test/tools/llvm-mca/X86/BtVer2/one-idioms.s

   1 # NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py
   2 # RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=btver2 -timeline -timeline-max-iterations=1 -register-file-stats < %s | FileCheck %s
   3
   4 # These are dependency-breaking one-idioms.
   5 # Much like zero-idioms, but they produce ones, and do consume resources.
   6
   7 # perf stats reports a throughput of 2.00 IPC.
   8
   9 pcmpeqb   %mm2, %mm2
  10 pcmpeqd   %mm2, %mm2
  11 pcmpeqw   %mm2, %mm2
  12
  13 pcmpeqb   %xmm2, %xmm2
  14 pcmpeqd   %xmm2, %xmm2
  15 pcmpeqq   %xmm2, %xmm2
  16 pcmpeqw   %xmm2, %xmm2
  17
  18 vpcmpeqb  %xmm3, %xmm3, %xmm3
  19 vpcmpeqd  %xmm3, %xmm3, %xmm3
  20 vpcmpeqq  %xmm3, %xmm3, %xmm3
  21 vpcmpeqw  %xmm3, %xmm3, %xmm3
  22
  23 vpcmpeqb  %xmm3, %xmm3, %xmm5
  24 vpcmpeqd  %xmm3, %xmm3, %xmm5
  25 vpcmpeqq  %xmm3, %xmm3, %xmm5
  26 vpcmpeqw  %xmm3, %xmm3, %xmm5
  27
  28 # FIXME: their handling is broken in llvm-mca.
  29
  30 # CHECK:      Iterations:        100
  31 # CHECK-NEXT: Instructions:      1500
  32 # CHECK-NEXT: Total Cycles:      753
  33 # CHECK-NEXT: Total uOps:        1500
  34
  35 # CHECK:      Dispatch Width:    2
  36 # CHECK-NEXT: uOps Per Cycle:    1.99
  37 # CHECK-NEXT: IPC:               1.99
  38 # CHECK-NEXT: Block RThroughput: 7.5
  39
  40 # CHECK:      Instruction Info:
  41 # CHECK-NEXT: [1]: #uOps
  42 # CHECK-NEXT: [2]: Latency
  43 # CHECK-NEXT: [3]: RThroughput
  44 # CHECK-NEXT: [4]: MayLoad
  45 # CHECK-NEXT: [5]: MayStore
  46 # CHECK-NEXT: [6]: HasSideEffects (U)
  47
  48 # CHECK:      [1]    [2]    [3]    [4]    [5]    [6]    Instructions:
  49 # CHECK-NEXT:  1      1     0.50                        pcmpeqb %mm2, %mm2
  50 # CHECK-NEXT:  1      1     0.50                        pcmpeqd %mm2, %mm2
  51 # CHECK-NEXT:  1      1     0.50                        pcmpeqw %mm2, %mm2
  52 # CHECK-NEXT:  1      1     0.50                        pcmpeqb %xmm2, %xmm2
  53 # CHECK-NEXT:  1      1     0.50                        pcmpeqd %xmm2, %xmm2
  54 # CHECK-NEXT:  1      1     0.50                        pcmpeqq %xmm2, %xmm2
  55 # CHECK-NEXT:  1      1     0.50                        pcmpeqw %xmm2, %xmm2
  56 # CHECK-NEXT:  1      1     0.50                        vpcmpeqb        %xmm3, %xmm3, %xmm3
  57 # CHECK-NEXT:  1      1     0.50                        vpcmpeqd        %xmm3, %xmm3, %xmm3
  58 # CHECK-NEXT:  1      1     0.50                        vpcmpeqq        %xmm3, %xmm3, %xmm3
  59 # CHECK-NEXT:  1      1     0.50                        vpcmpeqw        %xmm3, %xmm3, %xmm3
  60 # CHECK-NEXT:  1      1     0.50                        vpcmpeqb        %xmm3, %xmm3, %xmm5
  61 # CHECK-NEXT:  1      1     0.50                        vpcmpeqd        %xmm3, %xmm3, %xmm5
  62 # CHECK-NEXT:  1      1     0.50                        vpcmpeqq        %xmm3, %xmm3, %xmm5
