metadata\n",
	248	dev_name(&nd_region->dev));
	249	goto err;
	250	}
	251
	252	memset(pfn_sb, 0, sizeof(*pfn_sb));
	253	npfns = (pmem->size - SZ_8K) / SZ_4K;
	254	/*
	255	* Note, we use 64 here for the standard size of struct page,
	256	* debugging options may cause it to be larger in which case the
	257	* implementation will limit the pfns advertised through
	258	* ->direct_access() to those that are included in the memmap.
	259	*/
	260	if (nd_pfn->mode == PFN_MODE_PMEM)
	261	offset = ALIGN(SZ_8K + 64 * npfns, PMD_SIZE);
	262	else if (nd_pfn->mode == PFN_MODE_RAM)
	263	offset = SZ_8K;
	264	else
	265	goto err;
	266
	267	npfns = (pmem->size - offset) / SZ_4K;
	268	pfn_sb->mode = cpu_to_le32(nd_pfn->mode);
	269	pfn_sb->dataoff = cpu_to_le64(offset);
	270	pfn_sb->npfns = cpu_to_le64(npfns);
	271	memcpy(pfn_sb->signature, PFN_SIG, PFN_SIG_LEN);
	272	memcpy(pfn_sb->uuid, nd_pfn->uuid, 16);
	273	pfn_sb->version_major = cpu_to_le16(1);
	274	checksum = nd_sb_checksum((struct nd_gen_sb *) pfn_sb);
	275	pfn_sb->checksum = cpu_to_le64(checksum);
	276
	277	rc = nvdimm_write_bytes(ndns, SZ_4K, pfn_sb, sizeof(*pfn_sb));
	278	if (rc)
	279	goto err;
	280
	281	return 0;
	282	err:
	283	nd_pfn->pfn_sb = NULL;
	284	kfree(pfn_sb);
	285	return -ENXIO;
	286	}
	287
	288	static int nvdimm_namespace_detach_pfn(struct nd_namespace_common *ndns)
	289	{
	290	struct nd_pfn *nd_pfn = to_nd_pfn(ndns->claim);
	291	struct pmem_device *pmem;
	292
	293	/* free pmem disk */
	294	pmem = dev_get_drvdata(&nd_pfn->dev);
	295	pmem_detach_disk(pmem);
	296
	297	/* release nd_pfn resources */
	298	kfree(nd_pfn->pfn_sb);
	299	nd_pfn->pfn_sb = NULL;
	300
	301	return 0;
	302	}
	303
	304	static int nvdimm_namespace_attach_pfn(struct nd_namespace_common *ndns)
9e853f23	305	{
32ab0a3f DW	306	struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
	307	struct nd_pfn *nd_pfn = to_nd_pfn(ndns->claim);
	308	struct device *dev = &nd_pfn->dev;
	309	struct vmem_altmap *altmap;
	310	struct nd_region *nd_region;
	311	struct nd_pfn_sb *pfn_sb;
	312	struct pmem_device *pmem;
	313	phys_addr_t offset;
	314	int rc;
	315
	316	if (!nd_pfn->uuid \|\| !nd_pfn->ndns)
	317	return -ENODEV;
	318
	319	nd_region = to_nd_region(dev->parent);
	320	rc = nd_pfn_init(nd_pfn);
	321	if (rc)
	322	return rc;
	323
32ab0a3f DW	324	pfn_sb = nd_pfn->pfn_sb;
	325	offset = le64_to_cpu(pfn_sb->dataoff);
	326	nd_pfn->mode = le32_to_cpu(nd_pfn->pfn_sb->mode);
	327	if (nd_pfn->mode == PFN_MODE_RAM) {
	328	if (offset != SZ_8K)
	329	return -EINVAL;
	330	nd_pfn->npfns = le64_to_cpu(pfn_sb->npfns);
	331	altmap = NULL;
	332	} else {
	333	rc = -ENXIO;
	334	goto err;
	335	}
	336
	337	/* establish pfn range for lookup, and switch to direct map */
	338	pmem = dev_get_drvdata(dev);
a639315d DW	339	devm_memunmap(dev, (void __force *) pmem->virt_addr);
a639315d DW	340	pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, &nsio->res);
32ab0a3f DW	341	if (IS_ERR(pmem->virt_addr)) {
	342	rc = PTR_ERR(pmem->virt_addr);
	343	goto err;
	344	}
	345
	346	/* attach pmem disk in "pfn-mode" */
