aboutsummaryrefslogtreecommitdiff
path: root/drivers/net/ipa/gsi_trans.c
blob: 2fd21d75367d21ceb720fe9fdf1c36227c3723df (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
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
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
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
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
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
273
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
301
302
303
304
305
306
307
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
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
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
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
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
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
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
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
// SPDX-License-Identifier: GPL-2.0

/* Copyright (c) 2012-2018, The Linux Foundation. All rights reserved.
 * Copyright (C) 2019-2020 Linaro Ltd.
 */

#include <linux/types.h>
#include <linux/bits.h>
#include <linux/bitfield.h>
#include <linux/refcount.h>
#include <linux/scatterlist.h>
#include <linux/dma-direction.h>

#include "gsi.h"
#include "gsi_private.h"
#include "gsi_trans.h"
#include "ipa_gsi.h"
#include "ipa_data.h"
#include "ipa_cmd.h"

/**
 * DOC: GSI Transactions
 *
 * A GSI transaction abstracts the behavior of a GSI channel by representing
 * everything about a related group of IPA commands in a single structure.
 * (A "command" in this sense is either a data transfer or an IPA immediate
 * command.)  Most details of interaction with the GSI hardware are managed
 * by the GSI transaction core, allowing users to simply describe commands
 * to be performed.  When a transaction has completed a callback function
 * (dependent on the type of endpoint associated with the channel) allows
 * cleanup of resources associated with the transaction.
 *
 * To perform a command (or set of them), a user of the GSI transaction
 * interface allocates a transaction, indicating the number of TREs required
 * (one per command).  If sufficient TREs are available, they are reserved
 * for use in the transaction and the allocation succeeds.  This way
 * exhaustion of the available TREs in a channel ring is detected
 * as early as possible.  All resources required to complete a transaction
 * are allocated at transaction allocation time.
 *
 * Commands performed as part of a transaction are represented in an array
 * of Linux scatterlist structures.  This array is allocated with the
 * transaction, and its entries are initialized using standard scatterlist
 * functions (such as sg_set_buf() or skb_to_sgvec()).
 *
 * Once a transaction's scatterlist structures have been initialized, the
 * transaction is committed.  The caller is responsible for mapping buffers
 * for DMA if necessary, and this should be done *before* allocating
 * the transaction.  Between a successful allocation and commit of a
 * transaction no errors should occur.
 *
 * Committing transfers ownership of the entire transaction to the GSI
 * transaction core.  The GSI transaction code formats the content of
 * the scatterlist array into the channel ring buffer and informs the
 * hardware that new TREs are available to process.
 *
 * The last TRE in each transaction is marked to interrupt the AP when the
 * GSI hardware has completed it.  Because transfers described by TREs are
 * performed strictly in order, signaling the completion of just the last
 * TRE in the transaction is sufficient to indicate the full transaction
 * is complete.
 *
 * When a transaction is complete, ipa_gsi_trans_complete() is called by the
 * GSI code into the IPA layer, allowing it to perform any final cleanup
 * required before the transaction is freed.
 */

/* Hardware values representing a transfer element type */
enum gsi_tre_type {
	GSI_RE_XFER	= 0x2,
	GSI_RE_IMMD_CMD	= 0x3,
};

/* An entry in a channel ring */
struct gsi_tre {
	__le64 addr;		/* DMA address */
	__le16 len_opcode;	/* length in bytes or enum IPA_CMD_* */
	__le16 reserved;
	__le32 flags;		/* TRE_FLAGS_* */
};

/* gsi_tre->flags mask values (in CPU byte order) */
#define TRE_FLAGS_CHAIN_FMASK	GENMASK(0, 0)
#define TRE_FLAGS_IEOB_FMASK	GENMASK(8, 8)
#define TRE_FLAGS_IEOT_FMASK	GENMASK(9, 9)
#define TRE_FLAGS_BEI_FMASK	GENMASK(10, 10)
#define TRE_FLAGS_TYPE_FMASK	GENMASK(23, 16)

int gsi_trans_pool_init(struct gsi_trans_pool *pool, size_t size, u32 count,
			u32 max_alloc)
{
	void *virt;

