summaryrefslogtreecommitdiff
path: root/src/cpu/o3/fetch_impl.hh
blob: fe320fa79edf4a931e4c06139b9aba957d45814e (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
/*
 * Copyright (c) 2004-2006 The Regents of The University of Michigan
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions are
 * met: redistributions of source code must retain the above copyright
 * notice, this list of conditions and the following disclaimer;
 * redistributions in binary form must reproduce the above copyright
 * notice, this list of conditions and the following disclaimer in the
 * documentation and/or other materials provided with the distribution;
 * neither the name of the copyright holders nor the names of its
 * contributors may be used to endorse or promote products derived from
 * this software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
 * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
 * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 *
 * Authors: Kevin Lim
 *          Korey Sewell
 */

#include "config/use_checker.hh"

#include "arch/isa_traits.hh"
#include "arch/utility.hh"
#include "cpu/checker/cpu.hh"
#include "cpu/exetrace.hh"
#include "cpu/o3/fetch.hh"
#include "mem/packet.hh"
#include "mem/request.hh"
#include "sim/byteswap.hh"
#include "sim/host.hh"
#include "sim/root.hh"

#if FULL_SYSTEM
#include "arch/tlb.hh"
#include "arch/vtophys.hh"
#include "sim/system.hh"
#endif // FULL_SYSTEM

#include <algorithm>

template<class Impl>
Tick
DefaultFetch<Impl>::IcachePort::recvAtomic(PacketPtr pkt)
{
    panic("DefaultFetch doesn't expect recvAtomic callback!");
    return curTick;
}

template<class Impl>
void
DefaultFetch<Impl>::IcachePort::recvFunctional(PacketPtr pkt)
{
    DPRINTF(Fetch, "DefaultFetch doesn't update its state from a "
            "functional call.");
}

template<class Impl>
void
DefaultFetch<Impl>::IcachePort::recvStatusChange(Status status)
{
    if (status == RangeChange) {
        if (!snoopRangeSent) {
            snoopRangeSent = true;
            sendStatusChange(Port::RangeChange);
        }
        return;
    }

    panic("DefaultFetch doesn't expect recvStatusChange callback!");
}

template<class Impl>
bool
DefaultFetch<Impl>::IcachePort::recvTiming(PacketPtr pkt)
{
    DPRINTF(Fetch, "Received timing\n");
    if (pkt->isResponse()) {
        fetch->processCacheCompletion(pkt);
    }
    //else Snooped a coherence request, just return
    return true;
}

template<class Impl>
void
DefaultFetch<Impl>::IcachePort::recvRetry()
{
    fetch->recvRetry();
}

template<class Impl>
DefaultFetch<Impl>::DefaultFetch(Params *params)
    : branchPred(params),
      decodeToFetchDelay(params->decodeToFetchDelay),
      renameToFetchDelay(params->renameToFetchDelay),
      iewToFetchDelay(params->iewToFetchDelay),
      commitToFetchDelay(params->commitToFetchDelay),
      fetchWidth(params->fetchWidth),
      cacheBlocked(false),
      retryPkt(NULL),
      retryTid(-1),
      numThreads(params->numberOfThreads),
      numFetchingThreads(params->smtNumFetchingThreads),
      interruptPending(false),
      drainPending(false),
      switchedOut(false)
{
    if (numThreads > Impl::MaxThreads)
        fatal("numThreads is not a valid value\n");

    // Set fetch stage's status to inactive.
    _status = Inactive;

    std::string policy = params->smtFetchPolicy;

    // Convert string to lowercase
    std::transform(policy.begin(), policy.end(), policy.begin(),
                   (int(*)(int)) tolower);

    // Figure out fetch policy
    if (policy == "singlethread") {
        fetchPolicy = SingleThread;
        if (numThreads > 1)
            panic("Invalid Fetch Policy for a SMT workload.");
    } else if (policy == "roundrobin") {
        fetchPolicy = RoundRobin;
        DPRINTF(Fetch, "Fetch policy set to Round Robin\n");
    } else if (policy == "branch") {
        fetchPolicy = Branch;
        DPRINTF(Fetch, "Fetch policy set to Branch Count\n");
    } else if (policy == "iqcount") {
        fetchPolicy = IQ;
        DPRINTF(Fetch, "Fetch policy set to IQ count\n");
    } else if (policy == "lsqcount") {
        fetchPolicy = LSQ;
        DPRINTF(Fetch, "Fetch policy set to LSQ count\n");
    } else {
        fatal("Invalid Fetch Policy. Options Are: {SingleThread,"
              " RoundRobin,LSQcount,IQcount}\n");
    }

    // Get the size of an instruction.
    instSize = sizeof(TheISA::MachInst);
}

template <class Impl>
std::string
DefaultFetch<Impl>::name() const
{
    return cpu->name() + ".fetch";
}

template <class Impl>
void
DefaultFetch<Impl>::regStats()
{
    icacheStallCycles
        .name(name() + ".icacheStallCycles")
        .desc("Number of cycles fetch is stalled on an Icache miss")
        .prereq(icacheStallCycles);

    fetchedInsts
        .name(name() + ".Insts")
        .desc("Number of instructions fetch has processed")
        .prereq(fetchedInsts);

    fetchedBranches
        .name(name() + ".Branches")
        .desc("Number of branches that fetch encountered")
        .prereq(fetchedBranches);

    predictedBranches
        .name(name() + ".predictedBranches")
        .desc("Number of branches that fetch has predicted taken")
        .prereq(predictedBranches);

    fetchCycles
        .name(name() + ".Cycles")
        .desc("Number of cycles fetch has run and was not squashing or"
              " blocked")
        .prereq(fetchCycles);

