summaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/sfc/rx.c
blob: 864b6ffc9cb22f51e6271dfd0775bba327fe2c00 (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
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
/****************************************************************************
 * Driver for Solarflare Solarstorm network controllers and boards
 * Copyright 2005-2006 Fen Systems Ltd.
 * Copyright 2005-2011 Solarflare Communications Inc.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published
 * by the Free Software Foundation, incorporated herein by reference.
 */

#include <linux/socket.h>
#include <linux/in.h>
#include <linux/slab.h>
#include <linux/ip.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/prefetch.h>
#include <linux/moduleparam.h>
#include <linux/iommu.h>
#include <net/ip.h>
#include <net/checksum.h>
#include "net_driver.h"
#include "efx.h"
#include "filter.h"
#include "nic.h"
#include "selftest.h"
#include "workarounds.h"

/* Preferred number of descriptors to fill at once */
#define EFX_RX_PREFERRED_BATCH 8U

/* Number of RX buffers to recycle pages for.  When creating the RX page recycle
 * ring, this number is divided by the number of buffers per page to calculate
 * the number of pages to store in the RX page recycle ring.
 */
#define EFX_RECYCLE_RING_SIZE_IOMMU 4096
#define EFX_RECYCLE_RING_SIZE_NOIOMMU (2 * EFX_RX_PREFERRED_BATCH)

/* Size of buffer allocated for skb header area. */
#define EFX_SKB_HEADERS  128u

/* This is the percentage fill level below which new RX descriptors
 * will be added to the RX descriptor ring.
 */
static unsigned int rx_refill_threshold;

/* Each packet can consume up to ceil(max_frame_len / buffer_size) buffers */
#define EFX_RX_MAX_FRAGS DIV_ROUND_UP(EFX_MAX_FRAME_LEN(EFX_MAX_MTU), \
				      EFX_RX_USR_BUF_SIZE)

/*
 * RX maximum head room required.
 *
 * This must be at least 1 to prevent overflow, plus one packet-worth
 * to allow pipelined receives.
 */
#define EFX_RXD_HEAD_ROOM (1 + EFX_RX_MAX_FRAGS)

static inline u8 *efx_rx_buf_va(struct efx_rx_buffer *buf)
{
	return page_address(buf->page) + buf->page_offset;
}

static inline u32 efx_rx_buf_hash(struct efx_nic *efx, const u8 *eh)
{
#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS)
	return __le32_to_cpup((const __le32 *)(eh + efx->rx_packet_hash_offset));
#else
	const u8 *data = eh + efx->rx_packet_hash_offset;
	return (u32)data[0]	  |
	       (u32)data[1] << 8  |
	       (u32)data[2] << 16 |
	       (u32)data[3] << 24;
#endif
}

static inline struct efx_rx_buffer *
efx_rx_buf_next(struct efx_rx_queue *rx_queue, struct efx_rx_buffer *rx_buf)
{
	if (unlikely(rx_buf == efx_rx_buffer(rx_queue, rx_queue->ptr_mask)))
		return efx_rx_buffer(rx_queue, 0);
	else
		return rx_buf + 1;
}

static inline void efx_sync_rx_buffer(struct efx_nic *efx,
				      struct efx_rx_buffer *rx_buf,
				      unsigned int len)
{
	dma_sync_single_for_cpu(&efx->pci_dev->dev, rx_buf->dma_addr, len,
				DMA_FROM_DEVICE);
}

void efx_rx_config_page_split(struct efx_nic *efx)
{
	efx->rx_page_buf_step = ALIGN(efx->rx_dma_len + NET_IP_ALIGN,
				      EFX_RX_BUF_ALIGNMENT);
	efx->rx_bufs_per_page = efx->rx_buffer_order ? 1 :
		((PAGE_SIZE - sizeof(struct efx_rx_page_state)) /
		 efx->rx_page_buf_step);
	efx->rx_buffer_truesize = (PAGE_SIZE << efx->rx_buffer_order) /
		efx->rx_bufs_per_page;
	efx->rx_pages_per_batch = DIV_ROUND_UP(EFX_RX_PREFERRED_BATCH,
					       efx->rx_bufs_per_page);
}