#ifdef IPA_VALIDATE
	if (!size || size % 8)
		return -EINVAL;
	if (count < max_alloc)
		return -EINVAL;
	if (!max_alloc)
		return -EINVAL;
#endif /* IPA_VALIDATE */

	/* By allocating a few extra entries in our pool (one less
	 * than the maximum number that will be requested in a
	 * single allocation), we can always satisfy requests without
	 * ever worrying about straddling the end of the pool array.
	 * If there aren't enough entries starting at the free index,
	 * we just allocate free entries from the beginning of the pool.
	 */
	virt = kcalloc(count + max_alloc - 1, size, GFP_KERNEL);
	if (!virt)
		return -ENOMEM;

	pool->base = virt;
	/* If the allocator gave us any extra memory, use it */
	pool->count = ksize(pool->base) / size;
	pool->free = 0;
	pool->max_alloc = max_alloc;
	pool->size = size;
	pool->addr = 0;		/* Only used for DMA pools */

	return 0;
}

void gsi_trans_pool_exit(struct gsi_trans_pool *pool)
{
	kfree(pool->base);
	memset(pool, 0, sizeof(*pool));
}

/* Allocate the requested number of (zeroed) entries from the pool */
/* Home-grown DMA pool.  This way we can preallocate and use the tre_count
 * to guarantee allocations will succeed.  Even though we specify max_alloc
 * (and it can be more than one), we only allow allocation of a single
 * element from a DMA pool.
 */
int gsi_trans_pool_init_dma(struct device *dev, struct gsi_trans_pool *pool,
			    size_t size, u32 count, u32 max_alloc)
{
	size_t total_size;
	dma_addr_t addr;
	void *virt;

#ifdef IPA_VALIDATE
	if (!size || size % 8)
		return -EINVAL;
	if (count < max_alloc)
		return -EINVAL;
	if (!max_alloc)
		return -EINVAL;
#endif /* IPA_VALIDATE */

	/* Don't let allocations cross a power-of-two boundary */
	size = __roundup_pow_of_two(size);
	total_size = (count + max_alloc - 1) * size;

	/* The allocator will give us a power-of-2 number of pages.  But we
	 * can't guarantee that, so request it.  That way we won't waste any
	 * memory that would be available beyond the required space.
	 */
	total_size = get_order(total_size) << PAGE_SHIFT;

	virt = dma_alloc_coherent(dev, total_size, &addr, GFP_KERNEL);
	if (!virt)
		return -ENOMEM;

	pool->base = virt;
	pool->count = total_size / size;
	pool->free = 0;
	pool->size = size;
	pool->max_alloc = max_alloc;
	pool->addr = addr;

	return 0;
}

void gsi_trans_pool_exit_dma(struct device *dev, struct gsi_trans_pool *pool)
{
	dma_free_coherent(dev, pool->size, pool->base, pool->addr);
	memset(pool, 0, sizeof(*pool));
}

/* Return the byte offset of the next free entry in the pool */
static u32 gsi_trans_pool_alloc_common(struct gsi_trans_pool *pool, u32 count)
{
	u32 offset;

	/* assert(count > 0); */
	/* assert(count <= pool->max_alloc); */

	/* Allocate from beginning if wrap would occur */
	if (count > pool->count - pool->free)
		pool->free = 0;

	offset = pool->free * pool->size;
	pool->free += count;
	memset(pool->base + offset, 0, count * pool->size);

	return offset;
}

/* Allocate a contiguous block of zeroed entries from a pool */
void *gsi_trans_pool_alloc(struct gsi_trans_pool *pool, u32 count)
{
	return pool->base + gsi_trans_pool_alloc_common(pool, count);
}

/* Allocate a single zeroed entry from a DMA pool */
void *gsi_trans_pool_alloc_dma(struct gsi_trans_pool *pool, dma_addr_t *addr)
{
	u32 offset = gsi_trans_pool_alloc_common(pool, 1);