    fetchSquashCycles
        .name(name() + ".SquashCycles")
        .desc("Number of cycles fetch has spent squashing")
        .prereq(fetchSquashCycles);

    fetchIdleCycles
        .name(name() + ".IdleCycles")
        .desc("Number of cycles fetch was idle")
        .prereq(fetchIdleCycles);

    fetchBlockedCycles
        .name(name() + ".BlockedCycles")
        .desc("Number of cycles fetch has spent blocked")
        .prereq(fetchBlockedCycles);

    fetchedCacheLines
        .name(name() + ".CacheLines")
        .desc("Number of cache lines fetched")
        .prereq(fetchedCacheLines);

    fetchMiscStallCycles
        .name(name() + ".MiscStallCycles")
        .desc("Number of cycles fetch has spent waiting on interrupts, or "
              "bad addresses, or out of MSHRs")
        .prereq(fetchMiscStallCycles);

    fetchIcacheSquashes
        .name(name() + ".IcacheSquashes")
        .desc("Number of outstanding Icache misses that were squashed")
        .prereq(fetchIcacheSquashes);

    fetchNisnDist
        .init(/* base value */ 0,
              /* last value */ fetchWidth,
              /* bucket size */ 1)
        .name(name() + ".rateDist")
        .desc("Number of instructions fetched each cycle (Total)")
        .flags(Stats::pdf);

    idleRate
        .name(name() + ".idleRate")
        .desc("Percent of cycles fetch was idle")
        .prereq(idleRate);
    idleRate = fetchIdleCycles * 100 / cpu->numCycles;

    branchRate
        .name(name() + ".branchRate")
        .desc("Number of branch fetches per cycle")
        .flags(Stats::total);
    branchRate = fetchedBranches / cpu->numCycles;

    fetchRate
        .name(name() + ".rate")
        .desc("Number of inst fetches per cycle")
        .flags(Stats::total);
    fetchRate = fetchedInsts / cpu->numCycles;

    branchPred.regStats();
}

template<class Impl>
void
DefaultFetch<Impl>::setCPU(O3CPU *cpu_ptr)
{
    DPRINTF(Fetch, "Setting the CPU pointer.\n");
    cpu = cpu_ptr;

    // Name is finally available, so create the port.
    icachePort = new IcachePort(this);

    icachePort->snoopRangeSent = false;

#if USE_CHECKER
    if (cpu->checker) {
        cpu->checker->setIcachePort(icachePort);
    }
#endif

    // Schedule fetch to get the correct PC from the CPU
    // scheduleFetchStartupEvent(1);

    // Fetch needs to start fetching instructions at the very beginning,
    // so it must start up in active state.
    switchToActive();
}

template<class Impl>
void
DefaultFetch<Impl>::setTimeBuffer(TimeBuffer<TimeStruct> *time_buffer)
{
    DPRINTF(Fetch, "Setting the time buffer pointer.\n");
    timeBuffer = time_buffer;

    // Create wires to get information from proper places in time buffer.
    fromDecode = timeBuffer->getWire(-decodeToFetchDelay);
    fromRename = timeBuffer->getWire(-renameToFetchDelay);
    fromIEW = timeBuffer->getWire(-iewToFetchDelay);
    fromCommit = timeBuffer->getWire(-commitToFetchDelay);
}

template<class Impl>
void
DefaultFetch<Impl>::setActiveThreads(std::list<unsigned> *at_ptr)
{
    DPRINTF(Fetch, "Setting active threads list pointer.\n");
    activeThreads = at_ptr;
}

template<class Impl>
void
DefaultFetch<Impl>::setFetchQueue(TimeBuffer<FetchStruct> *fq_ptr)
{
    DPRINTF(Fetch, "Setting the fetch queue pointer.\n");
    fetchQueue = fq_ptr;

    // Create wire to write information to proper place in fetch queue.
    toDecode = fetchQueue->getWire(0);
}

template<class Impl>
void
DefaultFetch<Impl>::initStage()
{
    // Setup PC and nextPC with initial state.
    for (int tid = 0; tid < numThreads; tid++) {
        PC[tid] = cpu->readPC(tid);
        nextPC[tid] = cpu->readNextPC(tid);
#if ISA_HAS_DELAY_SLOT
        nextNPC[tid] = cpu->readNextNPC(tid);
#endif
    }

    // Size of cache block.
    cacheBlkSize = icachePort->peerBlockSize();

    // Create mask to get rid of offset bits.
    cacheBlkMask = (cacheBlkSize - 1);

    for (int tid=0; tid < numThreads; tid++) {

        fetchStatus[tid] = Running;

        priorityList.push_back(tid);

        memReq[tid] = NULL;

        // Create space to store a cache line.
        cacheData[tid] = new uint8_t[cacheBlkSize];
        cacheDataPC[tid] = 0;
        cacheDataValid[tid] = false;

        delaySlotInfo[tid].branchSeqNum = -1;
        delaySlotInfo[tid].numInsts = 0;
        delaySlotInfo[tid].targetAddr = 0;
        delaySlotInfo[tid].targetReady = false;

        stalls[tid].decode = false;
        stalls[tid].rename = false;
        stalls[tid].iew = false;
        stalls[tid].commit = false;
    }
}

template<class Impl>
void
DefaultFetch<Impl>::processCacheCompletion(PacketPtr pkt)
{
    unsigned tid = pkt->req->getThreadNum();

    DPRINTF(Fetch, "[tid:%u] Waking up from cache miss.\n",tid);

    // Only change the status if it's still waiting on the icache access
    // to return.
    if (fetchStatus[tid] != IcacheWaitResponse ||
        pkt->req != memReq[tid] ||
        isSwitchedOut()) {
        ++fetchIcacheSquashes;
        delete pkt->req;
        delete pkt;
        return;
    }

    memcpy(cacheData[tid], pkt->getPtr<uint8_t *>(), cacheBlkSize);
    cacheDataValid[tid] = true;

    if (!drainPending) {
        // Wake up the CPU (if it went to sleep and was waiting on
        // this completion event).
        cpu->wakeCPU();