/* Check the RX page recycle ring for a page that can be reused. */
static struct page *efx_reuse_page(struct efx_rx_queue *rx_queue)
{
	struct efx_nic *efx = rx_queue->efx;
	struct page *page;
	struct efx_rx_page_state *state;
	unsigned index;

	index = rx_queue->page_remove & rx_queue->page_ptr_mask;
	page = rx_queue->page_ring[index];
	if (page == NULL)
		return NULL;

	rx_queue->page_ring[index] = NULL;
	/* page_remove cannot exceed page_add. */
	if (rx_queue->page_remove != rx_queue->page_add)
		++rx_queue->page_remove;

	/* If page_count is 1 then we hold the only reference to this page. */
	if (page_count(page) == 1) {
		++rx_queue->page_recycle_count;
		return page;
	} else {
		state = page_address(page);
		dma_unmap_page(&efx->pci_dev->dev, state->dma_addr,
			       PAGE_SIZE << efx->rx_buffer_order,
			       DMA_FROM_DEVICE);
		put_page(page);
		++rx_queue->page_recycle_failed;
	}

	return NULL;
}

/**
 * efx_init_rx_buffers - create EFX_RX_BATCH page-based RX buffers
 *
 * @rx_queue:		Efx RX queue
 *
 * This allocates a batch of pages, maps them for DMA, and populates
 * struct efx_rx_buffers for each one. Return a negative error code or
 * 0 on success. If a single page can be used for multiple buffers,
 * then the page will either be inserted fully, or not at all.
 */
static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue)
{
	struct efx_nic *efx = rx_queue->efx;
	struct efx_rx_buffer *rx_buf;
	struct page *page;
	unsigned int page_offset;
	struct efx_rx_page_state *state;
	dma_addr_t dma_addr;
	unsigned index, count;

	count = 0;
	do {
		page = efx_reuse_page(rx_queue);
		if (page == NULL) {
			page = alloc_pages(__GFP_COLD | __GFP_COMP | GFP_ATOMIC,
					   efx->rx_buffer_order);
			if (unlikely(page == NULL))
				return -ENOMEM;
			dma_addr =
				dma_map_page(&efx->pci_dev->dev, page, 0,
					     PAGE_SIZE << efx->rx_buffer_order,
					     DMA_FROM_DEVICE);
			if (unlikely(dma_mapping_error(&efx->pci_dev->dev,
						       dma_addr))) {
				__free_pages(page, efx->rx_buffer_order);
				return -EIO;
			}
			state = page_address(page);
			state->dma_addr = dma_addr;
		} else {
			state = page_address(page);
			dma_addr = state->dma_addr;
		}

		dma_addr += sizeof(struct efx_rx_page_state);
		page_offset = sizeof(struct efx_rx_page_state);

		do {
			index = rx_queue->added_count & rx_queue->ptr_mask;
			rx_buf = efx_rx_buffer(rx_queue, index);
			rx_buf->dma_addr = dma_addr + NET_IP_ALIGN;
			rx_buf->page = page;
			rx_buf->page_offset = page_offset + NET_IP_ALIGN;
			rx_buf->len = efx->rx_dma_len;
			rx_buf->flags = 0;
			++rx_queue->added_count;
			get_page(page);
			dma_addr += efx->rx_page_buf_step;
			page_offset += efx->rx_page_buf_step;
		} while (page_offset + efx->rx_page_buf_step <= PAGE_SIZE);

		rx_buf->flags = EFX_RX_BUF_LAST_IN_PAGE;
	} while (++count < efx->rx_pages_per_batch);

	return 0;
}

/* Unmap a DMA-mapped page.  This function is only called for the final RX
 * buffer in a page.
 */
static void efx_unmap_rx_buffer(struct efx_nic *efx,
				struct efx_rx_buffer *rx_buf)
{
	struct page *page = rx_buf->page;

	if (page) {
		struct efx_rx_page_state *state = page_address(page);
		dma_unmap_page(&efx->pci_dev->dev,
			       state->dma_addr,
			       PAGE_SIZE << efx->rx_buffer_order,
			       DMA_FROM_DEVICE);
	}
}

static void efx_free_rx_buffer(struct efx_rx_buffer *rx_buf)
{
	if (rx_buf->page) {
		put_page(rx_buf->page);
		rx_buf->page = NULL;
	}
}

