summaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/i915/intel_lrc.c
blob: 15434cad543001317be875f1a266e3fef6636042 (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
/*
 * Copyright © 2014 Intel Corporation
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the "Software"),
 * to deal in the Software without restriction, including without limitation
 * the rights to use, copy, modify, merge, publish, distribute, sublicense,
 * and/or sell copies of the Software, and to permit persons to whom the
 * Software is furnished to do so, subject to the following conditions:
 *
 * The above copyright notice and this permission notice (including the next
 * paragraph) shall be included in all copies or substantial portions of the
 * Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 *
 * Authors:
 *    Ben Widawsky <ben@bwidawsk.net>
 *    Michel Thierry <michel.thierry@intel.com>
 *    Thomas Daniel <thomas.daniel@intel.com>
 *    Oscar Mateo <oscar.mateo@intel.com>
 *
 */

/**
 * DOC: Logical Rings, Logical Ring Contexts and Execlists
 *
 * Motivation:
 * GEN8 brings an expansion of the HW contexts: "Logical Ring Contexts".
 * These expanded contexts enable a number of new abilities, especially
 * "Execlists" (also implemented in this file).
 *
 * One of the main differences with the legacy HW contexts is that logical
 * ring contexts incorporate many more things to the context's state, like
 * PDPs or ringbuffer control registers:
 *
 * The reason why PDPs are included in the context is straightforward: as
 * PPGTTs (per-process GTTs) are actually per-context, having the PDPs
 * contained there mean you don't need to do a ppgtt->switch_mm yourself,
 * instead, the GPU will do it for you on the context switch.
 *
 * But, what about the ringbuffer control registers (head, tail, etc..)?
 * shouldn't we just need a set of those per engine command streamer? This is
 * where the name "Logical Rings" starts to make sense: by virtualizing the
 * rings, the engine cs shifts to a new "ring buffer" with every context
 * switch. When you want to submit a workload to the GPU you: A) choose your
 * context, B) find its appropriate virtualized ring, C) write commands to it
 * and then, finally, D) tell the GPU to switch to that context.
 *
 * Instead of the legacy MI_SET_CONTEXT, the way you tell the GPU to switch
 * to a contexts is via a context execution list, ergo "Execlists".
 *
 * LRC implementation:
 * Regarding the creation of contexts, we have:
 *
 * - One global default context.
 * - One local default context for each opened fd.
 * - One local extra context for each context create ioctl call.
 *
 * Now that ringbuffers belong per-context (and not per-engine, like before)
 * and that contexts are uniquely tied to a given engine (and not reusable,
 * like before) we need:
 *
 * - One ringbuffer per-engine inside each context.
 * - One backing object per-engine inside each context.
 *
 * The global default context starts its life with these new objects fully
 * allocated and populated. The local default context for each opened fd is
 * more complex, because we don't know at creation time which engine is going
 * to use them. To handle this, we have implemented a deferred creation of LR
 * contexts:
 *
 * The local context starts its life as a hollow or blank holder, that only
 * gets populated for a given engine once we receive an execbuffer. If later
 * on we receive another execbuffer ioctl for the same context but a different
 * engine, we allocate/populate a new ringbuffer and context backing object and
 * so on.
 *
 * Finally, regarding local contexts created using the ioctl call: as they are
 * only allowed with the render ring, we can allocate & populate them right
 * away (no need to defer anything, at least for now).
 *
 * Execlists implementation:
 * Execlists are the new method by which, on gen8+ hardware, workloads are
 * submitted for execution (as opposed to the legacy, ringbuffer-based, method).
 * This method works as follows:
 *
 * When a request is committed, its commands (the BB start and any leading or
 * trailing commands, like the seqno breadcrumbs) are placed in the ringbuffer
 * for the appropriate context. The tail pointer in the hardware context is not
 * updated at this time, but instead, kept by the driver in the ringbuffer
 * structure. A structure representing this request is added to a request queue
 * for the appropriate engine: this structure contains a copy of the context's
 * tail after the request was written to the ring buffer and a pointer to the
 * context itself.
 *
 * If the engine's request queue was empty before the request was added, the
 * queue is processed immediately. Otherwise the queue will be processed during
 * a context switch interrupt. In any case, elements on the queue will get sent
 * (in pairs) to the GPU's ExecLists Submit Port (ELSP, for short) with a
 * globally unique 20-bits submission ID.
 *
 * When execution of a request completes, the GPU updates the context status
 * buffer with a context complete event and generates a context switch interrupt.
 * During the interrupt handling, the driver examines the events in the buffer:
 * for each context complete event, if the announced ID matches that on the head
 * of the request queue, then that request is retired and removed from the queue.
 *
 * After processing, if any requests were retired and the queue is not empty
 * then a new execution list can be submitted. The two requests at the front of
 * the queue are next to be submitted but since a context may not occur twice in
 * an execution list, if subsequent requests have the same ID as the first then
 * the two requests must be combined. This is done simply by discarding requests
 * at the head of the queue until either only one requests is left (in which case
 * we use a NULL second context) or the first two requests have unique IDs.
 *
 * By always executing the first two requests in the queue the driver ensures
 * that the GPU is kept as busy as possible. In the case where a single context
 * completes but a second context is still executing, the request for this second
 * context will be at the head of the queue when we remove the first one. This
 * request will then be resubmitted along with a new request for a different context,
 * which will cause the hardware to continue executing the second request and queue
 * the new request (the GPU detects the condition of a context getting preempted
 * with the same context and optimizes the context switch flow by not doing
 * preemption, but just sampling the new tail pointer).
 *
 */
#include <linux/interrupt.h>

#include <drm/drmP.h>
#include <drm/i915_drm.h>
#include "i915_drv.h"
#include "i915_gem_render_state.h"
#include "intel_lrc_reg.h"
#include "intel_mocs.h"
#include "intel_workarounds.h"

#define RING_EXECLIST_QFULL		(1 << 0x2)
#define RING_EXECLIST1_VALID		(1 << 0x3)
#define RING_EXECLIST0_VALID		(1 << 0x4)
#define RING_EXECLIST_ACTIVE_STATUS	(3 << 0xE)
#define RING_EXECLIST1_ACTIVE		(1 << 0x11)
#define RING_EXECLIST0_ACTIVE		(1 << 0x12)

#define GEN8_CTX_STATUS_IDLE_ACTIVE	(1 << 0)
#define GEN8_CTX_STATUS_PREEMPTED	(1 << 1)
#define GEN8_CTX_STATUS_ELEMENT_SWITCH	(1 << 2)
#define GEN8_CTX_STATUS_ACTIVE_IDLE	(1 << 3)
#define GEN8_CTX_STATUS_COMPLETE	(1 << 4)
#define GEN8_CTX_STATUS_LITE_RESTORE	(1 << 15)

#define GEN8_CTX_STATUS_COMPLETED_MASK \
	 (GEN8_CTX_STATUS_COMPLETE | GEN8_CTX_STATUS_PREEMPTED)

/* Typical size of the average request (2 pipecontrols and a MI_BB) */
#define EXECLISTS_REQUEST_SIZE 64 /* bytes */
#define WA_TAIL_DWORDS 2
#define WA_TAIL_BYTES (sizeof(u32) * WA_TAIL_DWORDS)

static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
					    struct intel_engine_cs *engine);
static void execlists_init_reg_state(u32 *reg_state,
				     struct i915_gem_context *ctx,
				     struct intel_engine_cs *engine,
				     struct intel_ring *ring);

static inline struct i915_priolist *to_priolist(struct rb_node *rb)
{
	return rb_entry(rb, struct i915_priolist, node);
}

static inline int rq_prio(const struct i915_request *rq)
{
	return rq->sched.attr.priority;
}

static inline bool need_preempt(const struct intel_engine_cs *engine,
				const struct i915_request *last,
				int prio)
{
	return (intel_engine_has_preemption(engine) &&
		__execlists_need_preempt(prio, rq_prio(last)) &&
		!i915_request_completed(last));
}

/**
 * intel_lr_context_descriptor_update() - calculate & cache the descriptor
 * 					  descriptor for a pinned context
 * @ctx: Context to work on
 * @engine: Engine the descriptor will be used with
 *
 * The context descriptor encodes various attributes of a context,
 * including its GTT address and some flags. Because it's fairly
 * expensive to calculate, we'll just do it once and cache the result,
 * which remains valid until the context is unpinned.
 *
 * This is what a descriptor looks like, from LSB to MSB::
 *
 *      bits  0-11:    flags, GEN8_CTX_* (cached in ctx->desc_template)
 *      bits 12-31:    LRCA, GTT address of (the HWSP of) this context
 *      bits 32-52:    ctx ID, a globally unique tag
 *      bits 53-54:    mbz, reserved for use by hardware
 *      bits 55-63:    group ID, currently unused and set to 0
 *
 * Starting from Gen11, the upper dword of the descriptor has a new format:
 *
 *      bits 32-36:    reserved
 *      bits 37-47:    SW context ID
 *      bits 48:53:    engine instance
 *      bit 54:        mbz, reserved for use by hardware
 *      bits 55-60:    SW counter
 *      bits 61-63:    engine class
 *
 * engine info, SW context ID and SW counter need to form a unique number
 * (Context ID) per lrc.
 */
static void
intel_lr_context_descriptor_update(struct i915_gem_context *ctx,
				   struct intel_engine_cs *engine)
{
	struct intel_context *ce = to_intel_context(ctx, engine);
	u64 desc;

	BUILD_BUG_ON(MAX_CONTEXT_HW_ID > (BIT(GEN8_CTX_ID_WIDTH)));
	BUILD_BUG_ON(GEN11_MAX_CONTEXT_HW_ID > (BIT(GEN11_SW_CTX_ID_WIDTH)));

	desc = ctx->desc_template;				/* bits  0-11 */
	GEM_BUG_ON(desc & GENMASK_ULL(63, 12));

	desc |= i915_ggtt_offset(ce->state) + LRC_HEADER_PAGES * PAGE_SIZE;
								/* bits 12-31 */
	GEM_BUG_ON(desc & GENMASK_ULL(63, 32));

	if (INTEL_GEN(ctx->i915) >= 11) {
		GEM_BUG_ON(ctx->hw_id >= BIT(GEN11_SW_CTX_ID_WIDTH));
		desc |= (u64)ctx->hw_id << GEN11_SW_CTX_ID_SHIFT;
								/* bits 37-47 */

		desc |= (u64)engine->instance << GEN11_ENGINE_INSTANCE_SHIFT;
								/* bits 48-53 */

		/* TODO: decide what to do with SW counter (bits 55-60) */

		desc |= (u64)engine->class << GEN11_ENGINE_CLASS_SHIFT;
								/* bits 61-63 */
	} else {
		GEM_BUG_ON(ctx->hw_id >= BIT(GEN8_CTX_ID_WIDTH));
		desc |= (u64)ctx->hw_id << GEN8_CTX_ID_SHIFT;	/* bits 32-52 */
	}

	ce->lrc_desc = desc;
}

static struct i915_priolist *
lookup_priolist(struct intel_engine_cs *engine, int prio)
{
	struct intel_engine_execlists * const execlists = &engine->execlists;
	struct i915_priolist *p;
	struct rb_node **parent, *rb;
	bool first = true;

	if (unlikely(execlists->no_priolist))
		prio = I915_PRIORITY_NORMAL;

find_priolist:
	/* most positive priority is scheduled first, equal priorities fifo */
	rb = NULL;
	parent = &execlists->queue.rb_node;
	while (*parent) {
		rb = *parent;
		p = to_priolist(rb);
		if (prio > p->priority) {
			parent = &rb->rb_left;
		} else if (prio < p->priority) {
			parent = &rb->rb_right;
			first = false;
		} else {
			return p;
		}
	}

	if (prio == I915_PRIORITY_NORMAL) {
		p = &execlists->default_priolist;
	} else {
		p = kmem_cache_alloc(engine->i915->priorities, GFP_ATOMIC);
		/* Convert an allocation failure to a priority bump */
		if (unlikely(!p)) {
			prio = I915_PRIORITY_NORMAL; /* recurses just once */