	*addr = pool->addr + offset;

	return pool->base + offset;
}

/* Return the pool element that immediately follows the one given.
 * This only works done if elements are allocated one at a time.
 */
void *gsi_trans_pool_next(struct gsi_trans_pool *pool, void *element)
{
	void *end = pool->base + pool->count * pool->size;

	/* assert(element >= pool->base); */
	/* assert(element < end); */
	/* assert(pool->max_alloc == 1); */
	element += pool->size;

	return element < end ? element : pool->base;
}

/* Map a given ring entry index to the transaction associated with it */
static void gsi_channel_trans_map(struct gsi_channel *channel, u32 index,
				  struct gsi_trans *trans)
{
	/* Note: index *must* be used modulo the ring count here */
	channel->trans_info.map[index % channel->tre_ring.count] = trans;
}

/* Return the transaction mapped to a given ring entry */
struct gsi_trans *
gsi_channel_trans_mapped(struct gsi_channel *channel, u32 index)
{
	/* Note: index *must* be used modulo the ring count here */
	return channel->trans_info.map[index % channel->tre_ring.count];
}

/* Return the oldest completed transaction for a channel (or null) */
struct gsi_trans *gsi_channel_trans_complete(struct gsi_channel *channel)
{
	return list_first_entry_or_null(&channel->trans_info.complete,
					struct gsi_trans, links);
}

/* Move a transaction from the allocated list to the pending list */
static void gsi_trans_move_pending(struct gsi_trans *trans)
{
	struct gsi_channel *channel = &trans->gsi->channel[trans->channel_id];
	struct gsi_trans_info *trans_info = &channel->trans_info;

	spin_lock_bh(&trans_info->spinlock);

	list_move_tail(&trans->links, &trans_info->pending);

	spin_unlock_bh(&trans_info->spinlock);
}

/* Move a transaction and all of its predecessors from the pending list
 * to the completed list.
 */
void gsi_trans_move_complete(struct gsi_trans *trans)
{
	struct gsi_channel *channel = &trans->gsi->channel[trans->channel_id];
	struct gsi_trans_info *trans_info = &channel->trans_info;
	struct list_head list;

	spin_lock_bh(&trans_info->spinlock);

	/* Move this transaction and all predecessors to completed list */
	list_cut_position(&list, &trans_info->pending, &trans->links);
	list_splice_tail(&list, &trans_info->complete);

	spin_unlock_bh(&trans_info->spinlock);
}

/* Move a transaction from the completed list to the polled list */
void gsi_trans_move_polled(struct gsi_trans *trans)
{
	struct gsi_channel *channel = &trans->gsi->channel[trans->channel_id];
	struct gsi_trans_info *trans_info = &channel->trans_info;

	spin_lock_bh(&trans_info->spinlock);

	list_move_tail(&trans->links, &trans_info->polled);

	spin_unlock_bh(&trans_info->spinlock);
}

/* Reserve some number of TREs on a channel.  Returns true if successful */
static bool
gsi_trans_tre_reserve(struct gsi_trans_info *trans_info, u32 tre_count)
{
	int avail = atomic_read(&trans_info->tre_avail);
	int new;

	do {
		new = avail - (int)tre_count;
		if (unlikely(new < 0))
			return false;
	} while (!atomic_try_cmpxchg(&trans_info->tre_avail, &avail, new));

	return true;
}

/* Release previously-reserved TRE entries to a channel */
static void
gsi_trans_tre_release(struct gsi_trans_info *trans_info, u32 tre_count)
{
	atomic_add(tre_count, &trans_info->tre_avail);
}

/* Allocate a GSI transaction on a channel */
struct gsi_trans *gsi_channel_trans_alloc(struct gsi *gsi, u32 channel_id,
					  u32 tre_count,
					  enum dma_data_direction direction)
{
	struct gsi_channel *channel = &gsi->channel[channel_id];
	struct gsi_trans_info *trans_info;
	struct gsi_trans *trans;