/* Attempt to recycle the page if there is an RX recycle ring; the page can
 * only be added if this is the final RX buffer, to prevent pages being used in
 * the descriptor ring and appearing in the recycle ring simultaneously.
 */
static void efx_recycle_rx_page(struct efx_channel *channel,
				struct efx_rx_buffer *rx_buf)
{
	struct page *page = rx_buf->page;
	struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel);
	struct efx_nic *efx = rx_queue->efx;
	unsigned index;

	/* Only recycle the page after processing the final buffer. */
	if (!(rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE))
		return;

	index = rx_queue->page_add & rx_queue->page_ptr_mask;
	if (rx_queue->page_ring[index] == NULL) {
		unsigned read_index = rx_queue->page_remove &
			rx_queue->page_ptr_mask;

		/* The next slot in the recycle ring is available, but
		 * increment page_remove if the read pointer currently
		 * points here.
		 */
		if (read_index == index)
			++rx_queue->page_remove;
		rx_queue->page_ring[index] = page;
		++rx_queue->page_add;
		return;
	}
	++rx_queue->page_recycle_full;
	efx_unmap_rx_buffer(efx, rx_buf);
	put_page(rx_buf->page);
}

static void efx_fini_rx_buffer(struct efx_rx_queue *rx_queue,
			       struct efx_rx_buffer *rx_buf)
{
	/* Release the page reference we hold for the buffer. */
	if (rx_buf->page)
		put_page(rx_buf->page);

	/* If this is the last buffer in a page, unmap and free it. */
	if (rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE) {
		efx_unmap_rx_buffer(rx_queue->efx, rx_buf);
		efx_free_rx_buffer(rx_buf);
	}
	rx_buf->page = NULL;
}

/* Recycle the pages that are used by buffers that have just been received. */
static void efx_recycle_rx_pages(struct efx_channel *channel,
				 struct efx_rx_buffer *rx_buf,
				 unsigned int n_frags)
{
	struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel);

	do {
		efx_recycle_rx_page(channel, rx_buf);
		rx_buf = efx_rx_buf_next(rx_queue, rx_buf);
	} while (--n_frags);
}

static void efx_discard_rx_packet(struct efx_channel *channel,
				  struct efx_rx_buffer *rx_buf,
				  unsigned int n_frags)
{
	struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel);

	efx_recycle_rx_pages(channel, rx_buf, n_frags);

	do {
		efx_free_rx_buffer(rx_buf);
		rx_buf = efx_rx_buf_next(rx_queue, rx_buf);
	} while (--n_frags);
}

/**
 * efx_fast_push_rx_descriptors - push new RX descriptors quickly
 * @rx_queue:		RX descriptor queue
 *
 * This will aim to fill the RX descriptor queue up to
 * @rx_queue->@max_fill. If there is insufficient atomic
 * memory to do so, a slow fill will be scheduled.
 *
 * The caller must provide serialisation (none is used here). In practise,
 * this means this function must run from the NAPI handler, or be called
 * when NAPI is disabled.
 */
void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue)
{
	struct efx_nic *efx = rx_queue->efx;
	unsigned int fill_level, batch_size;
	int space, rc = 0;

	if (!rx_queue->refill_enabled)
		return;

	/* Calculate current fill level, and exit if we don't need to fill */
	fill_level = (rx_queue->added_count - rx_queue->removed_count);
	EFX_BUG_ON_PARANOID(fill_level > rx_queue->efx->rxq_entries);
	if (fill_level >= rx_queue->fast_fill_trigger)
		goto out;

	/* Record minimum fill level */
	if (unlikely(fill_level < rx_queue->min_fill)) {
		if (fill_level)
			rx_queue->min_fill = fill_level;
	}

	batch_size = efx->rx_pages_per_batch * efx->rx_bufs_per_page;
	space = rx_queue->max_fill - fill_level;
	EFX_BUG_ON_PARANOID(space < batch_size);

	netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev,
		   "RX queue %d fast-filling descriptor ring from"
		   " level %d to level %d\n",
		   efx_rx_queue_index(rx_queue), fill_level,
		   rx_queue->max_fill);