  63 # CHECK-NEXT:  1      1     0.50                        vpcmpeqw        %xmm3, %xmm3, %xmm5
  64
  65 # CHECK:      Register File statistics:
  66 # CHECK-NEXT: Total number of mappings created:    1500
  67 # CHECK-NEXT: Max number of mappings used:         6
  68
  69 # CHECK:      *  Register File #1 -- JFpuPRF:
  70 # CHECK-NEXT:    Number of physical registers:     72
  71 # CHECK-NEXT:    Total number of mappings created: 1500
  72 # CHECK-NEXT:    Max number of mappings used:      6
  73
  74 # CHECK:      *  Register File #2 -- JIntegerPRF:
  75 # CHECK-NEXT:    Number of physical registers:     64
  76 # CHECK-NEXT:    Total number of mappings created: 0
  77 # CHECK-NEXT:    Max number of mappings used:      0
  78
  79 # CHECK:      Resources:
  80 # CHECK-NEXT: [0]   - JALU0
  81 # CHECK-NEXT: [1]   - JALU1
  82 # CHECK-NEXT: [2]   - JDiv
  83 # CHECK-NEXT: [3]   - JFPA
  84 # CHECK-NEXT: [4]   - JFPM
  85 # CHECK-NEXT: [5]   - JFPU0
  86 # CHECK-NEXT: [6]   - JFPU1
  87 # CHECK-NEXT: [7]   - JLAGU
  88 # CHECK-NEXT: [8]   - JMul
  89 # CHECK-NEXT: [9]   - JSAGU
  90 # CHECK-NEXT: [10]  - JSTC
  91 # CHECK-NEXT: [11]  - JVALU0
  92 # CHECK-NEXT: [12]  - JVALU1
  93 # CHECK-NEXT: [13]  - JVIMUL
  94
  95 # CHECK:      Resource pressure per iteration:
  96 # CHECK-NEXT: [0]    [1]    [2]    [3]    [4]    [5]    [6]    [7]    [8]    [9]    [10]   [11]   [12]   [13]
  97 # CHECK-NEXT:  -      -      -      -      -     7.50   7.50    -      -      -      -     7.50   7.50    -
  98
  99 # CHECK:      Resource pressure by instruction:
 100 # CHECK-NEXT: [0]    [1]    [2]    [3]    [4]    [5]    [6]    [7]    [8]    [9]    [10]   [11]   [12]   [13]   Instructions:
 101 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     pcmpeqb %mm2, %mm2
 102 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     pcmpeqd %mm2, %mm2
 103 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     pcmpeqw %mm2, %mm2
 104 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     pcmpeqb %xmm2, %xmm2
 105 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     pcmpeqd %xmm2, %xmm2
 106 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     pcmpeqq %xmm2, %xmm2
 107 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     pcmpeqw %xmm2, %xmm2
 108 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqb        %xmm3, %xmm3, %xmm3
 109 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqd        %xmm3, %xmm3, %xmm3
 110 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqq        %xmm3, %xmm3, %xmm3
 111 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqw        %xmm3, %xmm3, %xmm3
 112 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqb        %xmm3, %xmm3, %xmm5
 113 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqd        %xmm3, %xmm3, %xmm5
 114 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqq        %xmm3, %xmm3, %xmm5
