[mirror_ubuntu-zesty-kernel.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 void 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);
}

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;
	resource_size_t size;

	if (pmem->data_offset) {
		/*
		 * Limit the direct_access() size to what is covered by
		 * the memmap
		 */
		size = (pmem->size - offset) & ~ND_PFN_MASK;
	} else
		size = pmem->size - offset;

	/* FIXME convert DAX to comprehend that this mapping has a lifetime */
	*kaddr = pmem->virt_addr + offset;
	*pfn = (pmem->phys_addr + offset) >> PAGE_SHIFT;

	return size;
}

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)) {
		void *addr = devm_memremap_pages(dev, res);

		if (IS_ERR(addr))
			return addr;
		pmem->virt_addr = (void __pmem *) addr;
	} else {
		pmem->virt_addr = (void __pmem *) devm_memremap(dev,
				pmem->phys_addr, pmem->size,
				ARCH_MEMREMAP_PMEM);
		if (!pmem->virt_addr)
			return ERR_PTR(-ENXIO);
	}

	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)
{
	struct gendisk *disk;

	pmem->pmem_queue = blk_alloc_queue(GFP_KERNEL);
	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(0);
	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;

	/* section alignment for simple hotplug */
	if (nvdimm_namespace_capacity(ndns) < ND_PFN_ALIGN
			|| pmem->phys_addr & ND_PFN_MASK)
		return -ENODEV;

	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;

	if (PAGE_SIZE != SZ_4K) {
		dev_err(dev, "only supported on systems with 4K PAGE_SIZE\n");
		return -ENXIO;
	}
	if (nsio->res.start & ND_PFN_MASK) {
		dev_err(dev, "%s not memory hotplug section aligned\n",
				dev_name(&ndns->dev));
		return -ENXIO;
	}