	/* assert(tre_count <= gsi_channel_trans_tre_max(gsi, channel_id)); */

	trans_info = &channel->trans_info;

	/* We reserve the TREs now, but consume them at commit time.
	 * If there aren't enough available, we're done.
	 */
	if (!gsi_trans_tre_reserve(trans_info, tre_count))
		return NULL;

	/* Allocate and initialize non-zero fields in the the transaction */
	trans = gsi_trans_pool_alloc(&trans_info->pool, 1);
	trans->gsi = gsi;
	trans->channel_id = channel_id;
	trans->tre_count = tre_count;
	init_completion(&trans->completion);

	/* Allocate the scatterlist and (if requested) info entries. */
	trans->sgl = gsi_trans_pool_alloc(&trans_info->sg_pool, tre_count);
	sg_init_marker(trans->sgl, tre_count);

	trans->direction = direction;

	spin_lock_bh(&trans_info->spinlock);

	list_add_tail(&trans->links, &trans_info->alloc);

	spin_unlock_bh(&trans_info->spinlock);

	refcount_set(&trans->refcount, 1);

	return trans;
}

/* Free a previously-allocated transaction (used only in case of error) */
void gsi_trans_free(struct gsi_trans *trans)
{
	struct gsi_trans_info *trans_info;

	if (!refcount_dec_and_test(&trans->refcount))
		return;

	trans_info = &trans->gsi->channel[trans->channel_id].trans_info;

	spin_lock_bh(&trans_info->spinlock);

	list_del(&trans->links);

	spin_unlock_bh(&trans_info->spinlock);

	ipa_gsi_trans_release(trans);

	/* Releasing the reserved TREs implicitly frees the sgl[] and
	 * (if present) info[] arrays, plus the transaction itself.
	 */
	gsi_trans_tre_release(trans_info, trans->tre_count);
}

/* Add an immediate command to a transaction */
void gsi_trans_cmd_add(struct gsi_trans *trans, void *buf, u32 size,
		       dma_addr_t addr, enum dma_data_direction direction,
		       enum ipa_cmd_opcode opcode)
{
	struct ipa_cmd_info *info;
	u32 which = trans->used++;
	struct scatterlist *sg;

	/* assert(which < trans->tre_count); */

	/* Set the page information for the buffer.  We also need to fill in
	 * the DMA address for the buffer (something dma_map_sg() normally
	 * does).
	 */
	sg = &trans->sgl[which];

	sg_set_buf(sg, buf, size);
	sg_dma_address(sg) = addr;

	info = &trans->info[which];
	info->opcode = opcode;
	info->direction = direction;
}

/* Add a page transfer to a transaction.  It will fill the only TRE. */
int gsi_trans_page_add(struct gsi_trans *trans, struct page *page, u32 size,
		       u32 offset)
{
	struct scatterlist *sg = &trans->sgl[0];
	int ret;

	/* assert(trans->tre_count == 1); */
	/* assert(!trans->used); */

	sg_set_page(sg, page, size, offset);
	ret = dma_map_sg(trans->gsi->dev, sg, 1, trans->direction);
	if (!ret)
		return -ENOMEM;

	trans->used++;	/* Transaction now owns the (DMA mapped) page */

	return 0;
}

/* Add an SKB transfer to a transaction.  No other TREs will be used. */
int gsi_trans_skb_add(struct gsi_trans *trans, struct sk_buff *skb)
{
	struct scatterlist *sg = &trans->sgl[0];
	u32 used;
	int ret;

