summaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c
blob: c9b7ad65e5633bc2a5147706f5c728f4d1cffa3c (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
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
1777
1778
1779
1780
1781
1782
1783
1784
1785
1786
1787
1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
1846
1847
1848
1849
1850
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
1891
1892
1893
1894
1895
1896
1897
1898
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
1910
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
2135
2136
2137
2138
2139
2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
2236
2237
2238
2239
2240
2241
2242
2243
2244
2245
2246
2247
2248
2249
2250
2251
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
2293
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
2304
2305
2306
2307
2308
2309
2310
2311
2312
2313
2314
2315
2316
2317
2318
2319
2320
2321
2322
2323
2324
2325
2326
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
2337
2338
2339
2340
2341
2342
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
2361
2362
2363
2364
2365
2366
2367
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
2399
2400
2401
2402
2403
2404
2405
2406
2407
2408
2409
2410
2411
2412
2413
2414
2415
2416
2417
2418
2419
2420
2421
2422
2423
2424
2425
2426
2427
2428
2429
2430
2431
2432
2433
2434
2435
2436
2437
2438
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
2451
2452
2453
2454
2455
2456
2457
2458
2459
2460
2461
2462
2463
2464
2465
2466
2467
2468
2469
2470
2471
2472
2473
2474
2475
2476
2477
2478
2479
2480
2481
2482
2483
2484
2485
2486
2487
2488
2489
2490
2491
2492
2493
2494
2495
2496
2497
2498
2499
2500
2501
2502
2503
2504
2505
2506
2507
2508
2509
2510
2511
2512
2513
2514
2515
2516
2517
2518
2519
2520
2521
2522
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
2534
2535
2536
2537
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
2566
2567
2568
2569
2570
2571
2572
2573
2574
2575
2576
2577
2578
2579
2580
2581
2582
2583
2584
2585
2586
2587
2588
2589
2590
2591
2592
2593
2594
2595
2596
2597
2598
2599
2600
2601
2602
2603
2604
2605
2606
2607
2608
2609
2610
2611
2612
2613
2614
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
2634
2635
2636
2637
2638
2639
2640
2641
2642
2643
2644
2645
2646
2647
2648
2649
2650
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
2662
2663
2664
2665
2666
2667
2668
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
2706
2707
2708
2709
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
/* Copyright 2008 - 2016 Freescale Semiconductor Inc.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are met:
 *     * Redistributions of source code must retain the above copyright
 *	 notice, this list of conditions and the following disclaimer.
 *     * Redistributions in binary form must reproduce the above copyright
 *	 notice, this list of conditions and the following disclaimer in the
 *	 documentation and/or other materials provided with the distribution.
 *     * Neither the name of Freescale Semiconductor nor the
 *	 names of its contributors may be used to endorse or promote products
 *	 derived from this software without specific prior written permission.
 *
 * ALTERNATIVELY, this software may be distributed under the terms of the
 * GNU General Public License ("GPL") as published by the Free Software
 * Foundation, either version 2 of that License or (at your option) any
 * later version.
 *
 * THIS SOFTWARE IS PROVIDED BY Freescale Semiconductor ``AS IS'' AND ANY
 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL Freescale Semiconductor BE LIABLE FOR ANY
 * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/init.h>
#include <linux/module.h>
#include <linux/of_platform.h>
#include <linux/of_mdio.h>
#include <linux/of_net.h>
#include <linux/io.h>
#include <linux/if_arp.h>
#include <linux/if_vlan.h>
#include <linux/icmp.h>
#include <linux/ip.h>
#include <linux/ipv6.h>
#include <linux/udp.h>
#include <linux/tcp.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <linux/etherdevice.h>
#include <linux/if_ether.h>
#include <linux/highmem.h>
#include <linux/percpu.h>
#include <linux/dma-mapping.h>
#include <linux/sort.h>
#include <soc/fsl/bman.h>
#include <soc/fsl/qman.h>

#include "fman.h"
#include "fman_port.h"
#include "mac.h"
#include "dpaa_eth.h"

/* CREATE_TRACE_POINTS only needs to be defined once. Other dpaa files
 * using trace events only need to #include <trace/events/sched.h>
 */
#define CREATE_TRACE_POINTS
#include "dpaa_eth_trace.h"

static int debug = -1;
module_param(debug, int, 0444);
MODULE_PARM_DESC(debug, "Module/Driver verbosity level (0=none,...,16=all)");

static u16 tx_timeout = 1000;
module_param(tx_timeout, ushort, 0444);
MODULE_PARM_DESC(tx_timeout, "The Tx timeout in ms");

#define FM_FD_STAT_RX_ERRORS						\
	(FM_FD_ERR_DMA | FM_FD_ERR_PHYSICAL	| \
	 FM_FD_ERR_SIZE | FM_FD_ERR_CLS_DISCARD | \
	 FM_FD_ERR_EXTRACTION | FM_FD_ERR_NO_SCHEME	| \
	 FM_FD_ERR_PRS_TIMEOUT | FM_FD_ERR_PRS_ILL_INSTRUCT | \
	 FM_FD_ERR_PRS_HDR_ERR)

#define FM_FD_STAT_TX_ERRORS \
	(FM_FD_ERR_UNSUPPORTED_FORMAT | \
	 FM_FD_ERR_LENGTH | FM_FD_ERR_DMA)

#define DPAA_MSG_DEFAULT (NETIF_MSG_DRV | NETIF_MSG_PROBE | \
			  NETIF_MSG_LINK | NETIF_MSG_IFUP | \
			  NETIF_MSG_IFDOWN)

#define DPAA_INGRESS_CS_THRESHOLD 0x10000000
/* Ingress congestion threshold on FMan ports
 * The size in bytes of the ingress tail-drop threshold on FMan ports.
 * Traffic piling up above this value will be rejected by QMan and discarded
 * by FMan.
 */

/* Size in bytes of the FQ taildrop threshold */
#define DPAA_FQ_TD 0x200000

#define DPAA_CS_THRESHOLD_1G 0x06000000
/* Egress congestion threshold on 1G ports, range 0x1000 .. 0x10000000
 * The size in bytes of the egress Congestion State notification threshold on
 * 1G ports. The 1G dTSECs can quite easily be flooded by cores doing Tx in a
 * tight loop (e.g. by sending UDP datagrams at "while(1) speed"),
 * and the larger the frame size, the more acute the problem.
 * So we have to find a balance between these factors:
 * - avoiding the device staying congested for a prolonged time (risking
 *   the netdev watchdog to fire - see also the tx_timeout module param);
 * - affecting performance of protocols such as TCP, which otherwise
 *   behave well under the congestion notification mechanism;
 * - preventing the Tx cores from tightly-looping (as if the congestion
 *   threshold was too low to be effective);
 * - running out of memory if the CS threshold is set too high.
 */

#define DPAA_CS_THRESHOLD_10G 0x10000000
/* The size in bytes of the egress Congestion State notification threshold on
 * 10G ports, range 0x1000 .. 0x10000000
 */

/* Largest value that the FQD's OAL field can hold */
#define FSL_QMAN_MAX_OAL	127

/* Default alignment for start of data in an Rx FD */
#define DPAA_FD_DATA_ALIGNMENT  16

/* Values for the L3R field of the FM Parse Results
 */
/* L3 Type field: First IP Present IPv4 */
#define FM_L3_PARSE_RESULT_IPV4	0x8000
/* L3 Type field: First IP Present IPv6 */
#define FM_L3_PARSE_RESULT_IPV6	0x4000
/* Values for the L4R field of the FM Parse Results */
/* L4 Type field: UDP */
#define FM_L4_PARSE_RESULT_UDP	0x40
/* L4 Type field: TCP */
#define FM_L4_PARSE_RESULT_TCP	0x20

#define DPAA_SGT_MAX_ENTRIES 16 /* maximum number of entries in SG Table */
#define DPAA_BUFF_RELEASE_MAX 8 /* maximum number of buffers released at once */

#define FSL_DPAA_BPID_INV		0xff
#define FSL_DPAA_ETH_MAX_BUF_COUNT	128
#define FSL_DPAA_ETH_REFILL_THRESHOLD	80

#define DPAA_TX_PRIV_DATA_SIZE	16
#define DPAA_PARSE_RESULTS_SIZE sizeof(struct fman_prs_result)
#define DPAA_TIME_STAMP_SIZE 8
#define DPAA_HASH_RESULTS_SIZE 8
#define DPAA_RX_PRIV_DATA_SIZE	(u16)(DPAA_TX_PRIV_DATA_SIZE + \
					dpaa_rx_extra_headroom)

#define DPAA_ETH_RX_QUEUES	128

#define DPAA_ENQUEUE_RETRIES	100000

enum port_type {RX, TX};

struct fm_port_fqs {
	struct dpaa_fq *tx_defq;
	struct dpaa_fq *tx_errq;
	struct dpaa_fq *rx_defq;
	struct dpaa_fq *rx_errq;
};

/* All the dpa bps in use at any moment */
static struct dpaa_bp *dpaa_bp_array[BM_MAX_NUM_OF_POOLS];

/* The raw buffer size must be cacheline aligned */
#define DPAA_BP_RAW_SIZE 4096
/* When using more than one buffer pool, the raw sizes are as follows:
 * 1 bp: 4KB
 * 2 bp: 2KB, 4KB
 * 3 bp: 1KB, 2KB, 4KB
 * 4 bp: 1KB, 2KB, 4KB, 8KB
 */
static inline size_t bpool_buffer_raw_size(u8 index, u8 cnt)
{
	size_t res = DPAA_BP_RAW_SIZE / 4;
	u8 i;

	for (i = (cnt < 3) ? cnt : 3; i < 3 + index; i++)
		res *= 2;
	return res;
}

/* FMan-DMA requires 16-byte alignment for Rx buffers, but SKB_DATA_ALIGN is
 * even stronger (SMP_CACHE_BYTES-aligned), so we just get away with that,
 * via SKB_WITH_OVERHEAD(). We can't rely on netdev_alloc_frag() giving us
 * half-page-aligned buffers, so we reserve some more space for start-of-buffer
 * alignment.
 */
#define dpaa_bp_size(raw_size) SKB_WITH_OVERHEAD((raw_size) - SMP_CACHE_BYTES)

static int dpaa_max_frm;

static int dpaa_rx_extra_headroom;

#define dpaa_get_max_mtu()	\
	(dpaa_max_frm - (VLAN_ETH_HLEN + ETH_FCS_LEN))

static int dpaa_netdev_init(struct net_device *net_dev,
			    const struct net_device_ops *dpaa_ops,
			    u16 tx_timeout)
{
	struct dpaa_priv *priv = netdev_priv(net_dev);
	struct device *dev = net_dev->dev.parent;
	struct dpaa_percpu_priv *percpu_priv;
	const u8 *mac_addr;
	int i, err;

	/* Although we access another CPU's private data here
	 * we do it at initialization so it is safe
	 */
	for_each_possible_cpu(i) {
		percpu_priv = per_cpu_ptr(priv->percpu_priv, i);
		percpu_priv->net_dev = net_dev;
	}

	net_dev->netdev_ops = dpaa_ops;
	mac_addr = priv->mac_dev->addr;

	net_dev->mem_start = priv->mac_dev->res->start;
	net_dev->mem_end = priv->mac_dev->res->end;

	net_dev->min_mtu = ETH_MIN_MTU;
	net_dev->max_mtu = dpaa_get_max_mtu();

	net_dev->hw_features |= (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
				 NETIF_F_LLTX);

	net_dev->hw_features |= NETIF_F_SG | NETIF_F_HIGHDMA;
	/* The kernels enables GSO automatically, if we declare NETIF_F_SG.
	 * For conformity, we'll still declare GSO explicitly.
	 */
	net_dev->features |= NETIF_F_GSO;

	net_dev->priv_flags |= IFF_LIVE_ADDR_CHANGE;
	/* we do not want shared skbs on TX */
	net_dev->priv_flags &= ~IFF_TX_SKB_SHARING;

	net_dev->features |= net_dev->hw_features;
	net_dev->vlan_features = net_dev->features;

	memcpy(net_dev->perm_addr, mac_addr, net_dev->addr_len);
	memcpy(net_dev->dev_addr, mac_addr, net_dev->addr_len);

	net_dev->ethtool_ops = &dpaa_ethtool_ops;

	net_dev->needed_headroom = priv->tx_headroom;
	net_dev->watchdog_timeo = msecs_to_jiffies(tx_timeout);

