blk-tag.c 9.79 KB
Newer Older
1 2 3 4 5 6 7 8
/*
 * Functions related to tagged command queuing
 */
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/bio.h>
#include <linux/blkdev.h>

9 10
#include "blk.h"

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
/**
 * blk_queue_find_tag - find a request by its tag and queue
 * @q:	 The request queue for the device
 * @tag: The tag of the request
 *
 * Notes:
 *    Should be used when a device returns a tag and you want to match
 *    it with a request.
 *
 *    no locks need be held.
 **/
struct request *blk_queue_find_tag(struct request_queue *q, int tag)
{
	return blk_map_queue_find_tag(q->queue_tags, tag);
}
EXPORT_SYMBOL(blk_queue_find_tag);

/**
 * __blk_free_tags - release a given set of tag maintenance info
 * @bqt:	the tag map to free
 *
32
 * Tries to free the specified @bqt.  Returns true if it was
33 34 35 36 37 38 39 40
 * actually freed and false if there are still references using it
 */
static int __blk_free_tags(struct blk_queue_tag *bqt)
{
	int retval;

	retval = atomic_dec_and_test(&bqt->refcnt);
	if (retval) {
41 42
		BUG_ON(find_first_bit(bqt->tag_map, bqt->max_depth) <
							bqt->max_depth);
43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73

		kfree(bqt->tag_index);
		bqt->tag_index = NULL;

		kfree(bqt->tag_map);
		bqt->tag_map = NULL;

		kfree(bqt);
	}

	return retval;
}

/**
 * __blk_queue_free_tags - release tag maintenance info
 * @q:  the request queue for the device
 *
 *  Notes:
 *    blk_cleanup_queue() will take care of calling this function, if tagging
 *    has been used. So there's no need to call this directly.
 **/
void __blk_queue_free_tags(struct request_queue *q)
{
	struct blk_queue_tag *bqt = q->queue_tags;

	if (!bqt)
		return;

	__blk_free_tags(bqt);

	q->queue_tags = NULL;
74
	queue_flag_clear_unlocked(QUEUE_FLAG_QUEUED, q);
75 76 77 78 79 80
}

/**
 * blk_free_tags - release a given set of tag maintenance info
 * @bqt:	the tag map to free
 *
81
 * For externally managed @bqt frees the map.  Callers of this
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96
 * function must guarantee to have released all the queues that
 * might have been using this tag map.
 */
void blk_free_tags(struct blk_queue_tag *bqt)
{
	if (unlikely(!__blk_free_tags(bqt)))
		BUG();
}
EXPORT_SYMBOL(blk_free_tags);

/**
 * blk_queue_free_tags - release tag maintenance info
 * @q:  the request queue for the device
 *
 *  Notes:
97
 *	This is used to disable tagged queuing to a device, yet leave
98 99 100 101
 *	queue in function.
 **/
void blk_queue_free_tags(struct request_queue *q)
{
102
	queue_flag_clear_unlocked(QUEUE_FLAG_QUEUED, q);
103 104 105 106 107 108 109 110 111 112 113 114 115
}
EXPORT_SYMBOL(blk_queue_free_tags);

static int
init_tag_map(struct request_queue *q, struct blk_queue_tag *tags, int depth)
{
	struct request **tag_index;
	unsigned long *tag_map;
	int nr_ulongs;

	if (q && depth > q->nr_requests * 2) {
		depth = q->nr_requests * 2;
		printk(KERN_ERR "%s: adjusted depth to %d\n",
116
		       __func__, depth);
117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172
	}

	tag_index = kzalloc(depth * sizeof(struct request *), GFP_ATOMIC);
	if (!tag_index)
		goto fail;

	nr_ulongs = ALIGN(depth, BITS_PER_LONG) / BITS_PER_LONG;
	tag_map = kzalloc(nr_ulongs * sizeof(unsigned long), GFP_ATOMIC);
	if (!tag_map)
		goto fail;

	tags->real_max_depth = depth;
	tags->max_depth = depth;
	tags->tag_index = tag_index;
	tags->tag_map = tag_map;

	return 0;
fail:
	kfree(tag_index);
	return -ENOMEM;
}

static struct blk_queue_tag *__blk_queue_init_tags(struct request_queue *q,
						   int depth)
{
	struct blk_queue_tag *tags;

	tags = kmalloc(sizeof(struct blk_queue_tag), GFP_ATOMIC);
	if (!tags)
		goto fail;

	if (init_tag_map(q, tags, depth))
		goto fail;

	atomic_set(&tags->refcnt, 1);
	return tags;
fail:
	kfree(tags);
	return NULL;
}

/**
 * blk_init_tags - initialize the tag info for an external tag map
 * @depth:	the maximum queue depth supported
 **/
struct blk_queue_tag *blk_init_tags(int depth)
{
	return __blk_queue_init_tags(NULL, depth);
}
EXPORT_SYMBOL(blk_init_tags);

