-
Notifications
You must be signed in to change notification settings - Fork 332
/
adl_grt_ratios.py
2270 lines (2004 loc) · 67.2 KB
/
adl_grt_ratios.py
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
# -*- coding: latin-1 -*-
#
# auto generated TopDown/TMA 3.51 description for Intel 12th gen Core (code name Alderlake) with GraceMont
# Please see http://ark.intel.com for more details on these CPUs.
#
# References:
# http://bit.ly/tma-ispass14
# http://halobates.de/blog/p/262
# https://sites.google.com/site/analysismethods/yasin-pubs
# https://download.01.org/perfmon/
# https://github.com/andikleen/pmu-tools/wiki/toplev-manual
#
# Helpers
print_error = lambda msg: False
version = "3.51"
base_frequency = -1.0
Memory = 0
Average_Frequency = 0.0
use_aux = False
def handle_error(obj, msg):
print_error(msg)
obj.errcount += 1
obj.val = 0
obj.thresh = False
def handle_error_metric(obj, msg):
print_error(msg)
obj.errcount += 1
obj.val = 0
# Constants
# Aux. formulas
# pipeline allocation width
def Pipeline_Width(self, EV, level):
return 5
def MEM_BOUND_STALLS_AT_RET_CORRECTION(self, EV, level):
return max((EV("MEM_BOUND_STALLS.LOAD", level) - EV("LD_HEAD.L1_MISS_AT_RET", level)) / CLKS(self, EV, level) , 0 )
def CLKS(self, EV, level):
return EV("CPU_CLK_UNHALTED.CORE", level)
def CLKS_P(self, EV, level):
return EV("CPU_CLK_UNHALTED.CORE_P", level)
def SLOTS(self, EV, level):
return Pipeline_Width(self, EV, level) * CLKS(self, EV, level)
# Instructions Per Cycle
def IPC(self, EV, level):
return EV("INST_RETIRED.ANY", level) / CLKS(self, EV, level)
# Cycles Per Instruction
def CPI(self, EV, level):
return CLKS(self, EV, level) / EV("INST_RETIRED.ANY", level)
# Uops Per Instruction
def UPI(self, EV, level):
return EV("UOPS_RETIRED.ALL", level) / EV("INST_RETIRED.ANY", level)
# Percentage of total non-speculative loads with a store forward or unknown store address block
def Store_Fwd_Blocks(self, EV, level):
return 100 * EV("LD_BLOCKS.DATA_UNKNOWN", level) / EV("MEM_UOPS_RETIRED.ALL_LOADS", level)
# Percentage of total non-speculative loads with a address aliasing block
def Address_Alias_Blocks(self, EV, level):
return 100 * EV("LD_BLOCKS.4K_ALIAS", level) / EV("MEM_UOPS_RETIRED.ALL_LOADS", level)
# Percentage of total non-speculative loads that are splits
def Load_Splits(self, EV, level):
return 100 * EV("MEM_UOPS_RETIRED.SPLIT_LOADS", level) / EV("MEM_UOPS_RETIRED.ALL_LOADS", level)
# Instructions per Branch (lower number means higher occurrence rate)
def IpBranch(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("BR_INST_RETIRED.ALL_BRANCHES", level)
# Instruction per (near) call (lower number means higher occurrence rate)
def IpCall(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("BR_INST_RETIRED.CALL", level)
# Instructions per Load
def IpLoad(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("MEM_UOPS_RETIRED.ALL_LOADS", level)
# Instructions per Store
def IpStore(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("MEM_UOPS_RETIRED.ALL_STORES", level)
# Instructions per retired Branch Misprediction
def IpMispredict(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("BR_MISP_RETIRED.ALL_BRANCHES", level)
# Instructions per retired conditional Branch Misprediction where the branch was not taken
def IpMisp_Cond_Ntaken(self, EV, level):
return EV("INST_RETIRED.ANY", level) / (EV("BR_MISP_RETIRED.COND", level) - EV("BR_MISP_RETIRED.COND_TAKEN", level))
# Instructions per retired conditional Branch Misprediction where the branch was taken
def IpMisp_Cond_Taken(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("BR_MISP_RETIRED.COND_TAKEN", level)