	do {
		rc = efx_init_rx_buffers(rx_queue);
		if (unlikely(rc)) {
			/* Ensure that we don't leave the rx queue empty */
			if (rx_queue->added_count == rx_queue->removed_count)
				efx_schedule_slow_fill(rx_queue);
			goto out;
		}
	} while ((space -= batch_size) >= batch_size);

	netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev,
		   "RX queue %d fast-filled descriptor ring "
		   "to level %d\n", efx_rx_queue_index(rx_queue),
		   rx_queue->added_count - rx_queue->removed_count);

 out:
	if (rx_queue->notified_count != rx_queue->added_count)
		efx_nic_notify_rx_desc(rx_queue);
}

void efx_rx_slow_fill(unsigned long context)
{
	struct efx_rx_queue *rx_queue = (struct efx_rx_queue *)context;

	/* Post an event to cause NAPI to run and refill the queue */
	efx_nic_generate_fill_event(rx_queue);
	++rx_queue->slow_fill_count;
}

static void efx_rx_packet__check_len(struct efx_rx_queue *rx_queue,
				     struct efx_rx_buffer *rx_buf,
				     int len)
{
	struct efx_nic *efx = rx_queue->efx;
	unsigned max_len = rx_buf->len - efx->type->rx_buffer_padding;

	if (likely(len <= max_len))
		return;

	/* The packet must be discarded, but this is only a fatal error
	 * if the caller indicated it was
	 */
	rx_buf->flags |= EFX_RX_PKT_DISCARD;

	if ((len > rx_buf->len) && EFX_WORKAROUND_8071(efx)) {
		if (net_ratelimit())
			netif_err(efx, rx_err, efx->net_dev,
				  " RX queue %d seriously overlength "
				  "RX event (0x%x > 0x%x+0x%x). Leaking\n",
				  efx_rx_queue_index(rx_queue), len, max_len,
				  efx->type->rx_buffer_padding);
		efx_schedule_reset(efx, RESET_TYPE_RX_RECOVERY);
	} else {
		if (net_ratelimit())
			netif_err(efx, rx_err, efx->net_dev,
				  " RX queue %d overlength RX event "
				  "(0x%x > 0x%x)\n",
				  efx_rx_queue_index(rx_queue), len, max_len);
	}

	efx_rx_queue_channel(rx_queue)->n_rx_overlength++;
}

/* Pass a received packet up through GRO.  GRO can handle pages
 * regardless of checksum state and skbs with a good checksum.
 */
static void
efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf,
		  unsigned int n_frags, u8 *eh)
{
	struct napi_struct *napi = &channel->napi_str;
	gro_result_t gro_result;
	struct efx_nic *efx = channel->efx;
	struct sk_buff *skb;

	skb = napi_get_frags(napi);
	if (unlikely(!skb)) {
		while (n_frags--) {
			put_page(rx_buf->page);
			rx_buf->page = NULL;
			rx_buf = efx_rx_buf_next(&channel->rx_queue, rx_buf);
		}
		return;
	}

	if (efx->net_dev->features & NETIF_F_RXHASH)
		skb->rxhash = efx_rx_buf_hash(efx, eh);
	skb->ip_summed = ((rx_buf->flags & EFX_RX_PKT_CSUMMED) ?
			  CHECKSUM_UNNECESSARY : CHECKSUM_NONE);

	for (;;) {
		skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags,
				   rx_buf->page, rx_buf->page_offset,
				   rx_buf->len);
		rx_buf->page = NULL;
		skb->len += rx_buf->len;
		if (skb_shinfo(skb)->nr_frags == n_frags)
			break;

		rx_buf = efx_rx_buf_next(&channel->rx_queue, rx_buf);
	}

	skb->data_len = skb->len;
	skb->truesize += n_frags * efx->rx_buffer_truesize;

	skb_record_rx_queue(skb, channel->rx_queue.core_index);

	gro_result = napi_gro_frags(napi);
	if (gro_result != GRO_DROP)
		channel->irq_mod_score += 2;
}

/* Allocate and construct an SKB around page fragments */
static struct sk_buff *efx_rx_mk_skb(struct efx_channel *channel,
				     struct efx_rx_buffer *rx_buf,
				     unsigned int n_frags,
				     u8 *eh, int hdr_len)
{
	struct efx_nic *efx = channel->efx;
	struct sk_buff *skb;