/**
 * blk_queue_init_tags - initialize the queue tag info
 * @q:  the request queue for the device
 * @depth:  the maximum queue depth supported
 * @tags: the tag to use
173 174 175
 *
 * Queue lock must be held here if the function is called to resize an
 * existing map.
176 177 178 179 180 181 182 183 184 185 186 187 188 189
 **/
int blk_queue_init_tags(struct request_queue *q, int depth,
			struct blk_queue_tag *tags)
{
	int rc;

	BUG_ON(tags && q->queue_tags && tags != q->queue_tags);

	if (!tags && !q->queue_tags) {
		tags = __blk_queue_init_tags(q, depth);

		if (!tags)
			goto fail;
	} else if (q->queue_tags) {
190 191
		rc = blk_queue_resize_tags(q, depth);
		if (rc)
192
			return rc;
193
		queue_flag_set(QUEUE_FLAG_QUEUED, q);
194 195 196 197 198 199 200 201
		return 0;
	} else
		atomic_inc(&tags->refcnt);

	/*
	 * assign it, all done
	 */
	q->queue_tags = tags;
202
	queue_flag_set_unlocked(QUEUE_FLAG_QUEUED, q);
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272
	INIT_LIST_HEAD(&q->tag_busy_list);
	return 0;
fail:
	kfree(tags);
	return -ENOMEM;
}
EXPORT_SYMBOL(blk_queue_init_tags);

/**
 * blk_queue_resize_tags - change the queueing depth
 * @q:  the request queue for the device
 * @new_depth: the new max command queueing depth
 *
 *  Notes:
 *    Must be called with the queue lock held.
 **/
int blk_queue_resize_tags(struct request_queue *q, int new_depth)
{
	struct blk_queue_tag *bqt = q->queue_tags;
	struct request **tag_index;
	unsigned long *tag_map;
	int max_depth, nr_ulongs;

	if (!bqt)
		return -ENXIO;

	/*
	 * if we already have large enough real_max_depth.  just
	 * adjust max_depth.  *NOTE* as requests with tag value
	 * between new_depth and real_max_depth can be in-flight, tag
	 * map can not be shrunk blindly here.
	 */
	if (new_depth <= bqt->real_max_depth) {
		bqt->max_depth = new_depth;
		return 0;
	}

	/*
	 * Currently cannot replace a shared tag map with a new
	 * one, so error out if this is the case
	 */
	if (atomic_read(&bqt->refcnt) != 1)
		return -EBUSY;

	/*
	 * save the old state info, so we can copy it back
	 */
	tag_index = bqt->tag_index;
	tag_map = bqt->tag_map;
	max_depth = bqt->real_max_depth;

	if (init_tag_map(q, bqt, new_depth))
		return -ENOMEM;

	memcpy(bqt->tag_index, tag_index, max_depth * sizeof(struct request *));
	nr_ulongs = ALIGN(max_depth, BITS_PER_LONG) / BITS_PER_LONG;
	memcpy(bqt->tag_map, tag_map, nr_ulongs * sizeof(unsigned long));

	kfree(tag_index);
	kfree(tag_map);
	return 0;
}
EXPORT_SYMBOL(blk_queue_resize_tags);

/**
 * blk_queue_end_tag - end tag operations for a request
 * @q:  the request queue for the device
 * @rq: the request that has completed
 *
 *  Description:
273
 *    Typically called when end_that_request_first() returns %0, meaning
274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300
 *    all transfers have been done for a request. It's important to call
 *    this function before end_that_request_last(), as that will put the
 *    request back on the free list thus corrupting the internal tag list.
 *
 *  Notes:
 *   queue lock must be held.
 **/
void blk_queue_end_tag(struct request_queue *q, struct request *rq)
{
	struct blk_queue_tag *bqt = q->queue_tags;
	int tag = rq->tag;