# Instructions per retired return Branch Misprediction
def IpMisp_Ret(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("BR_MISP_RETIRED.RETURN", level)
# Instructions per retired indirect call or jump Branch Misprediction
def IpMisp_Indirect(self, EV, level):
return EV("INST_RETIRED.ANY", level) / EV("BR_MISP_RETIRED.INDIRECT", level)
# Instructions per Far Branch
def IpFarBranch(self, EV, level):
return EV("INST_RETIRED.ANY", level) / (EV("BR_INST_RETIRED.FAR_BRANCH", level) / 2 )
# Ratio of all branches which mispredict
def Branch_Mispredict_Ratio(self, EV, level):
return EV("BR_MISP_RETIRED.ALL_BRANCHES", level) / EV("BR_INST_RETIRED.ALL_BRANCHES", level)
# Ratio between Mispredicted branches and unknown branches
def Branch_Mispredict_to_Unknown_Branch_Ratio(self, EV, level):
return EV("BR_MISP_RETIRED.ALL_BRANCHES", level) / EV("BACLEARS.ANY", level)
# Percentage of all uops which are ucode ops
def Microcode_Uop_Ratio(self, EV, level):
return 100 * EV("UOPS_RETIRED.MS", level) / EV("UOPS_RETIRED.ALL", level)
# Percentage of all uops which are FPDiv uops
def FPDiv_Uop_Ratio(self, EV, level):
return 100 * EV("UOPS_RETIRED.FPDIV", level) / EV("UOPS_RETIRED.ALL", level)
# Percentage of all uops which are IDiv uops
def IDiv_Uop_Ratio(self, EV, level):
return 100 * EV("UOPS_RETIRED.IDIV", level) / EV("UOPS_RETIRED.ALL", level)
# Percentage of all uops which are x87 uops
def X87_Uop_Ratio(self, EV, level):
return 100 * EV("UOPS_RETIRED.X87", level) / EV("UOPS_RETIRED.ALL", level)
# Average Frequency Utilization relative nominal frequency
def Turbo_Utilization(self, EV, level):
return CLKS(self, EV, level) / EV("CPU_CLK_UNHALTED.REF_TSC", level)
# Fraction of cycles spent in Kernel mode
def Kernel_Utilization(self, EV, level):
return EV("CPU_CLK_UNHALTED.CORE:sup", level) / EV("CPU_CLK_UNHALTED.CORE", level)
# Average CPU Utilization
def CPU_Utilization(self, EV, level):
return EV("CPU_CLK_UNHALTED.REF_TSC", level) / EV("msr/tsc/", 0)
# Cycle cost per L2 hit
def Cycles_per_Demand_Load_L2_Hit(self, EV, level):
return EV("MEM_BOUND_STALLS.LOAD_L2_HIT", level) / EV("MEM_LOAD_UOPS_RETIRED.L2_HIT", level)
# Cycle cost per LLC hit
def Cycles_per_Demand_Load_L3_Hit(self, EV, level):
return EV("MEM_BOUND_STALLS.LOAD_LLC_HIT", level) / EV("MEM_LOAD_UOPS_RETIRED.L3_HIT", level)
# Cycle cost per DRAM hit
def Cycles_per_Demand_Load_DRAM_Hit(self, EV, level):
return EV("MEM_BOUND_STALLS.LOAD_DRAM_HIT", level) / EV("MEM_LOAD_UOPS_RETIRED.DRAM_HIT", level)
# Percent of instruction miss cost that hit in the L2
def Inst_Miss_Cost_L2Hit_Percent(self, EV, level):
return 100 * EV("MEM_BOUND_STALLS.IFETCH_L2_HIT", level) / (EV("MEM_BOUND_STALLS.IFETCH", level))
# Percent of instruction miss cost that hit in the L3
def Inst_Miss_Cost_L3Hit_Percent(self, EV, level):
return 100 * EV("MEM_BOUND_STALLS.IFETCH_LLC_HIT", level) / (EV("MEM_BOUND_STALLS.IFETCH", level))
# Percent of instruction miss cost that hit in DRAM
def Inst_Miss_Cost_DRAMHit_Percent(self, EV, level):
return 100 * EV("MEM_BOUND_STALLS.IFETCH_DRAM_HIT", level) / (EV("MEM_BOUND_STALLS.IFETCH", level))
# load ops retired per 1000 instruction
def MemLoadPKI(self, EV, level):
return 1000 * EV("MEM_UOPS_RETIRED.ALL_LOADS", level) / EV("INST_RETIRED.ANY", level)
# Event groups
class Frontend_Bound:
name = "Frontend_Bound"
domain = "Slots"
area = "FE"
level = 1
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.ALL", 1) / SLOTS(self, EV, 1)
self.thresh = (self.val > 0.20)
except ZeroDivisionError:
handle_error(self, "Frontend_Bound zero division")
return self.val
desc = """
Counts the number of issue slots that were not consumed by
the backend due to frontend stalls."""