	/* start without the RUNNING flag, phylib controls it later */
	netif_carrier_off(net_dev);

	err = register_netdev(net_dev);
	if (err < 0) {
		dev_err(dev, "register_netdev() = %d\n", err);
		return err;
	}

	return 0;
}

static int dpaa_stop(struct net_device *net_dev)
{
	struct mac_device *mac_dev;
	struct dpaa_priv *priv;
	int i, err, error;

	priv = netdev_priv(net_dev);
	mac_dev = priv->mac_dev;

	netif_tx_stop_all_queues(net_dev);
	/* Allow the Fman (Tx) port to process in-flight frames before we
	 * try switching it off.
	 */
	usleep_range(5000, 10000);

	err = mac_dev->stop(mac_dev);
	if (err < 0)
		netif_err(priv, ifdown, net_dev, "mac_dev->stop() = %d\n",
			  err);

	for (i = 0; i < ARRAY_SIZE(mac_dev->port); i++) {
		error = fman_port_disable(mac_dev->port[i]);
		if (error)
			err = error;
	}

	if (net_dev->phydev)
		phy_disconnect(net_dev->phydev);
	net_dev->phydev = NULL;

	return err;
}

static void dpaa_tx_timeout(struct net_device *net_dev)
{
	struct dpaa_percpu_priv *percpu_priv;
	const struct dpaa_priv	*priv;

	priv = netdev_priv(net_dev);
	percpu_priv = this_cpu_ptr(priv->percpu_priv);

	netif_crit(priv, timer, net_dev, "Transmit timeout latency: %u ms\n",
		   jiffies_to_msecs(jiffies - dev_trans_start(net_dev)));

	percpu_priv->stats.tx_errors++;
}

/* Calculates the statistics for the given device by adding the statistics
 * collected by each CPU.
 */
static struct rtnl_link_stats64 *dpaa_get_stats64(struct net_device *net_dev,
						  struct rtnl_link_stats64 *s)
{
	int numstats = sizeof(struct rtnl_link_stats64) / sizeof(u64);
	struct dpaa_priv *priv = netdev_priv(net_dev);
	struct dpaa_percpu_priv *percpu_priv;
	u64 *netstats = (u64 *)s;
	u64 *cpustats;
	int i, j;

	for_each_possible_cpu(i) {
		percpu_priv = per_cpu_ptr(priv->percpu_priv, i);

		cpustats = (u64 *)&percpu_priv->stats;

		/* add stats from all CPUs */
		for (j = 0; j < numstats; j++)
			netstats[j] += cpustats[j];
	}

	return s;
}

static struct mac_device *dpaa_mac_dev_get(struct platform_device *pdev)
{
	struct platform_device *of_dev;
	struct dpaa_eth_data *eth_data;
	struct device *dpaa_dev, *dev;
	struct device_node *mac_node;
	struct mac_device *mac_dev;

	dpaa_dev = &pdev->dev;
	eth_data = dpaa_dev->platform_data;
	if (!eth_data)
		return ERR_PTR(-ENODEV);

	mac_node = eth_data->mac_node;

	of_dev = of_find_device_by_node(mac_node);
	if (!of_dev) {
		dev_err(dpaa_dev, "of_find_device_by_node(%s) failed\n",
			mac_node->full_name);
		of_node_put(mac_node);
		return ERR_PTR(-EINVAL);
	}
	of_node_put(mac_node);

	dev = &of_dev->dev;

	mac_dev = dev_get_drvdata(dev);
	if (!mac_dev) {
		dev_err(dpaa_dev, "dev_get_drvdata(%s) failed\n",
			dev_name(dev));
		return ERR_PTR(-EINVAL);
	}

	return mac_dev;
}

static int dpaa_set_mac_address(struct net_device *net_dev, void *addr)
{
	const struct dpaa_priv *priv;
	struct mac_device *mac_dev;
	struct sockaddr old_addr;
	int err;

	priv = netdev_priv(net_dev);

	memcpy(old_addr.sa_data, net_dev->dev_addr,  ETH_ALEN);

	err = eth_mac_addr(net_dev, addr);
	if (err < 0) {
		netif_err(priv, drv, net_dev, "eth_mac_addr() = %d\n", err);
		return err;
	}

	mac_dev = priv->mac_dev;

	err = mac_dev->change_addr(mac_dev->fman_mac,
				   (enet_addr_t *)net_dev->dev_addr);
	if (err < 0) {
		netif_err(priv, drv, net_dev, "mac_dev->change_addr() = %d\n",
			  err);
		/* reverting to previous address */
		eth_mac_addr(net_dev, &old_addr);

		return err;
	}

	return 0;
}

static void dpaa_set_rx_mode(struct net_device *net_dev)
{
	const struct dpaa_priv	*priv;
	int err;

	priv = netdev_priv(net_dev);

	if (!!(net_dev->flags & IFF_PROMISC) != priv->mac_dev->promisc) {
		priv->mac_dev->promisc = !priv->mac_dev->promisc;
		err = priv->mac_dev->set_promisc(priv->mac_dev->fman_mac,
						 priv->mac_dev->promisc);
		if (err < 0)
			netif_err(priv, drv, net_dev,
				  "mac_dev->set_promisc() = %d\n",
				  err);
	}

	err = priv->mac_dev->set_multi(net_dev, priv->mac_dev);
	if (err < 0)
		netif_err(priv, drv, net_dev, "mac_dev->set_multi() = %d\n",
			  err);
}

static struct dpaa_bp *dpaa_bpid2pool(int bpid)
{
	if (WARN_ON(bpid < 0 || bpid >= BM_MAX_NUM_OF_POOLS))
		return NULL;

	return dpaa_bp_array[bpid];
}

/* checks if this bpool is already allocated */
static bool dpaa_bpid2pool_use(int bpid)
{
	if (dpaa_bpid2pool(bpid)) {
		atomic_inc(&dpaa_bp_array[bpid]->refs);
		return true;
	}

	return false;
}

/* called only once per bpid by dpaa_bp_alloc_pool() */
static void dpaa_bpid2pool_map(int bpid, struct dpaa_bp *dpaa_bp)
{
	dpaa_bp_array[bpid] = dpaa_bp;
	atomic_set(&dpaa_bp->refs, 1);
}

static int dpaa_bp_alloc_pool(struct dpaa_bp *dpaa_bp)
{
	int err;

	if (dpaa_bp->size == 0 || dpaa_bp->config_count == 0) {
		pr_err("%s: Buffer pool is not properly initialized! Missing size or initial number of buffers\n",
		       __func__);
		return -EINVAL;
	}

	/* If the pool is already specified, we only create one per bpid */
	if (dpaa_bp->bpid != FSL_DPAA_BPID_INV &&
	    dpaa_bpid2pool_use(dpaa_bp->bpid))
		return 0;

	if (dpaa_bp->bpid == FSL_DPAA_BPID_INV) {
		dpaa_bp->pool = bman_new_pool();
		if (!dpaa_bp->pool) {
			pr_err("%s: bman_new_pool() failed\n",
			       __func__);
			return -ENODEV;
		}

		dpaa_bp->bpid = (u8)bman_get_bpid(dpaa_bp->pool);
	}

	if (dpaa_bp->seed_cb) {
		err = dpaa_bp->seed_cb(dpaa_bp);
		if (err)
			goto pool_seed_failed;
	}

	dpaa_bpid2pool_map(dpaa_bp->bpid, dpaa_bp);

	return 0;

pool_seed_failed:
	pr_err("%s: pool seeding failed\n", __func__);
	bman_free_pool(dpaa_bp->pool);

	return err;
}

/* remove and free all the buffers from the given buffer pool */
static void dpaa_bp_drain(struct dpaa_bp *bp)
{
	u8 num = 8;
	int ret;

	do {
		struct bm_buffer bmb[8];
		int i;

		ret = bman_acquire(bp->pool, bmb, num);
		if (ret < 0) {
			if (num == 8) {
				/* we have less than 8 buffers left;
				 * drain them one by one
				 */
				num = 1;
				ret = 1;
				continue;
			} else {
				/* Pool is fully drained */
				break;
			}
		}

		if (bp->free_buf_cb)
			for (i = 0; i < num; i++)
				bp->free_buf_cb(bp, &bmb[i]);
	} while (ret > 0);
}

static void dpaa_bp_free(struct dpaa_bp *dpaa_bp)
{
	struct dpaa_bp *bp = dpaa_bpid2pool(dpaa_bp->bpid);

	/* the mapping between bpid and dpaa_bp is done very late in the
	 * allocation procedure; if something failed before the mapping, the bp
	 * was not configured, therefore we don't need the below instructions
	 */
	if (!bp)
		return;

	if (!atomic_dec_and_test(&bp->refs))
		return;

	if (bp->free_buf_cb)
		dpaa_bp_drain(bp);

	dpaa_bp_array[bp->bpid] = NULL;
	bman_free_pool(bp->pool);
}

static void dpaa_bps_free(struct dpaa_priv *priv)
{
	int i;

	for (i = 0; i < DPAA_BPS_NUM; i++)
		dpaa_bp_free(priv->dpaa_bps[i]);
}

/* Use multiple WQs for FQ assignment:
 *	- Tx Confirmation queues go to WQ1.
 *	- Rx Error and Tx Error queues go to WQ2 (giving them a better chance
 *	  to be scheduled, in case there are many more FQs in WQ3).
 *	- Rx Default and Tx queues go to WQ3 (no differentiation between
 *	  Rx and Tx traffic).
 * This ensures that Tx-confirmed buffers are timely released. In particular,
 * it avoids congestion on the Tx Confirm FQs, which can pile up PFDRs if they
 * are greatly outnumbered by other FQs in the system, while
 * dequeue scheduling is round-robin.
 */
static inline void dpaa_assign_wq(struct dpaa_fq *fq)
{
	switch (fq->fq_type) {
	case FQ_TYPE_TX_CONFIRM:
	case FQ_TYPE_TX_CONF_MQ:
		fq->wq = 1;
		break;
	case FQ_TYPE_RX_ERROR:
	case FQ_TYPE_TX_ERROR:
		fq->wq = 2;
		break;
	case FQ_TYPE_RX_DEFAULT:
	case FQ_TYPE_TX:
		fq->wq = 3;
		break;
	default:
		WARN(1, "Invalid FQ type %d for FQID %d!\n",
		     fq->fq_type, fq->fqid);
	}
}

static struct dpaa_fq *dpaa_fq_alloc(struct device *dev,
				     u32 start, u32 count,
				     struct list_head *list,
				     enum dpaa_fq_type fq_type)
{
	struct dpaa_fq *dpaa_fq;
	int i;

	dpaa_fq = devm_kzalloc(dev, sizeof(*dpaa_fq) * count,
			       GFP_KERNEL);
	if (!dpaa_fq)
		return NULL;

	for (i = 0; i < count; i++) {
		dpaa_fq[i].fq_type = fq_type;
		dpaa_fq[i].fqid = start ? start + i : 0;
		list_add_tail(&dpaa_fq[i].list, list);
	}

	for (i = 0; i < count; i++)
		dpaa_assign_wq(dpaa_fq + i);

	return dpaa_fq;
}

static int dpaa_alloc_all_fqs(struct device *dev, struct list_head *list,
			      struct fm_port_fqs *port_fqs)
{
	struct dpaa_fq *dpaa_fq;

	dpaa_fq = dpaa_fq_alloc(dev, 0, 1, list, FQ_TYPE_RX_ERROR);
	if (!dpaa_fq)
		goto fq_alloc_failed;

	port_fqs->rx_errq = &dpaa_fq[0];

	dpaa_fq = dpaa_fq_alloc(dev, 0, 1, list, FQ_TYPE_RX_DEFAULT);
	if (!dpaa_fq)
		goto fq_alloc_failed;

	port_fqs->rx_defq = &dpaa_fq[0];

	if (!dpaa_fq_alloc(dev, 0, DPAA_ETH_TXQ_NUM, list, FQ_TYPE_TX_CONF_MQ))
		goto fq_alloc_failed;

	dpaa_fq = dpaa_fq_alloc(dev, 0, 1, list, FQ_TYPE_TX_ERROR);
	if (!dpaa_fq)
		goto fq_alloc_failed;

	port_fqs->tx_errq = &dpaa_fq[0];

	dpaa_fq = dpaa_fq_alloc(dev, 0, 1, list, FQ_TYPE_TX_CONFIRM);
	if (!dpaa_fq)
		goto fq_alloc_failed;

	port_fqs->tx_defq = &dpaa_fq[0];

	if (!dpaa_fq_alloc(dev, 0, DPAA_ETH_TXQ_NUM, list, FQ_TYPE_TX))
		goto fq_alloc_failed;

	return 0;

fq_alloc_failed:
	dev_err(dev, "dpaa_fq_alloc() failed\n");
	return -ENOMEM;
}

static u32 rx_pool_channel;
static DEFINE_SPINLOCK(rx_pool_channel_init);

static int dpaa_get_channel(void)
{
	spin_lock(&rx_pool_channel_init);
	if (!rx_pool_channel) {
		u32 pool;
		int ret;

		ret = qman_alloc_pool(&pool);

		if (!ret)
			rx_pool_channel = pool;
	}
	spin_unlock(&rx_pool_channel_init);
	if (!rx_pool_channel)
		return -ENOMEM;
	return rx_pool_channel;
}

static void dpaa_release_channel(void)
{
	qman_release_pool(rx_pool_channel);
}

static void dpaa_eth_add_channel(u16 channel)
{
	u32 pool = QM_SDQCR_CHANNELS_POOL_CONV(channel);
	const cpumask_t *cpus = qman_affine_cpus();
	struct qman_portal *portal;
	int cpu;

	for_each_cpu(cpu, cpus) {
		portal = qman_get_affine_portal(cpu);
		qman_p_static_dequeue_add(portal, pool);
	}
}

