null_blk.c 18.8 KB
Newer Older
1
#include <linux/module.h>
2

3 4 5 6 7 8 9 10
#include <linux/moduleparam.h>
#include <linux/sched.h>
#include <linux/fs.h>
#include <linux/blkdev.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/blk-mq.h>
#include <linux/hrtimer.h>
11
#include <linux/lightnvm.h>
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

struct nullb_cmd {
	struct list_head list;
	struct llist_node ll_list;
	struct call_single_data csd;
	struct request *rq;
	struct bio *bio;
	unsigned int tag;
	struct nullb_queue *nq;
};

struct nullb_queue {
	unsigned long *tag_map;
	wait_queue_head_t wait;
	unsigned int queue_depth;

	struct nullb_cmd *cmds;
};

struct nullb {
	struct list_head list;
	unsigned int index;
	struct request_queue *q;
	struct gendisk *disk;
36
	struct blk_mq_tag_set tag_set;
37 38 39 40 41 42
	struct hrtimer timer;
	unsigned int queue_depth;
	spinlock_t lock;

	struct nullb_queue *queues;
	unsigned int nr_queues;
43
	char disk_name[DISK_NAME_LEN];
44 45 46 47 48 49
};

static LIST_HEAD(nullb_list);
static struct mutex lock;
static int null_major;
static int nullb_indexes;
M
Matias Bjørling 已提交
50
static struct kmem_cache *ppa_cache;
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66

struct completion_queue {
	struct llist_head list;
	struct hrtimer timer;
};

/*
 * These are per-cpu for now, they will need to be configured by the
 * complete_queues parameter and appropriately mapped.
 */
static DEFINE_PER_CPU(struct completion_queue, completion_queues);

enum {
	NULL_IRQ_NONE		= 0,
	NULL_IRQ_SOFTIRQ	= 1,
	NULL_IRQ_TIMER		= 2,
67
};
68

69
enum {
70 71 72 73 74
	NULL_Q_BIO		= 0,
	NULL_Q_RQ		= 1,
	NULL_Q_MQ		= 2,
};

75
static int submit_queues;
76 77 78 79 80 81 82 83
module_param(submit_queues, int, S_IRUGO);
MODULE_PARM_DESC(submit_queues, "Number of submission queues");

static int home_node = NUMA_NO_NODE;
module_param(home_node, int, S_IRUGO);
MODULE_PARM_DESC(home_node, "Home node for the device");

static int queue_mode = NULL_Q_MQ;
84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104

static int null_param_store_val(const char *str, int *val, int min, int max)
{
	int ret, new_val;

	ret = kstrtoint(str, 10, &new_val);
	if (ret)
		return -EINVAL;

	if (new_val < min || new_val > max)
		return -EINVAL;

	*val = new_val;
	return 0;
}

static int null_set_queue_mode(const char *str, const struct kernel_param *kp)
{
	return null_param_store_val(str, &queue_mode, NULL_Q_BIO, NULL_Q_MQ);
}

105
static const struct kernel_param_ops null_queue_mode_param_ops = {
106 107 108 109 110
	.set	= null_set_queue_mode,
	.get	= param_get_int,
};

device_param_cb(queue_mode, &null_queue_mode_param_ops, &queue_mode, S_IRUGO);
111
MODULE_PARM_DESC(queue_mode, "Block interface to use (0=bio,1=rq,2=multiqueue)");
112 113 114 115 116 117 118 119 120 121 122 123 124

static int gb = 250;
module_param(gb, int, S_IRUGO);
MODULE_PARM_DESC(gb, "Size in GB");

static int bs = 512;
module_param(bs, int, S_IRUGO);
MODULE_PARM_DESC(bs, "Block size (in bytes)");

static int nr_devices = 2;
module_param(nr_devices, int, S_IRUGO);
MODULE_PARM_DESC(nr_devices, "Number of devices to register");

125 126 127 128
static bool use_lightnvm;
module_param(use_lightnvm, bool, S_IRUGO);
MODULE_PARM_DESC(use_lightnvm, "Register as a LightNVM device");

129
static int irqmode = NULL_IRQ_SOFTIRQ;
130 131 132 133 134 135 136

static int null_set_irqmode(const char *str, const struct kernel_param *kp)
{
	return null_param_store_val(str, &irqmode, NULL_IRQ_NONE,
					NULL_IRQ_TIMER);
}