class Fetch_Latency:
name = "Fetch_Latency"
domain = "Slots"
area = "FE"
level = 2
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.FRONTEND_LATENCY", 2) / SLOTS(self, EV, 2)
self.thresh = (self.val > 0.15)
except ZeroDivisionError:
handle_error(self, "Fetch_Latency zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to frontend bandwidth restrictions due to
decode, predecode, cisc, and other limitations."""
class ICache_Misses:
name = "ICache_Misses"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.ICACHE", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "ICache_Misses zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to instruction cache misses."""
class ITLB_Misses:
name = "ITLB_Misses"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.ITLB", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "ITLB_Misses zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to Instruction Table Lookaside Buffer
(ITLB) misses."""
class Branch_Detect:
name = "Branch_Detect"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.BRANCH_DETECT", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Branch_Detect zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to BACLEARS, which occurs when the Branch
Target Buffer (BTB) prediction or lack thereof, was
corrected by a later branch predictor in the frontend.
Includes BACLEARS due to all branch types including
conditional and unconditional jumps, returns, and indirect
branches."""
class Branch_Resteer:
name = "Branch_Resteer"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.BRANCH_RESTEER", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Branch_Resteer zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to BTCLEARS, which occurs when the Branch
Target Buffer (BTB) predicts a taken branch."""
class Fetch_Bandwidth:
name = "Fetch_Bandwidth"
domain = "Slots"
area = "FE"
level = 2
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.FRONTEND_BANDWIDTH", 2) / SLOTS(self, EV, 2)
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "Fetch_Bandwidth zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to frontend bandwidth restrictions due to
decode, predecode, cisc, and other limitations."""
class Cisc:
name = "Cisc"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.CISC", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Cisc zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to the microcode sequencer (MS)."""
class Decode:
name = "Decode"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.DECODE", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Decode zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to decode stalls."""
class Predecode:
name = "Predecode"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.PREDECODE", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Predecode zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to wrong predecodes."""
class Other_FB:
name = "Other_FB"
domain = "Slots"
area = "FE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_FE_BOUND.OTHER", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Other_FB zero division")
return self.val
desc = """
Counts the number of issue slots that were not delivered by
the frontend due to other common frontend stalls not
categorized."""
class Bad_Speculation:
name = "Bad_Speculation"
domain = "Slots"
area = "BAD"
level = 1
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = (SLOTS(self, EV, 1) - (EV("TOPDOWN_FE_BOUND.ALL", 1) + EV("TOPDOWN_BE_BOUND.ALL", 1) + EV("TOPDOWN_RETIRING.ALL", 1))) / SLOTS(self, EV, 1)
self.thresh = (self.val > 0.15)
except ZeroDivisionError:
handle_error(self, "Bad_Speculation zero division")
return self.val
desc = """
Counts the total number of issue slots that were not
consumed by the backend because allocation is stalled due to
a mispredicted jump or a machine clear. Only issue slots
wasted due to fast nukes such as memory ordering nukes are
counted. Other nukes are not accounted for. Counts all issue
slots blocked during this recovery window including relevant
microcode flows and while uops are not yet available in the
instruction queue (IQ). Also includes the issue slots that
were consumed by the backend but were thrown away because
they were younger than the mispredict or machine clear."""
class Branch_Mispredicts:
name = "Branch_Mispredicts"
domain = "Slots"
area = "BAD"
level = 2
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_BAD_SPECULATION.MISPREDICT", 2) / SLOTS(self, EV, 2)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Branch_Mispredicts zero division")
return self.val
desc = """
Counts the number of issue slots that were not consumed by
the backend due to branch mispredicts."""
class Machine_Clears:
name = "Machine_Clears"
domain = "Slots"
area = "BAD"
level = 2
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_BAD_SPECULATION.MACHINE_CLEARS", 2) / SLOTS(self, EV, 2)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Machine_Clears zero division")
return self.val
desc = """
Counts the total number of issue slots that were not
consumed by the backend because allocation is stalled due to
a machine clear (nuke) of any kind including memory ordering
and memory disambiguation."""
class Nuke:
name = "Nuke"
domain = "Slots"
area = "BAD"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_BAD_SPECULATION.NUKE", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Nuke zero division")
return self.val
desc = """
Counts the number of issue slots that were not consumed by
the backend due to a machine clear (slow nuke)."""
class SMC:
name = "SMC"
domain = "Count"
area = "BAD"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = self.Nuke.compute(EV) * (EV("MACHINE_CLEARS.SMC", 4) / EV("MACHINE_CLEARS.SLOW", 4))
self.thresh = (self.val > 0.02)
except ZeroDivisionError:
handle_error(self, "SMC zero division")
return self.val
desc = """
Counts the number of machine clears relative to the number
of nuke slots due to SMC."""