        DPRINTF(Activity, "[tid:%u] Activating fetch due to cache completion\n",
                tid);

        switchToActive();
    }

    // Only switch to IcacheAccessComplete if we're not stalled as well.
    if (checkStall(tid)) {
        fetchStatus[tid] = Blocked;
    } else {
        fetchStatus[tid] = IcacheAccessComplete;
    }

    // Reset the mem req to NULL.
    delete pkt->req;
    delete pkt;
    memReq[tid] = NULL;
}

template <class Impl>
bool
DefaultFetch<Impl>::drain()
{
    // Fetch is ready to drain at any time.
    cpu->signalDrained();
    drainPending = true;
    return true;
}

template <class Impl>
void
DefaultFetch<Impl>::resume()
{
    drainPending = false;
}

template <class Impl>
void
DefaultFetch<Impl>::switchOut()
{
    switchedOut = true;
    // Branch predictor needs to have its state cleared.
    branchPred.switchOut();
}

template <class Impl>
void
DefaultFetch<Impl>::takeOverFrom()
{
    // Reset all state
    for (int i = 0; i < Impl::MaxThreads; ++i) {
        stalls[i].decode = 0;
        stalls[i].rename = 0;
        stalls[i].iew = 0;
        stalls[i].commit = 0;
        PC[i] = cpu->readPC(i);
        nextPC[i] = cpu->readNextPC(i);
#if ISA_HAS_DELAY_SLOT
        nextNPC[i] = cpu->readNextNPC(i);
        delaySlotInfo[i].branchSeqNum = -1;
        delaySlotInfo[i].numInsts = 0;
        delaySlotInfo[i].targetAddr = 0;
        delaySlotInfo[i].targetReady = false;
#endif
        fetchStatus[i] = Running;
    }
    numInst = 0;
    wroteToTimeBuffer = false;
    _status = Inactive;
    switchedOut = false;
    interruptPending = false;
    branchPred.takeOverFrom();
}

template <class Impl>
void
DefaultFetch<Impl>::wakeFromQuiesce()
{
    DPRINTF(Fetch, "Waking up from quiesce\n");
    // Hopefully this is safe
    // @todo: Allow other threads to wake from quiesce.
    fetchStatus[0] = Running;
}

template <class Impl>
inline void
DefaultFetch<Impl>::switchToActive()
{
    if (_status == Inactive) {
        DPRINTF(Activity, "Activating stage.\n");

        cpu->activateStage(O3CPU::FetchIdx);

        _status = Active;
    }
}

template <class Impl>
inline void
DefaultFetch<Impl>::switchToInactive()
{
    if (_status == Active) {
        DPRINTF(Activity, "Deactivating stage.\n");

        cpu->deactivateStage(O3CPU::FetchIdx);

        _status = Inactive;
    }
}

template <class Impl>
bool
DefaultFetch<Impl>::lookupAndUpdateNextPC(DynInstPtr &inst, Addr &next_PC,
                                          Addr &next_NPC)
{
    // Do branch prediction check here.
    // A bit of a misnomer...next_PC is actually the current PC until
    // this function updates it.
    bool predict_taken;

    if (!inst->isControl()) {
#if ISA_HAS_DELAY_SLOT
        Addr cur_PC = next_PC;
        next_PC  = cur_PC + instSize;      //next_NPC;
        next_NPC = cur_PC + (2 * instSize);//next_NPC + instSize;
        inst->setPredTarg(next_NPC);
#else
        next_PC = next_PC + instSize;
        inst->setPredTarg(next_PC);
#endif
        return false;
    }

    int tid = inst->threadNumber;
#if ISA_HAS_DELAY_SLOT
    Addr pred_PC = next_PC;
    predict_taken = branchPred.predict(inst, pred_PC, tid);

    if (predict_taken) {
        DPRINTF(Fetch, "[tid:%i]: Branch predicted to be true.\n", tid);
    } else {
        DPRINTF(Fetch, "[tid:%i]: Branch predicted to be false.\n", tid);
    }

    if (predict_taken) {
        next_PC = next_NPC;
        next_NPC = pred_PC;

        // Update delay slot info
        ++delaySlotInfo[tid].numInsts;
        delaySlotInfo[tid].targetAddr = pred_PC;
        DPRINTF(Fetch, "[tid:%i]: %i delay slot inst(s) to process.\n", tid,
                delaySlotInfo[tid].numInsts);
    } else { // !predict_taken
        if (inst->isCondDelaySlot()) {
            next_PC = pred_PC;
            // The delay slot is skipped here if there is on
            // prediction
        } else {
            next_PC = next_NPC;
            // No need to declare a delay slot here since
            // there is no for the pred. target to jump
        }

        next_NPC = next_NPC + instSize;
    }
#else
    predict_taken = branchPred.predict(inst, next_PC, tid);
#endif

    ++fetchedBranches;

    if (predict_taken) {
        ++predictedBranches;
    }

    return predict_taken;
}

template <class Impl>
bool
DefaultFetch<Impl>::fetchCacheLine(Addr fetch_PC, Fault &ret_fault, unsigned tid)
{
    Fault fault = NoFault;