	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
	67	static void pmem_make_request(struct request_queue q, struct bio bio)
	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);
9e853f23 RZ	87	}
	88
	89	static int pmem_rw_page(struct block_device *bdev, sector_t sector,
	90	struct page *page, int rw)
	91	{
	92	struct pmem_device *pmem = bdev->bd_disk->private_data;
	93
	94	pmem_do_bvec(pmem, page, PAGE_CACHE_SIZE, 0, rw, sector);
ba8fe0f8 RZ	95	if (rw & WRITE)
ba8fe0f8 RZ	96	wmb_pmem();
9e853f23 RZ	97	page_endio(page, rw & WRITE, 0);
	98
	99	return 0;
	100	}
	101
	102	static long pmem_direct_access(struct block_device *bdev, sector_t sector,
cb389b9c	103	void __pmem *kaddr, unsigned long pfn)
9e853f23 RZ	104	{
9e853f23 RZ	105	struct pmem_device *pmem = bdev->bd_disk->private_data;
32ab0a3f DW	106	resource_size_t offset = sector * 512 + pmem->data_offset;
	107	resource_size_t size;
	108
	109	if (pmem->data_offset) {
	110	/*
	111	* Limit the direct_access() size to what is covered by
	112	* the memmap
	113	*/
	114	size = (pmem->size - offset) & ~ND_PFN_MASK;
	115	} else
	116	size = pmem->size - offset;
9e853f23	117
61031952	118	/* FIXME convert DAX to comprehend that this mapping has a lifetime */
e2e05394	119	*kaddr = pmem->virt_addr + offset;
9e853f23 RZ	120	*pfn = (pmem->phys_addr + offset) >> PAGE_SHIFT;
9e853f23 RZ	121
32ab0a3f	122	return size;
9e853f23 RZ	123	}
	124
	125	static const struct block_device_operations pmem_fops = {
	126	.owner = THIS_MODULE,
	127	.rw_page = pmem_rw_page,
	128	.direct_access = pmem_direct_access,
58138820	129	.revalidate_disk = nvdimm_revalidate_disk,
9e853f23 RZ	130	};
9e853f23 RZ	131
9f53f9fa DW	132	static struct pmem_device pmem_alloc(struct device dev,
9f53f9fa DW	133	struct resource *res, int id)
9e853f23 RZ	134	{
9e853f23 RZ	135	struct pmem_device *pmem;
9e853f23	136
708ab62b	137	pmem = devm_kzalloc(dev, sizeof(*pmem), GFP_KERNEL);
9e853f23	138	if (!pmem)
8c2f7e86	139	return ERR_PTR(-ENOMEM);
9e853f23 RZ	140
	141	pmem->phys_addr = res->start;
	142	pmem->size = resource_size(res);
96601adb	143	if (!arch_has_wmb_pmem())
61031952	144	dev_warn(dev, "unable to guarantee persistence of writes\n");
9e853f23	145
708ab62b CH	146	if (!devm_request_mem_region(dev, pmem->phys_addr, pmem->size,
708ab62b CH	147	dev_name(dev))) {
9f53f9fa DW	148	dev_warn(dev, "could not reserve region [0x%pa:0x%zx]\n",
9f53f9fa DW	149	&pmem->phys_addr, pmem->size);
8c2f7e86	150	return ERR_PTR(-EBUSY);
9e853f23 RZ	151	}
9e853f23 RZ	152
004f1afb DW	153	if (pmem_should_map_pages(dev)) {
	154	void *addr = devm_memremap_pages(dev, res);
	155
	156	if (IS_ERR(addr))
	157	return addr;
	158	pmem->virt_addr = (void __pmem *) addr;
	159	} else {
a639315d DW	160	pmem->virt_addr = (void __pmem *) devm_memremap(dev,
	161	pmem->phys_addr, pmem->size,
	162	ARCH_MEMREMAP_PMEM);
004f1afb DW	163	if (!pmem->virt_addr)
004f1afb DW	164	return ERR_PTR(-ENXIO);
8c2f7e86 DW	165	}
	166
	167	return pmem;
	168	}
	169
	170	static void pmem_detach_disk(struct pmem_device *pmem)
	171	{
32ab0a3f DW	172	if (!pmem->pmem_disk)
	173	return;
	174
8c2f7e86 DW	175	del_gendisk(pmem->pmem_disk);