class Memory_Ordering:
name = "Memory_Ordering"
domain = "Count"
area = "BAD"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = self.Nuke.compute(EV) * (EV("MACHINE_CLEARS.MEMORY_ORDERING", 4) / EV("MACHINE_CLEARS.SLOW", 4))
self.thresh = (self.val > 0.02)
except ZeroDivisionError:
handle_error(self, "Memory_Ordering zero division")
return self.val
desc = """
Counts the number of machine clears relative to the number
of nuke slots due to memory ordering."""
class FP_Assist:
name = "FP_Assist"
domain = "Count"
area = "BAD"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = self.Nuke.compute(EV) * (EV("MACHINE_CLEARS.FP_ASSIST", 4) / EV("MACHINE_CLEARS.SLOW", 4))
self.thresh = (self.val > 0.02)
except ZeroDivisionError:
handle_error(self, "FP_Assist zero division")
return self.val
desc = """
Counts the number of machine clears relative to the number
of nuke slots due to FP assists."""
class Disambiguation:
name = "Disambiguation"
domain = "Count"
area = "BAD"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = self.Nuke.compute(EV) * (EV("MACHINE_CLEARS.DISAMBIGUATION", 4) / EV("MACHINE_CLEARS.SLOW", 4))
self.thresh = (self.val > 0.02)
except ZeroDivisionError:
handle_error(self, "Disambiguation zero division")
return self.val
desc = """
Counts the number of machine clears relative to the number
of nuke slots due to memory disambiguation."""
class Page_Fault:
name = "Page_Fault"
domain = "Count"
area = "BAD"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = self.Nuke.compute(EV) * (EV("MACHINE_CLEARS.PAGE_FAULT", 4) / EV("MACHINE_CLEARS.SLOW", 4))
self.thresh = (self.val > 0.02)
except ZeroDivisionError:
handle_error(self, "Page_Fault zero division")
return self.val
desc = """
Counts the number of machine clears relative to the number
of nuke slots due to page faults."""
class Fast_Nuke:
name = "Fast_Nuke"
domain = "Slots"
area = "BAD"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_BAD_SPECULATION.FASTNUKE", 3) / SLOTS(self, EV, 3)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Fast_Nuke zero division")
return self.val
desc = """
Counts the number of issue slots that were not consumed by
the backend due to a machine clear classified as a fast nuke
due to memory ordering, memory disambiguation and memory
renaming."""
class Backend_Bound:
name = "Backend_Bound"
domain = "Slots"
area = "BE"
level = 1
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("TOPDOWN_BE_BOUND.ALL", 1) / SLOTS(self, EV, 1)
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "Backend_Bound zero division")
return self.val
desc = """
Counts the total number of issue slots that were not
consumed by the backend due to backend stalls. Note that
uops must be available for consumption in order for this
event to count. If a uop is not available (IQ is empty),
this event will not count. The rest of these subevents
count backend stalls, in cycles, due to an outstanding
request which is memory bound vs core bound. The subevents
are not slot based events and therefore can not be precisely
added or subtracted from the Backend_Bound_Aux subevents
which are slot based."""
class Core_Bound:
name = "Core_Bound"
domain = "Cycles"
area = "BE"
level = 2
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = max(0 , self.Backend_Bound.compute(EV) - self.Memory_Bound.compute(EV))
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "Core_Bound zero division")
return self.val
desc = """
Counts the number of cycles due to backend bound stalls that
are core execution bound and not attributed to outstanding
demand load or store stalls."""
class Memory_Bound:
name = "Memory_Bound"
domain = "Cycles"
area = "BE"
level = 2
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = min((EV("TOPDOWN_BE_BOUND.ALL", 2) / SLOTS(self, EV, 2)) , (EV("LD_HEAD.ANY_AT_RET", 2) / CLKS(self, EV, 2)) + self.Store_Bound.compute(EV))
self.thresh = (self.val > 0.20)
except ZeroDivisionError:
handle_error(self, "Memory_Bound zero division")
return self.val
desc = """
Counts the number of cycles the core is stalled due to
stores or loads."""
class Store_Bound:
name = "Store_Bound"
domain = "Cycles"
area = "BE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = self.Mem_Scheduler.compute(EV) * (EV("MEM_SCHEDULER_BLOCK.ST_BUF", 3) / EV("MEM_SCHEDULER_BLOCK.ALL", 3))
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "Store_Bound zero division")
return self.val
desc = """
Counts the number of cycles the core is stalled due to store
buffer full."""