137
static const struct kernel_param_ops null_irqmode_param_ops = {
138 139 140 141 142
	.set	= null_set_irqmode,
	.get	= param_get_int,
};

device_param_cb(irqmode, &null_irqmode_param_ops, &irqmode, S_IRUGO);
143 144 145 146 147 148 149 150 151 152
MODULE_PARM_DESC(irqmode, "IRQ completion handler. 0-none, 1-softirq, 2-timer");

static int completion_nsec = 10000;
module_param(completion_nsec, int, S_IRUGO);
MODULE_PARM_DESC(completion_nsec, "Time in ns to complete a request in hardware. Default: 10,000ns");

static int hw_queue_depth = 64;
module_param(hw_queue_depth, int, S_IRUGO);
MODULE_PARM_DESC(hw_queue_depth, "Queue depth for each hardware queue. Default: 64");

153
static bool use_per_node_hctx = false;
154
module_param(use_per_node_hctx, bool, S_IRUGO);
155
MODULE_PARM_DESC(use_per_node_hctx, "Use per-node allocation for hardware context queues. Default: false");
156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222

static void put_tag(struct nullb_queue *nq, unsigned int tag)
{
	clear_bit_unlock(tag, nq->tag_map);

	if (waitqueue_active(&nq->wait))
		wake_up(&nq->wait);
}

static unsigned int get_tag(struct nullb_queue *nq)
{
	unsigned int tag;

	do {
		tag = find_first_zero_bit(nq->tag_map, nq->queue_depth);
		if (tag >= nq->queue_depth)
			return -1U;
	} while (test_and_set_bit_lock(tag, nq->tag_map));

	return tag;
}

static void free_cmd(struct nullb_cmd *cmd)
{
	put_tag(cmd->nq, cmd->tag);
}

static struct nullb_cmd *__alloc_cmd(struct nullb_queue *nq)
{
	struct nullb_cmd *cmd;
	unsigned int tag;

	tag = get_tag(nq);
	if (tag != -1U) {
		cmd = &nq->cmds[tag];
		cmd->tag = tag;
		cmd->nq = nq;
		return cmd;
	}

	return NULL;
}

static struct nullb_cmd *alloc_cmd(struct nullb_queue *nq, int can_wait)
{
	struct nullb_cmd *cmd;
	DEFINE_WAIT(wait);

	cmd = __alloc_cmd(nq);
	if (cmd || !can_wait)
		return cmd;

	do {
		prepare_to_wait(&nq->wait, &wait, TASK_UNINTERRUPTIBLE);
		cmd = __alloc_cmd(nq);
		if (cmd)
			break;

		io_schedule();
	} while (1);

	finish_wait(&nq->wait, &wait);
	return cmd;
}

static void end_cmd(struct nullb_cmd *cmd)
{
223 224
	switch (queue_mode)  {
	case NULL_Q_MQ:
225
		blk_mq_end_request(cmd->rq, 0);
226 227 228 229 230 231
		return;
	case NULL_Q_RQ:
		INIT_LIST_HEAD(&cmd->rq->queuelist);
		blk_end_request_all(cmd->rq, 0);
		break;
	case NULL_Q_BIO:
232
		bio_endio(cmd->bio);
233 234
		break;
	}
235

236
	free_cmd(cmd);
237 238 239 240 241 242 243 244 245 246 247
}

static enum hrtimer_restart null_cmd_timer_expired(struct hrtimer *timer)
{
	struct completion_queue *cq;
	struct llist_node *entry;
	struct nullb_cmd *cmd;

	cq = &per_cpu(completion_queues, smp_processor_id());

	while ((entry = llist_del_all(&cq->list)) != NULL) {
248
		entry = llist_reverse_order(entry);
249
		do {
250 251
			struct request_queue *q = NULL;

252 253
			cmd = container_of(entry, struct nullb_cmd, ll_list);
			entry = entry->next;
254 255
			if (cmd->rq)
				q = cmd->rq->q;
M
Ming Lei 已提交
256
			end_cmd(cmd);
257