	176	put_disk(pmem->pmem_disk);
	177	blk_cleanup_queue(pmem->pmem_queue);
	178	}
	179
32ab0a3f DW	180	static int pmem_attach_disk(struct device *dev,
32ab0a3f DW	181	struct nd_namespace_common ndns, struct pmem_device pmem)
8c2f7e86	182	{
8c2f7e86	183	struct gendisk *disk;
9e853f23 RZ	184
	185	pmem->pmem_queue = blk_alloc_queue(GFP_KERNEL);
	186	if (!pmem->pmem_queue)
8c2f7e86	187	return -ENOMEM;
9e853f23 RZ	188
9e853f23 RZ	189	blk_queue_make_request(pmem->pmem_queue, pmem_make_request);
6b47496a	190	blk_queue_physical_block_size(pmem->pmem_queue, PAGE_SIZE);
43d3fa3a	191	blk_queue_max_hw_sectors(pmem->pmem_queue, UINT_MAX);
9e853f23	192	blk_queue_bounce_limit(pmem->pmem_queue, BLK_BOUNCE_ANY);
0f51c4fa	193	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, pmem->pmem_queue);
9e853f23	194
9f53f9fa	195	disk = alloc_disk(0);
8c2f7e86 DW	196	if (!disk) {
	197	blk_cleanup_queue(pmem->pmem_queue);
	198	return -ENOMEM;
	199	}
9e853f23	200
9e853f23	201	disk->major = pmem_major;
9f53f9fa	202	disk->first_minor = 0;
9e853f23 RZ	203	disk->fops = &pmem_fops;
	204	disk->private_data = pmem;
	205	disk->queue = pmem->pmem_queue;
	206	disk->flags = GENHD_FL_EXT_DEVT;
5212e11f	207	nvdimm_namespace_disk_name(ndns, disk->disk_name);
32ab0a3f DW	208	disk->driverfs_dev = dev;
32ab0a3f DW	209	set_capacity(disk, (pmem->size - pmem->data_offset) / 512);
9e853f23 RZ	210	pmem->pmem_disk = disk;
	211
	212	add_disk(disk);
58138820	213	revalidate_disk(disk);
9e853f23	214
8c2f7e86 DW	215	return 0;
8c2f7e86 DW	216	}
9e853f23	217
8c2f7e86 DW	218	static int pmem_rw_bytes(struct nd_namespace_common *ndns,
	219	resource_size_t offset, void *buf, size_t size, int rw)
	220	{
	221	struct pmem_device *pmem = dev_get_drvdata(ndns->claim);
	222
	223	if (unlikely(offset + size > pmem->size)) {
	224	dev_WARN_ONCE(&ndns->dev, 1, "request out of range\n");
	225	return -EFAULT;
	226	}
	227
	228	if (rw == READ)
61031952 RZ	229	memcpy_from_pmem(buf, pmem->virt_addr + offset, size);
	230	else {
	231	memcpy_to_pmem(pmem->virt_addr + offset, buf, size);
	232	wmb_pmem();
	233	}
8c2f7e86 DW	234
	235	return 0;
	236	}
	237
32ab0a3f DW	238	static int nd_pfn_init(struct nd_pfn *nd_pfn)
	239	{
	240	struct nd_pfn_sb pfn_sb = kzalloc(sizeof(pfn_sb), GFP_KERNEL);
	241	struct pmem_device *pmem = dev_get_drvdata(&nd_pfn->dev);
	242	struct nd_namespace_common *ndns = nd_pfn->ndns;
	243	struct nd_region *nd_region;
	244	unsigned long npfns;
	245	phys_addr_t offset;
	246	u64 checksum;
	247	int rc;
	248
	249	if (!pfn_sb)
	250	return -ENOMEM;
	251
	252	nd_pfn->pfn_sb = pfn_sb;
	253	rc = nd_pfn_validate(nd_pfn);
	254	if (rc == 0 \|\| rc == -EBUSY)
	255	return rc;
	256
	257	/* section alignment for simple hotplug */
	258	if (nvdimm_namespace_capacity(ndns) < ND_PFN_ALIGN
	259	\|\| pmem->phys_addr & ND_PFN_MASK)
	260	return -ENODEV;
	261
	262	nd_region = to_nd_region(nd_pfn->dev.parent);
	263	if (nd_region->ro) {
	264	dev_info(&nd_pfn->dev,
	265	"%s is read-only, unable to init metadata\n",
	266	dev_name(&nd_region->dev));
	267	goto err;
	268	}
	269
	270	memset(pfn_sb, 0, sizeof(*pfn_sb));