/* Congestion group state change notification callback.
 * Stops the device's egress queues while they are congested and
 * wakes them upon exiting congested state.
 * Also updates some CGR-related stats.
 */
static void dpaa_eth_cgscn(struct qman_portal *qm, struct qman_cgr *cgr,
			   int congested)
{
	struct dpaa_priv *priv = (struct dpaa_priv *)container_of(cgr,
		struct dpaa_priv, cgr_data.cgr);

	if (congested) {
		priv->cgr_data.congestion_start_jiffies = jiffies;
		netif_tx_stop_all_queues(priv->net_dev);
		priv->cgr_data.cgr_congested_count++;
	} else {
		priv->cgr_data.congested_jiffies +=
			(jiffies - priv->cgr_data.congestion_start_jiffies);
		netif_tx_wake_all_queues(priv->net_dev);
	}
}

static int dpaa_eth_cgr_init(struct dpaa_priv *priv)
{
	struct qm_mcc_initcgr initcgr;
	u32 cs_th;
	int err;

	err = qman_alloc_cgrid(&priv->cgr_data.cgr.cgrid);
	if (err < 0) {
		if (netif_msg_drv(priv))
			pr_err("%s: Error %d allocating CGR ID\n",
			       __func__, err);
		goto out_error;
	}
	priv->cgr_data.cgr.cb = dpaa_eth_cgscn;

	/* Enable Congestion State Change Notifications and CS taildrop */
	memset(&initcgr, 0, sizeof(initcgr));
	initcgr.we_mask = cpu_to_be16(QM_CGR_WE_CSCN_EN | QM_CGR_WE_CS_THRES);
	initcgr.cgr.cscn_en = QM_CGR_EN;

	/* Set different thresholds based on the MAC speed.
	 * This may turn suboptimal if the MAC is reconfigured at a speed
	 * lower than its max, e.g. if a dTSEC later negotiates a 100Mbps link.
	 * In such cases, we ought to reconfigure the threshold, too.
	 */
	if (priv->mac_dev->if_support & SUPPORTED_10000baseT_Full)
		cs_th = DPAA_CS_THRESHOLD_10G;
	else
		cs_th = DPAA_CS_THRESHOLD_1G;
	qm_cgr_cs_thres_set64(&initcgr.cgr.cs_thres, cs_th, 1);

	initcgr.we_mask |= cpu_to_be16(QM_CGR_WE_CSTD_EN);
	initcgr.cgr.cstd_en = QM_CGR_EN;

	err = qman_create_cgr(&priv->cgr_data.cgr, QMAN_CGR_FLAG_USE_INIT,
			      &initcgr);
	if (err < 0) {
		if (netif_msg_drv(priv))
			pr_err("%s: Error %d creating CGR with ID %d\n",
			       __func__, err, priv->cgr_data.cgr.cgrid);
		qman_release_cgrid(priv->cgr_data.cgr.cgrid);
		goto out_error;
	}
	if (netif_msg_drv(priv))
		pr_debug("Created CGR %d for netdev with hwaddr %pM on QMan channel %d\n",
			 priv->cgr_data.cgr.cgrid, priv->mac_dev->addr,
			 priv->cgr_data.cgr.chan);

out_error:
	return err;
}

static inline void dpaa_setup_ingress(const struct dpaa_priv *priv,
				      struct dpaa_fq *fq,
				      const struct qman_fq *template)
{
	fq->fq_base = *template;
	fq->net_dev = priv->net_dev;

	fq->flags = QMAN_FQ_FLAG_NO_ENQUEUE;
	fq->channel = priv->channel;
}

static inline void dpaa_setup_egress(const struct dpaa_priv *priv,
				     struct dpaa_fq *fq,
				     struct fman_port *port,
				     const struct qman_fq *template)
{
	fq->fq_base = *template;
	fq->net_dev = priv->net_dev;

	if (port) {
		fq->flags = QMAN_FQ_FLAG_TO_DCPORTAL;
		fq->channel = (u16)fman_port_get_qman_channel_id(port);
	} else {
		fq->flags = QMAN_FQ_FLAG_NO_MODIFY;
	}
}

static void dpaa_fq_setup(struct dpaa_priv *priv,
			  const struct dpaa_fq_cbs *fq_cbs,
			  struct fman_port *tx_port)
{
	int egress_cnt = 0, conf_cnt = 0, num_portals = 0, cpu;
	const cpumask_t *affine_cpus = qman_affine_cpus();
	u16 portals[NR_CPUS];
	struct dpaa_fq *fq;

	for_each_cpu(cpu, affine_cpus)
		portals[num_portals++] = qman_affine_channel(cpu);
	if (num_portals == 0)
		dev_err(priv->net_dev->dev.parent,
			"No Qman software (affine) channels found");

	/* Initialize each FQ in the list */
	list_for_each_entry(fq, &priv->dpaa_fq_list, list) {
		switch (fq->fq_type) {
		case FQ_TYPE_RX_DEFAULT:
			dpaa_setup_ingress(priv, fq, &fq_cbs->rx_defq);
			break;
		case FQ_TYPE_RX_ERROR:
			dpaa_setup_ingress(priv, fq, &fq_cbs->rx_errq);
			break;
		case FQ_TYPE_TX:
			dpaa_setup_egress(priv, fq, tx_port,
					  &fq_cbs->egress_ern);
			/* If we have more Tx queues than the number of cores,
			 * just ignore the extra ones.
			 */
			if (egress_cnt < DPAA_ETH_TXQ_NUM)
				priv->egress_fqs[egress_cnt++] = &fq->fq_base;
			break;
		case FQ_TYPE_TX_CONF_MQ:
			priv->conf_fqs[conf_cnt++] = &fq->fq_base;
			/* fall through */
		case FQ_TYPE_TX_CONFIRM:
			dpaa_setup_ingress(priv, fq, &fq_cbs->tx_defq);
			break;
		case FQ_TYPE_TX_ERROR:
			dpaa_setup_ingress(priv, fq, &fq_cbs->tx_errq);
			break;
		default:
			dev_warn(priv->net_dev->dev.parent,
				 "Unknown FQ type detected!\n");
			break;
		}
	}

	 /* Make sure all CPUs receive a corresponding Tx queue. */
	while (egress_cnt < DPAA_ETH_TXQ_NUM) {
		list_for_each_entry(fq, &priv->dpaa_fq_list, list) {
			if (fq->fq_type != FQ_TYPE_TX)
				continue;
			priv->egress_fqs[egress_cnt++] = &fq->fq_base;
			if (egress_cnt == DPAA_ETH_TXQ_NUM)
				break;
		}
	}
}

static inline int dpaa_tx_fq_to_id(const struct dpaa_priv *priv,
				   struct qman_fq *tx_fq)
{
	int i;

	for (i = 0; i < DPAA_ETH_TXQ_NUM; i++)
		if (priv->egress_fqs[i] == tx_fq)
			return i;

	return -EINVAL;
}

static int dpaa_fq_init(struct dpaa_fq *dpaa_fq, bool td_enable)
{
	const struct dpaa_priv	*priv;
	struct qman_fq *confq = NULL;
	struct qm_mcc_initfq initfq;
	struct device *dev;
	struct qman_fq *fq;
	int queue_id;
	int err;

	priv = netdev_priv(dpaa_fq->net_dev);
	dev = dpaa_fq->net_dev->dev.parent;

	if (dpaa_fq->fqid == 0)
		dpaa_fq->flags |= QMAN_FQ_FLAG_DYNAMIC_FQID;

	dpaa_fq->init = !(dpaa_fq->flags & QMAN_FQ_FLAG_NO_MODIFY);

	err = qman_create_fq(dpaa_fq->fqid, dpaa_fq->flags, &dpaa_fq->fq_base);
	if (err) {
		dev_err(dev, "qman_create_fq() failed\n");
		return err;
	}
	fq = &dpaa_fq->fq_base;

	if (dpaa_fq->init) {
		memset(&initfq, 0, sizeof(initfq));

		initfq.we_mask = cpu_to_be16(QM_INITFQ_WE_FQCTRL);
		/* Note: we may get to keep an empty FQ in cache */
		initfq.fqd.fq_ctrl = cpu_to_be16(QM_FQCTRL_PREFERINCACHE);

		/* Try to reduce the number of portal interrupts for
		 * Tx Confirmation FQs.
		 */
		if (dpaa_fq->fq_type == FQ_TYPE_TX_CONFIRM)
			initfq.fqd.fq_ctrl |= cpu_to_be16(QM_FQCTRL_HOLDACTIVE);

		/* FQ placement */
		initfq.we_mask |= cpu_to_be16(QM_INITFQ_WE_DESTWQ);

		qm_fqd_set_destwq(&initfq.fqd, dpaa_fq->channel, dpaa_fq->wq);

		/* Put all egress queues in a congestion group of their own.
		 * Sensu stricto, the Tx confirmation queues are Rx FQs,
		 * rather than Tx - but they nonetheless account for the
		 * memory footprint on behalf of egress traffic. We therefore
		 * place them in the netdev's CGR, along with the Tx FQs.
		 */
		if (dpaa_fq->fq_type == FQ_TYPE_TX ||
		    dpaa_fq->fq_type == FQ_TYPE_TX_CONFIRM ||
		    dpaa_fq->fq_type == FQ_TYPE_TX_CONF_MQ) {
			initfq.we_mask |= cpu_to_be16(QM_INITFQ_WE_CGID);
			initfq.fqd.fq_ctrl |= cpu_to_be16(QM_FQCTRL_CGE);
			initfq.fqd.cgid = (u8)priv->cgr_data.cgr.cgrid;
			/* Set a fixed overhead accounting, in an attempt to
			 * reduce the impact of fixed-size skb shells and the
			 * driver's needed headroom on system memory. This is
			 * especially the case when the egress traffic is
			 * composed of small datagrams.
			 * Unfortunately, QMan's OAL value is capped to an
			 * insufficient value, but even that is better than
			 * no overhead accounting at all.
			 */
			initfq.we_mask |= cpu_to_be16(QM_INITFQ_WE_OAC);
			qm_fqd_set_oac(&initfq.fqd, QM_OAC_CG);
			qm_fqd_set_oal(&initfq.fqd,
				       min(sizeof(struct sk_buff) +
				       priv->tx_headroom,
				       (size_t)FSL_QMAN_MAX_OAL));
		}

		if (td_enable) {
			initfq.we_mask |= cpu_to_be16(QM_INITFQ_WE_TDTHRESH);
			qm_fqd_set_taildrop(&initfq.fqd, DPAA_FQ_TD, 1);
			initfq.fqd.fq_ctrl = cpu_to_be16(QM_FQCTRL_TDE);
		}

		if (dpaa_fq->fq_type == FQ_TYPE_TX) {
			queue_id = dpaa_tx_fq_to_id(priv, &dpaa_fq->fq_base);
			if (queue_id >= 0)
				confq = priv->conf_fqs[queue_id];
			if (confq) {
				initfq.we_mask |=
					cpu_to_be16(QM_INITFQ_WE_CONTEXTA);
			/* ContextA: OVOM=1(use contextA2 bits instead of ICAD)
			 *	     A2V=1 (contextA A2 field is valid)
			 *	     A0V=1 (contextA A0 field is valid)
			 *	     B0V=1 (contextB field is valid)
			 * ContextA A2: EBD=1 (deallocate buffers inside FMan)
			 * ContextB B0(ASPID): 0 (absolute Virtual Storage ID)
			 */
				qm_fqd_context_a_set64(&initfq.fqd,
						       0x1e00000080000000ULL);
			}
		}

		/* Put all the ingress queues in our "ingress CGR". */
		if (priv->use_ingress_cgr &&
		    (dpaa_fq->fq_type == FQ_TYPE_RX_DEFAULT ||
		     dpaa_fq->fq_type == FQ_TYPE_RX_ERROR)) {
			initfq.we_mask |= cpu_to_be16(QM_INITFQ_WE_CGID);
			initfq.fqd.fq_ctrl |= cpu_to_be16(QM_FQCTRL_CGE);
			initfq.fqd.cgid = (u8)priv->ingress_cgr.cgrid;
			/* Set a fixed overhead accounting, just like for the
			 * egress CGR.
			 */
			initfq.we_mask |= cpu_to_be16(QM_INITFQ_WE_OAC);
			qm_fqd_set_oac(&initfq.fqd, QM_OAC_CG);
			qm_fqd_set_oal(&initfq.fqd,
				       min(sizeof(struct sk_buff) +
				       priv->tx_headroom,
				       (size_t)FSL_QMAN_MAX_OAL));
		}