    //AlphaDep
    if (cacheBlocked) {
        DPRINTF(Fetch, "[tid:%i] Can't fetch cache line, cache blocked\n",
                tid);
        return false;
    } else if (isSwitchedOut()) {
        DPRINTF(Fetch, "[tid:%i] Can't fetch cache line, switched out\n",
                tid);
        return false;
    } else if (interruptPending && !(fetch_PC & 0x3)) {
        // Hold off fetch from getting new instructions when:
        // Cache is blocked, or
        // while an interrupt is pending and we're not in PAL mode, or
        // fetch is switched out.
        DPRINTF(Fetch, "[tid:%i] Can't fetch cache line, interrupt pending\n",
                tid);
        return false;
    }

    // Align the fetch PC so it's at the start of a cache block.
    Addr block_PC = icacheBlockAlignPC(fetch_PC);

    // If we've already got the block, no need to try to fetch it again.
    if (cacheDataValid[tid] && block_PC == cacheDataPC[tid]) {
        return true;
    }

    // Setup the memReq to do a read of the first instruction's address.
    // Set the appropriate read size and flags as well.
    // Build request here.
    RequestPtr mem_req = new Request(tid, block_PC, cacheBlkSize, 0,
                                     fetch_PC, cpu->readCpuId(), tid);

    memReq[tid] = mem_req;

    // Translate the instruction request.
    fault = cpu->translateInstReq(mem_req, cpu->thread[tid]);

    // In the case of faults, the fetch stage may need to stall and wait
    // for the ITB miss to be handled.

    // If translation was successful, attempt to read the first
    // instruction.
    if (fault == NoFault) {
#if 0
        if (cpu->system->memctrl->badaddr(memReq[tid]->paddr) ||
            memReq[tid]->isUncacheable()) {
            DPRINTF(Fetch, "Fetch: Bad address %#x (hopefully on a "
                    "misspeculating path)!",
                    memReq[tid]->paddr);
            ret_fault = TheISA::genMachineCheckFault();
            return false;
        }
#endif

        // Build packet here.
        PacketPtr data_pkt = new Packet(mem_req,
                                        Packet::ReadReq, Packet::Broadcast);
        data_pkt->dataDynamicArray(new uint8_t[cacheBlkSize]);

        cacheDataPC[tid] = block_PC;
        cacheDataValid[tid] = false;

        DPRINTF(Fetch, "Fetch: Doing instruction read.\n");

        fetchedCacheLines++;

        // Now do the timing access to see whether or not the instruction
        // exists within the cache.
        if (!icachePort->sendTiming(data_pkt)) {
            if (data_pkt->result == Packet::BadAddress) {
                fault = TheISA::genMachineCheckFault();
                delete mem_req;
                memReq[tid] = NULL;
            }
            assert(retryPkt == NULL);
            assert(retryTid == -1);
            DPRINTF(Fetch, "[tid:%i] Out of MSHRs!\n", tid);
            fetchStatus[tid] = IcacheWaitRetry;
            retryPkt = data_pkt;
            retryTid = tid;
            cacheBlocked = true;
            return false;
        }

        DPRINTF(Fetch, "[tid:%i]: Doing cache access.\n", tid);

        lastIcacheStall[tid] = curTick;

        DPRINTF(Activity, "[tid:%i]: Activity: Waiting on I-cache "
                "response.\n", tid);

        fetchStatus[tid] = IcacheWaitResponse;
    } else {
        delete mem_req;
        memReq[tid] = NULL;
    }

    ret_fault = fault;
    return true;
}

template <class Impl>
inline void
DefaultFetch<Impl>::doSquash(const Addr &new_PC, unsigned tid)
{
    DPRINTF(Fetch, "[tid:%i]: Squashing, setting PC to: %#x.\n",
            tid, new_PC);

    PC[tid] = new_PC;
    nextPC[tid] = new_PC + instSize;
    nextNPC[tid] = new_PC + (2 * instSize);

    // Clear the icache miss if it's outstanding.
    if (fetchStatus[tid] == IcacheWaitResponse) {
        DPRINTF(Fetch, "[tid:%i]: Squashing outstanding Icache miss.\n",
                tid);
        memReq[tid] = NULL;
    }

    // Get rid of the retrying packet if it was from this thread.
    if (retryTid == tid) {
        assert(cacheBlocked);
        cacheBlocked = false;
        retryTid = -1;
        delete retryPkt->req;
        delete retryPkt;
        retryPkt = NULL;
    }

    fetchStatus[tid] = Squashing;

    ++fetchSquashCycles;
}

template<class Impl>
void
DefaultFetch<Impl>::squashFromDecode(const Addr &new_PC,
                                     const InstSeqNum &seq_num,
                                     unsigned tid)
{
    DPRINTF(Fetch, "[tid:%i]: Squashing from decode.\n",tid);

    doSquash(new_PC, tid);

#if ISA_HAS_DELAY_SLOT
    if (seq_num <=  delaySlotInfo[tid].branchSeqNum) {
        delaySlotInfo[tid].numInsts = 0;
        delaySlotInfo[tid].targetAddr = 0;
        delaySlotInfo[tid].targetReady = false;
    }
#endif

    // Tell the CPU to remove any instructions that are in flight between
    // fetch and decode.
    cpu->removeInstsUntil(seq_num, tid);
}

template<class Impl>
bool
DefaultFetch<Impl>::checkStall(unsigned tid) const
{
    bool ret_val = false;

    if (cpu->contextSwitch) {
        DPRINTF(Fetch,"[tid:%i]: Stalling for a context switch.\n",tid);
        ret_val = true;
    } else if (stalls[tid].decode) {
        DPRINTF(Fetch,"[tid:%i]: Stall from Decode stage detected.\n",tid);
        ret_val = true;
    } else if (stalls[tid].rename) {
        DPRINTF(Fetch,"[tid:%i]: Stall from Rename stage detected.\n",tid);
        ret_val = true;
    } else if (stalls[tid].iew) {
        DPRINTF(Fetch,"[tid:%i]: Stall from IEW stage detected.\n",tid);
        ret_val = true;
    } else if (stalls[tid].commit) {
        DPRINTF(Fetch,"[tid:%i]: Stall from Commit stage detected.\n",tid);
        ret_val = true;
    }