	271	npfns = (pmem->size - SZ_8K) / SZ_4K;
	272	/*
	273	* Note, we use 64 here for the standard size of struct page,
	274	* debugging options may cause it to be larger in which case the
	275	* implementation will limit the pfns advertised through
	276	* ->direct_access() to those that are included in the memmap.
	277	*/
	278	if (nd_pfn->mode == PFN_MODE_PMEM)
	279	offset = ALIGN(SZ_8K + 64 * npfns, PMD_SIZE);
	280	else if (nd_pfn->mode == PFN_MODE_RAM)
	281	offset = SZ_8K;
	282	else
	283	goto err;
	284
	285	npfns = (pmem->size - offset) / SZ_4K;
	286	pfn_sb->mode = cpu_to_le32(nd_pfn->mode);
	287	pfn_sb->dataoff = cpu_to_le64(offset);
	288	pfn_sb->npfns = cpu_to_le64(npfns);
	289	memcpy(pfn_sb->signature, PFN_SIG, PFN_SIG_LEN);
	290	memcpy(pfn_sb->uuid, nd_pfn->uuid, 16);
	291	pfn_sb->version_major = cpu_to_le16(1);
	292	checksum = nd_sb_checksum((struct nd_gen_sb *) pfn_sb);
	293	pfn_sb->checksum = cpu_to_le64(checksum);
	294
	295	rc = nvdimm_write_bytes(ndns, SZ_4K, pfn_sb, sizeof(*pfn_sb));
	296	if (rc)
	297	goto err;
	298
	299	return 0;
	300	err:
	301	nd_pfn->pfn_sb = NULL;
302	kfree(pfn_sb);
303	return -ENXIO;
304	}
305
306	static int nvdimm_namespace_detach_pfn(struct nd_namespace_common *ndns)
307	{
308	struct nd_pfn *nd_pfn = to_nd_pfn(ndns->claim);
309	struct pmem_device *pmem;
310
311	/* free pmem disk */
312	pmem = dev_get_drvdata(&nd_pfn->dev);
313	pmem_detach_disk(pmem);
314
315	/* release nd_pfn resources */
316	kfree(nd_pfn->pfn_sb);
317	nd_pfn->pfn_sb = NULL;
318
319	return 0;
320	}
321
322	static int nvdimm_namespace_attach_pfn(struct nd_namespace_common *ndns)
9e853f23	323	{
32ab0a3f DW	324	struct nd_namespace_io *nsio = to_nd_namespace_io(&ndns->dev);
	325	struct nd_pfn *nd_pfn = to_nd_pfn(ndns->claim);
	326	struct device *dev = &nd_pfn->dev;
	327	struct vmem_altmap *altmap;
	328	struct nd_region *nd_region;
	329	struct nd_pfn_sb *pfn_sb;
	330	struct pmem_device *pmem;
	331	phys_addr_t offset;
	332	int rc;
	333
	334	if (!nd_pfn->uuid \|\| !nd_pfn->ndns)
	335	return -ENODEV;
	336
	337	nd_region = to_nd_region(dev->parent);
	338	rc = nd_pfn_init(nd_pfn);
	339	if (rc)
	340	return rc;
	341
	342	if (PAGE_SIZE != SZ_4K) {
	343	dev_err(dev, "only supported on systems with 4K PAGE_SIZE\n");
	344	return -ENXIO;
	345	}
	346	if (nsio->res.start & ND_PFN_MASK) {
	347	dev_err(dev, "%s not memory hotplug section aligned\n",
	348	dev_name(&ndns->dev));
	349	return -ENXIO;
	350	}
	351
	352	pfn_sb = nd_pfn->pfn_sb;
	353	offset = le64_to_cpu(pfn_sb->dataoff);
	354	nd_pfn->mode = le32_to_cpu(nd_pfn->pfn_sb->mode);
	355	if (nd_pfn->mode == PFN_MODE_RAM) {
	356	if (offset != SZ_8K)
	357	return -EINVAL;
	358	nd_pfn->npfns = le64_to_cpu(pfn_sb->npfns);
	359	altmap = NULL;
	360	} else {
	361	rc = -ENXIO;
	362	goto err;
	363	}
	364
	365	/* establish pfn range for lookup, and switch to direct map */
	366	pmem = dev_get_drvdata(dev);
a639315d DW	367	devm_memunmap(dev, (void __force *) pmem->virt_addr);
a639315d DW	368	pmem->virt_addr = (void __pmem *) devm_memremap_pages(dev, &nsio->res);
32ab0a3f DW	369	if (IS_ERR(pmem->virt_addr)) {
	370	rc = PTR_ERR(pmem->virt_addr);
	371	goto err;