	347	pmem->data_offset = offset;
	348	rc = pmem_attach_disk(dev, ndns, pmem);
	349	if (rc)
	350	goto err;
	351
	352	return rc;
	353	err:
	354	nvdimm_namespace_detach_pfn(ndns);
	355	return rc;
9e853f23 RZ	356	}
9e853f23 RZ	357
9f53f9fa	358	static int nd_pmem_probe(struct device *dev)
9e853f23	359	{
9f53f9fa	360	struct nd_region *nd_region = to_nd_region(dev->parent);
8c2f7e86 DW	361	struct nd_namespace_common *ndns;
8c2f7e86 DW	362	struct nd_namespace_io *nsio;
9e853f23	363	struct pmem_device *pmem;
9e853f23	364
8c2f7e86 DW	365	ndns = nvdimm_namespace_common_probe(dev);
	366	if (IS_ERR(ndns))
	367	return PTR_ERR(ndns);
bf9bccc1	368
8c2f7e86	369	nsio = to_nd_namespace_io(&ndns->dev);
9f53f9fa	370	pmem = pmem_alloc(dev, &nsio->res, nd_region->id);
9e853f23 RZ	371	if (IS_ERR(pmem))
	372	return PTR_ERR(pmem);
	373
32ab0a3f	374	pmem->ndns = ndns;
9f53f9fa	375	dev_set_drvdata(dev, pmem);
8c2f7e86	376	ndns->rw_bytes = pmem_rw_bytes;
708ab62b	377
8c2f7e86	378	if (is_nd_btt(dev))
708ab62b CH	379	return nvdimm_namespace_attach_btt(ndns);
708ab62b CH	380
32ab0a3f DW	381	if (is_nd_pfn(dev))
	382	return nvdimm_namespace_attach_pfn(ndns);
	383
	384	if (nd_btt_probe(ndns, pmem) == 0) {
8c2f7e86	385	/* we'll come back as btt-pmem */
708ab62b	386	return -ENXIO;
32ab0a3f DW	387	}
	388
	389	if (nd_pfn_probe(ndns, pmem) == 0) {
	390	/* we'll come back as pfn-pmem */
	391	return -ENXIO;
	392	}
	393
	394	return pmem_attach_disk(dev, ndns, pmem);
9e853f23 RZ	395	}
9e853f23 RZ	396
9f53f9fa	397	static int nd_pmem_remove(struct device *dev)
9e853f23	398	{
9f53f9fa	399	struct pmem_device *pmem = dev_get_drvdata(dev);
9e853f23	400
8c2f7e86	401	if (is_nd_btt(dev))
32ab0a3f DW	402	nvdimm_namespace_detach_btt(pmem->ndns);
	403	else if (is_nd_pfn(dev))
	404	nvdimm_namespace_detach_pfn(pmem->ndns);
8c2f7e86 DW	405	else
8c2f7e86 DW	406	pmem_detach_disk(pmem);
8c2f7e86	407
9e853f23 RZ	408	return 0;
	409	}
	410
9f53f9fa DW	411	MODULE_ALIAS("pmem");
9f53f9fa DW	412	MODULE_ALIAS_ND_DEVICE(ND_DEVICE_NAMESPACE_IO);
bf9bccc1	413	MODULE_ALIAS_ND_DEVICE(ND_DEVICE_NAMESPACE_PMEM);
9f53f9fa DW	414	static struct nd_device_driver nd_pmem_driver = {
	415	.probe = nd_pmem_probe,
	416	.remove = nd_pmem_remove,
	417	.drv = {
	418	.name = "nd_pmem",
9e853f23	419	},
bf9bccc1	420	.type = ND_DRIVER_NAMESPACE_IO \| ND_DRIVER_NAMESPACE_PMEM,
9e853f23 RZ	421	};
	422
	423	static int __init pmem_init(void)
	424	{
	425	int error;
	426
	427	pmem_major = register_blkdev(0, "pmem");
	428	if (pmem_major < 0)
	429	return pmem_major;
	430
9f53f9fa DW	431	error = nd_driver_register(&nd_pmem_driver);
9f53f9fa DW	432	if (error) {
9e853f23	433	unregister_blkdev(pmem_major, "pmem");
9f53f9fa DW	434	return error;
	435	}
	436
	437	return 0;
9e853f23 RZ	438	}
	439	module_init(pmem_init);
	440
	441	static void pmem_exit(void)
	442	{
9f53f9fa	443	driver_unregister(&nd_pmem_driver.drv);
9e853f23 RZ	444	unregister_blkdev(pmem_major, "pmem");
	445	}
	446	module_exit(pmem_exit);
	447
	448	MODULE_AUTHOR("Ross Zwisler <ross.zwisler@linux.intel.com>");
	449	MODULE_LICENSE("GPL v2");