			/* To maintain ordering with all rendering, after an
			 * allocation failure we have to disable all scheduling.
			 * Requests will then be executed in fifo, and schedule
			 * will ensure that dependencies are emitted in fifo.
			 * There will be still some reordering with existing
			 * requests, so if userspace lied about their
			 * dependencies that reordering may be visible.
			 */
			execlists->no_priolist = true;
			goto find_priolist;
		}
	}

	p->priority = prio;
	INIT_LIST_HEAD(&p->requests);
	rb_link_node(&p->node, rb, parent);
	rb_insert_color(&p->node, &execlists->queue);

	if (first)
		execlists->first = &p->node;

	return p;
}

static void unwind_wa_tail(struct i915_request *rq)
{
	rq->tail = intel_ring_wrap(rq->ring, rq->wa_tail - WA_TAIL_BYTES);
	assert_ring_tail_valid(rq->ring, rq->tail);
}

static void __unwind_incomplete_requests(struct intel_engine_cs *engine)
{
	struct i915_request *rq, *rn;
	struct i915_priolist *uninitialized_var(p);
	int last_prio = I915_PRIORITY_INVALID;

	lockdep_assert_held(&engine->timeline.lock);

	list_for_each_entry_safe_reverse(rq, rn,
					 &engine->timeline.requests,
					 link) {
		if (i915_request_completed(rq))
			return;

		__i915_request_unsubmit(rq);
		unwind_wa_tail(rq);

		GEM_BUG_ON(rq_prio(rq) == I915_PRIORITY_INVALID);
		if (rq_prio(rq) != last_prio) {
			last_prio = rq_prio(rq);
			p = lookup_priolist(engine, last_prio);
		}

		GEM_BUG_ON(p->priority != rq_prio(rq));
		list_add(&rq->sched.link, &p->requests);
	}
}

void
execlists_unwind_incomplete_requests(struct intel_engine_execlists *execlists)
{
	struct intel_engine_cs *engine =
		container_of(execlists, typeof(*engine), execlists);
	unsigned long flags;

	spin_lock_irqsave(&engine->timeline.lock, flags);

	__unwind_incomplete_requests(engine);

	spin_unlock_irqrestore(&engine->timeline.lock, flags);
}

static inline void
execlists_context_status_change(struct i915_request *rq, unsigned long status)
{
	/*
	 * Only used when GVT-g is enabled now. When GVT-g is disabled,
	 * The compiler should eliminate this function as dead-code.
	 */
	if (!IS_ENABLED(CONFIG_DRM_I915_GVT))
		return;

	atomic_notifier_call_chain(&rq->engine->context_status_notifier,
				   status, rq);
}

inline void
execlists_user_begin(struct intel_engine_execlists *execlists,
		     const struct execlist_port *port)
{
	execlists_set_active_once(execlists, EXECLISTS_ACTIVE_USER);
}

inline void
execlists_user_end(struct intel_engine_execlists *execlists)
{
	execlists_clear_active(execlists, EXECLISTS_ACTIVE_USER);
}

static inline void
execlists_context_schedule_in(struct i915_request *rq)
{
	execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_IN);
	intel_engine_context_in(rq->engine);
}

static inline void
execlists_context_schedule_out(struct i915_request *rq, unsigned long status)
{
	intel_engine_context_out(rq->engine);
	execlists_context_status_change(rq, status);
	trace_i915_request_out(rq);
}

static void
execlists_update_context_pdps(struct i915_hw_ppgtt *ppgtt, u32 *reg_state)
{
	ASSIGN_CTX_PDP(ppgtt, reg_state, 3);
	ASSIGN_CTX_PDP(ppgtt, reg_state, 2);
	ASSIGN_CTX_PDP(ppgtt, reg_state, 1);
	ASSIGN_CTX_PDP(ppgtt, reg_state, 0);
}

static u64 execlists_update_context(struct i915_request *rq)
{
	struct intel_context *ce = to_intel_context(rq->ctx, rq->engine);
	struct i915_hw_ppgtt *ppgtt =
		rq->ctx->ppgtt ?: rq->i915->mm.aliasing_ppgtt;
	u32 *reg_state = ce->lrc_reg_state;

	reg_state[CTX_RING_TAIL+1] = intel_ring_set_tail(rq->ring, rq->tail);

	/* True 32b PPGTT with dynamic page allocation: update PDP
	 * registers and point the unallocated PDPs to scratch page.
	 * PML4 is allocated during ppgtt init, so this is not needed
	 * in 48-bit mode.
	 */
	if (ppgtt && !i915_vm_is_48bit(&ppgtt->base))
		execlists_update_context_pdps(ppgtt, reg_state);

	return ce->lrc_desc;
}

static inline void write_desc(struct intel_engine_execlists *execlists, u64 desc, u32 port)
{
	if (execlists->ctrl_reg) {
		writel(lower_32_bits(desc), execlists->submit_reg + port * 2);
		writel(upper_32_bits(desc), execlists->submit_reg + port * 2 + 1);
	} else {
		writel(upper_32_bits(desc), execlists->submit_reg);
		writel(lower_32_bits(desc), execlists->submit_reg);
	}
}

static void execlists_submit_ports(struct intel_engine_cs *engine)
{
	struct intel_engine_execlists *execlists = &engine->execlists;
	struct execlist_port *port = execlists->port;
	unsigned int n;

	/*
	 * ELSQ note: the submit queue is not cleared after being submitted
	 * to the HW so we need to make sure we always clean it up. This is
	 * currently ensured by the fact that we always write the same number
	 * of elsq entries, keep this in mind before changing the loop below.
	 */
	for (n = execlists_num_ports(execlists); n--; ) {
		struct i915_request *rq;
		unsigned int count;
		u64 desc;

		rq = port_unpack(&port[n], &count);
		if (rq) {
			GEM_BUG_ON(count > !n);
			if (!count++)
				execlists_context_schedule_in(rq);
			port_set(&port[n], port_pack(rq, count));
			desc = execlists_update_context(rq);
			GEM_DEBUG_EXEC(port[n].context_id = upper_32_bits(desc));

			GEM_TRACE("%s in[%d]:  ctx=%d.%d, global=%d (fence %llx:%d) (current %d), prio=%d\n",
				  engine->name, n,
				  port[n].context_id, count,
				  rq->global_seqno,
				  rq->fence.context, rq->fence.seqno,
				  intel_engine_get_seqno(engine),
				  rq_prio(rq));
		} else {
			GEM_BUG_ON(!n);
			desc = 0;
		}

		write_desc(execlists, desc, n);
	}

	/* we need to manually load the submit queue */
	if (execlists->ctrl_reg)
		writel(EL_CTRL_LOAD, execlists->ctrl_reg);

	execlists_clear_active(execlists, EXECLISTS_ACTIVE_HWACK);
}

static bool ctx_single_port_submission(const struct i915_gem_context *ctx)
{
	return (IS_ENABLED(CONFIG_DRM_I915_GVT) &&
		i915_gem_context_force_single_submission(ctx));
}

static bool can_merge_ctx(const struct i915_gem_context *prev,
			  const struct i915_gem_context *next)
{
	if (prev != next)
		return false;

	if (ctx_single_port_submission(prev))
		return false;

	return true;
}

static void port_assign(struct execlist_port *port, struct i915_request *rq)
{
	GEM_BUG_ON(rq == port_request(port));

	if (port_isset(port))
		i915_request_put(port_request(port));

	port_set(port, port_pack(i915_request_get(rq), port_count(port)));
}

static void inject_preempt_context(struct intel_engine_cs *engine)
{
	struct intel_engine_execlists *execlists = &engine->execlists;
	struct intel_context *ce =
		to_intel_context(engine->i915->preempt_context, engine);
	unsigned int n;

	GEM_BUG_ON(execlists->preempt_complete_status !=
		   upper_32_bits(ce->lrc_desc));
	GEM_BUG_ON((ce->lrc_reg_state[CTX_CONTEXT_CONTROL + 1] &
		    _MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT |
				       CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT)) !=
		   _MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT |
				      CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT));

	/*
	 * Switch to our empty preempt context so
	 * the state of the GPU is known (idle).
	 */
	GEM_TRACE("%s\n", engine->name);
	for (n = execlists_num_ports(execlists); --n; )
		write_desc(execlists, 0, n);

	write_desc(execlists, ce->lrc_desc, n);

	/* we need to manually load the submit queue */
	if (execlists->ctrl_reg)
		writel(EL_CTRL_LOAD, execlists->ctrl_reg);

	execlists_clear_active(&engine->execlists, EXECLISTS_ACTIVE_HWACK);
	execlists_set_active(&engine->execlists, EXECLISTS_ACTIVE_PREEMPT);
}

static bool __execlists_dequeue(struct intel_engine_cs *engine)
{
	struct intel_engine_execlists * const execlists = &engine->execlists;
	struct execlist_port *port = execlists->port;
	const struct execlist_port * const last_port =
		&execlists->port[execlists->port_mask];
	struct i915_request *last = port_request(port);
	struct rb_node *rb;
	bool submit = false;

	lockdep_assert_held(&engine->timeline.lock);

	/* Hardware submission is through 2 ports. Conceptually each port
	 * has a (RING_START, RING_HEAD, RING_TAIL) tuple. RING_START is
	 * static for a context, and unique to each, so we only execute
	 * requests belonging to a single context from each ring. RING_HEAD
	 * is maintained by the CS in the context image, it marks the place
	 * where it got up to last time, and through RING_TAIL we tell the CS
	 * where we want to execute up to this time.
	 *
	 * In this list the requests are in order of execution. Consecutive
	 * requests from the same context are adjacent in the ringbuffer. We
	 * can combine these requests into a single RING_TAIL update:
	 *
	 *              RING_HEAD...req1...req2
	 *                                    ^- RING_TAIL
	 * since to execute req2 the CS must first execute req1.
	 *
	 * Our goal then is to point each port to the end of a consecutive
	 * sequence of requests as being the most optimal (fewest wake ups
	 * and context switches) submission.
	 */

	rb = execlists->first;
	GEM_BUG_ON(rb_first(&execlists->queue) != rb);

	if (last) {
		/*
		 * Don't resubmit or switch until all outstanding
		 * preemptions (lite-restore) are seen. Then we
		 * know the next preemption status we see corresponds
		 * to this ELSP update.
		 */
		GEM_BUG_ON(!execlists_is_active(execlists,
						EXECLISTS_ACTIVE_USER));
		GEM_BUG_ON(!port_count(&port[0]));
		if (port_count(&port[0]) > 1)
			return false;

		/*
		 * If we write to ELSP a second time before the HW has had
		 * a chance to respond to the previous write, we can confuse
		 * the HW and hit "undefined behaviour". After writing to ELSP,
		 * we must then wait until we see a context-switch event from
		 * the HW to indicate that it has had a chance to respond.
		 */
		if (!execlists_is_active(execlists, EXECLISTS_ACTIVE_HWACK))
			return false;

		if (need_preempt(engine, last, execlists->queue_priority)) {
			inject_preempt_context(engine);
			return false;
		}

		/*
		 * In theory, we could coalesce more requests onto
		 * the second port (the first port is active, with
		 * no preemptions pending). However, that means we
		 * then have to deal with the possible lite-restore
		 * of the second port (as we submit the ELSP, there
		 * may be a context-switch) but also we may complete
		 * the resubmission before the context-switch. Ergo,
		 * coalescing onto the second port will cause a
		 * preemption event, but we cannot predict whether
		 * that will affect port[0] or port[1].
		 *
		 * If the second port is already active, we can wait
		 * until the next context-switch before contemplating
		 * new requests. The GPU will be busy and we should be
		 * able to resubmit the new ELSP before it idles,
		 * avoiding pipeline bubbles (momentary pauses where
		 * the driver is unable to keep up the supply of new
		 * work). However, we have to double check that the
		 * priorities of the ports haven't been switch.
		 */
		if (port_count(&port[1]))
			return false;