	/* Allocate an SKB to store the headers */
	skb = netdev_alloc_skb(efx->net_dev, hdr_len + EFX_PAGE_SKB_ALIGN);
	if (unlikely(skb == NULL))
		return NULL;

	EFX_BUG_ON_PARANOID(rx_buf->len < hdr_len);

	skb_reserve(skb, EFX_PAGE_SKB_ALIGN);
	memcpy(__skb_put(skb, hdr_len), eh, hdr_len);

	/* Append the remaining page(s) onto the frag list */
	if (rx_buf->len > hdr_len) {
		rx_buf->page_offset += hdr_len;
		rx_buf->len -= hdr_len;

		for (;;) {
			skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags,
					   rx_buf->page, rx_buf->page_offset,
					   rx_buf->len);
			rx_buf->page = NULL;
			skb->len += rx_buf->len;
			skb->data_len += rx_buf->len;
			if (skb_shinfo(skb)->nr_frags == n_frags)
				break;

			rx_buf = efx_rx_buf_next(&channel->rx_queue, rx_buf);
		}
	} else {
		__free_pages(rx_buf->page, efx->rx_buffer_order);
		rx_buf->page = NULL;
		n_frags = 0;
	}

	skb->truesize += n_frags * efx->rx_buffer_truesize;

	/* Move past the ethernet header */
	skb->protocol = eth_type_trans(skb, efx->net_dev);

	return skb;
}

void efx_rx_packet(struct efx_rx_queue *rx_queue, unsigned int index,
		   unsigned int n_frags, unsigned int len, u16 flags)
{
	struct efx_nic *efx = rx_queue->efx;
	struct efx_channel *channel = efx_rx_queue_channel(rx_queue);
	struct efx_rx_buffer *rx_buf;

	rx_buf = efx_rx_buffer(rx_queue, index);
	rx_buf->flags |= flags;

	/* Validate the number of fragments and completed length */
	if (n_frags == 1) {
		if (!(flags & EFX_RX_PKT_PREFIX_LEN))
			efx_rx_packet__check_len(rx_queue, rx_buf, len);
	} else if (unlikely(n_frags > EFX_RX_MAX_FRAGS) ||
		   unlikely(len <= (n_frags - 1) * EFX_RX_USR_BUF_SIZE) ||
		   unlikely(len > n_frags * EFX_RX_USR_BUF_SIZE) ||
		   unlikely(!efx->rx_scatter)) {
		/* If this isn't an explicit discard request, either
		 * the hardware or the driver is broken.
		 */
		WARN_ON(!(len == 0 && rx_buf->flags & EFX_RX_PKT_DISCARD));
		rx_buf->flags |= EFX_RX_PKT_DISCARD;
	}

	netif_vdbg(efx, rx_status, efx->net_dev,
		   "RX queue %d received ids %x-%x len %d %s%s\n",
		   efx_rx_queue_index(rx_queue), index,
		   (index + n_frags - 1) & rx_queue->ptr_mask, len,
		   (rx_buf->flags & EFX_RX_PKT_CSUMMED) ? " [SUMMED]" : "",
		   (rx_buf->flags & EFX_RX_PKT_DISCARD) ? " [DISCARD]" : "");

	/* Discard packet, if instructed to do so.  Process the
	 * previous receive first.
	 */
	if (unlikely(rx_buf->flags & EFX_RX_PKT_DISCARD)) {
		efx_rx_flush_packet(channel);
		efx_discard_rx_packet(channel, rx_buf, n_frags);
		return;
	}

	if (n_frags == 1 && !(flags & EFX_RX_PKT_PREFIX_LEN))
		rx_buf->len = len;

	/* Release and/or sync the DMA mapping - assumes all RX buffers
	 * consumed in-order per RX queue.
	 */
	efx_sync_rx_buffer(efx, rx_buf, rx_buf->len);