class L1_Bound:
name = "L1_Bound"
domain = "Cycles"
area = "BE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("LD_HEAD.L1_BOUND_AT_RET", 3) / CLKS(self, EV, 3)
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "L1_Bound zero division")
return self.val
desc = """
Counts the number of cycles that the oldest load of the load
buffer is stalled at retirement due to a load block."""
class Store_Fwd_Blk:
name = "Store_Fwd_Blk"
domain = "Cycles"
area = "BE"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("LD_HEAD.ST_ADDR_AT_RET", 4) / CLKS(self, EV, 4)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Store_Fwd_Blk zero division")
return self.val
desc = """
Counts the number of cycles that the oldest load of the load
buffer is stalled at retirement due to a store forward
block."""
class STLB_Hit:
name = "STLB_Hit"
domain = "Cycles"
area = "BE"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("LD_HEAD.DTLB_MISS_AT_RET", 4) / CLKS(self, EV, 4)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "STLB_Hit zero division")
return self.val
desc = """
Counts the number of cycles that the oldest load of the load
buffer is stalled at retirement due to a first level TLB
miss."""
class STLB_Miss:
name = "STLB_Miss"
domain = "Cycles"
area = "BE"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("LD_HEAD.PGWALK_AT_RET", 4) / CLKS(self, EV, 4)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "STLB_Miss zero division")
return self.val
desc = """
Counts the number of cycles that the oldest load of the load
buffer is stalled at retirement due to a second level TLB
miss requiring a page walk."""
class Other_L1:
name = "Other_L1"
domain = "Cycles"
area = "BE"
level = 4
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = EV("LD_HEAD.OTHER_AT_RET", 4) / CLKS(self, EV, 4)
self.thresh = (self.val > 0.05)
except ZeroDivisionError:
handle_error(self, "Other_L1 zero division")
return self.val
desc = """
Counts the number of cycles that the oldest load of the load
buffer is stalled at retirement due to a number of other
load blocks."""
class L2_Bound:
name = "L2_Bound"
domain = "Cycles"
area = "BE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = (EV("MEM_BOUND_STALLS.LOAD_L2_HIT", 3) / CLKS(self, EV, 3)) - (MEM_BOUND_STALLS_AT_RET_CORRECTION(self, EV, 3) * EV("MEM_BOUND_STALLS.LOAD_L2_HIT", 3) / EV("MEM_BOUND_STALLS.LOAD", 3))
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "L2_Bound zero division")
return self.val
desc = """
Counts the number of cycles a core is stalled due to a
demand load which hit in the L2 Cache."""
class L3_Bound:
name = "L3_Bound"
domain = "Cycles"
area = "BE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = (EV("MEM_BOUND_STALLS.LOAD_LLC_HIT", 3) / CLKS(self, EV, 3)) - (MEM_BOUND_STALLS_AT_RET_CORRECTION(self, EV, 3) * EV("MEM_BOUND_STALLS.LOAD_LLC_HIT", 3) / EV("MEM_BOUND_STALLS.LOAD", 3))
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "L3_Bound zero division")
return self.val
desc = """
Counts the number of cycles a core is stalled due to a
demand load which hit in the Last Level Cache (LLC) or other
core with HITE/F/M."""
class DRAM_Bound:
name = "DRAM_Bound"
domain = "Cycles"
area = "BE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = (EV("MEM_BOUND_STALLS.LOAD_DRAM_HIT", 3) / CLKS(self, EV, 3)) - (MEM_BOUND_STALLS_AT_RET_CORRECTION(self, EV, 3) * EV("MEM_BOUND_STALLS.LOAD_DRAM_HIT", 3) / EV("MEM_BOUND_STALLS.LOAD", 3))
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "DRAM_Bound zero division")
return self.val
desc = """
Counts the number of cycles the core is stalled due to a
demand load miss which hit in DRAM or MMIO (Non-DRAM)."""
class Other_Load_Store:
name = "Other_Load_Store"
domain = "Cycles"
area = "BE"
level = 3
htoff = False
sample = []
errcount = 0
sibling = None
metricgroup = frozenset([])
maxval = None
def compute(self, EV):
try:
self.val = max(0 , self.Memory_Bound.compute(EV) - (self.Store_Bound.compute(EV) + self.L1_Bound.compute(EV) + self.L2_Bound.compute(EV) + self.L3_Bound.compute(EV) + self.DRAM_Bound.compute(EV)))
self.thresh = (self.val > 0.10)
except ZeroDivisionError:
handle_error(self, "Other_Load_Store zero division")