    return ret_val;
}

template<class Impl>
typename DefaultFetch<Impl>::FetchStatus
DefaultFetch<Impl>::updateFetchStatus()
{
    //Check Running
    std::list<unsigned>::iterator threads = activeThreads->begin();
    std::list<unsigned>::iterator end = activeThreads->end();

    while (threads != end) {
        unsigned tid = *threads++;

        if (fetchStatus[tid] == Running ||
            fetchStatus[tid] == Squashing ||
            fetchStatus[tid] == IcacheAccessComplete) {

            if (_status == Inactive) {
                DPRINTF(Activity, "[tid:%i]: Activating stage.\n",tid);

                if (fetchStatus[tid] == IcacheAccessComplete) {
                    DPRINTF(Activity, "[tid:%i]: Activating fetch due to cache"
                            "completion\n",tid);
                }

                cpu->activateStage(O3CPU::FetchIdx);
            }

            return Active;
        }
    }

    // Stage is switching from active to inactive, notify CPU of it.
    if (_status == Active) {
        DPRINTF(Activity, "Deactivating stage.\n");

        cpu->deactivateStage(O3CPU::FetchIdx);
    }

    return Inactive;
}

template <class Impl>
void
DefaultFetch<Impl>::squash(const Addr &new_PC, const InstSeqNum &seq_num,
                           bool squash_delay_slot, unsigned tid)
{
    DPRINTF(Fetch, "[tid:%u]: Squash from commit.\n",tid);

    doSquash(new_PC, tid);

#if ISA_HAS_DELAY_SLOT
    if (seq_num <=  delaySlotInfo[tid].branchSeqNum) {
        delaySlotInfo[tid].numInsts = 0;
        delaySlotInfo[tid].targetAddr = 0;
        delaySlotInfo[tid].targetReady = false;
    }

    // Tell the CPU to remove any instructions that are not in the ROB.
    cpu->removeInstsNotInROB(tid, squash_delay_slot, seq_num);
#else
    // Tell the CPU to remove any instructions that are not in the ROB.
    cpu->removeInstsNotInROB(tid, true, 0);
#endif
}

template <class Impl>
void
DefaultFetch<Impl>::tick()
{
    std::list<unsigned>::iterator threads = activeThreads->begin();
    std::list<unsigned>::iterator end = activeThreads->end();
    bool status_change = false;

    wroteToTimeBuffer = false;

    while (threads != end) {
        unsigned tid = *threads++;

        // Check the signals for each thread to determine the proper status
        // for each thread.
        bool updated_status = checkSignalsAndUpdate(tid);
        status_change =  status_change || updated_status;
    }

    DPRINTF(Fetch, "Running stage.\n");

    // Reset the number of the instruction we're fetching.
    numInst = 0;

#if FULL_SYSTEM
    if (fromCommit->commitInfo[0].interruptPending) {
        interruptPending = true;
    }

    if (fromCommit->commitInfo[0].clearInterrupt) {
        interruptPending = false;
    }
#endif

    for (threadFetched = 0; threadFetched < numFetchingThreads;
         threadFetched++) {
        // Fetch each of the actively fetching threads.
        fetch(status_change);
    }

    // Record number of instructions fetched this cycle for distribution.
    fetchNisnDist.sample(numInst);

    if (status_change) {
        // Change the fetch stage status if there was a status change.
        _status = updateFetchStatus();
    }

    // If there was activity this cycle, inform the CPU of it.
    if (wroteToTimeBuffer || cpu->contextSwitch) {
        DPRINTF(Activity, "Activity this cycle.\n");

        cpu->activityThisCycle();
    }
}

template <class Impl>
bool
DefaultFetch<Impl>::checkSignalsAndUpdate(unsigned tid)
{
    // Update the per thread stall statuses.
    if (fromDecode->decodeBlock[tid]) {
        stalls[tid].decode = true;
    }

    if (fromDecode->decodeUnblock[tid]) {
        assert(stalls[tid].decode);
        assert(!fromDecode->decodeBlock[tid]);
        stalls[tid].decode = false;
    }

    if (fromRename->renameBlock[tid]) {
        stalls[tid].rename = true;
    }

    if (fromRename->renameUnblock[tid]) {
        assert(stalls[tid].rename);
        assert(!fromRename->renameBlock[tid]);
        stalls[tid].rename = false;
    }

    if (fromIEW->iewBlock[tid]) {
        stalls[tid].iew = true;
    }

    if (fromIEW->iewUnblock[tid]) {
        assert(stalls[tid].iew);
        assert(!fromIEW->iewBlock[tid]);
        stalls[tid].iew = false;
    }

    if (fromCommit->commitBlock[tid]) {
        stalls[tid].commit = true;
    }

    if (fromCommit->commitUnblock[tid]) {
        assert(stalls[tid].commit);
        assert(!fromCommit->commitBlock[tid]);
        stalls[tid].commit = false;
    }

    // Check squash signals from commit.
    if (fromCommit->commitInfo[tid].squash) {

        DPRINTF(Fetch, "[tid:%u]: Squashing instructions due to squash "
                "from commit.\n",tid);

#if ISA_HAS_DELAY_SLOT
    InstSeqNum doneSeqNum = fromCommit->commitInfo[tid].bdelayDoneSeqNum;
#else
    InstSeqNum doneSeqNum = fromCommit->commitInfo[tid].doneSeqNum;
#endif
        // In any case, squash.
        squash(fromCommit->commitInfo[tid].nextPC,
               doneSeqNum,
               fromCommit->commitInfo[tid].squashDelaySlot,
               tid);