258 259 260 261 262
			if (q && !q->mq_ops && blk_queue_stopped(q)) {
				spin_lock(q->queue_lock);
				if (blk_queue_stopped(q))
					blk_start_queue(q);
				spin_unlock(q->queue_lock);
263
			}
264 265 266 267 268 269 270 271 272 273 274 275 276 277
		} while (entry);
	}

	return HRTIMER_NORESTART;
}

static void null_cmd_end_timer(struct nullb_cmd *cmd)
{
	struct completion_queue *cq = &per_cpu(completion_queues, get_cpu());

	cmd->ll_list.next = NULL;
	if (llist_add(&cmd->ll_list, &cq->list)) {
		ktime_t kt = ktime_set(0, completion_nsec);

278
		hrtimer_start(&cq->timer, kt, HRTIMER_MODE_REL_PINNED);
279 280 281 282 283 284 285
	}

	put_cpu();
}

static void null_softirq_done_fn(struct request *rq)
{
286 287 288 289
	if (queue_mode == NULL_Q_MQ)
		end_cmd(blk_mq_rq_to_pdu(rq));
	else
		end_cmd(rq->special);
290 291 292 293 294 295 296
}

static inline void null_handle_cmd(struct nullb_cmd *cmd)
{
	/* Complete IO by inline, softirq or timer */
	switch (irqmode) {
	case NULL_IRQ_SOFTIRQ:
297 298
		switch (queue_mode)  {
		case NULL_Q_MQ:
299
			blk_mq_complete_request(cmd->rq, cmd->rq->errors);
300 301 302 303 304 305 306 307 308 309 310 311 312
			break;
		case NULL_Q_RQ:
			blk_complete_request(cmd->rq);
			break;
		case NULL_Q_BIO:
			/*
			 * XXX: no proper submitting cpu information available.
			 */
			end_cmd(cmd);
			break;
		}
		break;
	case NULL_IRQ_NONE:
313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330
		end_cmd(cmd);
		break;
	case NULL_IRQ_TIMER:
		null_cmd_end_timer(cmd);
		break;
	}
}

static struct nullb_queue *nullb_to_queue(struct nullb *nullb)
{
	int index = 0;

	if (nullb->nr_queues != 1)
		index = raw_smp_processor_id() / ((nr_cpu_ids + nullb->nr_queues - 1) / nullb->nr_queues);

	return &nullb->queues[index];
}

331
static blk_qc_t null_queue_bio(struct request_queue *q, struct bio *bio)
332 333 334 335 336 337 338 339 340
{
	struct nullb *nullb = q->queuedata;
	struct nullb_queue *nq = nullb_to_queue(nullb);
	struct nullb_cmd *cmd;

	cmd = alloc_cmd(nq, 1);
	cmd->bio = bio;

	null_handle_cmd(cmd);
341
	return BLK_QC_T_NONE;
342 343 344 345 346 347 348 349 350 351 352 353 354 355
}

static int null_rq_prep_fn(struct request_queue *q, struct request *req)
{
	struct nullb *nullb = q->queuedata;
	struct nullb_queue *nq = nullb_to_queue(nullb);
	struct nullb_cmd *cmd;

	cmd = alloc_cmd(nq, 0);
	if (cmd) {
		cmd->rq = req;
		req->special = cmd;
		return BLKPREP_OK;
	}
356
	blk_stop_queue(q);
357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373

	return BLKPREP_DEFER;
}

static void null_request_fn(struct request_queue *q)
{
	struct request *rq;

	while ((rq = blk_fetch_request(q)) != NULL) {
		struct nullb_cmd *cmd = rq->special;

		spin_unlock_irq(q->queue_lock);
		null_handle_cmd(cmd);
		spin_lock_irq(q->queue_lock);
	}
}

374 375
static int null_queue_rq(struct blk_mq_hw_ctx *hctx,
			 const struct blk_mq_queue_data *bd)
376
{
377
	struct nullb_cmd *cmd = blk_mq_rq_to_pdu(bd->rq);
378

379
	cmd->rq = bd->rq;
380 381
	cmd->nq = hctx->driver_data;