	/* Prefetch nice and early so data will (hopefully) be in cache by
	 * the time we look at it.
	 */
	prefetch(efx_rx_buf_va(rx_buf));

	rx_buf->page_offset += efx->rx_prefix_size;
	rx_buf->len -= efx->rx_prefix_size;

	if (n_frags > 1) {
		/* Release/sync DMA mapping for additional fragments.
		 * Fix length for last fragment.
		 */
		unsigned int tail_frags = n_frags - 1;

		for (;;) {
			rx_buf = efx_rx_buf_next(rx_queue, rx_buf);
			if (--tail_frags == 0)
				break;
			efx_sync_rx_buffer(efx, rx_buf, EFX_RX_USR_BUF_SIZE);
		}
		rx_buf->len = len - (n_frags - 1) * EFX_RX_USR_BUF_SIZE;
		efx_sync_rx_buffer(efx, rx_buf, rx_buf->len);
	}

	/* All fragments have been DMA-synced, so recycle pages. */
	rx_buf = efx_rx_buffer(rx_queue, index);
	efx_recycle_rx_pages(channel, rx_buf, n_frags);

	/* Pipeline receives so that we give time for packet headers to be
	 * prefetched into cache.
	 */
	efx_rx_flush_packet(channel);
	channel->rx_pkt_n_frags = n_frags;
	channel->rx_pkt_index = index;
}

static void efx_rx_deliver(struct efx_channel *channel, u8 *eh,
			   struct efx_rx_buffer *rx_buf,
			   unsigned int n_frags)
{
	struct sk_buff *skb;
	u16 hdr_len = min_t(u16, rx_buf->len, EFX_SKB_HEADERS);

	skb = efx_rx_mk_skb(channel, rx_buf, n_frags, eh, hdr_len);
	if (unlikely(skb == NULL)) {
		efx_free_rx_buffer(rx_buf);
		return;
	}
	skb_record_rx_queue(skb, channel->rx_queue.core_index);

	/* Set the SKB flags */
	skb_checksum_none_assert(skb);
	if (likely(rx_buf->flags & EFX_RX_PKT_CSUMMED))
		skb->ip_summed = CHECKSUM_UNNECESSARY;

	if (channel->type->receive_skb)
		if (channel->type->receive_skb(channel, skb))
			return;

	/* Pass the packet up */
	netif_receive_skb(skb);
}

/* Handle a received packet.  Second half: Touches packet payload. */
void __efx_rx_packet(struct efx_channel *channel)
{
	struct efx_nic *efx = channel->efx;
	struct efx_rx_buffer *rx_buf =
		efx_rx_buffer(&channel->rx_queue, channel->rx_pkt_index);
	u8 *eh = efx_rx_buf_va(rx_buf);

	/* Read length from the prefix if necessary.  This already
	 * excludes the length of the prefix itself.
	 */
	if (rx_buf->flags & EFX_RX_PKT_PREFIX_LEN)
		rx_buf->len = le16_to_cpup((__le16 *)
					   (eh + efx->rx_packet_len_offset));

	/* If we're in loopback test, then pass the packet directly to the
	 * loopback layer, and free the rx_buf here
	 */
	if (unlikely(efx->loopback_selftest)) {
		efx_loopback_rx_packet(efx, eh, rx_buf->len);
		efx_free_rx_buffer(rx_buf);
		goto out;
	}

	if (unlikely(!(efx->net_dev->features & NETIF_F_RXCSUM)))
		rx_buf->flags &= ~EFX_RX_PKT_CSUMMED;

	if ((rx_buf->flags & EFX_RX_PKT_TCP) && !channel->type->receive_skb)
		efx_rx_packet_gro(channel, rx_buf, channel->rx_pkt_n_frags, eh);
	else
		efx_rx_deliver(channel, eh, rx_buf, channel->rx_pkt_n_frags);
out:
	channel->rx_pkt_n_frags = 0;
}

int efx_probe_rx_queue(struct efx_rx_queue *rx_queue)
{
	struct efx_nic *efx = rx_queue->efx;
	unsigned int entries;
	int rc;

	/* Create the smallest power-of-two aligned ring */
	entries = max(roundup_pow_of_two(efx->rxq_entries), EFX_MIN_DMAQ_SIZE);
	EFX_BUG_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE);
	rx_queue->ptr_mask = entries - 1;

	netif_dbg(efx, probe, efx->net_dev,
		  "creating RX queue %d size %#x mask %#x\n",
		  efx_rx_queue_index(rx_queue), efx->rxq_entries,
		  rx_queue->ptr_mask);