 115 # CHECK-NEXT:  -      -      -      -      -     0.50   0.50    -      -      -      -     0.50   0.50    -     vpcmpeqw        %xmm3, %xmm3, %xmm5
 116
 117 # CHECK:      Timeline view:
 118 # CHECK-NEXT:                     0
 119 # CHECK-NEXT: Index     0123456789
 120
 121 # CHECK:      [0,0]     DeER .    .   pcmpeqb   %mm2, %mm2
 122 # CHECK-NEXT: [0,1]     DeER .    .   pcmpeqd   %mm2, %mm2
 123 # CHECK-NEXT: [0,2]     .DeER.    .   pcmpeqw   %mm2, %mm2
 124 # CHECK-NEXT: [0,3]     .DeER.    .   pcmpeqb   %xmm2, %xmm2
 125 # CHECK-NEXT: [0,4]     . DeER    .   pcmpeqd   %xmm2, %xmm2
 126 # CHECK-NEXT: [0,5]     . DeER    .   pcmpeqq   %xmm2, %xmm2
 127 # CHECK-NEXT: [0,6]     .  DeER   .   pcmpeqw   %xmm2, %xmm2
 128 # CHECK-NEXT: [0,7]     .  DeER   .   vpcmpeqb  %xmm3, %xmm3, %xmm3
 129 # CHECK-NEXT: [0,8]     .   DeER  .   vpcmpeqd  %xmm3, %xmm3, %xmm3
 130 # CHECK-NEXT: [0,9]     .   DeER  .   vpcmpeqq  %xmm3, %xmm3, %xmm3
 131 # CHECK-NEXT: [0,10]    .    DeER .   vpcmpeqw  %xmm3, %xmm3, %xmm3
 132 # CHECK-NEXT: [0,11]    .    DeER .   vpcmpeqb  %xmm3, %xmm3, %xmm5
 133 # CHECK-NEXT: [0,12]    .    .DeER.   vpcmpeqd  %xmm3, %xmm3, %xmm5
 134 # CHECK-NEXT: [0,13]    .    .DeER.   vpcmpeqq  %xmm3, %xmm3, %xmm5
 135 # CHECK-NEXT: [0,14]    .    . DeER   vpcmpeqw  %xmm3, %xmm3, %xmm5
 136
 137 # CHECK:      Average Wait times (based on the timeline view):
 138 # CHECK-NEXT: [0]: Executions
 139 # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue
 140 # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready
 141 # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage
 142
 143 # CHECK:            [0]    [1]    [2]    [3]
 144 # CHECK-NEXT: 0.     1     1.0    1.0    0.0       pcmpeqb      %mm2, %mm2
 145 # CHECK-NEXT: 1.     1     1.0    1.0    0.0       pcmpeqd      %mm2, %mm2
 146 # CHECK-NEXT: 2.     1     1.0    1.0    0.0       pcmpeqw      %mm2, %mm2
 147 # CHECK-NEXT: 3.     1     1.0    1.0    0.0       pcmpeqb      %xmm2, %xmm2
 148 # CHECK-NEXT: 4.     1     1.0    1.0    0.0       pcmpeqd      %xmm2, %xmm2
 149 # CHECK-NEXT: 5.     1     1.0    1.0    0.0       pcmpeqq      %xmm2, %xmm2
 150 # CHECK-NEXT: 6.     1     1.0    1.0    0.0       pcmpeqw      %xmm2, %xmm2
 151 # CHECK-NEXT: 7.     1     1.0    1.0    0.0       vpcmpeqb     %xmm3, %xmm3, %xmm3
 152 # CHECK-NEXT: 8.     1     1.0    1.0    0.0       vpcmpeqd     %xmm3, %xmm3, %xmm3
 153 # CHECK-NEXT: 9.     1     1.0    1.0    0.0       vpcmpeqq     %xmm3, %xmm3, %xmm3
 154 # CHECK-NEXT: 10.    1     1.0    1.0    0.0       vpcmpeqw     %xmm3, %xmm3, %xmm3
 155 # CHECK-NEXT: 11.    1     1.0    1.0    0.0       vpcmpeqb     %xmm3, %xmm3, %xmm5
 156 # CHECK-NEXT: 12.    1     1.0    1.0    0.0       vpcmpeqd     %xmm3, %xmm3, %xmm5
 157 # CHECK-NEXT: 13.    1     1.0    1.0    0.0       vpcmpeqq     %xmm3, %xmm3, %xmm5
 158 # CHECK-NEXT: 14.    1     1.0    1.0    0.0       vpcmpeqw     %xmm3, %xmm3, %xmm5