        // Also check if there's a mispredict that happened.
        if (fromCommit->commitInfo[tid].branchMispredict) {
            branchPred.squash(fromCommit->commitInfo[tid].doneSeqNum,
                              fromCommit->commitInfo[tid].nextPC,
                              fromCommit->commitInfo[tid].branchTaken,
                              tid);
        } else {
            branchPred.squash(fromCommit->commitInfo[tid].doneSeqNum,
                              tid);
        }

        return true;
    } else if (fromCommit->commitInfo[tid].doneSeqNum) {
        // Update the branch predictor if it wasn't a squashed instruction
        // that was broadcasted.
        branchPred.update(fromCommit->commitInfo[tid].doneSeqNum, tid);
    }

    // Check ROB squash signals from commit.
    if (fromCommit->commitInfo[tid].robSquashing) {
        DPRINTF(Fetch, "[tid:%u]: ROB is still squashing.\n", tid);

        // Continue to squash.
        fetchStatus[tid] = Squashing;

        return true;
    }

    // Check squash signals from decode.
    if (fromDecode->decodeInfo[tid].squash) {
        DPRINTF(Fetch, "[tid:%u]: Squashing instructions due to squash "
                "from decode.\n",tid);

        // Update the branch predictor.
        if (fromDecode->decodeInfo[tid].branchMispredict) {
            branchPred.squash(fromDecode->decodeInfo[tid].doneSeqNum,
                              fromDecode->decodeInfo[tid].nextPC,
                              fromDecode->decodeInfo[tid].branchTaken,
                              tid);
        } else {
            branchPred.squash(fromDecode->decodeInfo[tid].doneSeqNum,
                              tid);
        }

        if (fetchStatus[tid] != Squashing) {

#if ISA_HAS_DELAY_SLOT
            InstSeqNum doneSeqNum = fromDecode->decodeInfo[tid].bdelayDoneSeqNum;
#else
            InstSeqNum doneSeqNum = fromDecode->decodeInfo[tid].doneSeqNum;
#endif
            // Squash unless we're already squashing
            squashFromDecode(fromDecode->decodeInfo[tid].nextPC,
                             doneSeqNum,
                             tid);

            return true;
        }
    }

    if (checkStall(tid) &&
        fetchStatus[tid] != IcacheWaitResponse &&
        fetchStatus[tid] != IcacheWaitRetry) {
        DPRINTF(Fetch, "[tid:%i]: Setting to blocked\n",tid);

        fetchStatus[tid] = Blocked;

        return true;
    }

    if (fetchStatus[tid] == Blocked ||
        fetchStatus[tid] == Squashing) {
        // Switch status to running if fetch isn't being told to block or
        // squash this cycle.
        DPRINTF(Fetch, "[tid:%i]: Done squashing, switching to running.\n",
                tid);

        fetchStatus[tid] = Running;

        return true;
    }

    // If we've reached this point, we have not gotten any signals that
    // cause fetch to change its status.  Fetch remains the same as before.
    return false;
}

template<class Impl>
void
DefaultFetch<Impl>::fetch(bool &status_change)
{
    //////////////////////////////////////////
    // Start actual fetch
    //////////////////////////////////////////
    int tid = getFetchingThread(fetchPolicy);

    if (tid == -1 || drainPending) {
        DPRINTF(Fetch,"There are no more threads available to fetch from.\n");

        // Breaks looping condition in tick()
        threadFetched = numFetchingThreads;
        return;
    }

    DPRINTF(Fetch, "Attempting to fetch from [tid:%i]\n", tid);

    // The current PC.
    Addr &fetch_PC = PC[tid];

    // Fault code for memory access.
    Fault fault = NoFault;

    // If returning from the delay of a cache miss, then update the status
    // to running, otherwise do the cache access.  Possibly move this up
    // to tick() function.
    if (fetchStatus[tid] == IcacheAccessComplete) {
        DPRINTF(Fetch, "[tid:%i]: Icache miss is complete.\n",
                tid);

        fetchStatus[tid] = Running;
        status_change = true;
    } else if (fetchStatus[tid] == Running) {
        DPRINTF(Fetch, "[tid:%i]: Attempting to translate and read "
                "instruction, starting at PC %08p.\n",
                tid, fetch_PC);

        bool fetch_success = fetchCacheLine(fetch_PC, fault, tid);
        if (!fetch_success) {
            if (cacheBlocked) {
                ++icacheStallCycles;
            } else {
                ++fetchMiscStallCycles;
            }
            return;
        }
    } else {
        if (fetchStatus[tid] == Idle) {
            ++fetchIdleCycles;
            DPRINTF(Fetch, "[tid:%i]: Fetch is idle!\n", tid);
        } else if (fetchStatus[tid] == Blocked) {
            ++fetchBlockedCycles;
            DPRINTF(Fetch, "[tid:%i]: Fetch is blocked!\n", tid);
        } else if (fetchStatus[tid] == Squashing) {
            ++fetchSquashCycles;
            DPRINTF(Fetch, "[tid:%i]: Fetch is squashing!\n", tid);
        } else if (fetchStatus[tid] == IcacheWaitResponse) {
            ++icacheStallCycles;
            DPRINTF(Fetch, "[tid:%i]: Fetch is waiting cache response!\n", tid);
        }

        // Status is Idle, Squashing, Blocked, or IcacheWaitResponse, so
        // fetch should do nothing.
        return;
    }

    ++fetchCycles;

    // If we had a stall due to an icache miss, then return.
    if (fetchStatus[tid] == IcacheWaitResponse) {
        ++icacheStallCycles;
        status_change = true;
        return;
    }

    Addr next_PC = fetch_PC;
    Addr next_NPC = next_PC + instSize;
    InstSeqNum inst_seq;
    MachInst inst;
    ExtMachInst ext_inst;
    // @todo: Fix this hack.
    unsigned offset = (fetch_PC & cacheBlkMask) & ~3;

    if (fault == NoFault) {
        // If the read of the first instruction was successful, then grab the
        // instructions from the rest of the cache line and put them into the
        // queue heading to decode.