382
	blk_mq_start_request(bd->rq);
383

384 385 386 387
	null_handle_cmd(cmd);
	return BLK_MQ_RQ_QUEUE_OK;
}

388 389 390 391 392 393 394 395 396
static void null_init_queue(struct nullb *nullb, struct nullb_queue *nq)
{
	BUG_ON(!nullb);
	BUG_ON(!nq);

	init_waitqueue_head(&nq->wait);
	nq->queue_depth = nullb->queue_depth;
}

397 398 399 400 401 402 403
static int null_init_hctx(struct blk_mq_hw_ctx *hctx, void *data,
			  unsigned int index)
{
	struct nullb *nullb = data;
	struct nullb_queue *nq = &nullb->queues[index];

	hctx->driver_data = nq;
404 405
	null_init_queue(nullb, nq);
	nullb->nr_queues++;
406 407 408 409 410 411 412 413

	return 0;
}

static struct blk_mq_ops null_mq_ops = {
	.queue_rq       = null_queue_rq,
	.map_queue      = blk_mq_map_queue,
	.init_hctx	= null_init_hctx,
414
	.complete	= null_softirq_done_fn,
415 416
};

417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
static void cleanup_queue(struct nullb_queue *nq)
{
	kfree(nq->tag_map);
	kfree(nq->cmds);
}

static void cleanup_queues(struct nullb *nullb)
{
	int i;

	for (i = 0; i < nullb->nr_queues; i++)
		cleanup_queue(&nullb->queues[i]);

	kfree(nullb->queues);
}

433 434 435 436
static void null_del_dev(struct nullb *nullb)
{
	list_del_init(&nullb->list);

437
	if (use_lightnvm)
438 439 440
		nvm_unregister(nullb->disk_name);
	else
		del_gendisk(nullb->disk);
441
	blk_cleanup_queue(nullb->q);
442 443
	if (queue_mode == NULL_Q_MQ)
		blk_mq_free_tag_set(&nullb->tag_set);
444 445
	if (!use_lightnvm)
		put_disk(nullb->disk);
446
	cleanup_queues(nullb);
447 448 449
	kfree(nullb);
}

450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490
#ifdef CONFIG_NVM

static void null_lnvm_end_io(struct request *rq, int error)
{
	struct nvm_rq *rqd = rq->end_io_data;
	struct nvm_dev *dev = rqd->dev;

	dev->mt->end_io(rqd, error);

	blk_put_request(rq);
}

static int null_lnvm_submit_io(struct request_queue *q, struct nvm_rq *rqd)
{
	struct request *rq;
	struct bio *bio = rqd->bio;

	rq = blk_mq_alloc_request(q, bio_rw(bio), GFP_KERNEL, 0);
	if (IS_ERR(rq))
		return -ENOMEM;

	rq->cmd_type = REQ_TYPE_DRV_PRIV;
	rq->__sector = bio->bi_iter.bi_sector;
	rq->ioprio = bio_prio(bio);

	if (bio_has_data(bio))
		rq->nr_phys_segments = bio_phys_segments(q, bio);

	rq->__data_len = bio->bi_iter.bi_size;
	rq->bio = rq->biotail = bio;

	rq->end_io_data = rqd;

	blk_execute_rq_nowait(q, NULL, rq, 0, null_lnvm_end_io);

	return 0;
}

static int null_lnvm_id(struct request_queue *q, struct nvm_id *id)
{
	sector_t size = gb * 1024 * 1024 * 1024ULL;
491
	sector_t blksize;
492 493 494 495 496 497 498
	struct nvm_id_group *grp;

	id->ver_id = 0x1;
	id->vmnt = 0;
	id->cgrps = 1;
	id->cap = 0x3;
	id->dom = 0x1;
499 500 501 502 503 504 505 506 507 508 509 510 511

	id->ppaf.blk_offset = 0;
	id->ppaf.blk_len = 16;
	id->ppaf.pg_offset = 16;
	id->ppaf.pg_len = 16;
	id->ppaf.sect_offset = 32;
	id->ppaf.sect_len = 8;
	id->ppaf.pln_offset = 40;
	id->ppaf.pln_len = 8;
	id->ppaf.lun_offset = 48;
	id->ppaf.lun_len = 8;
	id->ppaf.ch_offset = 56;
	id->ppaf.ch_len = 8;
512 513

	do_div(size, bs); /* convert size to pages */
514
	do_div(size, 256); /* concert size to pgs pr blk */
515 516
	grp = &id->groups[0];
	grp->mtype = 0;
517
	grp->fmtype = 0;
518 519
	grp->num_ch = 1;
	grp->num_pg = 256;
520 521 522 523 524 525 526
	blksize = size;
	do_div(size, (1 << 16));
	grp->num_lun = size + 1;
	do_div(blksize, grp->num_lun);
	grp->num_blk = blksize;
	grp->num_pln = 1;