		/*
		 * WaIdleLiteRestore:bdw,skl
		 * Apply the wa NOOPs to prevent
		 * ring:HEAD == rq:TAIL as we resubmit the
		 * request. See gen8_emit_breadcrumb() for
		 * where we prepare the padding after the
		 * end of the request.
		 */
		last->tail = last->wa_tail;
	}

	while (rb) {
		struct i915_priolist *p = to_priolist(rb);
		struct i915_request *rq, *rn;

		list_for_each_entry_safe(rq, rn, &p->requests, sched.link) {
			/*
			 * Can we combine this request with the current port?
			 * It has to be the same context/ringbuffer and not
			 * have any exceptions (e.g. GVT saying never to
			 * combine contexts).
			 *
			 * If we can combine the requests, we can execute both
			 * by updating the RING_TAIL to point to the end of the
			 * second request, and so we never need to tell the
			 * hardware about the first.
			 */
			if (last && !can_merge_ctx(rq->ctx, last->ctx)) {
				/*
				 * If we are on the second port and cannot
				 * combine this request with the last, then we
				 * are done.
				 */
				if (port == last_port) {
					__list_del_many(&p->requests,
							&rq->sched.link);
					goto done;
				}

				/*
				 * If GVT overrides us we only ever submit
				 * port[0], leaving port[1] empty. Note that we
				 * also have to be careful that we don't queue
				 * the same context (even though a different
				 * request) to the second port.
				 */
				if (ctx_single_port_submission(last->ctx) ||
				    ctx_single_port_submission(rq->ctx)) {
					__list_del_many(&p->requests,
							&rq->sched.link);
					goto done;
				}

				GEM_BUG_ON(last->ctx == rq->ctx);

				if (submit)
					port_assign(port, last);
				port++;

				GEM_BUG_ON(port_isset(port));
			}

			INIT_LIST_HEAD(&rq->sched.link);
			__i915_request_submit(rq);
			trace_i915_request_in(rq, port_index(port, execlists));
			last = rq;
			submit = true;
		}

		rb = rb_next(rb);
		rb_erase(&p->node, &execlists->queue);
		INIT_LIST_HEAD(&p->requests);
		if (p->priority != I915_PRIORITY_NORMAL)
			kmem_cache_free(engine->i915->priorities, p);
	}

done:
	/*
	 * Here be a bit of magic! Or sleight-of-hand, whichever you prefer.
	 *
	 * We choose queue_priority such that if we add a request of greater
	 * priority than this, we kick the submission tasklet to decide on
	 * the right order of submitting the requests to hardware. We must
	 * also be prepared to reorder requests as they are in-flight on the
	 * HW. We derive the queue_priority then as the first "hole" in
	 * the HW submission ports and if there are no available slots,
	 * the priority of the lowest executing request, i.e. last.
	 *
	 * When we do receive a higher priority request ready to run from the
	 * user, see queue_request(), the queue_priority is bumped to that
	 * request triggering preemption on the next dequeue (or subsequent
	 * interrupt for secondary ports).
	 */
	execlists->queue_priority =
		port != execlists->port ? rq_prio(last) : INT_MIN;

	execlists->first = rb;
	if (submit)
		port_assign(port, last);

	/* We must always keep the beast fed if we have work piled up */
	GEM_BUG_ON(execlists->first && !port_isset(execlists->port));

	/* Re-evaluate the executing context setup after each preemptive kick */
	if (last)
		execlists_user_begin(execlists, execlists->port);

	return submit;
}

static void execlists_dequeue(struct intel_engine_cs *engine)
{
	struct intel_engine_execlists * const execlists = &engine->execlists;
	unsigned long flags;
	bool submit;

	spin_lock_irqsave(&engine->timeline.lock, flags);
	submit = __execlists_dequeue(engine);
	spin_unlock_irqrestore(&engine->timeline.lock, flags);

	if (submit)
		execlists_submit_ports(engine);

	GEM_BUG_ON(port_isset(execlists->port) &&
		   !execlists_is_active(execlists, EXECLISTS_ACTIVE_USER));
}

void
execlists_cancel_port_requests(struct intel_engine_execlists * const execlists)
{
	struct execlist_port *port = execlists->port;
	unsigned int num_ports = execlists_num_ports(execlists);

	while (num_ports-- && port_isset(port)) {
		struct i915_request *rq = port_request(port);

		GEM_TRACE("%s:port%u global=%d (fence %llx:%d), (current %d)\n",
			  rq->engine->name,
			  (unsigned int)(port - execlists->port),
			  rq->global_seqno,
			  rq->fence.context, rq->fence.seqno,
			  intel_engine_get_seqno(rq->engine));

		GEM_BUG_ON(!execlists->active);
		execlists_context_schedule_out(rq,
					       i915_request_completed(rq) ?
					       INTEL_CONTEXT_SCHEDULE_OUT :
					       INTEL_CONTEXT_SCHEDULE_PREEMPTED);

		i915_request_put(rq);

		memset(port, 0, sizeof(*port));
		port++;
	}

	execlists_user_end(execlists);
}

static void clear_gtiir(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;
	int i;

	/*
	 * Clear any pending interrupt state.
	 *
	 * We do it twice out of paranoia that some of the IIR are
	 * double buffered, and so if we only reset it once there may
	 * still be an interrupt pending.
	 */
	if (INTEL_GEN(dev_priv) >= 11) {
		static const struct {
			u8 bank;
			u8 bit;
		} gen11_gtiir[] = {
			[RCS] = {0, GEN11_RCS0},
			[BCS] = {0, GEN11_BCS},
			[_VCS(0)] = {1, GEN11_VCS(0)},
			[_VCS(1)] = {1, GEN11_VCS(1)},
			[_VCS(2)] = {1, GEN11_VCS(2)},
			[_VCS(3)] = {1, GEN11_VCS(3)},
			[_VECS(0)] = {1, GEN11_VECS(0)},
			[_VECS(1)] = {1, GEN11_VECS(1)},
		};
		unsigned long irqflags;

		GEM_BUG_ON(engine->id >= ARRAY_SIZE(gen11_gtiir));

		spin_lock_irqsave(&dev_priv->irq_lock, irqflags);
		for (i = 0; i < 2; i++) {
			gen11_reset_one_iir(dev_priv,
					    gen11_gtiir[engine->id].bank,
					    gen11_gtiir[engine->id].bit);
		}
		spin_unlock_irqrestore(&dev_priv->irq_lock, irqflags);
	} else {
		static const u8 gtiir[] = {
			[RCS]  = 0,
			[BCS]  = 0,
			[VCS]  = 1,
			[VCS2] = 1,
			[VECS] = 3,
		};

		GEM_BUG_ON(engine->id >= ARRAY_SIZE(gtiir));

		for (i = 0; i < 2; i++) {
			I915_WRITE(GEN8_GT_IIR(gtiir[engine->id]),
				   engine->irq_keep_mask);
			POSTING_READ(GEN8_GT_IIR(gtiir[engine->id]));
		}
		GEM_BUG_ON(I915_READ(GEN8_GT_IIR(gtiir[engine->id])) &
			   engine->irq_keep_mask);
	}
}

static void reset_irq(struct intel_engine_cs *engine)
{
	/* Mark all CS interrupts as complete */
	smp_store_mb(engine->execlists.active, 0);
	synchronize_hardirq(engine->i915->drm.irq);

	clear_gtiir(engine);

	/*
	 * The port is checked prior to scheduling a tasklet, but
	 * just in case we have suspended the tasklet to do the
	 * wedging make sure that when it wakes, it decides there
	 * is no work to do by clearing the irq_posted bit.
	 */
	clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
}

static void execlists_cancel_requests(struct intel_engine_cs *engine)
{
	struct intel_engine_execlists * const execlists = &engine->execlists;
	struct i915_request *rq, *rn;
	struct rb_node *rb;
	unsigned long flags;

	GEM_TRACE("%s current %d\n",
		  engine->name, intel_engine_get_seqno(engine));

	/*
	 * Before we call engine->cancel_requests(), we should have exclusive
	 * access to the submission state. This is arranged for us by the
	 * caller disabling the interrupt generation, the tasklet and other
	 * threads that may then access the same state, giving us a free hand
	 * to reset state. However, we still need to let lockdep be aware that
	 * we know this state may be accessed in hardirq context, so we
	 * disable the irq around this manipulation and we want to keep
	 * the spinlock focused on its duties and not accidentally conflate
	 * coverage to the submission's irq state. (Similarly, although we
	 * shouldn't need to disable irq around the manipulation of the
	 * submission's irq state, we also wish to remind ourselves that
	 * it is irq state.)
	 */
	local_irq_save(flags);

	/* Cancel the requests on the HW and clear the ELSP tracker. */
	execlists_cancel_port_requests(execlists);
	reset_irq(engine);

	spin_lock(&engine->timeline.lock);

	/* Mark all executing requests as skipped. */
	list_for_each_entry(rq, &engine->timeline.requests, link) {
		GEM_BUG_ON(!rq->global_seqno);
		if (!i915_request_completed(rq))
			dma_fence_set_error(&rq->fence, -EIO);
	}

	/* Flush the queued requests to the timeline list (for retiring). */
	rb = execlists->first;
	while (rb) {
		struct i915_priolist *p = to_priolist(rb);

		list_for_each_entry_safe(rq, rn, &p->requests, sched.link) {
			INIT_LIST_HEAD(&rq->sched.link);

			dma_fence_set_error(&rq->fence, -EIO);
			__i915_request_submit(rq);
		}

		rb = rb_next(rb);
		rb_erase(&p->node, &execlists->queue);
		INIT_LIST_HEAD(&p->requests);
		if (p->priority != I915_PRIORITY_NORMAL)
			kmem_cache_free(engine->i915->priorities, p);
	}

	/* Remaining _unready_ requests will be nop'ed when submitted */

	execlists->queue_priority = INT_MIN;
	execlists->queue = RB_ROOT;
	execlists->first = NULL;
	GEM_BUG_ON(port_isset(execlists->port));

	spin_unlock(&engine->timeline.lock);

	local_irq_restore(flags);
}

/*
 * Check the unread Context Status Buffers and manage the submission of new
 * contexts to the ELSP accordingly.
 */
static void execlists_submission_tasklet(unsigned long data)
{
	struct intel_engine_cs * const engine = (struct intel_engine_cs *)data;
	struct intel_engine_execlists * const execlists = &engine->execlists;
	struct execlist_port *port = execlists->port;
	struct drm_i915_private *dev_priv = engine->i915;
	bool fw = false;

	/*
	 * We can skip acquiring intel_runtime_pm_get() here as it was taken
	 * on our behalf by the request (see i915_gem_mark_busy()) and it will
	 * not be relinquished until the device is idle (see
	 * i915_gem_idle_work_handler()). As a precaution, we make sure
	 * that all ELSP are drained i.e. we have processed the CSB,
	 * before allowing ourselves to idle and calling intel_runtime_pm_put().
	 */
	GEM_BUG_ON(!dev_priv->gt.awake);

	/*
	 * Prefer doing test_and_clear_bit() as a two stage operation to avoid
	 * imposing the cost of a locked atomic transaction when submitting a
	 * new request (outside of the context-switch interrupt).
	 */
	while (test_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted)) {
		/* The HWSP contains a (cacheable) mirror of the CSB */
		const u32 *buf =
			&engine->status_page.page_addr[I915_HWS_CSB_BUF0_INDEX];
		unsigned int head, tail;

		if (unlikely(execlists->csb_use_mmio)) {
			buf = (u32 * __force)
				(dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_BUF_LO(engine, 0)));
			execlists->csb_head = -1; /* force mmio read of CSB ptrs */
		}

		/* Clear before reading to catch new interrupts */
		clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted);
		smp_mb__after_atomic();

		if (unlikely(execlists->csb_head == -1)) { /* following a reset */
			if (!fw) {
				intel_uncore_forcewake_get(dev_priv,
							   execlists->fw_domains);
				fw = true;
			}

			head = readl(dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine)));
			tail = GEN8_CSB_WRITE_PTR(head);
			head = GEN8_CSB_READ_PTR(head);
			execlists->csb_head = head;
		} else {
			const int write_idx =
				intel_hws_csb_write_index(dev_priv) -
				I915_HWS_CSB_BUF0_INDEX;

			head = execlists->csb_head;
			tail = READ_ONCE(buf[write_idx]);
			rmb(); /* Hopefully paired with a wmb() in HW */
		}
		GEM_TRACE("%s cs-irq head=%d [%d%s], tail=%d [%d%s]\n",
			  engine->name,
			  head, GEN8_CSB_READ_PTR(readl(dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine)))), fw ? "" : "?",
			  tail, GEN8_CSB_WRITE_PTR(readl(dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine)))), fw ? "" : "?");

		while (head != tail) {
			struct i915_request *rq;
			unsigned int status;
			unsigned int count;

			if (++head == GEN8_CSB_ENTRIES)
				head = 0;