		/* Initialization common to all ingress queues */
		if (dpaa_fq->flags & QMAN_FQ_FLAG_NO_ENQUEUE) {
			initfq.we_mask |= cpu_to_be16(QM_INITFQ_WE_CONTEXTA);
			initfq.fqd.fq_ctrl |= cpu_to_be16(QM_FQCTRL_HOLDACTIVE);
			initfq.fqd.context_a.stashing.exclusive =
				QM_STASHING_EXCL_DATA | QM_STASHING_EXCL_CTX |
				QM_STASHING_EXCL_ANNOTATION;
			qm_fqd_set_stashing(&initfq.fqd, 1, 2,
					    DIV_ROUND_UP(sizeof(struct qman_fq),
							 64));
		}

		err = qman_init_fq(fq, QMAN_INITFQ_FLAG_SCHED, &initfq);
		if (err < 0) {
			dev_err(dev, "qman_init_fq(%u) = %d\n",
				qman_fq_fqid(fq), err);
			qman_destroy_fq(fq);
			return err;
		}
	}

	dpaa_fq->fqid = qman_fq_fqid(fq);

	return 0;
}

static int dpaa_fq_free_entry(struct device *dev, struct qman_fq *fq)
{
	const struct dpaa_priv  *priv;
	struct dpaa_fq *dpaa_fq;
	int err, error;

	err = 0;

	dpaa_fq = container_of(fq, struct dpaa_fq, fq_base);
	priv = netdev_priv(dpaa_fq->net_dev);

	if (dpaa_fq->init) {
		err = qman_retire_fq(fq, NULL);
		if (err < 0 && netif_msg_drv(priv))
			dev_err(dev, "qman_retire_fq(%u) = %d\n",
				qman_fq_fqid(fq), err);

		error = qman_oos_fq(fq);
		if (error < 0 && netif_msg_drv(priv)) {
			dev_err(dev, "qman_oos_fq(%u) = %d\n",
				qman_fq_fqid(fq), error);
			if (err >= 0)
				err = error;
		}
	}

	qman_destroy_fq(fq);
	list_del(&dpaa_fq->list);

	return err;
}

static int dpaa_fq_free(struct device *dev, struct list_head *list)
{
	struct dpaa_fq *dpaa_fq, *tmp;
	int err, error;

	err = 0;
	list_for_each_entry_safe(dpaa_fq, tmp, list, list) {
		error = dpaa_fq_free_entry(dev, (struct qman_fq *)dpaa_fq);
		if (error < 0 && err >= 0)
			err = error;
	}

	return err;
}

static void dpaa_eth_init_tx_port(struct fman_port *port, struct dpaa_fq *errq,
				  struct dpaa_fq *defq,
				  struct dpaa_buffer_layout *buf_layout)
{
	struct fman_buffer_prefix_content buf_prefix_content;
	struct fman_port_params params;
	int err;

	memset(&params, 0, sizeof(params));
	memset(&buf_prefix_content, 0, sizeof(buf_prefix_content));

	buf_prefix_content.priv_data_size = buf_layout->priv_data_size;
	buf_prefix_content.pass_prs_result = true;
	buf_prefix_content.pass_hash_result = true;
	buf_prefix_content.pass_time_stamp = false;
	buf_prefix_content.data_align = DPAA_FD_DATA_ALIGNMENT;

	params.specific_params.non_rx_params.err_fqid = errq->fqid;
	params.specific_params.non_rx_params.dflt_fqid = defq->fqid;

	err = fman_port_config(port, &params);
	if (err)
		pr_err("%s: fman_port_config failed\n", __func__);

	err = fman_port_cfg_buf_prefix_content(port, &buf_prefix_content);
	if (err)
		pr_err("%s: fman_port_cfg_buf_prefix_content failed\n",
		       __func__);

	err = fman_port_init(port);
	if (err)
		pr_err("%s: fm_port_init failed\n", __func__);
}

static void dpaa_eth_init_rx_port(struct fman_port *port, struct dpaa_bp **bps,
				  size_t count, struct dpaa_fq *errq,
				  struct dpaa_fq *defq,
				  struct dpaa_buffer_layout *buf_layout)
{
	struct fman_buffer_prefix_content buf_prefix_content;
	struct fman_port_rx_params *rx_p;
	struct fman_port_params params;
	int i, err;

	memset(&params, 0, sizeof(params));
	memset(&buf_prefix_content, 0, sizeof(buf_prefix_content));

	buf_prefix_content.priv_data_size = buf_layout->priv_data_size;
	buf_prefix_content.pass_prs_result = true;
	buf_prefix_content.pass_hash_result = true;
	buf_prefix_content.pass_time_stamp = false;
	buf_prefix_content.data_align = DPAA_FD_DATA_ALIGNMENT;

	rx_p = &params.specific_params.rx_params;
	rx_p->err_fqid = errq->fqid;
	rx_p->dflt_fqid = defq->fqid;

	count = min(ARRAY_SIZE(rx_p->ext_buf_pools.ext_buf_pool), count);
	rx_p->ext_buf_pools.num_of_pools_used = (u8)count;
	for (i = 0; i < count; i++) {
		rx_p->ext_buf_pools.ext_buf_pool[i].id =  bps[i]->bpid;
		rx_p->ext_buf_pools.ext_buf_pool[i].size = (u16)bps[i]->size;
	}

	err = fman_port_config(port, &params);
	if (err)
		pr_err("%s: fman_port_config failed\n", __func__);

	err = fman_port_cfg_buf_prefix_content(port, &buf_prefix_content);
	if (err)
		pr_err("%s: fman_port_cfg_buf_prefix_content failed\n",
		       __func__);

	err = fman_port_init(port);
	if (err)
		pr_err("%s: fm_port_init failed\n", __func__);
}

static void dpaa_eth_init_ports(struct mac_device *mac_dev,
				struct dpaa_bp **bps, size_t count,
				struct fm_port_fqs *port_fqs,
				struct dpaa_buffer_layout *buf_layout,
				struct device *dev)
{
	struct fman_port *rxport = mac_dev->port[RX];
	struct fman_port *txport = mac_dev->port[TX];

	dpaa_eth_init_tx_port(txport, port_fqs->tx_errq,
			      port_fqs->tx_defq, &buf_layout[TX]);
	dpaa_eth_init_rx_port(rxport, bps, count, port_fqs->rx_errq,
			      port_fqs->rx_defq, &buf_layout[RX]);
}

static int dpaa_bman_release(const struct dpaa_bp *dpaa_bp,
			     struct bm_buffer *bmb, int cnt)
{
	int err;

	err = bman_release(dpaa_bp->pool, bmb, cnt);
	/* Should never occur, address anyway to avoid leaking the buffers */
	if (unlikely(WARN_ON(err)) && dpaa_bp->free_buf_cb)
		while (cnt-- > 0)
			dpaa_bp->free_buf_cb(dpaa_bp, &bmb[cnt]);

	return cnt;
}

static void dpaa_release_sgt_members(struct qm_sg_entry *sgt)
{
	struct bm_buffer bmb[DPAA_BUFF_RELEASE_MAX];
	struct dpaa_bp *dpaa_bp;
	int i = 0, j;

	memset(bmb, 0, sizeof(bmb));

	do {
		dpaa_bp = dpaa_bpid2pool(sgt[i].bpid);
		if (!dpaa_bp)
			return;

		j = 0;
		do {
			WARN_ON(qm_sg_entry_is_ext(&sgt[i]));

			bm_buffer_set64(&bmb[j], qm_sg_entry_get64(&sgt[i]));

			j++; i++;
		} while (j < ARRAY_SIZE(bmb) &&
				!qm_sg_entry_is_final(&sgt[i - 1]) &&
				sgt[i - 1].bpid == sgt[i].bpid);

		dpaa_bman_release(dpaa_bp, bmb, j);
	} while (!qm_sg_entry_is_final(&sgt[i - 1]));
}

static void dpaa_fd_release(const struct net_device *net_dev,
			    const struct qm_fd *fd)
{
	struct qm_sg_entry *sgt;
	struct dpaa_bp *dpaa_bp;
	struct bm_buffer bmb;
	dma_addr_t addr;
	void *vaddr;

	bmb.data = 0;
	bm_buffer_set64(&bmb, qm_fd_addr(fd));

	dpaa_bp = dpaa_bpid2pool(fd->bpid);
	if (!dpaa_bp)
		return;

	if (qm_fd_get_format(fd) == qm_fd_sg) {
		vaddr = phys_to_virt(qm_fd_addr(fd));
		sgt = vaddr + qm_fd_get_offset(fd);

		dma_unmap_single(dpaa_bp->dev, qm_fd_addr(fd), dpaa_bp->size,
				 DMA_FROM_DEVICE);

		dpaa_release_sgt_members(sgt);

		addr = dma_map_single(dpaa_bp->dev, vaddr, dpaa_bp->size,
				      DMA_FROM_DEVICE);
		if (dma_mapping_error(dpaa_bp->dev, addr)) {
			dev_err(dpaa_bp->dev, "DMA mapping failed");
			return;
		}
		bm_buffer_set64(&bmb, addr);
	}

	dpaa_bman_release(dpaa_bp, &bmb, 1);
}

static void count_ern(struct dpaa_percpu_priv *percpu_priv,
		      const union qm_mr_entry *msg)
{
	switch (msg->ern.rc & QM_MR_RC_MASK) {
	case QM_MR_RC_CGR_TAILDROP:
		percpu_priv->ern_cnt.cg_tdrop++;
		break;
	case QM_MR_RC_WRED:
		percpu_priv->ern_cnt.wred++;
		break;
	case QM_MR_RC_ERROR:
		percpu_priv->ern_cnt.err_cond++;
		break;
	case QM_MR_RC_ORPWINDOW_EARLY:
		percpu_priv->ern_cnt.early_window++;
		break;
	case QM_MR_RC_ORPWINDOW_LATE:
		percpu_priv->ern_cnt.late_window++;
		break;
	case QM_MR_RC_FQ_TAILDROP:
		percpu_priv->ern_cnt.fq_tdrop++;
		break;
	case QM_MR_RC_ORPWINDOW_RETIRED:
		percpu_priv->ern_cnt.fq_retired++;
		break;
	case QM_MR_RC_ORP_ZERO:
		percpu_priv->ern_cnt.orp_zero++;
		break;
	}
}

/* Turn on HW checksum computation for this outgoing frame.
 * If the current protocol is not something we support in this regard
 * (or if the stack has already computed the SW checksum), we do nothing.
 *
 * Returns 0 if all goes well (or HW csum doesn't apply), and a negative value
 * otherwise.
 *
 * Note that this function may modify the fd->cmd field and the skb data buffer
 * (the Parse Results area).
 */
static int dpaa_enable_tx_csum(struct dpaa_priv *priv,
			       struct sk_buff *skb,
			       struct qm_fd *fd,
			       char *parse_results)
{
	struct fman_prs_result *parse_result;
	u16 ethertype = ntohs(skb->protocol);
	struct ipv6hdr *ipv6h = NULL;
	struct iphdr *iph;
	int retval = 0;
	u8 l4_proto;

	if (skb->ip_summed != CHECKSUM_PARTIAL)
		return 0;

	/* Note: L3 csum seems to be already computed in sw, but we can't choose
	 * L4 alone from the FM configuration anyway.
	 */

	/* Fill in some fields of the Parse Results array, so the FMan
	 * can find them as if they came from the FMan Parser.
	 */
	parse_result = (struct fman_prs_result *)parse_results;

	/* If we're dealing with VLAN, get the real Ethernet type */
	if (ethertype == ETH_P_8021Q) {
		/* We can't always assume the MAC header is set correctly
		 * by the stack, so reset to beginning of skb->data
		 */
		skb_reset_mac_header(skb);
		ethertype = ntohs(vlan_eth_hdr(skb)->h_vlan_encapsulated_proto);
	}

	/* Fill in the relevant L3 parse result fields
	 * and read the L4 protocol type
	 */
	switch (ethertype) {
	case ETH_P_IP:
		parse_result->l3r = cpu_to_be16(FM_L3_PARSE_RESULT_IPV4);
		iph = ip_hdr(skb);
		WARN_ON(!iph);
		l4_proto = iph->protocol;
		break;
	case ETH_P_IPV6:
		parse_result->l3r = cpu_to_be16(FM_L3_PARSE_RESULT_IPV6);
		ipv6h = ipv6_hdr(skb);
		WARN_ON(!ipv6h);
		l4_proto = ipv6h->nexthdr;
		break;
	default:
		/* We shouldn't even be here */
		if (net_ratelimit())
			netif_alert(priv, tx_err, priv->net_dev,
				    "Can't compute HW csum for L3 proto 0x%x\n",
				    ntohs(skb->protocol));
		retval = -EIO;
		goto return_error;
	}