527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544
	grp->fpg_sz = bs;
	grp->csecs = bs;
	grp->trdt = 25000;
	grp->trdm = 25000;
	grp->tprt = 500000;
	grp->tprm = 500000;
	grp->tbet = 1500000;
	grp->tbem = 1500000;
	grp->mpos = 0x010101; /* single plane rwe */
	grp->cpar = hw_queue_depth;

	return 0;
}

static void *null_lnvm_create_dma_pool(struct request_queue *q, char *name)
{
	mempool_t *virtmem_pool;

M
Matias Bjørling 已提交
545
	virtmem_pool = mempool_create_slab_pool(64, ppa_cache);
546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586
	if (!virtmem_pool) {
		pr_err("null_blk: Unable to create virtual memory pool\n");
		return NULL;
	}

	return virtmem_pool;
}

static void null_lnvm_destroy_dma_pool(void *pool)
{
	mempool_destroy(pool);
}

static void *null_lnvm_dev_dma_alloc(struct request_queue *q, void *pool,
				gfp_t mem_flags, dma_addr_t *dma_handler)
{
	return mempool_alloc(pool, mem_flags);
}

static void null_lnvm_dev_dma_free(void *pool, void *entry,
							dma_addr_t dma_handler)
{
	mempool_free(entry, pool);
}

static struct nvm_dev_ops null_lnvm_dev_ops = {
	.identity		= null_lnvm_id,
	.submit_io		= null_lnvm_submit_io,

	.create_dma_pool	= null_lnvm_create_dma_pool,
	.destroy_dma_pool	= null_lnvm_destroy_dma_pool,
	.dev_dma_alloc		= null_lnvm_dev_dma_alloc,
	.dev_dma_free		= null_lnvm_dev_dma_free,

	/* Simulate nvme protocol restriction */
	.max_phys_sect		= 64,
};
#else
static struct nvm_dev_ops null_lnvm_dev_ops;
#endif /* CONFIG_NVM */

587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608
static int null_open(struct block_device *bdev, fmode_t mode)
{
	return 0;
}

static void null_release(struct gendisk *disk, fmode_t mode)
{
}

static const struct block_device_operations null_fops = {
	.owner =	THIS_MODULE,
	.open =		null_open,
	.release =	null_release,
};

static int setup_commands(struct nullb_queue *nq)
{
	struct nullb_cmd *cmd;
	int i, tag_size;

	nq->cmds = kzalloc(nq->queue_depth * sizeof(*cmd), GFP_KERNEL);
	if (!nq->cmds)
609
		return -ENOMEM;
610 611 612 613 614

	tag_size = ALIGN(nq->queue_depth, BITS_PER_LONG) / BITS_PER_LONG;
	nq->tag_map = kzalloc(tag_size * sizeof(unsigned long), GFP_KERNEL);
	if (!nq->tag_map) {
		kfree(nq->cmds);
615
		return -ENOMEM;
616 617 618 619 620 621 622 623 624 625 626 627 628 629
	}

	for (i = 0; i < nq->queue_depth; i++) {
		cmd = &nq->cmds[i];
		INIT_LIST_HEAD(&cmd->list);
		cmd->ll_list.next = NULL;
		cmd->tag = -1U;
	}

	return 0;
}

static int setup_queues(struct nullb *nullb)
{
630 631
	nullb->queues = kzalloc(submit_queues * sizeof(struct nullb_queue),
								GFP_KERNEL);
632
	if (!nullb->queues)
633
		return -ENOMEM;
634 635 636 637

	nullb->nr_queues = 0;
	nullb->queue_depth = hw_queue_depth;