	372	}
	373
	374	/* attach pmem disk in "pfn-mode" */
	375	pmem->data_offset = offset;
	376	rc = pmem_attach_disk(dev, ndns, pmem);
	377	if (rc)
	378	goto err;
	379
	380	return rc;
	381	err:
	382	nvdimm_namespace_detach_pfn(ndns);
	383	return rc;
9e853f23 RZ	384	}
9e853f23 RZ	385
9f53f9fa	386	static int nd_pmem_probe(struct device *dev)
9e853f23	387	{
9f53f9fa	388	struct nd_region *nd_region = to_nd_region(dev->parent);
8c2f7e86 DW	389	struct nd_namespace_common *ndns;
8c2f7e86 DW	390	struct nd_namespace_io *nsio;
9e853f23	391	struct pmem_device *pmem;
9e853f23	392
8c2f7e86 DW	393	ndns = nvdimm_namespace_common_probe(dev);
	394	if (IS_ERR(ndns))
	395	return PTR_ERR(ndns);
bf9bccc1	396
8c2f7e86	397	nsio = to_nd_namespace_io(&ndns->dev);
9f53f9fa	398	pmem = pmem_alloc(dev, &nsio->res, nd_region->id);
9e853f23 RZ	399	if (IS_ERR(pmem))
	400	return PTR_ERR(pmem);
	401
32ab0a3f	402	pmem->ndns = ndns;
9f53f9fa	403	dev_set_drvdata(dev, pmem);
8c2f7e86	404	ndns->rw_bytes = pmem_rw_bytes;
708ab62b	405
8c2f7e86	406	if (is_nd_btt(dev))
708ab62b CH	407	return nvdimm_namespace_attach_btt(ndns);
708ab62b CH	408
32ab0a3f DW	409	if (is_nd_pfn(dev))
	410	return nvdimm_namespace_attach_pfn(ndns);
	411
	412	if (nd_btt_probe(ndns, pmem) == 0) {
8c2f7e86	413	/* we'll come back as btt-pmem */
708ab62b	414	return -ENXIO;
32ab0a3f DW	415	}
	416
	417	if (nd_pfn_probe(ndns, pmem) == 0) {
	418	/* we'll come back as pfn-pmem */
	419	return -ENXIO;
	420	}
	421
	422	return pmem_attach_disk(dev, ndns, pmem);
9e853f23 RZ	423	}
9e853f23 RZ	424
9f53f9fa	425	static int nd_pmem_remove(struct device *dev)
9e853f23	426	{
9f53f9fa	427	struct pmem_device *pmem = dev_get_drvdata(dev);
9e853f23	428
8c2f7e86	429	if (is_nd_btt(dev))
32ab0a3f DW	430	nvdimm_namespace_detach_btt(pmem->ndns);
	431	else if (is_nd_pfn(dev))
	432	nvdimm_namespace_detach_pfn(pmem->ndns);
8c2f7e86 DW	433	else
8c2f7e86 DW	434	pmem_detach_disk(pmem);
8c2f7e86	435
9e853f23 RZ	436	return 0;
	437	}
	438
9f53f9fa DW	439	MODULE_ALIAS("pmem");
9f53f9fa DW	440	MODULE_ALIAS_ND_DEVICE(ND_DEVICE_NAMESPACE_IO);
bf9bccc1	441	MODULE_ALIAS_ND_DEVICE(ND_DEVICE_NAMESPACE_PMEM);
9f53f9fa DW	442	static struct nd_device_driver nd_pmem_driver = {
	443	.probe = nd_pmem_probe,
	444	.remove = nd_pmem_remove,
	445	.drv = {
	446	.name = "nd_pmem",
9e853f23	447	},
bf9bccc1	448	.type = ND_DRIVER_NAMESPACE_IO \| ND_DRIVER_NAMESPACE_PMEM,
9e853f23 RZ	449	};
	450
	451	static int __init pmem_init(void)
	452	{
	453	int error;
	454
	455	pmem_major = register_blkdev(0, "pmem");
	456	if (pmem_major < 0)
	457	return pmem_major;
	458
9f53f9fa DW	459	error = nd_driver_register(&nd_pmem_driver);
9f53f9fa DW	460	if (error) {
9e853f23	461	unregister_blkdev(pmem_major, "pmem");
9f53f9fa DW	462	return error;
	463	}
	464
	465	return 0;
9e853f23 RZ	466	}
	467	module_init(pmem_init);
	468
	469	static void pmem_exit(void)
	470	{
9f53f9fa	471	driver_unregister(&nd_pmem_driver.drv);
9e853f23 RZ	472	unregister_blkdev(pmem_major, "pmem");
	473	}
	474	module_exit(pmem_exit);
	475
	476	MODULE_AUTHOR("Ross Zwisler <ross.zwisler@linux.intel.com>");
	477	MODULE_LICENSE("GPL v2");