			/* We are flying near dragons again.
			 *
			 * We hold a reference to the request in execlist_port[]
			 * but no more than that. We are operating in softirq
			 * context and so cannot hold any mutex or sleep. That
			 * prevents us stopping the requests we are processing
			 * in port[] from being retired simultaneously (the
			 * breadcrumb will be complete before we see the
			 * context-switch). As we only hold the reference to the
			 * request, any pointer chasing underneath the request
			 * is subject to a potential use-after-free. Thus we
			 * store all of the bookkeeping within port[] as
			 * required, and avoid using unguarded pointers beneath
			 * request itself. The same applies to the atomic
			 * status notifier.
			 */

			status = READ_ONCE(buf[2 * head]); /* maybe mmio! */
			GEM_TRACE("%s csb[%d]: status=0x%08x:0x%08x, active=0x%x\n",
				  engine->name, head,
				  status, buf[2*head + 1],
				  execlists->active);

			if (status & (GEN8_CTX_STATUS_IDLE_ACTIVE |
				      GEN8_CTX_STATUS_PREEMPTED))
				execlists_set_active(execlists,
						     EXECLISTS_ACTIVE_HWACK);
			if (status & GEN8_CTX_STATUS_ACTIVE_IDLE)
				execlists_clear_active(execlists,
						       EXECLISTS_ACTIVE_HWACK);

			if (!(status & GEN8_CTX_STATUS_COMPLETED_MASK))
				continue;

			/* We should never get a COMPLETED | IDLE_ACTIVE! */
			GEM_BUG_ON(status & GEN8_CTX_STATUS_IDLE_ACTIVE);

			if (status & GEN8_CTX_STATUS_COMPLETE &&
			    buf[2*head + 1] == execlists->preempt_complete_status) {
				GEM_TRACE("%s preempt-idle\n", engine->name);

				execlists_cancel_port_requests(execlists);
				execlists_unwind_incomplete_requests(execlists);

				GEM_BUG_ON(!execlists_is_active(execlists,
								EXECLISTS_ACTIVE_PREEMPT));
				execlists_clear_active(execlists,
						       EXECLISTS_ACTIVE_PREEMPT);
				continue;
			}

			if (status & GEN8_CTX_STATUS_PREEMPTED &&
			    execlists_is_active(execlists,
						EXECLISTS_ACTIVE_PREEMPT))
				continue;

			GEM_BUG_ON(!execlists_is_active(execlists,
							EXECLISTS_ACTIVE_USER));

			rq = port_unpack(port, &count);
			GEM_TRACE("%s out[0]: ctx=%d.%d, global=%d (fence %llx:%d) (current %d), prio=%d\n",
				  engine->name,
				  port->context_id, count,
				  rq ? rq->global_seqno : 0,
				  rq ? rq->fence.context : 0,
				  rq ? rq->fence.seqno : 0,
				  intel_engine_get_seqno(engine),
				  rq ? rq_prio(rq) : 0);

			/* Check the context/desc id for this event matches */
			GEM_DEBUG_BUG_ON(buf[2 * head + 1] != port->context_id);

			GEM_BUG_ON(count == 0);
			if (--count == 0) {
				/*
				 * On the final event corresponding to the
				 * submission of this context, we expect either
				 * an element-switch event or a completion
				 * event (and on completion, the active-idle
				 * marker). No more preemptions, lite-restore
				 * or otherwise.
				 */
				GEM_BUG_ON(status & GEN8_CTX_STATUS_PREEMPTED);
				GEM_BUG_ON(port_isset(&port[1]) &&
					   !(status & GEN8_CTX_STATUS_ELEMENT_SWITCH));
				GEM_BUG_ON(!port_isset(&port[1]) &&
					   !(status & GEN8_CTX_STATUS_ACTIVE_IDLE));

				/*
				 * We rely on the hardware being strongly
				 * ordered, that the breadcrumb write is
				 * coherent (visible from the CPU) before the
				 * user interrupt and CSB is processed.
				 */
				GEM_BUG_ON(!i915_request_completed(rq));

				execlists_context_schedule_out(rq,
							       INTEL_CONTEXT_SCHEDULE_OUT);
				i915_request_put(rq);

				GEM_TRACE("%s completed ctx=%d\n",
					  engine->name, port->context_id);

				port = execlists_port_complete(execlists, port);
				if (port_isset(port))
					execlists_user_begin(execlists, port);
				else
					execlists_user_end(execlists);
			} else {
				port_set(port, port_pack(rq, count));
			}
		}

		if (head != execlists->csb_head) {
			execlists->csb_head = head;
			writel(_MASKED_FIELD(GEN8_CSB_READ_PTR_MASK, head << 8),
			       dev_priv->regs + i915_mmio_reg_offset(RING_CONTEXT_STATUS_PTR(engine)));
		}
	}

	if (!execlists_is_active(execlists, EXECLISTS_ACTIVE_PREEMPT))
		execlists_dequeue(engine);

	if (fw)
		intel_uncore_forcewake_put(dev_priv, execlists->fw_domains);

	/* If the engine is now idle, so should be the flag; and vice versa. */
	GEM_BUG_ON(execlists_is_active(&engine->execlists,
				       EXECLISTS_ACTIVE_USER) ==
		   !port_isset(engine->execlists.port));
}

static void queue_request(struct intel_engine_cs *engine,
			  struct i915_sched_node *node,
			  int prio)
{
	list_add_tail(&node->link,
		      &lookup_priolist(engine, prio)->requests);
}

static void __submit_queue(struct intel_engine_cs *engine, int prio)
{
	engine->execlists.queue_priority = prio;
	tasklet_hi_schedule(&engine->execlists.tasklet);
}

static void submit_queue(struct intel_engine_cs *engine, int prio)
{
	if (prio > engine->execlists.queue_priority)
		__submit_queue(engine, prio);
}

static void execlists_submit_request(struct i915_request *request)
{
	struct intel_engine_cs *engine = request->engine;
	unsigned long flags;

	/* Will be called from irq-context when using foreign fences. */
	spin_lock_irqsave(&engine->timeline.lock, flags);

	queue_request(engine, &request->sched, rq_prio(request));
	submit_queue(engine, rq_prio(request));

	GEM_BUG_ON(!engine->execlists.first);
	GEM_BUG_ON(list_empty(&request->sched.link));

	spin_unlock_irqrestore(&engine->timeline.lock, flags);
}

static struct i915_request *sched_to_request(struct i915_sched_node *node)
{
	return container_of(node, struct i915_request, sched);
}

static struct intel_engine_cs *
sched_lock_engine(struct i915_sched_node *node, struct intel_engine_cs *locked)
{
	struct intel_engine_cs *engine = sched_to_request(node)->engine;

	GEM_BUG_ON(!locked);

	if (engine != locked) {
		spin_unlock(&locked->timeline.lock);
		spin_lock(&engine->timeline.lock);
	}

	return engine;
}

static void execlists_schedule(struct i915_request *request,
			       const struct i915_sched_attr *attr)
{
	struct i915_priolist *uninitialized_var(pl);
	struct intel_engine_cs *engine, *last;
	struct i915_dependency *dep, *p;
	struct i915_dependency stack;
	const int prio = attr->priority;
	LIST_HEAD(dfs);

	GEM_BUG_ON(prio == I915_PRIORITY_INVALID);

	if (i915_request_completed(request))
		return;

	if (prio <= READ_ONCE(request->sched.attr.priority))
		return;

	/* Need BKL in order to use the temporary link inside i915_dependency */
	lockdep_assert_held(&request->i915->drm.struct_mutex);

	stack.signaler = &request->sched;
	list_add(&stack.dfs_link, &dfs);

	/*
	 * Recursively bump all dependent priorities to match the new request.
	 *
	 * A naive approach would be to use recursion:
	 * static void update_priorities(struct i915_sched_node *node, prio) {
	 *	list_for_each_entry(dep, &node->signalers_list, signal_link)
	 *		update_priorities(dep->signal, prio)
	 *	queue_request(node);
	 * }
	 * but that may have unlimited recursion depth and so runs a very
	 * real risk of overunning the kernel stack. Instead, we build
	 * a flat list of all dependencies starting with the current request.
	 * As we walk the list of dependencies, we add all of its dependencies
	 * to the end of the list (this may include an already visited
	 * request) and continue to walk onwards onto the new dependencies. The
	 * end result is a topological list of requests in reverse order, the
	 * last element in the list is the request we must execute first.
	 */
	list_for_each_entry(dep, &dfs, dfs_link) {
		struct i915_sched_node *node = dep->signaler;

		/*
		 * Within an engine, there can be no cycle, but we may
		 * refer to the same dependency chain multiple times
		 * (redundant dependencies are not eliminated) and across
		 * engines.
		 */
		list_for_each_entry(p, &node->signalers_list, signal_link) {
			GEM_BUG_ON(p == dep); /* no cycles! */

			if (i915_sched_node_signaled(p->signaler))
				continue;

			GEM_BUG_ON(p->signaler->attr.priority < node->attr.priority);
			if (prio > READ_ONCE(p->signaler->attr.priority))
				list_move_tail(&p->dfs_link, &dfs);
		}
	}

	/*
	 * If we didn't need to bump any existing priorities, and we haven't
	 * yet submitted this request (i.e. there is no potential race with
	 * execlists_submit_request()), we can set our own priority and skip
	 * acquiring the engine locks.
	 */
	if (request->sched.attr.priority == I915_PRIORITY_INVALID) {
		GEM_BUG_ON(!list_empty(&request->sched.link));
		request->sched.attr = *attr;
		if (stack.dfs_link.next == stack.dfs_link.prev)
			return;
		__list_del_entry(&stack.dfs_link);
	}

	last = NULL;
	engine = request->engine;
	spin_lock_irq(&engine->timeline.lock);

	/* Fifo and depth-first replacement ensure our deps execute before us */
	list_for_each_entry_safe_reverse(dep, p, &dfs, dfs_link) {
		struct i915_sched_node *node = dep->signaler;

		INIT_LIST_HEAD(&dep->dfs_link);

		engine = sched_lock_engine(node, engine);

		if (prio <= node->attr.priority)
			continue;

		node->attr.priority = prio;
		if (!list_empty(&node->link)) {
			if (last != engine) {
				pl = lookup_priolist(engine, prio);
				last = engine;
			}
			GEM_BUG_ON(pl->priority != prio);
			list_move_tail(&node->link, &pl->requests);
		}

		if (prio > engine->execlists.queue_priority &&
		    i915_sw_fence_done(&sched_to_request(node)->submit))
			__submit_queue(engine, prio);
	}

	spin_unlock_irq(&engine->timeline.lock);
}

static int __context_pin(struct i915_gem_context *ctx, struct i915_vma *vma)
{
	unsigned int flags;
	int err;