	/* assert(trans->tre_count == 1); */
	/* assert(!trans->used); */

	/* skb->len will not be 0 (checked early) */
	ret = skb_to_sgvec(skb, sg, 0, skb->len);
	if (ret < 0)
		return ret;
	used = ret;

	ret = dma_map_sg(trans->gsi->dev, sg, used, trans->direction);
	if (!ret)
		return -ENOMEM;

	trans->used += used;	/* Transaction now owns the (DMA mapped) skb */

	return 0;
}

/* Compute the length/opcode value to use for a TRE */
static __le16 gsi_tre_len_opcode(enum ipa_cmd_opcode opcode, u32 len)
{
	return opcode == IPA_CMD_NONE ? cpu_to_le16((u16)len)
				      : cpu_to_le16((u16)opcode);
}

/* Compute the flags value to use for a given TRE */
static __le32 gsi_tre_flags(bool last_tre, bool bei, enum ipa_cmd_opcode opcode)
{
	enum gsi_tre_type tre_type;
	u32 tre_flags;

	tre_type = opcode == IPA_CMD_NONE ? GSI_RE_XFER : GSI_RE_IMMD_CMD;
	tre_flags = u32_encode_bits(tre_type, TRE_FLAGS_TYPE_FMASK);

	/* Last TRE contains interrupt flags */
	if (last_tre) {
		/* All transactions end in a transfer completion interrupt */
		tre_flags |= TRE_FLAGS_IEOT_FMASK;
		/* Don't interrupt when outbound commands are acknowledged */
		if (bei)
			tre_flags |= TRE_FLAGS_BEI_FMASK;
	} else {	/* All others indicate there's more to come */
		tre_flags |= TRE_FLAGS_CHAIN_FMASK;
	}

	return cpu_to_le32(tre_flags);
}

static void gsi_trans_tre_fill(struct gsi_tre *dest_tre, dma_addr_t addr,
			       u32 len, bool last_tre, bool bei,
			       enum ipa_cmd_opcode opcode)
{
	struct gsi_tre tre;

	tre.addr = cpu_to_le64(addr);
	tre.len_opcode = gsi_tre_len_opcode(opcode, len);
	tre.reserved = 0;
	tre.flags = gsi_tre_flags(last_tre, bei, opcode);

	/* ARM64 can write 16 bytes as a unit with a single instruction.
	 * Doing the assignment this way is an attempt to make that happen.
	 */
	*dest_tre = tre;
}

/**
 * __gsi_trans_commit() - Common GSI transaction commit code
 * @trans:	Transaction to commit
 * @ring_db:	Whether to tell the hardware about these queued transfers
 *
 * Formats channel ring TRE entries based on the content of the scatterlist.
 * Maps a transaction pointer to the last ring entry used for the transaction,
 * so it can be recovered when it completes.  Moves the transaction to the
 * pending list.  Finally, updates the channel ring pointer and optionally
 * rings the doorbell.
 */
static void __gsi_trans_commit(struct gsi_trans *trans, bool ring_db)
{
	struct gsi_channel *channel = &trans->gsi->channel[trans->channel_id];
	struct gsi_ring *ring = &channel->tre_ring;
	enum ipa_cmd_opcode opcode = IPA_CMD_NONE;
	bool bei = channel->toward_ipa;
	struct ipa_cmd_info *info;
	struct gsi_tre *dest_tre;
	struct scatterlist *sg;
	u32 byte_count = 0;
	u32 avail;
	u32 i;

	/* assert(trans->used > 0); */

	/* Consume the entries.  If we cross the end of the ring while
	 * filling them we'll switch to the beginning to finish.
	 * If there is no info array we're doing a simple data
	 * transfer request, whose opcode is IPA_CMD_NONE.
	 */
	info = trans->info ? &trans->info[0] : NULL;
	avail = ring->count - ring->index % ring->count;
	dest_tre = gsi_ring_virt(ring, ring->index);
	for_each_sg(trans->sgl, sg, trans->used, i) {
		bool last_tre = i == trans->used - 1;
		dma_addr_t addr = sg_dma_address(sg);
		u32 len = sg_dma_len(sg);

		byte_count += len;
		if (!avail--)
			dest_tre = gsi_ring_virt(ring, 0);
		if (info)
			opcode = info++->opcode;

		gsi_trans_tre_fill(dest_tre, addr, len, last_tre, bei, opcode);
		dest_tre++;
	}
	ring->index += trans->used;

	if (channel->toward_ipa) {
		/* We record TX bytes when they are sent */
		trans->len = byte_count;
		trans->trans_count = channel->trans_count;
		trans->byte_count = channel->byte_count;
		channel->trans_count++;
		channel->byte_count += byte_count;
	}