638 639 640 641 642 643 644
	return 0;
}

static int init_driver_queues(struct nullb *nullb)
{
	struct nullb_queue *nq;
	int i, ret = 0;
645 646 647

	for (i = 0; i < submit_queues; i++) {
		nq = &nullb->queues[i];
648 649 650 651 652

		null_init_queue(nullb, nq);

		ret = setup_commands(nq);
		if (ret)
653
			return ret;
654 655
		nullb->nr_queues++;
	}
656
	return 0;
657 658 659 660 661 662 663
}

static int null_add_dev(void)
{
	struct gendisk *disk;
	struct nullb *nullb;
	sector_t size;
664
	int rv;
665 666

	nullb = kzalloc_node(sizeof(*nullb), GFP_KERNEL, home_node);
667 668
	if (!nullb) {
		rv = -ENOMEM;
669
		goto out;
670
	}
671 672 673

	spin_lock_init(&nullb->lock);

674 675 676
	if (queue_mode == NULL_Q_MQ && use_per_node_hctx)
		submit_queues = nr_online_nodes;

677 678
	rv = setup_queues(nullb);
	if (rv)
679
		goto out_free_nullb;
680 681

	if (queue_mode == NULL_Q_MQ) {
682
		nullb->tag_set.ops = &null_mq_ops;
683 684 685 686 687 688 689
		nullb->tag_set.nr_hw_queues = submit_queues;
		nullb->tag_set.queue_depth = hw_queue_depth;
		nullb->tag_set.numa_node = home_node;
		nullb->tag_set.cmd_size	= sizeof(struct nullb_cmd);
		nullb->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
		nullb->tag_set.driver_data = nullb;

690 691
		rv = blk_mq_alloc_tag_set(&nullb->tag_set);
		if (rv)
692 693 694
			goto out_cleanup_queues;

		nullb->q = blk_mq_init_queue(&nullb->tag_set);
695
		if (IS_ERR(nullb->q)) {
696
			rv = -ENOMEM;
697
			goto out_cleanup_tags;
698
		}
699 700
	} else if (queue_mode == NULL_Q_BIO) {
		nullb->q = blk_alloc_queue_node(GFP_KERNEL, home_node);
701 702
		if (!nullb->q) {
			rv = -ENOMEM;
703
			goto out_cleanup_queues;
704
		}
705
		blk_queue_make_request(nullb->q, null_queue_bio);
706 707 708
		rv = init_driver_queues(nullb);
		if (rv)
			goto out_cleanup_blk_queue;
709 710
	} else {
		nullb->q = blk_init_queue_node(null_request_fn, &nullb->lock, home_node);
711 712
		if (!nullb->q) {
			rv = -ENOMEM;
713
			goto out_cleanup_queues;
714
		}
715
		blk_queue_prep_rq(nullb->q, null_rq_prep_fn);
716
		blk_queue_softirq_done(nullb->q, null_softirq_done_fn);
717 718 719
		rv = init_driver_queues(nullb);
		if (rv)
			goto out_cleanup_blk_queue;
720 721 722 723
	}

	nullb->q->queuedata = nullb;
	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, nullb->q);
724
	queue_flag_clear_unlocked(QUEUE_FLAG_ADD_RANDOM, nullb->q);
725 726 727 728 729 730 731 732 733 734


	mutex_lock(&lock);
	list_add_tail(&nullb->list, &nullb_list);
	nullb->index = nullb_indexes++;
	mutex_unlock(&lock);

	blk_queue_logical_block_size(nullb->q, bs);
	blk_queue_physical_block_size(nullb->q, bs);

735 736 737 738 739 740 741 742 743 744 745 746 747 748 749
	sprintf(nullb->disk_name, "nullb%d", nullb->index);

	if (use_lightnvm) {
		rv = nvm_register(nullb->q, nullb->disk_name,
							&null_lnvm_dev_ops);
		if (rv)
			goto out_cleanup_blk_queue;
		goto done;
	}

	disk = nullb->disk = alloc_disk_node(1, home_node);
	if (!disk) {
		rv = -ENOMEM;
		goto out_cleanup_lightnvm;
	}
750
	size = gb * 1024 * 1024 * 1024ULL;
751
	set_capacity(disk, size >> 9);
752

753
	disk->flags |= GENHD_FL_EXT_DEVT | GENHD_FL_SUPPRESS_PARTITION_INFO;
754 755 756 757 758
	disk->major		= null_major;
	disk->first_minor	= nullb->index;
	disk->fops		= &null_fops;
	disk->private_data	= nullb;
	disk->queue		= nullb->q;
759 760
	strncpy(disk->disk_name, nullb->disk_name, DISK_NAME_LEN);