	BUG_ON(tag == -1);

	if (unlikely(tag >= bqt->real_max_depth))
		/*
		 * This can happen after tag depth has been reduced.
		 * FIXME: how about a warning or info message here?
		 */
		return;

	list_del_init(&rq->queuelist);
	rq->cmd_flags &= ~REQ_QUEUED;
	rq->tag = -1;

	if (unlikely(bqt->tag_index[tag] == NULL))
		printk(KERN_ERR "%s: tag %d is missing\n",
301
		       __func__, tag);
302 303 304 305 306

	bqt->tag_index[tag] = NULL;

	if (unlikely(!test_bit(tag, bqt->tag_map))) {
		printk(KERN_ERR "%s: attempt to clear non-busy tag (%d)\n",
307
		       __func__, tag);
308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338
		return;
	}
	/*
	 * The tag_map bit acts as a lock for tag_index[bit], so we need
	 * unlock memory barrier semantics.
	 */
	clear_bit_unlock(tag, bqt->tag_map);
}
EXPORT_SYMBOL(blk_queue_end_tag);

/**
 * blk_queue_start_tag - find a free tag and assign it
 * @q:  the request queue for the device
 * @rq:  the block request that needs tagging
 *
 *  Description:
 *    This can either be used as a stand-alone helper, or possibly be
 *    assigned as the queue &prep_rq_fn (in which case &struct request
 *    automagically gets a tag assigned). Note that this function
 *    assumes that any type of request can be queued! if this is not
 *    true for your device, you must check the request type before
 *    calling this function.  The request will also be removed from
 *    the request queue, so it's the drivers responsibility to readd
 *    it if it should need to be restarted for some reason.
 *
 *  Notes:
 *   queue lock must be held.
 **/
int blk_queue_start_tag(struct request_queue *q, struct request *rq)
{
	struct blk_queue_tag *bqt = q->queue_tags;
339
	unsigned max_depth, offset;
340 341 342
	int tag;

	if (unlikely((rq->cmd_flags & REQ_QUEUED))) {
343
		printk(KERN_ERR
344
		       "%s: request %p for device [%s] already tagged %d",
345
		       __func__, rq,
346 347 348 349 350 351 352
		       rq->rq_disk ? rq->rq_disk->disk_name : "?", rq->tag);
		BUG();
	}

	/*
	 * Protect against shared tag maps, as we may not have exclusive
	 * access to the tag map.
353 354 355
	 *
	 * We reserve a few tags just for sync IO, since we don't want
	 * to starve sync IO on behalf of flooding async IO.
356
	 */
357 358 359 360 361 362
	max_depth = bqt->max_depth;
	if (rq_is_sync(rq))
		offset = 0;
	else
		offset = max_depth >> 2;

363
	do {
364 365
		tag = find_next_zero_bit(bqt->tag_map, max_depth, offset);
		if (tag >= max_depth)
366 367 368 369 370 371 372 373 374 375 376
			return 1;

	} while (test_and_set_bit_lock(tag, bqt->tag_map));
	/*
	 * We need lock ordering semantics given by test_and_set_bit_lock.
	 * See blk_queue_end_tag for details.
	 */

	rq->cmd_flags |= REQ_QUEUED;
	rq->tag = tag;
	bqt->tag_index[tag] = rq;
377
	blk_start_request(rq);
378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402
	list_add(&rq->queuelist, &q->tag_busy_list);
	return 0;
}
EXPORT_SYMBOL(blk_queue_start_tag);

/**
 * blk_queue_invalidate_tags - invalidate all pending tags
 * @q:  the request queue for the device
 *
 *  Description:
 *   Hardware conditions may dictate a need to stop all pending requests.
 *   In this case, we will safely clear the block side of the tag queue and
 *   readd all requests to the request queue in the right order.
 *
 *  Notes:
 *   queue lock must be held.
 **/
void blk_queue_invalidate_tags(struct request_queue *q)
{
	struct list_head *tmp, *n;

	list_for_each_safe(tmp, n, &q->tag_busy_list)
		blk_requeue_request(q, list_entry_rq(tmp));
}
EXPORT_SYMBOL(blk_queue_invalidate_tags);