	/*
	 * Clear this page out of any CPU caches for coherent swap-in/out.
	 * We only want to do this on the first bind so that we do not stall
	 * on an active context (which by nature is already on the GPU).
	 */
	if (!(vma->flags & I915_VMA_GLOBAL_BIND)) {
		err = i915_gem_object_set_to_gtt_domain(vma->obj, true);
		if (err)
			return err;
	}

	flags = PIN_GLOBAL | PIN_HIGH;
	if (ctx->ggtt_offset_bias)
		flags |= PIN_OFFSET_BIAS | ctx->ggtt_offset_bias;

	return i915_vma_pin(vma, 0, GEN8_LR_CONTEXT_ALIGN, flags);
}

static struct intel_ring *
execlists_context_pin(struct intel_engine_cs *engine,
		      struct i915_gem_context *ctx)
{
	struct intel_context *ce = to_intel_context(ctx, engine);
	void *vaddr;
	int ret;

	lockdep_assert_held(&ctx->i915->drm.struct_mutex);

	if (likely(ce->pin_count++))
		goto out;
	GEM_BUG_ON(!ce->pin_count); /* no overflow please! */

	ret = execlists_context_deferred_alloc(ctx, engine);
	if (ret)
		goto err;
	GEM_BUG_ON(!ce->state);

	ret = __context_pin(ctx, ce->state);
	if (ret)
		goto err;

	vaddr = i915_gem_object_pin_map(ce->state->obj, I915_MAP_WB);
	if (IS_ERR(vaddr)) {
		ret = PTR_ERR(vaddr);
		goto unpin_vma;
	}

	ret = intel_ring_pin(ce->ring, ctx->i915, ctx->ggtt_offset_bias);
	if (ret)
		goto unpin_map;

	intel_lr_context_descriptor_update(ctx, engine);

	ce->lrc_reg_state = vaddr + LRC_STATE_PN * PAGE_SIZE;
	ce->lrc_reg_state[CTX_RING_BUFFER_START+1] =
		i915_ggtt_offset(ce->ring->vma);
	ce->lrc_reg_state[CTX_RING_HEAD+1] = ce->ring->head;

	ce->state->obj->pin_global++;
	i915_gem_context_get(ctx);
out:
	return ce->ring;

unpin_map:
	i915_gem_object_unpin_map(ce->state->obj);
unpin_vma:
	__i915_vma_unpin(ce->state);
err:
	ce->pin_count = 0;
	return ERR_PTR(ret);
}

static void execlists_context_unpin(struct intel_engine_cs *engine,
				    struct i915_gem_context *ctx)
{
	struct intel_context *ce = to_intel_context(ctx, engine);

	lockdep_assert_held(&ctx->i915->drm.struct_mutex);
	GEM_BUG_ON(ce->pin_count == 0);

	if (--ce->pin_count)
		return;

	intel_ring_unpin(ce->ring);

	ce->state->obj->pin_global--;
	i915_gem_object_unpin_map(ce->state->obj);
	i915_vma_unpin(ce->state);

	i915_gem_context_put(ctx);
}

static int execlists_request_alloc(struct i915_request *request)
{
	struct intel_context *ce =
		to_intel_context(request->ctx, request->engine);
	int ret;

	GEM_BUG_ON(!ce->pin_count);

	/* Flush enough space to reduce the likelihood of waiting after
	 * we start building the request - in which case we will just
	 * have to repeat work.
	 */
	request->reserved_space += EXECLISTS_REQUEST_SIZE;

	ret = intel_ring_wait_for_space(request->ring, request->reserved_space);
	if (ret)
		return ret;

	/* Note that after this point, we have committed to using
	 * this request as it is being used to both track the
	 * state of engine initialisation and liveness of the
	 * golden renderstate above. Think twice before you try
	 * to cancel/unwind this request now.
	 */

	request->reserved_space -= EXECLISTS_REQUEST_SIZE;
	return 0;
}

/*
 * In this WA we need to set GEN8_L3SQCREG4[21:21] and reset it after
 * PIPE_CONTROL instruction. This is required for the flush to happen correctly
 * but there is a slight complication as this is applied in WA batch where the
 * values are only initialized once so we cannot take register value at the
 * beginning and reuse it further; hence we save its value to memory, upload a
 * constant value with bit21 set and then we restore it back with the saved value.
 * To simplify the WA, a constant value is formed by using the default value
 * of this register. This shouldn't be a problem because we are only modifying
 * it for a short period and this batch in non-premptible. We can ofcourse
 * use additional instructions that read the actual value of the register
 * at that time and set our bit of interest but it makes the WA complicated.
 *
 * This WA is also required for Gen9 so extracting as a function avoids
 * code duplication.
 */
static u32 *
gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch)
{
	*batch++ = MI_STORE_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT;
	*batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4);
	*batch++ = i915_ggtt_offset(engine->scratch) + 256;
	*batch++ = 0;

	*batch++ = MI_LOAD_REGISTER_IMM(1);
	*batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4);
	*batch++ = 0x40400000 | GEN8_LQSC_FLUSH_COHERENT_LINES;

	batch = gen8_emit_pipe_control(batch,
				       PIPE_CONTROL_CS_STALL |
				       PIPE_CONTROL_DC_FLUSH_ENABLE,
				       0);

	*batch++ = MI_LOAD_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT;
	*batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4);
	*batch++ = i915_ggtt_offset(engine->scratch) + 256;
	*batch++ = 0;

	return batch;
}

/*
 * Typically we only have one indirect_ctx and per_ctx batch buffer which are
 * initialized at the beginning and shared across all contexts but this field
 * helps us to have multiple batches at different offsets and select them based
 * on a criteria. At the moment this batch always start at the beginning of the page
 * and at this point we don't have multiple wa_ctx batch buffers.
 *
 * The number of WA applied are not known at the beginning; we use this field
 * to return the no of DWORDS written.
 *
 * It is to be noted that this batch does not contain MI_BATCH_BUFFER_END
 * so it adds NOOPs as padding to make it cacheline aligned.
 * MI_BATCH_BUFFER_END will be added to perctx batch and both of them together
 * makes a complete batch buffer.
 */
static u32 *gen8_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)
{
	/* WaDisableCtxRestoreArbitration:bdw,chv */
	*batch++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;

	/* WaFlushCoherentL3CacheLinesAtContextSwitch:bdw */
	if (IS_BROADWELL(engine->i915))
		batch = gen8_emit_flush_coherentl3_wa(engine, batch);

	/* WaClearSlmSpaceAtContextSwitch:bdw,chv */
	/* Actual scratch location is at 128 bytes offset */
	batch = gen8_emit_pipe_control(batch,
				       PIPE_CONTROL_FLUSH_L3 |
				       PIPE_CONTROL_GLOBAL_GTT_IVB |
				       PIPE_CONTROL_CS_STALL |
				       PIPE_CONTROL_QW_WRITE,
				       i915_ggtt_offset(engine->scratch) +
				       2 * CACHELINE_BYTES);

	*batch++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;

	/* Pad to end of cacheline */
	while ((unsigned long)batch % CACHELINE_BYTES)
		*batch++ = MI_NOOP;

	/*
	 * MI_BATCH_BUFFER_END is not required in Indirect ctx BB because
	 * execution depends on the length specified in terms of cache lines
	 * in the register CTX_RCS_INDIRECT_CTX
	 */

	return batch;
}

static u32 *gen9_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)
{
	*batch++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;

	/* WaFlushCoherentL3CacheLinesAtContextSwitch:skl,bxt,glk */
	batch = gen8_emit_flush_coherentl3_wa(engine, batch);

	/* WaDisableGatherAtSetShaderCommonSlice:skl,bxt,kbl,glk */
	*batch++ = MI_LOAD_REGISTER_IMM(1);
	*batch++ = i915_mmio_reg_offset(COMMON_SLICE_CHICKEN2);
	*batch++ = _MASKED_BIT_DISABLE(
			GEN9_DISABLE_GATHER_AT_SET_SHADER_COMMON_SLICE);
	*batch++ = MI_NOOP;

	/* WaClearSlmSpaceAtContextSwitch:kbl */
	/* Actual scratch location is at 128 bytes offset */
	if (IS_KBL_REVID(engine->i915, 0, KBL_REVID_A0)) {
		batch = gen8_emit_pipe_control(batch,
					       PIPE_CONTROL_FLUSH_L3 |
					       PIPE_CONTROL_GLOBAL_GTT_IVB |
					       PIPE_CONTROL_CS_STALL |
					       PIPE_CONTROL_QW_WRITE,
					       i915_ggtt_offset(engine->scratch)
					       + 2 * CACHELINE_BYTES);
	}

	/* WaMediaPoolStateCmdInWABB:bxt,glk */
	if (HAS_POOLED_EU(engine->i915)) {
		/*
		 * EU pool configuration is setup along with golden context
		 * during context initialization. This value depends on
		 * device type (2x6 or 3x6) and needs to be updated based
		 * on which subslice is disabled especially for 2x6
		 * devices, however it is safe to load default
		 * configuration of 3x6 device instead of masking off
		 * corresponding bits because HW ignores bits of a disabled
		 * subslice and drops down to appropriate config. Please
		 * see render_state_setup() in i915_gem_render_state.c for
		 * possible configurations, to avoid duplication they are
		 * not shown here again.
		 */
		*batch++ = GEN9_MEDIA_POOL_STATE;
		*batch++ = GEN9_MEDIA_POOL_ENABLE;
		*batch++ = 0x00777000;
		*batch++ = 0;
		*batch++ = 0;
		*batch++ = 0;
	}

	*batch++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;

	/* Pad to end of cacheline */
	while ((unsigned long)batch % CACHELINE_BYTES)
		*batch++ = MI_NOOP;

	return batch;
}

static u32 *
gen10_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)
{
	int i;

	/*
	 * WaPipeControlBefore3DStateSamplePattern: cnl
	 *
	 * Ensure the engine is idle prior to programming a
	 * 3DSTATE_SAMPLE_PATTERN during a context restore.
	 */
	batch = gen8_emit_pipe_control(batch,
				       PIPE_CONTROL_CS_STALL,
				       0);
	/*
	 * WaPipeControlBefore3DStateSamplePattern says we need 4 dwords for
	 * the PIPE_CONTROL followed by 12 dwords of 0x0, so 16 dwords in
	 * total. However, a PIPE_CONTROL is 6 dwords long, not 4, which is
	 * confusing. Since gen8_emit_pipe_control() already advances the
	 * batch by 6 dwords, we advance the other 10 here, completing a
	 * cacheline. It's not clear if the workaround requires this padding
	 * before other commands, or if it's just the regular padding we would
	 * already have for the workaround bb, so leave it here for now.
	 */
	for (i = 0; i < 10; i++)
		*batch++ = MI_NOOP;

	/* Pad to end of cacheline */
	while ((unsigned long)batch % CACHELINE_BYTES)
		*batch++ = MI_NOOP;

	return batch;
}

#define CTX_WA_BB_OBJ_SIZE (PAGE_SIZE)

static int lrc_setup_wa_ctx(struct intel_engine_cs *engine)
{
	struct drm_i915_gem_object *obj;
	struct i915_vma *vma;
	int err;

	obj = i915_gem_object_create(engine->i915, CTX_WA_BB_OBJ_SIZE);
	if (IS_ERR(obj))
		return PTR_ERR(obj);

	vma = i915_vma_instance(obj, &engine->i915->ggtt.base, NULL);
	if (IS_ERR(vma)) {
		err = PTR_ERR(vma);
		goto err;
	}

	err = i915_vma_pin(vma, 0, PAGE_SIZE, PIN_GLOBAL | PIN_HIGH);
	if (err)
		goto err;

	engine->wa_ctx.vma = vma;
	return 0;

err:
	i915_gem_object_put(obj);
	return err;
}

static void lrc_destroy_wa_ctx(struct intel_engine_cs *engine)
{
	i915_vma_unpin_and_release(&engine->wa_ctx.vma);
}

typedef u32 *(*wa_bb_func_t)(struct intel_engine_cs *engine, u32 *batch);

static int intel_init_workaround_bb(struct intel_engine_cs *engine)
{
	struct i915_ctx_workarounds *wa_ctx = &engine->wa_ctx;
	struct i915_wa_ctx_bb *wa_bb[2] = { &wa_ctx->indirect_ctx,
					    &wa_ctx->per_ctx };
	wa_bb_func_t wa_bb_fn[2];
	struct page *page;
	void *batch, *batch_ptr;
	unsigned int i;
	int ret;

	if (GEM_WARN_ON(engine->id != RCS))
		return -EINVAL;

	switch (INTEL_GEN(engine->i915)) {
	case 11:
		return 0;
	case 10:
		wa_bb_fn[0] = gen10_init_indirectctx_bb;
		wa_bb_fn[1] = NULL;
		break;
	case 9:
		wa_bb_fn[0] = gen9_init_indirectctx_bb;
		wa_bb_fn[1] = NULL;
		break;
	case 8:
		wa_bb_fn[0] = gen8_init_indirectctx_bb;
		wa_bb_fn[1] = NULL;
		break;
	default:
		MISSING_CASE(INTEL_GEN(engine->i915));
		return 0;
	}

	ret = lrc_setup_wa_ctx(engine);
	if (ret) {
		DRM_DEBUG_DRIVER("Failed to setup context WA page: %d\n", ret);
		return ret;
	}

	page = i915_gem_object_get_dirty_page(wa_ctx->vma->obj, 0);
	batch = batch_ptr = kmap_atomic(page);