	/* Fill in the relevant L4 parse result fields */
	switch (l4_proto) {
	case IPPROTO_UDP:
		parse_result->l4r = FM_L4_PARSE_RESULT_UDP;
		break;
	case IPPROTO_TCP:
		parse_result->l4r = FM_L4_PARSE_RESULT_TCP;
		break;
	default:
		if (net_ratelimit())
			netif_alert(priv, tx_err, priv->net_dev,
				    "Can't compute HW csum for L4 proto 0x%x\n",
				    l4_proto);
		retval = -EIO;
		goto return_error;
	}

	/* At index 0 is IPOffset_1 as defined in the Parse Results */
	parse_result->ip_off[0] = (u8)skb_network_offset(skb);
	parse_result->l4_off = (u8)skb_transport_offset(skb);

	/* Enable L3 (and L4, if TCP or UDP) HW checksum. */
	fd->cmd |= cpu_to_be32(FM_FD_CMD_RPD | FM_FD_CMD_DTC);

	/* On P1023 and similar platforms fd->cmd interpretation could
	 * be disabled by setting CONTEXT_A bit ICMD; currently this bit
	 * is not set so we do not need to check; in the future, if/when
	 * using context_a we need to check this bit
	 */

return_error:
	return retval;
}

static int dpaa_bp_add_8_bufs(const struct dpaa_bp *dpaa_bp)
{
	struct device *dev = dpaa_bp->dev;
	struct bm_buffer bmb[8];
	dma_addr_t addr;
	void *new_buf;
	u8 i;

	for (i = 0; i < 8; i++) {
		new_buf = netdev_alloc_frag(dpaa_bp->raw_size);
		if (unlikely(!new_buf)) {
			dev_err(dev, "netdev_alloc_frag() failed, size %zu\n",
				dpaa_bp->raw_size);
			goto release_previous_buffs;
		}
		new_buf = PTR_ALIGN(new_buf, SMP_CACHE_BYTES);

		addr = dma_map_single(dev, new_buf,
				      dpaa_bp->size, DMA_FROM_DEVICE);
		if (unlikely(dma_mapping_error(dev, addr))) {
			dev_err(dpaa_bp->dev, "DMA map failed");
			goto release_previous_buffs;
		}

		bmb[i].data = 0;
		bm_buffer_set64(&bmb[i], addr);
	}

release_bufs:
	return dpaa_bman_release(dpaa_bp, bmb, i);

release_previous_buffs:
	WARN_ONCE(1, "dpaa_eth: failed to add buffers on Rx\n");

	bm_buffer_set64(&bmb[i], 0);
	/* Avoid releasing a completely null buffer; bman_release() requires
	 * at least one buffer.
	 */
	if (likely(i))
		goto release_bufs;

	return 0;
}

static int dpaa_bp_seed(struct dpaa_bp *dpaa_bp)
{
	int i;

	/* Give each CPU an allotment of "config_count" buffers */
	for_each_possible_cpu(i) {
		int *count_ptr = per_cpu_ptr(dpaa_bp->percpu_count, i);
		int j;

		/* Although we access another CPU's counters here
		 * we do it at boot time so it is safe
		 */
		for (j = 0; j < dpaa_bp->config_count; j += 8)
			*count_ptr += dpaa_bp_add_8_bufs(dpaa_bp);
	}
	return 0;
}

/* Add buffers/(pages) for Rx processing whenever bpool count falls below
 * REFILL_THRESHOLD.
 */
static int dpaa_eth_refill_bpool(struct dpaa_bp *dpaa_bp, int *countptr)
{
	int count = *countptr;
	int new_bufs;

	if (unlikely(count < FSL_DPAA_ETH_REFILL_THRESHOLD)) {
		do {
			new_bufs = dpaa_bp_add_8_bufs(dpaa_bp);
			if (unlikely(!new_bufs)) {
				/* Avoid looping forever if we've temporarily
				 * run out of memory. We'll try again at the
				 * next NAPI cycle.
				 */
				break;
			}
			count += new_bufs;
		} while (count < FSL_DPAA_ETH_MAX_BUF_COUNT);

		*countptr = count;
		if (unlikely(count < FSL_DPAA_ETH_MAX_BUF_COUNT))
			return -ENOMEM;
	}

	return 0;
}

static int dpaa_eth_refill_bpools(struct dpaa_priv *priv)
{
	struct dpaa_bp *dpaa_bp;
	int *countptr;
	int res, i;

	for (i = 0; i < DPAA_BPS_NUM; i++) {
		dpaa_bp = priv->dpaa_bps[i];
		if (!dpaa_bp)
			return -EINVAL;
		countptr = this_cpu_ptr(dpaa_bp->percpu_count);
		res  = dpaa_eth_refill_bpool(dpaa_bp, countptr);
		if (res)
			return res;
	}
	return 0;
}

/* Cleanup function for outgoing frame descriptors that were built on Tx path,
 * either contiguous frames or scatter/gather ones.
 * Skb freeing is not handled here.
 *
 * This function may be called on error paths in the Tx function, so guard
 * against cases when not all fd relevant fields were filled in.
 *
 * Return the skb backpointer, since for S/G frames the buffer containing it
 * gets freed here.
 */
static struct sk_buff *dpaa_cleanup_tx_fd(const struct dpaa_priv *priv,
					  const struct qm_fd *fd)
{
	const enum dma_data_direction dma_dir = DMA_TO_DEVICE;
	struct device *dev = priv->net_dev->dev.parent;
	dma_addr_t addr = qm_fd_addr(fd);
	const struct qm_sg_entry *sgt;
	struct sk_buff **skbh, *skb;
	int nr_frags, i;

	skbh = (struct sk_buff **)phys_to_virt(addr);
	skb = *skbh;

	if (unlikely(qm_fd_get_format(fd) == qm_fd_sg)) {
		nr_frags = skb_shinfo(skb)->nr_frags;
		dma_unmap_single(dev, addr, qm_fd_get_offset(fd) +
				 sizeof(struct qm_sg_entry) * (1 + nr_frags),
				 dma_dir);

		/* The sgt buffer has been allocated with netdev_alloc_frag(),
		 * it's from lowmem.
		 */
		sgt = phys_to_virt(addr + qm_fd_get_offset(fd));

		/* sgt[0] is from lowmem, was dma_map_single()-ed */
		dma_unmap_single(dev, qm_sg_addr(&sgt[0]),
				 qm_sg_entry_get_len(&sgt[0]), dma_dir);

		/* remaining pages were mapped with skb_frag_dma_map() */
		for (i = 1; i < nr_frags; i++) {
			WARN_ON(qm_sg_entry_is_ext(&sgt[i]));

			dma_unmap_page(dev, qm_sg_addr(&sgt[i]),
				       qm_sg_entry_get_len(&sgt[i]), dma_dir);
		}

		/* Free the page frag that we allocated on Tx */
		skb_free_frag(phys_to_virt(addr));
	} else {
		dma_unmap_single(dev, addr,
				 skb_tail_pointer(skb) - (u8 *)skbh, dma_dir);
	}

	return skb;
}

/* Build a linear skb around the received buffer.
 * We are guaranteed there is enough room at the end of the data buffer to
 * accommodate the shared info area of the skb.
 */
static struct sk_buff *contig_fd_to_skb(const struct dpaa_priv *priv,
					const struct qm_fd *fd)
{
	ssize_t fd_off = qm_fd_get_offset(fd);
	dma_addr_t addr = qm_fd_addr(fd);
	struct dpaa_bp *dpaa_bp;
	struct sk_buff *skb;
	void *vaddr;

	vaddr = phys_to_virt(addr);
	WARN_ON(!IS_ALIGNED((unsigned long)vaddr, SMP_CACHE_BYTES));

	dpaa_bp = dpaa_bpid2pool(fd->bpid);
	if (!dpaa_bp)
		goto free_buffer;

	skb = build_skb(vaddr, dpaa_bp->size +
			SKB_DATA_ALIGN(sizeof(struct skb_shared_info)));
	if (unlikely(!skb)) {
		WARN_ONCE(1, "Build skb failure on Rx\n");
		goto free_buffer;
	}
	WARN_ON(fd_off != priv->rx_headroom);
	skb_reserve(skb, fd_off);
	skb_put(skb, qm_fd_get_length(fd));

	skb->ip_summed = CHECKSUM_NONE;

	return skb;

free_buffer:
	skb_free_frag(vaddr);
	return NULL;
}

/* Build an skb with the data of the first S/G entry in the linear portion and
 * the rest of the frame as skb fragments.
 *
 * The page fragment holding the S/G Table is recycled here.
 */
static struct sk_buff *sg_fd_to_skb(const struct dpaa_priv *priv,
				    const struct qm_fd *fd)
{
	ssize_t fd_off = qm_fd_get_offset(fd);
	dma_addr_t addr = qm_fd_addr(fd);
	const struct qm_sg_entry *sgt;
	struct page *page, *head_page;
	struct dpaa_bp *dpaa_bp;
	void *vaddr, *sg_vaddr;
	int frag_off, frag_len;
	struct sk_buff *skb;
	dma_addr_t sg_addr;
	int page_offset;
	unsigned int sz;
	int *count_ptr;
	int i;

	vaddr = phys_to_virt(addr);
	WARN_ON(!IS_ALIGNED((unsigned long)vaddr, SMP_CACHE_BYTES));

	/* Iterate through the SGT entries and add data buffers to the skb */
	sgt = vaddr + fd_off;
	for (i = 0; i < DPAA_SGT_MAX_ENTRIES; i++) {
		/* Extension bit is not supported */
		WARN_ON(qm_sg_entry_is_ext(&sgt[i]));

		sg_addr = qm_sg_addr(&sgt[i]);
		sg_vaddr = phys_to_virt(sg_addr);
		WARN_ON(!IS_ALIGNED((unsigned long)sg_vaddr,
				    SMP_CACHE_BYTES));

		/* We may use multiple Rx pools */
		dpaa_bp = dpaa_bpid2pool(sgt[i].bpid);
		if (!dpaa_bp)
			goto free_buffers;

		count_ptr = this_cpu_ptr(dpaa_bp->percpu_count);
		dma_unmap_single(dpaa_bp->dev, sg_addr, dpaa_bp->size,
				 DMA_FROM_DEVICE);
		if (i == 0) {
			sz = dpaa_bp->size +
				SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
			skb = build_skb(sg_vaddr, sz);
			if (WARN_ON(unlikely(!skb)))
				goto free_buffers;

			skb->ip_summed = CHECKSUM_NONE;

			/* Make sure forwarded skbs will have enough space
			 * on Tx, if extra headers are added.
			 */
			WARN_ON(fd_off != priv->rx_headroom);
			skb_reserve(skb, fd_off);
			skb_put(skb, qm_sg_entry_get_len(&sgt[i]));
		} else {
			/* Not the first S/G entry; all data from buffer will
			 * be added in an skb fragment; fragment index is offset
			 * by one since first S/G entry was incorporated in the
			 * linear part of the skb.
			 *
			 * Caution: 'page' may be a tail page.
			 */
			page = virt_to_page(sg_vaddr);
			head_page = virt_to_head_page(sg_vaddr);

			/* Compute offset in (possibly tail) page */
			page_offset = ((unsigned long)sg_vaddr &
					(PAGE_SIZE - 1)) +
				(page_address(page) - page_address(head_page));
			/* page_offset only refers to the beginning of sgt[i];
			 * but the buffer itself may have an internal offset.
			 */
			frag_off = qm_sg_entry_get_off(&sgt[i]) + page_offset;
			frag_len = qm_sg_entry_get_len(&sgt[i]);
			/* skb_add_rx_frag() does no checking on the page; if
			 * we pass it a tail page, we'll end up with
			 * bad page accounting and eventually with segafults.
			 */
			skb_add_rx_frag(skb, i - 1, head_page, frag_off,
					frag_len, dpaa_bp->size);
		}
		/* Update the pool count for the current {cpu x bpool} */
		(*count_ptr)--;

		if (qm_sg_entry_is_final(&sgt[i]))
			break;
	}
	WARN_ONCE(i == DPAA_SGT_MAX_ENTRIES, "No final bit on SGT\n");

	/* free the SG table buffer */
	skb_free_frag(vaddr);

	return skb;

free_buffers:
	/* compensate sw bpool counter changes */
	for (i--; i > 0; i--) {
		dpaa_bp = dpaa_bpid2pool(sgt[i].bpid);
		if (dpaa_bp) {
			count_ptr = this_cpu_ptr(dpaa_bp->percpu_count);
			(*count_ptr)++;
		}
	}
	/* free all the SG entries */
	for (i = 0; i < DPAA_SGT_MAX_ENTRIES ; i++) {
		sg_addr = qm_sg_addr(&sgt[i]);
		sg_vaddr = phys_to_virt(sg_addr);
		skb_free_frag(sg_vaddr);
		dpaa_bp = dpaa_bpid2pool(sgt[i].bpid);
		if (dpaa_bp) {
			count_ptr = this_cpu_ptr(dpaa_bp->percpu_count);
			(*count_ptr)--;
		}

		if (qm_sg_entry_is_final(&sgt[i]))
			break;
	}
	/* free the SGT fragment */
	skb_free_frag(vaddr);

	return NULL;
}

static int skb_to_contig_fd(struct dpaa_priv *priv,
			    struct sk_buff *skb, struct qm_fd *fd,
			    int *offset)
{
	struct net_device *net_dev = priv->net_dev;
	struct device *dev = net_dev->dev.parent;
	enum dma_data_direction dma_dir;
	unsigned char *buffer_start;
	struct sk_buff **skbh;
	dma_addr_t addr;
	int err;