761
	add_disk(disk);
762
done:
763
	return 0;
764

765 766 767
out_cleanup_lightnvm:
	if (use_lightnvm)
		nvm_unregister(nullb->disk_name);
768 769 770 771 772 773 774 775 776 777
out_cleanup_blk_queue:
	blk_cleanup_queue(nullb->q);
out_cleanup_tags:
	if (queue_mode == NULL_Q_MQ)
		blk_mq_free_tag_set(&nullb->tag_set);
out_cleanup_queues:
	cleanup_queues(nullb);
out_free_nullb:
	kfree(nullb);
out:
778
	return rv;
779 780 781 782 783 784
}

static int __init null_init(void)
{
	unsigned int i;

785 786 787 788 789
	if (bs > PAGE_SIZE) {
		pr_warn("null_blk: invalid block size\n");
		pr_warn("null_blk: defaults block size to %lu\n", PAGE_SIZE);
		bs = PAGE_SIZE;
	}
790

M
Matias Bjørling 已提交
791 792 793 794 795 796
	if (use_lightnvm && bs != 4096) {
		pr_warn("null_blk: LightNVM only supports 4k block size\n");
		pr_warn("null_blk: defaults block size to 4k\n");
		bs = 4096;
	}

797 798 799 800 801 802
	if (use_lightnvm && queue_mode != NULL_Q_MQ) {
		pr_warn("null_blk: LightNVM only supported for blk-mq\n");
		pr_warn("null_blk: defaults queue mode to blk-mq\n");
		queue_mode = NULL_Q_MQ;
	}

803
	if (queue_mode == NULL_Q_MQ && use_per_node_hctx) {
804
		if (submit_queues < nr_online_nodes) {
805 806
			pr_warn("null_blk: submit_queues param is set to %u.",
							nr_online_nodes);
807 808
			submit_queues = nr_online_nodes;
		}
809
	} else if (submit_queues > nr_cpu_ids)
810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832
		submit_queues = nr_cpu_ids;
	else if (!submit_queues)
		submit_queues = 1;

	mutex_init(&lock);

	/* Initialize a separate list for each CPU for issuing softirqs */
	for_each_possible_cpu(i) {
		struct completion_queue *cq = &per_cpu(completion_queues, i);

		init_llist_head(&cq->list);

		if (irqmode != NULL_IRQ_TIMER)
			continue;

		hrtimer_init(&cq->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
		cq->timer.function = null_cmd_timer_expired;
	}

	null_major = register_blkdev(0, "nullb");
	if (null_major < 0)
		return null_major;

M
Matias Bjørling 已提交
833 834 835 836 837 838 839 840 841
	if (use_lightnvm) {
		ppa_cache = kmem_cache_create("ppa_cache", 64 * sizeof(u64),
								0, 0, NULL);
		if (!ppa_cache) {
			pr_err("null_blk: unable to create ppa cache\n");
			return -ENOMEM;
		}
	}

842 843 844
	for (i = 0; i < nr_devices; i++) {
		if (null_add_dev()) {
			unregister_blkdev(null_major, "nullb");
M
Matias Bjørling 已提交
845
			goto err_ppa;
846 847 848 849 850
		}
	}

	pr_info("null: module loaded\n");
	return 0;
M
Matias Bjørling 已提交
851 852 853
err_ppa:
	kmem_cache_destroy(ppa_cache);
	return -EINVAL;
854 855 856 857 858 859 860 861 862 863 864 865 866 867
}

static void __exit null_exit(void)
{
	struct nullb *nullb;

	unregister_blkdev(null_major, "nullb");

	mutex_lock(&lock);
	while (!list_empty(&nullb_list)) {
		nullb = list_entry(nullb_list.next, struct nullb, list);
		null_del_dev(nullb);
	}
	mutex_unlock(&lock);
M
Matias Bjørling 已提交
868 869

	kmem_cache_destroy(ppa_cache);
870 871 872 873 874 875 876
}

module_init(null_init);
module_exit(null_exit);

MODULE_AUTHOR("Jens Axboe <jaxboe@fusionio.com>");
MODULE_LICENSE("GPL");