	/*
	 * Emit the two workaround batch buffers, recording the offset from the
	 * start of the workaround batch buffer object for each and their
	 * respective sizes.
	 */
	for (i = 0; i < ARRAY_SIZE(wa_bb_fn); i++) {
		wa_bb[i]->offset = batch_ptr - batch;
		if (GEM_WARN_ON(!IS_ALIGNED(wa_bb[i]->offset,
					    CACHELINE_BYTES))) {
			ret = -EINVAL;
			break;
		}
		if (wa_bb_fn[i])
			batch_ptr = wa_bb_fn[i](engine, batch_ptr);
		wa_bb[i]->size = batch_ptr - (batch + wa_bb[i]->offset);
	}

	BUG_ON(batch_ptr - batch > CTX_WA_BB_OBJ_SIZE);

	kunmap_atomic(batch);
	if (ret)
		lrc_destroy_wa_ctx(engine);

	return ret;
}

static void enable_execlists(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;

	I915_WRITE(RING_HWSTAM(engine->mmio_base), 0xffffffff);

	/*
	 * Make sure we're not enabling the new 12-deep CSB
	 * FIFO as that requires a slightly updated handling
	 * in the ctx switch irq. Since we're currently only
	 * using only 2 elements of the enhanced execlists the
	 * deeper FIFO it's not needed and it's not worth adding
	 * more statements to the irq handler to support it.
	 */
	if (INTEL_GEN(dev_priv) >= 11)
		I915_WRITE(RING_MODE_GEN7(engine),
			   _MASKED_BIT_DISABLE(GEN11_GFX_DISABLE_LEGACY_MODE));
	else
		I915_WRITE(RING_MODE_GEN7(engine),
			   _MASKED_BIT_ENABLE(GFX_RUN_LIST_ENABLE));

	I915_WRITE(RING_HWS_PGA(engine->mmio_base),
		   engine->status_page.ggtt_offset);
	POSTING_READ(RING_HWS_PGA(engine->mmio_base));

	/* Following the reset, we need to reload the CSB read/write pointers */
	engine->execlists.csb_head = -1;
}

static int gen8_init_common_ring(struct intel_engine_cs *engine)
{
	struct intel_engine_execlists * const execlists = &engine->execlists;
	int ret;

	ret = intel_mocs_init_engine(engine);
	if (ret)
		return ret;

	intel_engine_reset_breadcrumbs(engine);
	intel_engine_init_hangcheck(engine);

	enable_execlists(engine);

	/* After a GPU reset, we may have requests to replay */
	if (execlists->first)
		tasklet_schedule(&execlists->tasklet);

	return 0;
}

static int gen8_init_render_ring(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;
	int ret;

	ret = gen8_init_common_ring(engine);
	if (ret)
		return ret;

	intel_whitelist_workarounds_apply(engine);

	/* We need to disable the AsyncFlip performance optimisations in order
	 * to use MI_WAIT_FOR_EVENT within the CS. It should already be
	 * programmed to '1' on all products.
	 *
	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv,bdw,chv
	 */
	I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));

	I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING));

	return 0;
}

static int gen9_init_render_ring(struct intel_engine_cs *engine)
{
	int ret;

	ret = gen8_init_common_ring(engine);
	if (ret)
		return ret;

	intel_whitelist_workarounds_apply(engine);

	return 0;
}

static void reset_common_ring(struct intel_engine_cs *engine,
			      struct i915_request *request)
{
	struct intel_engine_execlists * const execlists = &engine->execlists;
	unsigned long flags;
	u32 *regs;

	GEM_TRACE("%s request global=%x, current=%d\n",
		  engine->name, request ? request->global_seqno : 0,
		  intel_engine_get_seqno(engine));

	/* See execlists_cancel_requests() for the irq/spinlock split. */
	local_irq_save(flags);

	/*
	 * Catch up with any missed context-switch interrupts.
	 *
	 * Ideally we would just read the remaining CSB entries now that we
	 * know the gpu is idle. However, the CSB registers are sometimes^W
	 * often trashed across a GPU reset! Instead we have to rely on
	 * guessing the missed context-switch events by looking at what
	 * requests were completed.
	 */
	execlists_cancel_port_requests(execlists);
	reset_irq(engine);

	/* Push back any incomplete requests for replay after the reset. */
	spin_lock(&engine->timeline.lock);
	__unwind_incomplete_requests(engine);
	spin_unlock(&engine->timeline.lock);

	local_irq_restore(flags);

	/*
	 * If the request was innocent, we leave the request in the ELSP
	 * and will try to replay it on restarting. The context image may
	 * have been corrupted by the reset, in which case we may have
	 * to service a new GPU hang, but more likely we can continue on
	 * without impact.
	 *
	 * If the request was guilty, we presume the context is corrupt
	 * and have to at least restore the RING register in the context
	 * image back to the expected values to skip over the guilty request.
	 */
	if (!request || request->fence.error != -EIO)
		return;

	/*
	 * We want a simple context + ring to execute the breadcrumb update.
	 * We cannot rely on the context being intact across the GPU hang,
	 * so clear it and rebuild just what we need for the breadcrumb.
	 * All pending requests for this context will be zapped, and any
	 * future request will be after userspace has had the opportunity
	 * to recreate its own state.
	 */
	regs = to_intel_context(request->ctx, engine)->lrc_reg_state;
	if (engine->default_state) {
		void *defaults;

		defaults = i915_gem_object_pin_map(engine->default_state,
						   I915_MAP_WB);
		if (!IS_ERR(defaults)) {
			memcpy(regs, /* skip restoring the vanilla PPHWSP */
			       defaults + LRC_STATE_PN * PAGE_SIZE,
			       engine->context_size - PAGE_SIZE);
			i915_gem_object_unpin_map(engine->default_state);
		}
	}
	execlists_init_reg_state(regs, request->ctx, engine, request->ring);

	/* Move the RING_HEAD onto the breadcrumb, past the hanging batch */
	regs[CTX_RING_BUFFER_START + 1] = i915_ggtt_offset(request->ring->vma);
	regs[CTX_RING_HEAD + 1] = request->postfix;

	request->ring->head = request->postfix;
	intel_ring_update_space(request->ring);

	/* Reset WaIdleLiteRestore:bdw,skl as well */
	unwind_wa_tail(request);
}

static int intel_logical_ring_emit_pdps(struct i915_request *rq)
{
	struct i915_hw_ppgtt *ppgtt = rq->ctx->ppgtt;
	struct intel_engine_cs *engine = rq->engine;
	const int num_lri_cmds = GEN8_3LVL_PDPES * 2;
	u32 *cs;
	int i;

	cs = intel_ring_begin(rq, num_lri_cmds * 2 + 2);
	if (IS_ERR(cs))
		return PTR_ERR(cs);

	*cs++ = MI_LOAD_REGISTER_IMM(num_lri_cmds);
	for (i = GEN8_3LVL_PDPES - 1; i >= 0; i--) {
		const dma_addr_t pd_daddr = i915_page_dir_dma_addr(ppgtt, i);

		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_UDW(engine, i));
		*cs++ = upper_32_bits(pd_daddr);
		*cs++ = i915_mmio_reg_offset(GEN8_RING_PDP_LDW(engine, i));
		*cs++ = lower_32_bits(pd_daddr);
	}

	*cs++ = MI_NOOP;
	intel_ring_advance(rq, cs);

	return 0;
}

static int gen8_emit_bb_start(struct i915_request *rq,
			      u64 offset, u32 len,
			      const unsigned int flags)
{
	u32 *cs;
	int ret;

	/* Don't rely in hw updating PDPs, specially in lite-restore.
	 * Ideally, we should set Force PD Restore in ctx descriptor,
	 * but we can't. Force Restore would be a second option, but
	 * it is unsafe in case of lite-restore (because the ctx is
	 * not idle). PML4 is allocated during ppgtt init so this is
	 * not needed in 48-bit.*/
	if (rq->ctx->ppgtt &&
	    (intel_engine_flag(rq->engine) & rq->ctx->ppgtt->pd_dirty_rings) &&
	    !i915_vm_is_48bit(&rq->ctx->ppgtt->base) &&
	    !intel_vgpu_active(rq->i915)) {
		ret = intel_logical_ring_emit_pdps(rq);
		if (ret)
			return ret;

		rq->ctx->ppgtt->pd_dirty_rings &= ~intel_engine_flag(rq->engine);
	}

	cs = intel_ring_begin(rq, 6);
	if (IS_ERR(cs))
		return PTR_ERR(cs);

	/*
	 * WaDisableCtxRestoreArbitration:bdw,chv
	 *
	 * We don't need to perform MI_ARB_ENABLE as often as we do (in
	 * particular all the gen that do not need the w/a at all!), if we
	 * took care to make sure that on every switch into this context
	 * (both ordinary and for preemption) that arbitrartion was enabled
	 * we would be fine. However, there doesn't seem to be a downside to
	 * being paranoid and making sure it is set before each batch and
	 * every context-switch.
	 *
	 * Note that if we fail to enable arbitration before the request
	 * is complete, then we do not see the context-switch interrupt and
	 * the engine hangs (with RING_HEAD == RING_TAIL).
	 *
	 * That satisfies both the GPGPU w/a and our heavy-handed paranoia.
	 */
	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;

	/* FIXME(BDW): Address space and security selectors. */
	*cs++ = MI_BATCH_BUFFER_START_GEN8 |
		(flags & I915_DISPATCH_SECURE ? 0 : BIT(8)) |
		(flags & I915_DISPATCH_RS ? MI_BATCH_RESOURCE_STREAMER : 0);
	*cs++ = lower_32_bits(offset);
	*cs++ = upper_32_bits(offset);

	*cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE;
	*cs++ = MI_NOOP;
	intel_ring_advance(rq, cs);

	return 0;
}

static void gen8_logical_ring_enable_irq(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;
	I915_WRITE_IMR(engine,
		       ~(engine->irq_enable_mask | engine->irq_keep_mask));
	POSTING_READ_FW(RING_IMR(engine->mmio_base));
}

static void gen8_logical_ring_disable_irq(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;
	I915_WRITE_IMR(engine, ~engine->irq_keep_mask);
}

static int gen8_emit_flush(struct i915_request *request, u32 mode)
{
	u32 cmd, *cs;

	cs = intel_ring_begin(request, 4);
	if (IS_ERR(cs))
		return PTR_ERR(cs);

	cmd = MI_FLUSH_DW + 1;

	/* We always require a command barrier so that subsequent
	 * commands, such as breadcrumb interrupts, are strictly ordered
	 * wrt the contents of the write cache being flushed to memory
	 * (and thus being coherent from the CPU).
	 */
	cmd |= MI_FLUSH_DW_STORE_INDEX | MI_FLUSH_DW_OP_STOREDW;

	if (mode & EMIT_INVALIDATE) {
		cmd |= MI_INVALIDATE_TLB;
		if (request->engine->id == VCS)
			cmd |= MI_INVALIDATE_BSD;
	}

	*cs++ = cmd;
	*cs++ = I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT;
	*cs++ = 0; /* upper addr */
	*cs++ = 0; /* value */
	intel_ring_advance(request, cs);

	return 0;
}

static int gen8_emit_flush_render(struct i915_request *request,
				  u32 mode)
{
	struct intel_engine_cs *engine = request->engine;
	u32 scratch_addr =
		i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES;
	bool vf_flush_wa = false, dc_flush_wa = false;
	u32 *cs, flags = 0;
	int len;

	flags |= PIPE_CONTROL_CS_STALL;

	if (mode & EMIT_FLUSH) {
		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
		flags |= PIPE_CONTROL_DC_FLUSH_ENABLE;
		flags |= PIPE_CONTROL_FLUSH_ENABLE;
	}

	if (mode & EMIT_INVALIDATE) {
		flags |= PIPE_CONTROL_TLB_INVALIDATE;
		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
		flags |= PIPE_CONTROL_VF_CACHE_INVALIDATE;
		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
		flags |= PIPE_CONTROL_QW_WRITE;
		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;