	/* Allocate RX buffers */
	rx_queue->buffer = kcalloc(entries, sizeof(*rx_queue->buffer),
				   GFP_KERNEL);
	if (!rx_queue->buffer)
		return -ENOMEM;

	rc = efx_nic_probe_rx(rx_queue);
	if (rc) {
		kfree(rx_queue->buffer);
		rx_queue->buffer = NULL;
	}

	return rc;
}

static void efx_init_rx_recycle_ring(struct efx_nic *efx,
				     struct efx_rx_queue *rx_queue)
{
	unsigned int bufs_in_recycle_ring, page_ring_size;

	/* Set the RX recycle ring size */
#ifdef CONFIG_PPC64
	bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU;
#else
	if (iommu_present(&pci_bus_type))
		bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU;
	else
		bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_NOIOMMU;
#endif /* CONFIG_PPC64 */

	page_ring_size = roundup_pow_of_two(bufs_in_recycle_ring /
					    efx->rx_bufs_per_page);
	rx_queue->page_ring = kcalloc(page_ring_size,
				      sizeof(*rx_queue->page_ring), GFP_KERNEL);
	rx_queue->page_ptr_mask = page_ring_size - 1;
}

void efx_init_rx_queue(struct efx_rx_queue *rx_queue)
{
	struct efx_nic *efx = rx_queue->efx;
	unsigned int max_fill, trigger, max_trigger;

	netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev,
		  "initialising RX queue %d\n", efx_rx_queue_index(rx_queue));

	/* Initialise ptr fields */
	rx_queue->added_count = 0;
	rx_queue->notified_count = 0;
	rx_queue->removed_count = 0;
	rx_queue->min_fill = -1U;
	efx_init_rx_recycle_ring(efx, rx_queue);

	rx_queue->page_remove = 0;
	rx_queue->page_add = rx_queue->page_ptr_mask + 1;
	rx_queue->page_recycle_count = 0;
	rx_queue->page_recycle_failed = 0;
	rx_queue->page_recycle_full = 0;

	/* Initialise limit fields */
	max_fill = efx->rxq_entries - EFX_RXD_HEAD_ROOM;
	max_trigger =
		max_fill - efx->rx_pages_per_batch * efx->rx_bufs_per_page;
	if (rx_refill_threshold != 0) {
		trigger = max_fill * min(rx_refill_threshold, 100U) / 100U;
		if (trigger > max_trigger)
			trigger = max_trigger;
	} else {
		trigger = max_trigger;
	}

	rx_queue->max_fill = max_fill;
	rx_queue->fast_fill_trigger = trigger;
	rx_queue->refill_enabled = true;

	/* Set up RX descriptor ring */
	efx_nic_init_rx(rx_queue);
}

void efx_fini_rx_queue(struct efx_rx_queue *rx_queue)
{
	int i;
	struct efx_nic *efx = rx_queue->efx;
	struct efx_rx_buffer *rx_buf;

	netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev,
		  "shutting down RX queue %d\n", efx_rx_queue_index(rx_queue));

	del_timer_sync(&rx_queue->slow_fill);

	/* Release RX buffers from the current read ptr to the write ptr */
	if (rx_queue->buffer) {
		for (i = rx_queue->removed_count; i < rx_queue->added_count;
		     i++) {
			unsigned index = i & rx_queue->ptr_mask;
			rx_buf = efx_rx_buffer(rx_queue, index);
			efx_fini_rx_buffer(rx_queue, rx_buf);
		}
	}

	/* Unmap and release the pages in the recycle ring. Remove the ring. */
	for (i = 0; i <= rx_queue->page_ptr_mask; i++) {
		struct page *page = rx_queue->page_ring[i];
		struct efx_rx_page_state *state;

		if (page == NULL)
			continue;

		state = page_address(page);
		dma_unmap_page(&efx->pci_dev->dev, state->dma_addr,
			       PAGE_SIZE << efx->rx_buffer_order,
			       DMA_FROM_DEVICE);
		put_page(page);
	}
	kfree(rx_queue->page_ring);
	rx_queue->page_ring = NULL;
}

void efx_remove_rx_queue(struct efx_rx_queue *rx_queue)
{
	netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev,
		  "destroying RX queue %d\n", efx_rx_queue_index(rx_queue));

	efx_nic_remove_rx(rx_queue);

	kfree(rx_queue->buffer);
	rx_queue->buffer = NULL;
}


module_param(rx_refill_threshold, uint, 0444);
MODULE_PARM_DESC(rx_refill_threshold,
		 "RX descriptor ring refill threshold (%)");