	/* We are guaranteed to have at least tx_headroom bytes
	 * available, so just use that for offset.
	 */
	fd->bpid = FSL_DPAA_BPID_INV;
	buffer_start = skb->data - priv->tx_headroom;
	dma_dir = DMA_TO_DEVICE;

	skbh = (struct sk_buff **)buffer_start;
	*skbh = skb;

	/* Enable L3/L4 hardware checksum computation.
	 *
	 * We must do this before dma_map_single(DMA_TO_DEVICE), because we may
	 * need to write into the skb.
	 */
	err = dpaa_enable_tx_csum(priv, skb, fd,
				  ((char *)skbh) + DPAA_TX_PRIV_DATA_SIZE);
	if (unlikely(err < 0)) {
		if (net_ratelimit())
			netif_err(priv, tx_err, net_dev, "HW csum error: %d\n",
				  err);
		return err;
	}

	/* Fill in the rest of the FD fields */
	qm_fd_set_contig(fd, priv->tx_headroom, skb->len);
	fd->cmd |= cpu_to_be32(FM_FD_CMD_FCO);

	/* Map the entire buffer size that may be seen by FMan, but no more */
	addr = dma_map_single(dev, skbh,
			      skb_tail_pointer(skb) - buffer_start, dma_dir);
	if (unlikely(dma_mapping_error(dev, addr))) {
		if (net_ratelimit())
			netif_err(priv, tx_err, net_dev, "dma_map_single() failed\n");
		return -EINVAL;
	}
	qm_fd_addr_set64(fd, addr);

	return 0;
}

static int skb_to_sg_fd(struct dpaa_priv *priv,
			struct sk_buff *skb, struct qm_fd *fd)
{
	const enum dma_data_direction dma_dir = DMA_TO_DEVICE;
	const int nr_frags = skb_shinfo(skb)->nr_frags;
	struct net_device *net_dev = priv->net_dev;
	struct device *dev = net_dev->dev.parent;
	struct qm_sg_entry *sgt;
	struct sk_buff **skbh;
	int i, j, err, sz;
	void *buffer_start;
	skb_frag_t *frag;
	dma_addr_t addr;
	size_t frag_len;
	void *sgt_buf;

	/* get a page frag to store the SGTable */
	sz = SKB_DATA_ALIGN(priv->tx_headroom +
		sizeof(struct qm_sg_entry) * (1 + nr_frags));
	sgt_buf = netdev_alloc_frag(sz);
	if (unlikely(!sgt_buf)) {
		netdev_err(net_dev, "netdev_alloc_frag() failed for size %d\n",
			   sz);
		return -ENOMEM;
	}

	/* Enable L3/L4 hardware checksum computation.
	 *
	 * We must do this before dma_map_single(DMA_TO_DEVICE), because we may
	 * need to write into the skb.
	 */
	err = dpaa_enable_tx_csum(priv, skb, fd,
				  sgt_buf + DPAA_TX_PRIV_DATA_SIZE);
	if (unlikely(err < 0)) {
		if (net_ratelimit())
			netif_err(priv, tx_err, net_dev, "HW csum error: %d\n",
				  err);
		goto csum_failed;
	}

	sgt = (struct qm_sg_entry *)(sgt_buf + priv->tx_headroom);
	qm_sg_entry_set_len(&sgt[0], skb_headlen(skb));
	sgt[0].bpid = FSL_DPAA_BPID_INV;
	sgt[0].offset = 0;
	addr = dma_map_single(dev, skb->data,
			      skb_headlen(skb), dma_dir);
	if (unlikely(dma_mapping_error(dev, addr))) {
		dev_err(dev, "DMA mapping failed");
		err = -EINVAL;
		goto sg0_map_failed;
	}
	qm_sg_entry_set64(&sgt[0], addr);

	/* populate the rest of SGT entries */
	frag = &skb_shinfo(skb)->frags[0];
	frag_len = frag->size;
	for (i = 1; i <= nr_frags; i++, frag++) {
		WARN_ON(!skb_frag_page(frag));
		addr = skb_frag_dma_map(dev, frag, 0,
					frag_len, dma_dir);
		if (unlikely(dma_mapping_error(dev, addr))) {
			dev_err(dev, "DMA mapping failed");
			err = -EINVAL;
			goto sg_map_failed;
		}

		qm_sg_entry_set_len(&sgt[i], frag_len);
		sgt[i].bpid = FSL_DPAA_BPID_INV;
		sgt[i].offset = 0;

		/* keep the offset in the address */
		qm_sg_entry_set64(&sgt[i], addr);
		frag_len = frag->size;
	}
	qm_sg_entry_set_f(&sgt[i - 1], frag_len);

	qm_fd_set_sg(fd, priv->tx_headroom, skb->len);

	/* DMA map the SGT page */
	buffer_start = (void *)sgt - priv->tx_headroom;
	skbh = (struct sk_buff **)buffer_start;
	*skbh = skb;

	addr = dma_map_single(dev, buffer_start, priv->tx_headroom +
			      sizeof(struct qm_sg_entry) * (1 + nr_frags),
			      dma_dir);
	if (unlikely(dma_mapping_error(dev, addr))) {
		dev_err(dev, "DMA mapping failed");
		err = -EINVAL;
		goto sgt_map_failed;
	}

	fd->bpid = FSL_DPAA_BPID_INV;
	fd->cmd |= cpu_to_be32(FM_FD_CMD_FCO);
	qm_fd_addr_set64(fd, addr);

	return 0;

sgt_map_failed:
sg_map_failed:
	for (j = 0; j < i; j++)
		dma_unmap_page(dev, qm_sg_addr(&sgt[j]),
			       qm_sg_entry_get_len(&sgt[j]), dma_dir);
sg0_map_failed:
csum_failed:
	skb_free_frag(sgt_buf);

	return err;
}

static inline int dpaa_xmit(struct dpaa_priv *priv,
			    struct rtnl_link_stats64 *percpu_stats,
			    int queue,
			    struct qm_fd *fd)
{
	struct qman_fq *egress_fq;
	int err, i;

	egress_fq = priv->egress_fqs[queue];
	if (fd->bpid == FSL_DPAA_BPID_INV)
		fd->cmd |= cpu_to_be32(qman_fq_fqid(priv->conf_fqs[queue]));

	/* Trace this Tx fd */
	trace_dpaa_tx_fd(priv->net_dev, egress_fq, fd);

	for (i = 0; i < DPAA_ENQUEUE_RETRIES; i++) {
		err = qman_enqueue(egress_fq, fd);
		if (err != -EBUSY)
			break;
	}

	if (unlikely(err < 0)) {
		percpu_stats->tx_errors++;
		percpu_stats->tx_fifo_errors++;
		return err;
	}

	percpu_stats->tx_packets++;
	percpu_stats->tx_bytes += qm_fd_get_length(fd);

	return 0;
}

static int dpaa_start_xmit(struct sk_buff *skb, struct net_device *net_dev)
{
	const int queue_mapping = skb_get_queue_mapping(skb);
	bool nonlinear = skb_is_nonlinear(skb);
	struct rtnl_link_stats64 *percpu_stats;
	struct dpaa_percpu_priv *percpu_priv;
	struct dpaa_priv *priv;
	struct qm_fd fd;
	int offset = 0;
	int err = 0;

	priv = netdev_priv(net_dev);
	percpu_priv = this_cpu_ptr(priv->percpu_priv);
	percpu_stats = &percpu_priv->stats;

	qm_fd_clear_fd(&fd);

	if (!nonlinear) {
		/* We're going to store the skb backpointer at the beginning
		 * of the data buffer, so we need a privately owned skb
		 *
		 * We've made sure skb is not shared in dev->priv_flags,
		 * we need to verify the skb head is not cloned
		 */
		if (skb_cow_head(skb, priv->tx_headroom))
			goto enomem;

		WARN_ON(skb_is_nonlinear(skb));
	}

	/* MAX_SKB_FRAGS is equal or larger than our dpaa_SGT_MAX_ENTRIES;
	 * make sure we don't feed FMan with more fragments than it supports.
	 */
	if (nonlinear &&
	    likely(skb_shinfo(skb)->nr_frags < DPAA_SGT_MAX_ENTRIES)) {
		/* Just create a S/G fd based on the skb */
		err = skb_to_sg_fd(priv, skb, &fd);
		percpu_priv->tx_frag_skbuffs++;
	} else {
		/* If the egress skb contains more fragments than we support
		 * we have no choice but to linearize it ourselves.
		 */
		if (unlikely(nonlinear) && __skb_linearize(skb))
			goto enomem;

		/* Finally, create a contig FD from this skb */
		err = skb_to_contig_fd(priv, skb, &fd, &offset);
	}
	if (unlikely(err < 0))
		goto skb_to_fd_failed;

	if (likely(dpaa_xmit(priv, percpu_stats, queue_mapping, &fd) == 0))
		return NETDEV_TX_OK;

	dpaa_cleanup_tx_fd(priv, &fd);
skb_to_fd_failed:
enomem:
	percpu_stats->tx_errors++;
	dev_kfree_skb(skb);
	return NETDEV_TX_OK;
}

static void dpaa_rx_error(struct net_device *net_dev,
			  const struct dpaa_priv *priv,
			  struct dpaa_percpu_priv *percpu_priv,
			  const struct qm_fd *fd,
			  u32 fqid)
{
	if (net_ratelimit())
		netif_err(priv, hw, net_dev, "Err FD status = 0x%08x\n",
			  be32_to_cpu(fd->status) & FM_FD_STAT_RX_ERRORS);

	percpu_priv->stats.rx_errors++;

	if (be32_to_cpu(fd->status) & FM_FD_ERR_DMA)
		percpu_priv->rx_errors.dme++;
	if (be32_to_cpu(fd->status) & FM_FD_ERR_PHYSICAL)
		percpu_priv->rx_errors.fpe++;
	if (be32_to_cpu(fd->status) & FM_FD_ERR_SIZE)
		percpu_priv->rx_errors.fse++;
	if (be32_to_cpu(fd->status) & FM_FD_ERR_PRS_HDR_ERR)
		percpu_priv->rx_errors.phe++;

	dpaa_fd_release(net_dev, fd);
}

static void dpaa_tx_error(struct net_device *net_dev,
			  const struct dpaa_priv *priv,
			  struct dpaa_percpu_priv *percpu_priv,
			  const struct qm_fd *fd,
			  u32 fqid)
{
	struct sk_buff *skb;

	if (net_ratelimit())
		netif_warn(priv, hw, net_dev, "FD status = 0x%08x\n",
			   be32_to_cpu(fd->status) & FM_FD_STAT_TX_ERRORS);

	percpu_priv->stats.tx_errors++;

	skb = dpaa_cleanup_tx_fd(priv, fd);
	dev_kfree_skb(skb);
}

static int dpaa_eth_poll(struct napi_struct *napi, int budget)
{
	struct dpaa_napi_portal *np =
			container_of(napi, struct dpaa_napi_portal, napi);

	int cleaned = qman_p_poll_dqrr(np->p, budget);

	if (cleaned < budget) {
		napi_complete(napi);
		qman_p_irqsource_add(np->p, QM_PIRQ_DQRI);