	/* Associate the last TRE with the transaction */
	gsi_channel_trans_map(channel, ring->index - 1, trans);

	gsi_trans_move_pending(trans);

	/* Ring doorbell if requested, or if all TREs are allocated */
	if (ring_db || !atomic_read(&channel->trans_info.tre_avail)) {
		/* Report what we're handing off to hardware for TX channels */
		if (channel->toward_ipa)
			gsi_channel_tx_queued(channel);
		gsi_channel_doorbell(channel);
	}
}

/* Commit a GSI transaction */
void gsi_trans_commit(struct gsi_trans *trans, bool ring_db)
{
	if (trans->used)
		__gsi_trans_commit(trans, ring_db);
	else
		gsi_trans_free(trans);
}

/* Commit a GSI transaction and wait for it to complete */
void gsi_trans_commit_wait(struct gsi_trans *trans)
{
	if (!trans->used)
		goto out_trans_free;

	refcount_inc(&trans->refcount);

	__gsi_trans_commit(trans, true);

	wait_for_completion(&trans->completion);

out_trans_free:
	gsi_trans_free(trans);
}

/* Commit a GSI transaction and wait for it to complete, with timeout */
int gsi_trans_commit_wait_timeout(struct gsi_trans *trans,
				  unsigned long timeout)
{
	unsigned long timeout_jiffies = msecs_to_jiffies(timeout);
	unsigned long remaining = 1;	/* In case of empty transaction */

	if (!trans->used)
		goto out_trans_free;

	refcount_inc(&trans->refcount);

	__gsi_trans_commit(trans, true);

	remaining = wait_for_completion_timeout(&trans->completion,
						timeout_jiffies);
out_trans_free:
	gsi_trans_free(trans);

	return remaining ? 0 : -ETIMEDOUT;
}

/* Process the completion of a transaction; called while polling */
void gsi_trans_complete(struct gsi_trans *trans)
{
	/* If the entire SGL was mapped when added, unmap it now */
	if (trans->direction != DMA_NONE)
		dma_unmap_sg(trans->gsi->dev, trans->sgl, trans->used,
			     trans->direction);

	ipa_gsi_trans_complete(trans);

	complete(&trans->completion);

	gsi_trans_free(trans);
}

/* Cancel a channel's pending transactions */
void gsi_channel_trans_cancel_pending(struct gsi_channel *channel)
{
	struct gsi_trans_info *trans_info = &channel->trans_info;
	struct gsi_trans *trans;
	bool cancelled;

	/* channel->gsi->mutex is held by caller */
	spin_lock_bh(&trans_info->spinlock);

	cancelled = !list_empty(&trans_info->pending);
	list_for_each_entry(trans, &trans_info->pending, links)
		trans->cancelled = true;

	list_splice_tail_init(&trans_info->pending, &trans_info->complete);

	spin_unlock_bh(&trans_info->spinlock);

	/* Schedule NAPI polling to complete the cancelled transactions */
	if (cancelled)
		napi_schedule(&channel->napi);
}

/* Issue a command to read a single byte from a channel */
int gsi_trans_read_byte(struct gsi *gsi, u32 channel_id, dma_addr_t addr)
{
	struct gsi_channel *channel = &gsi->channel[channel_id];
	struct gsi_ring *ring = &channel->tre_ring;
	struct gsi_trans_info *trans_info;
	struct gsi_tre *dest_tre;

	trans_info = &channel->trans_info;

	/* First reserve the TRE, if possible */
	if (!gsi_trans_tre_reserve(trans_info, 1))
		return -EBUSY;