#ifdef CONFIG_RFS_ACCEL

int efx_filter_rfs(struct net_device *net_dev, const struct sk_buff *skb,
		   u16 rxq_index, u32 flow_id)
{
	struct efx_nic *efx = netdev_priv(net_dev);
	struct efx_channel *channel;
	struct efx_filter_spec spec;
	const struct iphdr *ip;
	const __be16 *ports;
	int nhoff;
	int rc;

	nhoff = skb_network_offset(skb);

	if (skb->protocol == htons(ETH_P_8021Q)) {
		EFX_BUG_ON_PARANOID(skb_headlen(skb) <
				    nhoff + sizeof(struct vlan_hdr));
		if (((const struct vlan_hdr *)skb->data + nhoff)->
		    h_vlan_encapsulated_proto != htons(ETH_P_IP))
			return -EPROTONOSUPPORT;

		/* This is IP over 802.1q VLAN.  We can't filter on the
		 * IP 5-tuple and the vlan together, so just strip the
		 * vlan header and filter on the IP part.
		 */
		nhoff += sizeof(struct vlan_hdr);
	} else if (skb->protocol != htons(ETH_P_IP)) {
		return -EPROTONOSUPPORT;
	}

	/* RFS must validate the IP header length before calling us */
	EFX_BUG_ON_PARANOID(skb_headlen(skb) < nhoff + sizeof(*ip));
	ip = (const struct iphdr *)(skb->data + nhoff);
	if (ip_is_fragment(ip))
		return -EPROTONOSUPPORT;
	EFX_BUG_ON_PARANOID(skb_headlen(skb) < nhoff + 4 * ip->ihl + 4);
	ports = (const __be16 *)(skb->data + nhoff + 4 * ip->ihl);

	efx_filter_init_rx(&spec, EFX_FILTER_PRI_HINT,
			   efx->rx_scatter ? EFX_FILTER_FLAG_RX_SCATTER : 0,
			   rxq_index);
	rc = efx_filter_set_ipv4_full(&spec, ip->protocol,
				      ip->daddr, ports[1], ip->saddr, ports[0]);
	if (rc)
		return rc;

	rc = efx->type->filter_rfs_insert(efx, &spec);
	if (rc < 0)
		return rc;

	/* Remember this so we can check whether to expire the filter later */
	efx->rps_flow_id[rc] = flow_id;
	channel = efx_get_channel(efx, skb_get_rx_queue(skb));
	++channel->rfs_filters_added;

	netif_info(efx, rx_status, efx->net_dev,
		   "steering %s %pI4:%u:%pI4:%u to queue %u [flow %u filter %d]\n",
		   (ip->protocol == IPPROTO_TCP) ? "TCP" : "UDP",
		   &ip->saddr, ntohs(ports[0]), &ip->daddr, ntohs(ports[1]),
		   rxq_index, flow_id, rc);

	return rc;
}

bool __efx_filter_rfs_expire(struct efx_nic *efx, unsigned int quota)
{
	bool (*expire_one)(struct efx_nic *efx, u32 flow_id, unsigned int index);
	unsigned int index, size;
	u32 flow_id;

	if (!spin_trylock_bh(&efx->filter_lock))
		return false;

	expire_one = efx->type->filter_rfs_expire_one;
	index = efx->rps_expire_index;
	size = efx->type->max_rx_ip_filters;
	while (quota--) {
		flow_id = efx->rps_flow_id[index];
		if (expire_one(efx, flow_id, index))
			netif_info(efx, rx_status, efx->net_dev,
				   "expired filter %d [flow %u]\n",
				   index, flow_id);
		if (++index == size)
			index = 0;
	}
	efx->rps_expire_index = index;

	spin_unlock_bh(&efx->filter_lock);
	return true;
}

#endif /* CONFIG_RFS_ACCEL */