		/*
		 * On GEN9: before VF_CACHE_INVALIDATE we need to emit a NULL
		 * pipe control.
		 */
		if (IS_GEN9(request->i915))
			vf_flush_wa = true;

		/* WaForGAMHang:kbl */
		if (IS_KBL_REVID(request->i915, 0, KBL_REVID_B0))
			dc_flush_wa = true;
	}

	len = 6;

	if (vf_flush_wa)
		len += 6;

	if (dc_flush_wa)
		len += 12;

	cs = intel_ring_begin(request, len);
	if (IS_ERR(cs))
		return PTR_ERR(cs);

	if (vf_flush_wa)
		cs = gen8_emit_pipe_control(cs, 0, 0);

	if (dc_flush_wa)
		cs = gen8_emit_pipe_control(cs, PIPE_CONTROL_DC_FLUSH_ENABLE,
					    0);

	cs = gen8_emit_pipe_control(cs, flags, scratch_addr);

	if (dc_flush_wa)
		cs = gen8_emit_pipe_control(cs, PIPE_CONTROL_CS_STALL, 0);

	intel_ring_advance(request, cs);

	return 0;
}

/*
 * Reserve space for 2 NOOPs at the end of each request to be
 * used as a workaround for not being allowed to do lite
 * restore with HEAD==TAIL (WaIdleLiteRestore).
 */
static void gen8_emit_wa_tail(struct i915_request *request, u32 *cs)
{
	/* Ensure there's always at least one preemption point per-request. */
	*cs++ = MI_ARB_CHECK;
	*cs++ = MI_NOOP;
	request->wa_tail = intel_ring_offset(request, cs);
}

static void gen8_emit_breadcrumb(struct i915_request *request, u32 *cs)
{
	/* w/a: bit 5 needs to be zero for MI_FLUSH_DW address. */
	BUILD_BUG_ON(I915_GEM_HWS_INDEX_ADDR & (1 << 5));

	cs = gen8_emit_ggtt_write(cs, request->global_seqno,
				  intel_hws_seqno_address(request->engine));
	*cs++ = MI_USER_INTERRUPT;
	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
	request->tail = intel_ring_offset(request, cs);
	assert_ring_tail_valid(request->ring, request->tail);

	gen8_emit_wa_tail(request, cs);
}
static const int gen8_emit_breadcrumb_sz = 6 + WA_TAIL_DWORDS;

static void gen8_emit_breadcrumb_rcs(struct i915_request *request, u32 *cs)
{
	/* We're using qword write, seqno should be aligned to 8 bytes. */
	BUILD_BUG_ON(I915_GEM_HWS_INDEX & 1);

	cs = gen8_emit_ggtt_write_rcs(cs, request->global_seqno,
				      intel_hws_seqno_address(request->engine));
	*cs++ = MI_USER_INTERRUPT;
	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
	request->tail = intel_ring_offset(request, cs);
	assert_ring_tail_valid(request->ring, request->tail);

	gen8_emit_wa_tail(request, cs);
}
static const int gen8_emit_breadcrumb_rcs_sz = 8 + WA_TAIL_DWORDS;

static int gen8_init_rcs_context(struct i915_request *rq)
{
	int ret;

	ret = intel_ctx_workarounds_emit(rq);
	if (ret)
		return ret;

	ret = intel_rcs_context_init_mocs(rq);
	/*
	 * Failing to program the MOCS is non-fatal.The system will not
	 * run at peak performance. So generate an error and carry on.
	 */
	if (ret)
		DRM_ERROR("MOCS failed to program: expect performance issues.\n");

	return i915_gem_render_state_emit(rq);
}

/**
 * intel_logical_ring_cleanup() - deallocate the Engine Command Streamer
 * @engine: Engine Command Streamer.
 */
void intel_logical_ring_cleanup(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv;

	/*
	 * Tasklet cannot be active at this point due intel_mark_active/idle
	 * so this is just for documentation.
	 */
	if (WARN_ON(test_bit(TASKLET_STATE_SCHED,
			     &engine->execlists.tasklet.state)))
		tasklet_kill(&engine->execlists.tasklet);

	dev_priv = engine->i915;

	if (engine->buffer) {
		WARN_ON((I915_READ_MODE(engine) & MODE_IDLE) == 0);
	}

	if (engine->cleanup)
		engine->cleanup(engine);

	intel_engine_cleanup_common(engine);

	lrc_destroy_wa_ctx(engine);

	engine->i915 = NULL;
	dev_priv->engine[engine->id] = NULL;
	kfree(engine);
}

static void execlists_set_default_submission(struct intel_engine_cs *engine)
{
	engine->submit_request = execlists_submit_request;
	engine->cancel_requests = execlists_cancel_requests;
	engine->schedule = execlists_schedule;
	engine->execlists.tasklet.func = execlists_submission_tasklet;

	engine->park = NULL;
	engine->unpark = NULL;

	engine->flags |= I915_ENGINE_SUPPORTS_STATS;
	if (engine->i915->preempt_context)
		engine->flags |= I915_ENGINE_HAS_PREEMPTION;

	engine->i915->caps.scheduler =
		I915_SCHEDULER_CAP_ENABLED |
		I915_SCHEDULER_CAP_PRIORITY;
	if (intel_engine_has_preemption(engine))
		engine->i915->caps.scheduler |= I915_SCHEDULER_CAP_PREEMPTION;
}

static void
logical_ring_default_vfuncs(struct intel_engine_cs *engine)
{
	/* Default vfuncs which can be overriden by each engine. */
	engine->init_hw = gen8_init_common_ring;
	engine->reset_hw = reset_common_ring;

	engine->context_pin = execlists_context_pin;
	engine->context_unpin = execlists_context_unpin;

	engine->request_alloc = execlists_request_alloc;

	engine->emit_flush = gen8_emit_flush;
	engine->emit_breadcrumb = gen8_emit_breadcrumb;
	engine->emit_breadcrumb_sz = gen8_emit_breadcrumb_sz;

	engine->set_default_submission = execlists_set_default_submission;

	if (INTEL_GEN(engine->i915) < 11) {
		engine->irq_enable = gen8_logical_ring_enable_irq;
		engine->irq_disable = gen8_logical_ring_disable_irq;
	} else {
		/*
		 * TODO: On Gen11 interrupt masks need to be clear
		 * to allow C6 entry. Keep interrupts enabled at
		 * and take the hit of generating extra interrupts
		 * until a more refined solution exists.
		 */
	}
	engine->emit_bb_start = gen8_emit_bb_start;
}

static inline void
logical_ring_default_irqs(struct intel_engine_cs *engine)
{
	unsigned int shift = 0;

	if (INTEL_GEN(engine->i915) < 11) {
		const u8 irq_shifts[] = {
			[RCS]  = GEN8_RCS_IRQ_SHIFT,
			[BCS]  = GEN8_BCS_IRQ_SHIFT,
			[VCS]  = GEN8_VCS1_IRQ_SHIFT,
			[VCS2] = GEN8_VCS2_IRQ_SHIFT,
			[VECS] = GEN8_VECS_IRQ_SHIFT,
		};

		shift = irq_shifts[engine->id];
	}

	engine->irq_enable_mask = GT_RENDER_USER_INTERRUPT << shift;
	engine->irq_keep_mask = GT_CONTEXT_SWITCH_INTERRUPT << shift;
}

static void
logical_ring_setup(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;
	enum forcewake_domains fw_domains;

	intel_engine_setup_common(engine);

	/* Intentionally left blank. */
	engine->buffer = NULL;

	fw_domains = intel_uncore_forcewake_for_reg(dev_priv,
						    RING_ELSP(engine),
						    FW_REG_WRITE);

	fw_domains |= intel_uncore_forcewake_for_reg(dev_priv,
						     RING_CONTEXT_STATUS_PTR(engine),
						     FW_REG_READ | FW_REG_WRITE);

	fw_domains |= intel_uncore_forcewake_for_reg(dev_priv,
						     RING_CONTEXT_STATUS_BUF_BASE(engine),
						     FW_REG_READ);

	engine->execlists.fw_domains = fw_domains;

	tasklet_init(&engine->execlists.tasklet,
		     execlists_submission_tasklet, (unsigned long)engine);

	logical_ring_default_vfuncs(engine);
	logical_ring_default_irqs(engine);
}

static int logical_ring_init(struct intel_engine_cs *engine)
{
	int ret;

	ret = intel_engine_init_common(engine);
	if (ret)
		goto error;

	if (HAS_LOGICAL_RING_ELSQ(engine->i915)) {
		engine->execlists.submit_reg = engine->i915->regs +
			i915_mmio_reg_offset(RING_EXECLIST_SQ_CONTENTS(engine));
		engine->execlists.ctrl_reg = engine->i915->regs +
			i915_mmio_reg_offset(RING_EXECLIST_CONTROL(engine));
	} else {
		engine->execlists.submit_reg = engine->i915->regs +
			i915_mmio_reg_offset(RING_ELSP(engine));
	}

	engine->execlists.preempt_complete_status = ~0u;
	if (engine->i915->preempt_context) {
		struct intel_context *ce =
			to_intel_context(engine->i915->preempt_context, engine);

		engine->execlists.preempt_complete_status =
			upper_32_bits(ce->lrc_desc);
	}

	return 0;

error:
	intel_logical_ring_cleanup(engine);
	return ret;
}

int logical_render_ring_init(struct intel_engine_cs *engine)
{
	struct drm_i915_private *dev_priv = engine->i915;
	int ret;

	logical_ring_setup(engine);

	if (HAS_L3_DPF(dev_priv))
		engine->irq_keep_mask |= GT_RENDER_L3_PARITY_ERROR_INTERRUPT;

	/* Override some for render ring. */
	if (INTEL_GEN(dev_priv) >= 9)
		engine->init_hw = gen9_init_render_ring;
	else
		engine->init_hw = gen8_init_render_ring;
	engine->init_context = gen8_init_rcs_context;
	engine->emit_flush = gen8_emit_flush_render;
	engine->emit_breadcrumb = gen8_emit_breadcrumb_rcs;
	engine->emit_breadcrumb_sz = gen8_emit_breadcrumb_rcs_sz;

	ret = intel_engine_create_scratch(engine, PAGE_SIZE);
	if (ret)
		return ret;

	ret = intel_init_workaround_bb(engine);
	if (ret) {
		/*
		 * We continue even if we fail to initialize WA batch
		 * because we only expect rare glitches but nothing
		 * critical to prevent us from using GPU
		 */
		DRM_ERROR("WA batch buffer initialization failed: %d\n",
			  ret);
	}

	return logical_ring_init(engine);
}

int logical_xcs_ring_init(struct intel_engine_cs *engine)
{
	logical_ring_setup(engine);

	return logical_ring_init(engine);
}

static u32
make_rpcs(struct drm_i915_private *dev_priv)
{
	u32 rpcs = 0;

	/*
	 * No explicit RPCS request is needed to ensure full
	 * slice/subslice/EU enablement prior to Gen9.
	*/
	if (INTEL_GEN(dev_priv) < 9)
		return 0;