	} else if (np->down) {
		qman_p_irqsource_add(np->p, QM_PIRQ_DQRI);
	}

	return cleaned;
}

static void dpaa_tx_conf(struct net_device *net_dev,
			 const struct dpaa_priv *priv,
			 struct dpaa_percpu_priv *percpu_priv,
			 const struct qm_fd *fd,
			 u32 fqid)
{
	struct sk_buff	*skb;

	if (unlikely(be32_to_cpu(fd->status) & FM_FD_STAT_TX_ERRORS)) {
		if (net_ratelimit())
			netif_warn(priv, hw, net_dev, "FD status = 0x%08x\n",
				   be32_to_cpu(fd->status) &
				   FM_FD_STAT_TX_ERRORS);

		percpu_priv->stats.tx_errors++;
	}

	percpu_priv->tx_confirm++;

	skb = dpaa_cleanup_tx_fd(priv, fd);

	consume_skb(skb);
}

static inline int dpaa_eth_napi_schedule(struct dpaa_percpu_priv *percpu_priv,
					 struct qman_portal *portal)
{
	if (unlikely(in_irq() || !in_serving_softirq())) {
		/* Disable QMan IRQ and invoke NAPI */
		qman_p_irqsource_remove(portal, QM_PIRQ_DQRI);

		percpu_priv->np.p = portal;
		napi_schedule(&percpu_priv->np.napi);
		percpu_priv->in_interrupt++;
		return 1;
	}
	return 0;
}

static enum qman_cb_dqrr_result rx_error_dqrr(struct qman_portal *portal,
					      struct qman_fq *fq,
					      const struct qm_dqrr_entry *dq)
{
	struct dpaa_fq *dpaa_fq = container_of(fq, struct dpaa_fq, fq_base);
	struct dpaa_percpu_priv *percpu_priv;
	struct net_device *net_dev;
	struct dpaa_bp *dpaa_bp;
	struct dpaa_priv *priv;

	net_dev = dpaa_fq->net_dev;
	priv = netdev_priv(net_dev);
	dpaa_bp = dpaa_bpid2pool(dq->fd.bpid);
	if (!dpaa_bp)
		return qman_cb_dqrr_consume;

	percpu_priv = this_cpu_ptr(priv->percpu_priv);

	if (dpaa_eth_napi_schedule(percpu_priv, portal))
		return qman_cb_dqrr_stop;

	if (dpaa_eth_refill_bpools(priv))
		/* Unable to refill the buffer pool due to insufficient
		 * system memory. Just release the frame back into the pool,
		 * otherwise we'll soon end up with an empty buffer pool.
		 */
		dpaa_fd_release(net_dev, &dq->fd);
	else
		dpaa_rx_error(net_dev, priv, percpu_priv, &dq->fd, fq->fqid);

	return qman_cb_dqrr_consume;
}

static enum qman_cb_dqrr_result rx_default_dqrr(struct qman_portal *portal,
						struct qman_fq *fq,
						const struct qm_dqrr_entry *dq)
{
	struct rtnl_link_stats64 *percpu_stats;
	struct dpaa_percpu_priv *percpu_priv;
	const struct qm_fd *fd = &dq->fd;
	dma_addr_t addr = qm_fd_addr(fd);
	enum qm_fd_format fd_format;
	struct net_device *net_dev;
	u32 fd_status = fd->status;
	struct dpaa_bp *dpaa_bp;
	struct dpaa_priv *priv;
	unsigned int skb_len;
	struct sk_buff *skb;
	int *count_ptr;

	fd_status = be32_to_cpu(fd->status);
	fd_format = qm_fd_get_format(fd);
	net_dev = ((struct dpaa_fq *)fq)->net_dev;
	priv = netdev_priv(net_dev);
	dpaa_bp = dpaa_bpid2pool(dq->fd.bpid);
	if (!dpaa_bp)
		return qman_cb_dqrr_consume;

	/* Trace the Rx fd */
	trace_dpaa_rx_fd(net_dev, fq, &dq->fd);

	percpu_priv = this_cpu_ptr(priv->percpu_priv);
	percpu_stats = &percpu_priv->stats;

	if (unlikely(dpaa_eth_napi_schedule(percpu_priv, portal)))
		return qman_cb_dqrr_stop;

	/* Make sure we didn't run out of buffers */
	if (unlikely(dpaa_eth_refill_bpools(priv))) {
		/* Unable to refill the buffer pool due to insufficient
		 * system memory. Just release the frame back into the pool,
		 * otherwise we'll soon end up with an empty buffer pool.
		 */
		dpaa_fd_release(net_dev, &dq->fd);
		return qman_cb_dqrr_consume;
	}

	if (unlikely(fd_status & FM_FD_STAT_RX_ERRORS) != 0) {
		if (net_ratelimit())
			netif_warn(priv, hw, net_dev, "FD status = 0x%08x\n",
				   fd_status & FM_FD_STAT_RX_ERRORS);

		percpu_stats->rx_errors++;
		dpaa_fd_release(net_dev, fd);
		return qman_cb_dqrr_consume;
	}

	dpaa_bp = dpaa_bpid2pool(fd->bpid);
	if (!dpaa_bp)
		return qman_cb_dqrr_consume;

	dma_unmap_single(dpaa_bp->dev, addr, dpaa_bp->size, DMA_FROM_DEVICE);

	/* prefetch the first 64 bytes of the frame or the SGT start */
	prefetch(phys_to_virt(addr) + qm_fd_get_offset(fd));

	fd_format = qm_fd_get_format(fd);
	/* The only FD types that we may receive are contig and S/G */
	WARN_ON((fd_format != qm_fd_contig) && (fd_format != qm_fd_sg));

	/* Account for either the contig buffer or the SGT buffer (depending on
	 * which case we were in) having been removed from the pool.
	 */
	count_ptr = this_cpu_ptr(dpaa_bp->percpu_count);
	(*count_ptr)--;

	if (likely(fd_format == qm_fd_contig))
		skb = contig_fd_to_skb(priv, fd);
	else
		skb = sg_fd_to_skb(priv, fd);
	if (!skb)
		return qman_cb_dqrr_consume;

	skb->protocol = eth_type_trans(skb, net_dev);

	skb_len = skb->len;

	if (unlikely(netif_receive_skb(skb) == NET_RX_DROP))
		return qman_cb_dqrr_consume;

	percpu_stats->rx_packets++;
	percpu_stats->rx_bytes += skb_len;

	return qman_cb_dqrr_consume;
}

static enum qman_cb_dqrr_result conf_error_dqrr(struct qman_portal *portal,
						struct qman_fq *fq,
						const struct qm_dqrr_entry *dq)
{
	struct dpaa_percpu_priv *percpu_priv;
	struct net_device *net_dev;
	struct dpaa_priv *priv;

	net_dev = ((struct dpaa_fq *)fq)->net_dev;
	priv = netdev_priv(net_dev);

	percpu_priv = this_cpu_ptr(priv->percpu_priv);

	if (dpaa_eth_napi_schedule(percpu_priv, portal))
		return qman_cb_dqrr_stop;

	dpaa_tx_error(net_dev, priv, percpu_priv, &dq->fd, fq->fqid);

	return qman_cb_dqrr_consume;
}

static enum qman_cb_dqrr_result conf_dflt_dqrr(struct qman_portal *portal,
					       struct qman_fq *fq,
					       const struct qm_dqrr_entry *dq)
{
	struct dpaa_percpu_priv *percpu_priv;
	struct net_device *net_dev;
	struct dpaa_priv *priv;

	net_dev = ((struct dpaa_fq *)fq)->net_dev;
	priv = netdev_priv(net_dev);

	/* Trace the fd */
	trace_dpaa_tx_conf_fd(net_dev, fq, &dq->fd);

	percpu_priv = this_cpu_ptr(priv->percpu_priv);

	if (dpaa_eth_napi_schedule(percpu_priv, portal))
		return qman_cb_dqrr_stop;

	dpaa_tx_conf(net_dev, priv, percpu_priv, &dq->fd, fq->fqid);

	return qman_cb_dqrr_consume;
}

static void egress_ern(struct qman_portal *portal,
		       struct qman_fq *fq,
		       const union qm_mr_entry *msg)
{
	const struct qm_fd *fd = &msg->ern.fd;
	struct dpaa_percpu_priv *percpu_priv;
	const struct dpaa_priv *priv;
	struct net_device *net_dev;
	struct sk_buff *skb;

	net_dev = ((struct dpaa_fq *)fq)->net_dev;
	priv = netdev_priv(net_dev);
	percpu_priv = this_cpu_ptr(priv->percpu_priv);

	percpu_priv->stats.tx_dropped++;
	percpu_priv->stats.tx_fifo_errors++;
	count_ern(percpu_priv, msg);

	skb = dpaa_cleanup_tx_fd(priv, fd);
	dev_kfree_skb_any(skb);
}

static const struct dpaa_fq_cbs dpaa_fq_cbs = {
	.rx_defq = { .cb = { .dqrr = rx_default_dqrr } },
	.tx_defq = { .cb = { .dqrr = conf_dflt_dqrr } },
	.rx_errq = { .cb = { .dqrr = rx_error_dqrr } },
	.tx_errq = { .cb = { .dqrr = conf_error_dqrr } },
	.egress_ern = { .cb = { .ern = egress_ern } }
};

static void dpaa_eth_napi_enable(struct dpaa_priv *priv)
{
	struct dpaa_percpu_priv *percpu_priv;
	int i;

	for_each_possible_cpu(i) {
		percpu_priv = per_cpu_ptr(priv->percpu_priv, i);

		percpu_priv->np.down = 0;
		napi_enable(&percpu_priv->np.napi);
	}
}

static void dpaa_eth_napi_disable(struct dpaa_priv *priv)
{
	struct dpaa_percpu_priv *percpu_priv;
	int i;

	for_each_possible_cpu(i) {
		percpu_priv = per_cpu_ptr(priv->percpu_priv, i);

		percpu_priv->np.down = 1;
		napi_disable(&percpu_priv->np.napi);
	}
}

static int dpaa_open(struct net_device *net_dev)
{
	struct mac_device *mac_dev;
	struct dpaa_priv *priv;
	int err, i;

	priv = netdev_priv(net_dev);
	mac_dev = priv->mac_dev;
	dpaa_eth_napi_enable(priv);

	net_dev->phydev = mac_dev->init_phy(net_dev, priv->mac_dev);
	if (!net_dev->phydev) {
		netif_err(priv, ifup, net_dev, "init_phy() failed\n");
		err = -ENODEV;
		goto phy_init_failed;
	}

	for (i = 0; i < ARRAY_SIZE(mac_dev->port); i++) {
		err = fman_port_enable(mac_dev->port[i]);
		if (err)
			goto mac_start_failed;
	}

	err = priv->mac_dev->start(mac_dev);
	if (err < 0) {
		netif_err(priv, ifup, net_dev, "mac_dev->start() = %d\n", err);
		goto mac_start_failed;
	}

	netif_tx_start_all_queues(net_dev);

	return 0;

mac_start_failed:
	for (i = 0; i < ARRAY_SIZE(mac_dev->port); i++)
		fman_port_disable(mac_dev->port[i]);

phy_init_failed:
	dpaa_eth_napi_disable(priv);

	return err;
}

static int dpaa_eth_stop(struct net_device *net_dev)
{
	struct dpaa_priv *priv;
	int err;

	err = dpaa_stop(net_dev);

	priv = netdev_priv(net_dev);
	dpaa_eth_napi_disable(priv);

	return err;
}

static const struct net_device_ops dpaa_ops = {
	.ndo_open = dpaa_open,
	.ndo_start_xmit = dpaa_start_xmit,
	.ndo_stop = dpaa_eth_stop,
	.ndo_tx_timeout = dpaa_tx_timeout,
	.ndo_get_stats64 = dpaa_get_stats64,
	.ndo_set_mac_address = dpaa_set_mac_address,
	.ndo_validate_addr = eth_validate_addr,
	.ndo_set_rx_mode = dpaa_set_rx_mode,
};

static int dpaa_napi_add(struct net_device *net_dev)
{
	struct dpaa_priv *priv = netdev_priv(net_dev);
	struct dpaa_percpu_priv *percpu_priv;
	int cpu;

	for_each_possible_cpu(cpu) {
		percpu_priv = per_cpu_ptr(priv->percpu_priv, cpu);

		netif_napi_add(net_dev, &percpu_priv->np.napi,
			       dpaa_eth_poll, NAPI_POLL_WEIGHT);
	}

	return 0;
}

static void dpaa_napi_del(struct net_device *net_dev)
{
	struct dpaa_priv *priv = netdev_priv(net_dev);
	struct dpaa_percpu_priv *percpu_priv;
	int cpu;

	for_each_possible_cpu(cpu) {
		percpu_priv = per_cpu_ptr(priv->percpu_priv, cpu);

		netif_napi_del(&percpu_priv->np.napi);
	}
}

static inline void dpaa_bp_free_pf(const struct dpaa_bp *bp,
				   struct bm_buffer *bmb)
{
	dma_addr_t addr = bm_buf_addr(bmb);

	dma_unmap_single(bp->dev, addr, bp->size, DMA_FROM_DEVICE);

	skb_free_frag(phys_to_virt(addr));
}

/* Alloc the dpaa_bp struct and configure default values */
static struct dpaa_bp *dpaa_bp_alloc(struct device *dev)
{
	struct dpaa_bp *dpaa_bp;

	dpaa_bp = devm_kzalloc(dev, sizeof(*dpaa_bp), GFP_KERNEL);
	if (!dpaa_bp)
		return ERR_PTR(-ENOMEM);

	dpaa_bp->bpid = FSL_DPAA_BPID_INV;
	dpaa_bp->percpu_count = devm_alloc_percpu(dev, *dpaa_bp->percpu_count);
	dpaa_bp->config_count = FSL_DPAA_ETH_MAX_BUF_COUNT;

	dpaa_bp->seed_cb = dpaa_bp_seed;
	dpaa_bp->free_buf_cb = dpaa_bp_free_pf;

	return dpaa_bp;
}

/* Place all ingress FQs (Rx Default, Rx Error) in a dedicated CGR.
 * We won't be sending congestion notifications to FMan; for now, we just use
 * this CGR to generate enqueue rejections to FMan in order to drop the frames
 * before they reach our ingress queues and eat up memory.
 */
static int dpaa_ingress_cgr_init(struct dpaa_priv *priv)
{
	struct qm_mcc_initcgr initcgr;
	u32 cs_th;
	int err;

	err = qman_alloc_cgrid(&priv->ingress_cgr.cgrid);
	if (err < 0) {
		if (netif_msg_drv(priv))
			pr_err("Error %d allocating CGR ID\n", err);
		goto out_error;
	}