	/* Now fill the the reserved TRE and tell the hardware */

	dest_tre = gsi_ring_virt(ring, ring->index);
	gsi_trans_tre_fill(dest_tre, addr, 1, true, false, IPA_CMD_NONE);

	ring->index++;
	gsi_channel_doorbell(channel);

	return 0;
}

/* Mark a gsi_trans_read_byte() request done */
void gsi_trans_read_byte_done(struct gsi *gsi, u32 channel_id)
{
	struct gsi_channel *channel = &gsi->channel[channel_id];

	gsi_trans_tre_release(&channel->trans_info, 1);
}

/* Initialize a channel's GSI transaction info */
int gsi_channel_trans_init(struct gsi *gsi, u32 channel_id)
{
	struct gsi_channel *channel = &gsi->channel[channel_id];
	struct gsi_trans_info *trans_info;
	u32 tre_max;
	int ret;

	/* Ensure the size of a channel element is what's expected */
	BUILD_BUG_ON(sizeof(struct gsi_tre) != GSI_RING_ELEMENT_SIZE);

	/* The map array is used to determine what transaction is associated
	 * with a TRE that the hardware reports has completed.  We need one
	 * map entry per TRE.
	 */
	trans_info = &channel->trans_info;
	trans_info->map = kcalloc(channel->tre_count, sizeof(*trans_info->map),
				  GFP_KERNEL);
	if (!trans_info->map)
		return -ENOMEM;

	/* We can't use more TREs than there are available in the ring.
	 * This limits the number of transactions that can be oustanding.
	 * Worst case is one TRE per transaction (but we actually limit
	 * it to something a little less than that).  We allocate resources
	 * for transactions (including transaction structures) based on
	 * this maximum number.
	 */
	tre_max = gsi_channel_tre_max(channel->gsi, channel_id);

	/* Transactions are allocated one at a time. */
	ret = gsi_trans_pool_init(&trans_info->pool, sizeof(struct gsi_trans),
				  tre_max, 1);
	if (ret)
		goto err_kfree;

	/* A transaction uses a scatterlist array to represent the data
	 * transfers implemented by the transaction.  Each scatterlist
	 * element is used to fill a single TRE when the transaction is
	 * committed.  So we need as many scatterlist elements as the
	 * maximum number of TREs that can be outstanding.
	 *
	 * All TREs in a transaction must fit within the channel's TLV FIFO.
	 * A transaction on a channel can allocate as many TREs as that but
	 * no more.
	 */
	ret = gsi_trans_pool_init(&trans_info->sg_pool,
				  sizeof(struct scatterlist),
				  tre_max, channel->tlv_count);
	if (ret)
		goto err_trans_pool_exit;

	/* Finally, the tre_avail field is what ultimately limits the number
	 * of outstanding transactions and their resources.  A transaction
	 * allocation succeeds only if the TREs available are sufficient for
	 * what the transaction might need.  Transaction resource pools are
	 * sized based on the maximum number of outstanding TREs, so there
	 * will always be resources available if there are TREs available.
	 */
	atomic_set(&trans_info->tre_avail, tre_max);

	spin_lock_init(&trans_info->spinlock);
	INIT_LIST_HEAD(&trans_info->alloc);
	INIT_LIST_HEAD(&trans_info->pending);
	INIT_LIST_HEAD(&trans_info->complete);
	INIT_LIST_HEAD(&trans_info->polled);

	return 0;

err_trans_pool_exit:
	gsi_trans_pool_exit(&trans_info->pool);
err_kfree:
	kfree(trans_info->map);

	dev_err(gsi->dev, "error %d initializing channel %u transactions\n",
		ret, channel_id);

	return ret;
}

/* Inverse of gsi_channel_trans_init() */
void gsi_channel_trans_exit(struct gsi_channel *channel)
{
	struct gsi_trans_info *trans_info = &channel->trans_info;

	gsi_trans_pool_exit(&trans_info->sg_pool);
	gsi_trans_pool_exit(&trans_info->pool);
	kfree(trans_info->map);
}