	/*
	 * Starting in Gen9, render power gating can leave
	 * slice/subslice/EU in a partially enabled state. We
	 * must make an explicit request through RPCS for full
	 * enablement.
	*/
	if (INTEL_INFO(dev_priv)->sseu.has_slice_pg) {
		rpcs |= GEN8_RPCS_S_CNT_ENABLE;
		rpcs |= hweight8(INTEL_INFO(dev_priv)->sseu.slice_mask) <<
			GEN8_RPCS_S_CNT_SHIFT;
		rpcs |= GEN8_RPCS_ENABLE;
	}

	if (INTEL_INFO(dev_priv)->sseu.has_subslice_pg) {
		rpcs |= GEN8_RPCS_SS_CNT_ENABLE;
		rpcs |= hweight8(INTEL_INFO(dev_priv)->sseu.subslice_mask[0]) <<
			GEN8_RPCS_SS_CNT_SHIFT;
		rpcs |= GEN8_RPCS_ENABLE;
	}

	if (INTEL_INFO(dev_priv)->sseu.has_eu_pg) {
		rpcs |= INTEL_INFO(dev_priv)->sseu.eu_per_subslice <<
			GEN8_RPCS_EU_MIN_SHIFT;
		rpcs |= INTEL_INFO(dev_priv)->sseu.eu_per_subslice <<
			GEN8_RPCS_EU_MAX_SHIFT;
		rpcs |= GEN8_RPCS_ENABLE;
	}

	return rpcs;
}

static u32 intel_lr_indirect_ctx_offset(struct intel_engine_cs *engine)
{
	u32 indirect_ctx_offset;

	switch (INTEL_GEN(engine->i915)) {
	default:
		MISSING_CASE(INTEL_GEN(engine->i915));
		/* fall through */
	case 11:
		indirect_ctx_offset =
			GEN11_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT;
		break;
	case 10:
		indirect_ctx_offset =
			GEN10_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT;
		break;
	case 9:
		indirect_ctx_offset =
			GEN9_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT;
		break;
	case 8:
		indirect_ctx_offset =
			GEN8_CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT;
		break;
	}

	return indirect_ctx_offset;
}

static void execlists_init_reg_state(u32 *regs,
				     struct i915_gem_context *ctx,
				     struct intel_engine_cs *engine,
				     struct intel_ring *ring)
{
	struct drm_i915_private *dev_priv = engine->i915;
	struct i915_hw_ppgtt *ppgtt = ctx->ppgtt ?: dev_priv->mm.aliasing_ppgtt;
	u32 base = engine->mmio_base;
	bool rcs = engine->id == RCS;

	/* A context is actually a big batch buffer with several
	 * MI_LOAD_REGISTER_IMM commands followed by (reg, value) pairs. The
	 * values we are setting here are only for the first context restore:
	 * on a subsequent save, the GPU will recreate this batchbuffer with new
	 * values (including all the missing MI_LOAD_REGISTER_IMM commands that
	 * we are not initializing here).
	 */
	regs[CTX_LRI_HEADER_0] = MI_LOAD_REGISTER_IMM(rcs ? 14 : 11) |
				 MI_LRI_FORCE_POSTED;

	CTX_REG(regs, CTX_CONTEXT_CONTROL, RING_CONTEXT_CONTROL(engine),
		_MASKED_BIT_DISABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT |
				    CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT) |
		_MASKED_BIT_ENABLE(CTX_CTRL_INHIBIT_SYN_CTX_SWITCH |
				   (HAS_RESOURCE_STREAMER(dev_priv) ?
				   CTX_CTRL_RS_CTX_ENABLE : 0)));
	CTX_REG(regs, CTX_RING_HEAD, RING_HEAD(base), 0);
	CTX_REG(regs, CTX_RING_TAIL, RING_TAIL(base), 0);
	CTX_REG(regs, CTX_RING_BUFFER_START, RING_START(base), 0);
	CTX_REG(regs, CTX_RING_BUFFER_CONTROL, RING_CTL(base),
		RING_CTL_SIZE(ring->size) | RING_VALID);
	CTX_REG(regs, CTX_BB_HEAD_U, RING_BBADDR_UDW(base), 0);
	CTX_REG(regs, CTX_BB_HEAD_L, RING_BBADDR(base), 0);
	CTX_REG(regs, CTX_BB_STATE, RING_BBSTATE(base), RING_BB_PPGTT);
	CTX_REG(regs, CTX_SECOND_BB_HEAD_U, RING_SBBADDR_UDW(base), 0);
	CTX_REG(regs, CTX_SECOND_BB_HEAD_L, RING_SBBADDR(base), 0);
	CTX_REG(regs, CTX_SECOND_BB_STATE, RING_SBBSTATE(base), 0);
	if (rcs) {
		struct i915_ctx_workarounds *wa_ctx = &engine->wa_ctx;

		CTX_REG(regs, CTX_RCS_INDIRECT_CTX, RING_INDIRECT_CTX(base), 0);
		CTX_REG(regs, CTX_RCS_INDIRECT_CTX_OFFSET,
			RING_INDIRECT_CTX_OFFSET(base), 0);
		if (wa_ctx->indirect_ctx.size) {
			u32 ggtt_offset = i915_ggtt_offset(wa_ctx->vma);

			regs[CTX_RCS_INDIRECT_CTX + 1] =
				(ggtt_offset + wa_ctx->indirect_ctx.offset) |
				(wa_ctx->indirect_ctx.size / CACHELINE_BYTES);

			regs[CTX_RCS_INDIRECT_CTX_OFFSET + 1] =
				intel_lr_indirect_ctx_offset(engine) << 6;
		}

		CTX_REG(regs, CTX_BB_PER_CTX_PTR, RING_BB_PER_CTX_PTR(base), 0);
		if (wa_ctx->per_ctx.size) {
			u32 ggtt_offset = i915_ggtt_offset(wa_ctx->vma);

			regs[CTX_BB_PER_CTX_PTR + 1] =
				(ggtt_offset + wa_ctx->per_ctx.offset) | 0x01;
		}
	}

	regs[CTX_LRI_HEADER_1] = MI_LOAD_REGISTER_IMM(9) | MI_LRI_FORCE_POSTED;

	CTX_REG(regs, CTX_CTX_TIMESTAMP, RING_CTX_TIMESTAMP(base), 0);
	/* PDP values well be assigned later if needed */
	CTX_REG(regs, CTX_PDP3_UDW, GEN8_RING_PDP_UDW(engine, 3), 0);
	CTX_REG(regs, CTX_PDP3_LDW, GEN8_RING_PDP_LDW(engine, 3), 0);
	CTX_REG(regs, CTX_PDP2_UDW, GEN8_RING_PDP_UDW(engine, 2), 0);
	CTX_REG(regs, CTX_PDP2_LDW, GEN8_RING_PDP_LDW(engine, 2), 0);
	CTX_REG(regs, CTX_PDP1_UDW, GEN8_RING_PDP_UDW(engine, 1), 0);
	CTX_REG(regs, CTX_PDP1_LDW, GEN8_RING_PDP_LDW(engine, 1), 0);
	CTX_REG(regs, CTX_PDP0_UDW, GEN8_RING_PDP_UDW(engine, 0), 0);
	CTX_REG(regs, CTX_PDP0_LDW, GEN8_RING_PDP_LDW(engine, 0), 0);

	if (ppgtt && i915_vm_is_48bit(&ppgtt->base)) {
		/* 64b PPGTT (48bit canonical)
		 * PDP0_DESCRIPTOR contains the base address to PML4 and
		 * other PDP Descriptors are ignored.
		 */
		ASSIGN_CTX_PML4(ppgtt, regs);
	}

	if (rcs) {
		regs[CTX_LRI_HEADER_2] = MI_LOAD_REGISTER_IMM(1);
		CTX_REG(regs, CTX_R_PWR_CLK_STATE, GEN8_R_PWR_CLK_STATE,
			make_rpcs(dev_priv));

		i915_oa_init_reg_state(engine, ctx, regs);
	}
}

static int
populate_lr_context(struct i915_gem_context *ctx,
		    struct drm_i915_gem_object *ctx_obj,
		    struct intel_engine_cs *engine,
		    struct intel_ring *ring)
{
	void *vaddr;
	u32 *regs;
	int ret;

	ret = i915_gem_object_set_to_cpu_domain(ctx_obj, true);
	if (ret) {
		DRM_DEBUG_DRIVER("Could not set to CPU domain\n");
		return ret;
	}

	vaddr = i915_gem_object_pin_map(ctx_obj, I915_MAP_WB);
	if (IS_ERR(vaddr)) {
		ret = PTR_ERR(vaddr);
		DRM_DEBUG_DRIVER("Could not map object pages! (%d)\n", ret);
		return ret;
	}
	ctx_obj->mm.dirty = true;

	if (engine->default_state) {
		/*
		 * We only want to copy over the template context state;
		 * skipping over the headers reserved for GuC communication,
		 * leaving those as zero.
		 */
		const unsigned long start = LRC_HEADER_PAGES * PAGE_SIZE;
		void *defaults;

		defaults = i915_gem_object_pin_map(engine->default_state,
						   I915_MAP_WB);
		if (IS_ERR(defaults)) {
			ret = PTR_ERR(defaults);
			goto err_unpin_ctx;
		}

		memcpy(vaddr + start, defaults + start, engine->context_size);
		i915_gem_object_unpin_map(engine->default_state);
	}

	/* The second page of the context object contains some fields which must
	 * be set up prior to the first execution. */
	regs = vaddr + LRC_STATE_PN * PAGE_SIZE;
	execlists_init_reg_state(regs, ctx, engine, ring);
	if (!engine->default_state)
		regs[CTX_CONTEXT_CONTROL + 1] |=
			_MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT);
	if (ctx == ctx->i915->preempt_context && INTEL_GEN(engine->i915) < 11)
		regs[CTX_CONTEXT_CONTROL + 1] |=
			_MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT |
					   CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT);

err_unpin_ctx:
	i915_gem_object_unpin_map(ctx_obj);
	return ret;
}

static int execlists_context_deferred_alloc(struct i915_gem_context *ctx,
					    struct intel_engine_cs *engine)
{
	struct drm_i915_gem_object *ctx_obj;
	struct intel_context *ce = to_intel_context(ctx, engine);
	struct i915_vma *vma;
	uint32_t context_size;
	struct intel_ring *ring;
	struct i915_timeline *timeline;
	int ret;

	if (ce->state)
		return 0;

	context_size = round_up(engine->context_size, I915_GTT_PAGE_SIZE);

	/*
	 * Before the actual start of the context image, we insert a few pages
	 * for our own use and for sharing with the GuC.
	 */
	context_size += LRC_HEADER_PAGES * PAGE_SIZE;

	ctx_obj = i915_gem_object_create(ctx->i915, context_size);
	if (IS_ERR(ctx_obj)) {
		ret = PTR_ERR(ctx_obj);
		goto error_deref_obj;
	}

	vma = i915_vma_instance(ctx_obj, &ctx->i915->ggtt.base, NULL);
	if (IS_ERR(vma)) {
		ret = PTR_ERR(vma);
		goto error_deref_obj;
	}

	timeline = i915_timeline_create(ctx->i915, ctx->name);
	if (IS_ERR(timeline)) {
		ret = PTR_ERR(timeline);
		goto error_deref_obj;
	}

	ring = intel_engine_create_ring(engine, timeline, ctx->ring_size);
	i915_timeline_put(timeline);
	if (IS_ERR(ring)) {
		ret = PTR_ERR(ring);
		goto error_deref_obj;
	}

	ret = populate_lr_context(ctx, ctx_obj, engine, ring);
	if (ret) {
		DRM_DEBUG_DRIVER("Failed to populate LRC: %d\n", ret);
		goto error_ring_free;
	}

	ce->ring = ring;
	ce->state = vma;

	return 0;

error_ring_free:
	intel_ring_free(ring);
error_deref_obj:
	i915_gem_object_put(ctx_obj);
	return ret;
}

void intel_lr_context_resume(struct drm_i915_private *dev_priv)
{
	struct intel_engine_cs *engine;
	struct i915_gem_context *ctx;
	enum intel_engine_id id;

	/* Because we emit WA_TAIL_DWORDS there may be a disparity
	 * between our bookkeeping in ce->ring->head and ce->ring->tail and
	 * that stored in context. As we only write new commands from
	 * ce->ring->tail onwards, everything before that is junk. If the GPU
	 * starts reading from its RING_HEAD from the context, it may try to
	 * execute that junk and die.
	 *
	 * So to avoid that we reset the context images upon resume. For
	 * simplicity, we just zero everything out.
	 */
	list_for_each_entry(ctx, &dev_priv->contexts.list, link) {
		for_each_engine(engine, dev_priv, id) {
			struct intel_context *ce =
				to_intel_context(ctx, engine);
			u32 *reg;

			if (!ce->state)
				continue;

			reg = i915_gem_object_pin_map(ce->state->obj,
						      I915_MAP_WB);
			if (WARN_ON(IS_ERR(reg)))
				continue;

			reg += LRC_STATE_PN * PAGE_SIZE / sizeof(*reg);
			reg[CTX_RING_HEAD+1] = 0;
			reg[CTX_RING_TAIL+1] = 0;

			ce->state->obj->mm.dirty = true;
			i915_gem_object_unpin_map(ce->state->obj);

			intel_ring_reset(ce->ring, 0);
		}
	}
}

#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
#include "selftests/intel_lrc.c"
#endif