        DPRINTF(Fetch, "[tid:%i]: Adding instructions to queue to "
                "decode.\n",tid);

        // Need to keep track of whether or not a predicted branch
        // ended this fetch block.
        bool predicted_branch = false;

        // Need to keep track of whether or not a delay slot
        // instruction has been fetched

        for (;
             offset < cacheBlkSize &&
                 numInst < fetchWidth &&
                 (!predicted_branch || delaySlotInfo[tid].numInsts > 0);
             ++numInst) {

            // Get a sequence number.
            inst_seq = cpu->getAndIncrementInstSeq();

            // Make sure this is a valid index.
            assert(offset <= cacheBlkSize - instSize);

            // Get the instruction from the array of the cache line.
            inst = TheISA::gtoh(*reinterpret_cast<TheISA::MachInst *>
                        (&cacheData[tid][offset]));

#if THE_ISA == ALPHA_ISA
            ext_inst = TheISA::makeExtMI(inst, fetch_PC);
#elif THE_ISA == SPARC_ISA
            ext_inst = TheISA::makeExtMI(inst, cpu->thread[tid]->getTC());
#elif THE_ISA == MIPS_ISA
            ext_inst = TheISA::makeExtMI(inst, cpu->thread[tid]->getTC());
#endif

            // Create a new DynInst from the instruction fetched.
            DynInstPtr instruction = new DynInst(ext_inst, fetch_PC,
                                                 next_PC,
                                                 inst_seq, cpu);
            instruction->setTid(tid);

            instruction->setASID(tid);

            instruction->setThreadState(cpu->thread[tid]);

            DPRINTF(Fetch, "[tid:%i]: Instruction PC %#x created "
                    "[sn:%lli]\n",
                    tid, instruction->readPC(), inst_seq);

            DPRINTF(Fetch, "[tid:%i]: Instruction is: %s\n",
                    tid, instruction->staticInst->disassemble(fetch_PC));

            instruction->traceData =
                Trace::getInstRecord(curTick, cpu->tcBase(tid),
                                     instruction->staticInst,
                                     instruction->readPC());

            predicted_branch = lookupAndUpdateNextPC(instruction, next_PC,
                                                     next_NPC);

            // Add instruction to the CPU's list of instructions.
            instruction->setInstListIt(cpu->addInst(instruction));

            // Write the instruction to the first slot in the queue
            // that heads to decode.
            toDecode->insts[numInst] = instruction;

            toDecode->size++;

            // Increment stat of fetched instructions.
            ++fetchedInsts;

            // Move to the next instruction, unless we have a branch.
            fetch_PC = next_PC;

            if (instruction->isQuiesce()) {
                DPRINTF(Fetch, "Quiesce instruction encountered, halting fetch!",
                        curTick);
                fetchStatus[tid] = QuiescePending;
                ++numInst;
                status_change = true;
                break;
            }

            offset += instSize;

#if ISA_HAS_DELAY_SLOT
            if (predicted_branch) {
                delaySlotInfo[tid].branchSeqNum = inst_seq;

                DPRINTF(Fetch, "[tid:%i]: Delay slot branch set to [sn:%i]\n",
                        tid, inst_seq);
                continue;
            } else if (delaySlotInfo[tid].numInsts > 0) {
                --delaySlotInfo[tid].numInsts;

                // It's OK to set PC to target of branch
                if (delaySlotInfo[tid].numInsts == 0) {
                    delaySlotInfo[tid].targetReady = true;

                    // Break the looping condition
                    predicted_branch = true;
                }

                DPRINTF(Fetch, "[tid:%i]: %i delay slot inst(s) left to"
                        " process.\n", tid, delaySlotInfo[tid].numInsts);
            }
#endif
        }

        if (offset >= cacheBlkSize) {
            DPRINTF(Fetch, "[tid:%i]: Done fetching, reached the end of cache "
                    "block.\n", tid);
        } else if (numInst >= fetchWidth) {
            DPRINTF(Fetch, "[tid:%i]: Done fetching, reached fetch bandwidth "
                    "for this cycle.\n", tid);
        } else if (predicted_branch && delaySlotInfo[tid].numInsts <= 0) {
            DPRINTF(Fetch, "[tid:%i]: Done fetching, predicted branch "
                    "instruction encountered.\n", tid);
        }
    }

    if (numInst > 0) {
        wroteToTimeBuffer = true;
    }

    // Now that fetching is completed, update the PC to signify what the next
    // cycle will be.
    if (fault == NoFault) {
#if ISA_HAS_DELAY_SLOT
        if (delaySlotInfo[tid].targetReady &&
            delaySlotInfo[tid].numInsts == 0) {
            // Set PC to target
            PC[tid] = delaySlotInfo[tid].targetAddr; //next_PC
            nextPC[tid] = next_PC + instSize;        //next_NPC
            nextNPC[tid] = next_PC + (2 * instSize);

            delaySlotInfo[tid].targetReady = false;
        } else {
            PC[tid] = next_PC;
            nextPC[tid] = next_NPC;
            nextNPC[tid] = next_NPC + instSize;
        }