	/* Enable CS TD, but disable Congestion State Change Notifications. */
	memset(&initcgr, 0, sizeof(initcgr));
	initcgr.we_mask = cpu_to_be16(QM_CGR_WE_CS_THRES);
	initcgr.cgr.cscn_en = QM_CGR_EN;
	cs_th = DPAA_INGRESS_CS_THRESHOLD;
	qm_cgr_cs_thres_set64(&initcgr.cgr.cs_thres, cs_th, 1);

	initcgr.we_mask |= cpu_to_be16(QM_CGR_WE_CSTD_EN);
	initcgr.cgr.cstd_en = QM_CGR_EN;

	/* This CGR will be associated with the SWP affined to the current CPU.
	 * However, we'll place all our ingress FQs in it.
	 */
	err = qman_create_cgr(&priv->ingress_cgr, QMAN_CGR_FLAG_USE_INIT,
			      &initcgr);
	if (err < 0) {
		if (netif_msg_drv(priv))
			pr_err("Error %d creating ingress CGR with ID %d\n",
			       err, priv->ingress_cgr.cgrid);
		qman_release_cgrid(priv->ingress_cgr.cgrid);
		goto out_error;
	}
	if (netif_msg_drv(priv))
		pr_debug("Created ingress CGR %d for netdev with hwaddr %pM\n",
			 priv->ingress_cgr.cgrid, priv->mac_dev->addr);

	priv->use_ingress_cgr = true;

out_error:
	return err;
}

static const struct of_device_id dpaa_match[];

static inline u16 dpaa_get_headroom(struct dpaa_buffer_layout *bl)
{
	u16 headroom;

	/* The frame headroom must accommodate:
	 * - the driver private data area
	 * - parse results, hash results, timestamp if selected
	 * If either hash results or time stamp are selected, both will
	 * be copied to/from the frame headroom, as TS is located between PR and
	 * HR in the IC and IC copy size has a granularity of 16bytes
	 * (see description of FMBM_RICP and FMBM_TICP registers in DPAARM)
	 *
	 * Also make sure the headroom is a multiple of data_align bytes
	 */
	headroom = (u16)(bl->priv_data_size + DPAA_PARSE_RESULTS_SIZE +
		DPAA_TIME_STAMP_SIZE + DPAA_HASH_RESULTS_SIZE);

	return DPAA_FD_DATA_ALIGNMENT ? ALIGN(headroom,
					      DPAA_FD_DATA_ALIGNMENT) :
					headroom;
}

static int dpaa_eth_probe(struct platform_device *pdev)
{
	struct dpaa_bp *dpaa_bps[DPAA_BPS_NUM] = {NULL};
	struct dpaa_percpu_priv *percpu_priv;
	struct net_device *net_dev = NULL;
	struct dpaa_fq *dpaa_fq, *tmp;
	struct dpaa_priv *priv = NULL;
	struct fm_port_fqs port_fqs;
	struct mac_device *mac_dev;
	int err = 0, i, channel;
	struct device *dev;

	dev = &pdev->dev;

	/* Allocate this early, so we can store relevant information in
	 * the private area
	 */
	net_dev = alloc_etherdev_mq(sizeof(*priv), DPAA_ETH_TXQ_NUM);
	if (!net_dev) {
		dev_err(dev, "alloc_etherdev_mq() failed\n");
		goto alloc_etherdev_mq_failed;
	}

	/* Do this here, so we can be verbose early */
	SET_NETDEV_DEV(net_dev, dev);
	dev_set_drvdata(dev, net_dev);

	priv = netdev_priv(net_dev);
	priv->net_dev = net_dev;

	priv->msg_enable = netif_msg_init(debug, DPAA_MSG_DEFAULT);

	mac_dev = dpaa_mac_dev_get(pdev);
	if (IS_ERR(mac_dev)) {
		dev_err(dev, "dpaa_mac_dev_get() failed\n");
		err = PTR_ERR(mac_dev);
		goto mac_probe_failed;
	}

	/* If fsl_fm_max_frm is set to a higher value than the all-common 1500,
	 * we choose conservatively and let the user explicitly set a higher
	 * MTU via ifconfig. Otherwise, the user may end up with different MTUs
	 * in the same LAN.
	 * If on the other hand fsl_fm_max_frm has been chosen below 1500,
	 * start with the maximum allowed.
	 */
	net_dev->mtu = min(dpaa_get_max_mtu(), ETH_DATA_LEN);

	netdev_dbg(net_dev, "Setting initial MTU on net device: %d\n",
		   net_dev->mtu);

	priv->buf_layout[RX].priv_data_size = DPAA_RX_PRIV_DATA_SIZE; /* Rx */
	priv->buf_layout[TX].priv_data_size = DPAA_TX_PRIV_DATA_SIZE; /* Tx */

	/* device used for DMA mapping */
	arch_setup_dma_ops(dev, 0, 0, NULL, false);
	err = dma_coerce_mask_and_coherent(dev, DMA_BIT_MASK(40));
	if (err) {
		dev_err(dev, "dma_coerce_mask_and_coherent() failed\n");
		goto dev_mask_failed;
	}

	/* bp init */
	for (i = 0; i < DPAA_BPS_NUM; i++) {
		int err;

		dpaa_bps[i] = dpaa_bp_alloc(dev);
		if (IS_ERR(dpaa_bps[i]))
			return PTR_ERR(dpaa_bps[i]);
		/* the raw size of the buffers used for reception */
		dpaa_bps[i]->raw_size = bpool_buffer_raw_size(i, DPAA_BPS_NUM);
		/* avoid runtime computations by keeping the usable size here */
		dpaa_bps[i]->size = dpaa_bp_size(dpaa_bps[i]->raw_size);
		dpaa_bps[i]->dev = dev;

		err = dpaa_bp_alloc_pool(dpaa_bps[i]);
		if (err < 0) {
			dpaa_bps_free(priv);
			priv->dpaa_bps[i] = NULL;
			goto bp_create_failed;
		}
		priv->dpaa_bps[i] = dpaa_bps[i];
	}

	INIT_LIST_HEAD(&priv->dpaa_fq_list);

	memset(&port_fqs, 0, sizeof(port_fqs));

	err = dpaa_alloc_all_fqs(dev, &priv->dpaa_fq_list, &port_fqs);
	if (err < 0) {
		dev_err(dev, "dpaa_alloc_all_fqs() failed\n");
		goto fq_probe_failed;
	}

	priv->mac_dev = mac_dev;

	channel = dpaa_get_channel();
	if (channel < 0) {
		dev_err(dev, "dpaa_get_channel() failed\n");
		err = channel;
		goto get_channel_failed;
	}

	priv->channel = (u16)channel;

	/* Start a thread that will walk the CPUs with affine portals
	 * and add this pool channel to each's dequeue mask.
	 */
	dpaa_eth_add_channel(priv->channel);

	dpaa_fq_setup(priv, &dpaa_fq_cbs, priv->mac_dev->port[TX]);

	/* Create a congestion group for this netdev, with
	 * dynamically-allocated CGR ID.
	 * Must be executed after probing the MAC, but before
	 * assigning the egress FQs to the CGRs.
	 */
	err = dpaa_eth_cgr_init(priv);
	if (err < 0) {
		dev_err(dev, "Error initializing CGR\n");
		goto tx_cgr_init_failed;
	}

	err = dpaa_ingress_cgr_init(priv);
	if (err < 0) {
		dev_err(dev, "Error initializing ingress CGR\n");
		goto rx_cgr_init_failed;
	}

	/* Add the FQs to the interface, and make them active */
	list_for_each_entry_safe(dpaa_fq, tmp, &priv->dpaa_fq_list, list) {
		err = dpaa_fq_init(dpaa_fq, false);
		if (err < 0)
			goto fq_alloc_failed;
	}

	priv->tx_headroom = dpaa_get_headroom(&priv->buf_layout[TX]);
	priv->rx_headroom = dpaa_get_headroom(&priv->buf_layout[RX]);

	/* All real interfaces need their ports initialized */
	dpaa_eth_init_ports(mac_dev, dpaa_bps, DPAA_BPS_NUM, &port_fqs,
			    &priv->buf_layout[0], dev);

	priv->percpu_priv = devm_alloc_percpu(dev, *priv->percpu_priv);
	if (!priv->percpu_priv) {
		dev_err(dev, "devm_alloc_percpu() failed\n");
		err = -ENOMEM;
		goto alloc_percpu_failed;
	}
	for_each_possible_cpu(i) {
		percpu_priv = per_cpu_ptr(priv->percpu_priv, i);
		memset(percpu_priv, 0, sizeof(*percpu_priv));
	}

	/* Initialize NAPI */
	err = dpaa_napi_add(net_dev);
	if (err < 0)
		goto napi_add_failed;

	err = dpaa_netdev_init(net_dev, &dpaa_ops, tx_timeout);
	if (err < 0)
		goto netdev_init_failed;

	dpaa_eth_sysfs_init(&net_dev->dev);

	netif_info(priv, probe, net_dev, "Probed interface %s\n",
		   net_dev->name);

	return 0;

netdev_init_failed:
napi_add_failed:
	dpaa_napi_del(net_dev);
alloc_percpu_failed:
	dpaa_fq_free(dev, &priv->dpaa_fq_list);
fq_alloc_failed:
	qman_delete_cgr_safe(&priv->ingress_cgr);
	qman_release_cgrid(priv->ingress_cgr.cgrid);
rx_cgr_init_failed:
	qman_delete_cgr_safe(&priv->cgr_data.cgr);
	qman_release_cgrid(priv->cgr_data.cgr.cgrid);
tx_cgr_init_failed:
get_channel_failed:
	dpaa_bps_free(priv);
bp_create_failed:
fq_probe_failed:
dev_mask_failed:
mac_probe_failed:
	dev_set_drvdata(dev, NULL);
	free_netdev(net_dev);
alloc_etherdev_mq_failed:
	for (i = 0; i < DPAA_BPS_NUM && dpaa_bps[i]; i++) {
		if (atomic_read(&dpaa_bps[i]->refs) == 0)
			devm_kfree(dev, dpaa_bps[i]);
	}
	return err;
}

static int dpaa_remove(struct platform_device *pdev)
{
	struct net_device *net_dev;
	struct dpaa_priv *priv;
	struct device *dev;
	int err;

	dev = &pdev->dev;
	net_dev = dev_get_drvdata(dev);

	priv = netdev_priv(net_dev);

	dpaa_eth_sysfs_remove(dev);

	dev_set_drvdata(dev, NULL);
	unregister_netdev(net_dev);

	err = dpaa_fq_free(dev, &priv->dpaa_fq_list);

	qman_delete_cgr_safe(&priv->ingress_cgr);
	qman_release_cgrid(priv->ingress_cgr.cgrid);
	qman_delete_cgr_safe(&priv->cgr_data.cgr);
	qman_release_cgrid(priv->cgr_data.cgr.cgrid);

	dpaa_napi_del(net_dev);

	dpaa_bps_free(priv);

	free_netdev(net_dev);

	return err;
}

static struct platform_device_id dpaa_devtype[] = {
	{
		.name = "dpaa-ethernet",
		.driver_data = 0,
	}, {
	}
};
MODULE_DEVICE_TABLE(platform, dpaa_devtype);

static struct platform_driver dpaa_driver = {
	.driver = {
		.name = KBUILD_MODNAME,
	},
	.id_table = dpaa_devtype,
	.probe = dpaa_eth_probe,
	.remove = dpaa_remove
};

static int __init dpaa_load(void)
{
	int err;

	pr_debug("FSL DPAA Ethernet driver\n");

	/* initialize dpaa_eth mirror values */
	dpaa_rx_extra_headroom = fman_get_rx_extra_headroom();
	dpaa_max_frm = fman_get_max_frm();

	err = platform_driver_register(&dpaa_driver);
	if (err < 0)
		pr_err("Error, platform_driver_register() = %d\n", err);

	return err;
}
module_init(dpaa_load);

static void __exit dpaa_unload(void)
{
	platform_driver_unregister(&dpaa_driver);

	/* Only one channel is used and needs to be released after all
	 * interfaces are removed
	 */
	dpaa_release_channel();
}
module_exit(dpaa_unload);

MODULE_LICENSE("Dual BSD/GPL");
MODULE_DESCRIPTION("FSL DPAA Ethernet driver");