        DPRINTF(Fetch, "[tid:%i]: Setting PC to %08p.\n", tid, PC[tid]);
#else
        DPRINTF(Fetch, "[tid:%i]: Setting PC to %08p.\n",tid, next_PC);
        PC[tid] = next_PC;
        nextPC[tid] = next_PC + instSize;
#endif
    } else {
        // We shouldn't be in an icache miss and also have a fault (an ITB
        // miss)
        if (fetchStatus[tid] == IcacheWaitResponse) {
            panic("Fetch should have exited prior to this!");
        }

        // Send the fault to commit.  This thread will not do anything
        // until commit handles the fault.  The only other way it can
        // wake up is if a squash comes along and changes the PC.
#if FULL_SYSTEM
        assert(numInst != fetchWidth);
        // Get a sequence number.
        inst_seq = cpu->getAndIncrementInstSeq();
        // We will use a nop in order to carry the fault.
        ext_inst = TheISA::NoopMachInst;

        // Create a new DynInst from the dummy nop.
        DynInstPtr instruction = new DynInst(ext_inst, fetch_PC,
                                             next_PC,
                                             inst_seq, cpu);
        instruction->setPredTarg(next_PC + instSize);
        instruction->setTid(tid);

        instruction->setASID(tid);

        instruction->setThreadState(cpu->thread[tid]);

        instruction->traceData = NULL;

        instruction->setInstListIt(cpu->addInst(instruction));

        instruction->fault = fault;

        toDecode->insts[numInst] = instruction;
        toDecode->size++;

        DPRINTF(Fetch, "[tid:%i]: Blocked, need to handle the trap.\n",tid);

        fetchStatus[tid] = TrapPending;
        status_change = true;
#else // !FULL_SYSTEM
        fetchStatus[tid] = TrapPending;
        status_change = true;

#endif // FULL_SYSTEM
        DPRINTF(Fetch, "[tid:%i]: fault (%s) detected @ PC %08p",
                tid, fault->name(), PC[tid]);
    }
}

template<class Impl>
void
DefaultFetch<Impl>::recvRetry()
{
    if (retryPkt != NULL) {
        assert(cacheBlocked);
        assert(retryTid != -1);
        assert(fetchStatus[retryTid] == IcacheWaitRetry);

        if (icachePort->sendTiming(retryPkt)) {
            fetchStatus[retryTid] = IcacheWaitResponse;
            retryPkt = NULL;
            retryTid = -1;
            cacheBlocked = false;
        }
    } else {
        assert(retryTid == -1);
        // Access has been squashed since it was sent out.  Just clear
        // the cache being blocked.
        cacheBlocked = false;
    }
}

///////////////////////////////////////
//                                   //
//  SMT FETCH POLICY MAINTAINED HERE //
//                                   //
///////////////////////////////////////
template<class Impl>
int
DefaultFetch<Impl>::getFetchingThread(FetchPriority &fetch_priority)
{
    if (numThreads > 1) {
        switch (fetch_priority) {

          case SingleThread:
            return 0;

          case RoundRobin:
            return roundRobin();

          case IQ:
            return iqCount();

          case LSQ:
            return lsqCount();

          case Branch:
            return branchCount();

          default:
            return -1;
        }
    } else {
        std::list<unsigned>::iterator thread = activeThreads->begin();
        assert(thread != activeThreads->end());
        int tid = *thread;

        if (fetchStatus[tid] == Running ||
            fetchStatus[tid] == IcacheAccessComplete ||
            fetchStatus[tid] == Idle) {
            return tid;
        } else {
            return -1;
        }
    }

}


template<class Impl>
int
DefaultFetch<Impl>::roundRobin()
{
    std::list<unsigned>::iterator pri_iter = priorityList.begin();
    std::list<unsigned>::iterator end      = priorityList.end();

    int high_pri;

    while (pri_iter != end) {
        high_pri = *pri_iter;

        assert(high_pri <= numThreads);

        if (fetchStatus[high_pri] == Running ||
            fetchStatus[high_pri] == IcacheAccessComplete ||
            fetchStatus[high_pri] == Idle) {

            priorityList.erase(pri_iter);
            priorityList.push_back(high_pri);

            return high_pri;
        }

        pri_iter++;
    }

    return -1;
}

template<class Impl>
int
DefaultFetch<Impl>::iqCount()
{
    std::priority_queue<unsigned> PQ;

    std::list<unsigned>::iterator threads = activeThreads->begin();
    std::list<unsigned>::iterator end = activeThreads->end();

    while (threads != end) {
        unsigned tid = *threads++;

        PQ.push(fromIEW->iewInfo[tid].iqCount);
    }

    while (!PQ.empty()) {

        unsigned high_pri = PQ.top();

        if (fetchStatus[high_pri] == Running ||
            fetchStatus[high_pri] == IcacheAccessComplete ||
            fetchStatus[high_pri] == Idle)
            return high_pri;
        else
            PQ.pop();

    }

    return -1;
}

template<class Impl>
int
DefaultFetch<Impl>::lsqCount()
{
    std::priority_queue<unsigned> PQ;

    std::list<unsigned>::iterator threads = activeThreads->begin();
    std::list<unsigned>::iterator end = activeThreads->end();

    while (threads != end) {
        unsigned tid = *threads++;

        PQ.push(fromIEW->iewInfo[tid].ldstqCount);
    }

    while (!PQ.empty()) {

        unsigned high_pri = PQ.top();

        if (fetchStatus[high_pri] == Running ||
            fetchStatus[high_pri] == IcacheAccessComplete ||
            fetchStatus[high_pri] == Idle)
            return high_pri;
        else
            PQ.pop();

    }

    return -1;
}

template<class Impl>
int
DefaultFetch<Impl>::branchCount()
{
    std::list<unsigned>::iterator thread = activeThreads->begin();
    assert(thread != activeThreads->end());
    unsigned tid = *thread;

    panic("Branch Count Fetch policy unimplemented\n");
    return 0 * tid;
}