160 lines · plain
1# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py2# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=btver2 -timeline -timeline-max-iterations=1 -register-file-stats < %s | FileCheck %s3 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 9pcmpeqb %mm2, %mm210pcmpeqd %mm2, %mm211pcmpeqw %mm2, %mm212 13pcmpeqb %xmm2, %xmm214pcmpeqd %xmm2, %xmm215pcmpeqq %xmm2, %xmm216pcmpeqw %xmm2, %xmm217 18vpcmpeqb %xmm3, %xmm3, %xmm319vpcmpeqd %xmm3, %xmm3, %xmm320vpcmpeqq %xmm3, %xmm3, %xmm321vpcmpeqw %xmm3, %xmm3, %xmm322 23vpcmpeqb %xmm3, %xmm3, %xmm524vpcmpeqd %xmm3, %xmm3, %xmm525vpcmpeqq %xmm3, %xmm3, %xmm526vpcmpeqw %xmm3, %xmm3, %xmm527 28# FIXME: their handling is broken in llvm-mca.29 30# CHECK: Iterations: 10031# CHECK-NEXT: Instructions: 150032# CHECK-NEXT: Total Cycles: 75333# CHECK-NEXT: Total uOps: 150034 35# CHECK: Dispatch Width: 236# CHECK-NEXT: uOps Per Cycle: 1.9937# CHECK-NEXT: IPC: 1.9938# CHECK-NEXT: Block RThroughput: 7.539 40# CHECK: Instruction Info:41# CHECK-NEXT: [1]: #uOps42# CHECK-NEXT: [2]: Latency43# CHECK-NEXT: [3]: RThroughput44# CHECK-NEXT: [4]: MayLoad45# CHECK-NEXT: [5]: MayStore46# CHECK-NEXT: [6]: HasSideEffects (U)47 48# CHECK: [1] [2] [3] [4] [5] [6] Instructions:49# CHECK-NEXT: 1 1 0.50 pcmpeqb %mm2, %mm250# CHECK-NEXT: 1 1 0.50 pcmpeqd %mm2, %mm251# CHECK-NEXT: 1 1 0.50 pcmpeqw %mm2, %mm252# CHECK-NEXT: 1 1 0.50 pcmpeqb %xmm2, %xmm253# CHECK-NEXT: 1 1 0.50 pcmpeqd %xmm2, %xmm254# CHECK-NEXT: 1 1 0.50 pcmpeqq %xmm2, %xmm255# CHECK-NEXT: 1 1 0.50 pcmpeqw %xmm2, %xmm256# CHECK-NEXT: 1 1 0.50 vpcmpeqb %xmm3, %xmm3, %xmm357# CHECK-NEXT: 1 1 0.50 vpcmpeqd %xmm3, %xmm3, %xmm358# CHECK-NEXT: 1 1 0.50 vpcmpeqq %xmm3, %xmm3, %xmm359# CHECK-NEXT: 1 1 0.50 vpcmpeqw %xmm3, %xmm3, %xmm360# CHECK-NEXT: 1 1 0.50 vpcmpeqb %xmm3, %xmm3, %xmm561# CHECK-NEXT: 1 1 0.50 vpcmpeqd %xmm3, %xmm3, %xmm562# CHECK-NEXT: 1 1 0.50 vpcmpeqq %xmm3, %xmm3, %xmm563# CHECK-NEXT: 1 1 0.50 vpcmpeqw %xmm3, %xmm3, %xmm564 65# CHECK: Register File statistics:66# CHECK-NEXT: Total number of mappings created: 150067# CHECK-NEXT: Max number of mappings used: 668 69# CHECK: * Register File #1 -- JFpuPRF:70# CHECK-NEXT: Number of physical registers: 7271# CHECK-NEXT: Total number of mappings created: 150072# CHECK-NEXT: Max number of mappings used: 673 74# CHECK: * Register File #2 -- JIntegerPRF:75# CHECK-NEXT: Number of physical registers: 6476# CHECK-NEXT: Total number of mappings created: 077# CHECK-NEXT: Max number of mappings used: 078 79# CHECK: Resources:80# CHECK-NEXT: [0] - JALU081# CHECK-NEXT: [1] - JALU182# CHECK-NEXT: [2] - JDiv83# CHECK-NEXT: [3] - JFPA84# CHECK-NEXT: [4] - JFPM85# CHECK-NEXT: [5] - JFPU086# CHECK-NEXT: [6] - JFPU187# CHECK-NEXT: [7] - JLAGU88# CHECK-NEXT: [8] - JMul89# CHECK-NEXT: [9] - JSAGU90# CHECK-NEXT: [10] - JSTC91# CHECK-NEXT: [11] - JVALU092# CHECK-NEXT: [12] - JVALU193# CHECK-NEXT: [13] - JVIMUL94 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, %mm2102# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - pcmpeqd %mm2, %mm2103# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - pcmpeqw %mm2, %mm2104# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - pcmpeqb %xmm2, %xmm2105# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - pcmpeqd %xmm2, %xmm2106# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - pcmpeqq %xmm2, %xmm2107# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - pcmpeqw %xmm2, %xmm2108# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqb %xmm3, %xmm3, %xmm3109# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqd %xmm3, %xmm3, %xmm3110# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqq %xmm3, %xmm3, %xmm3111# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqw %xmm3, %xmm3, %xmm3112# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqb %xmm3, %xmm3, %xmm5113# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqd %xmm3, %xmm3, %xmm5114# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqq %xmm3, %xmm3, %xmm5115# CHECK-NEXT: - - - - - 0.50 0.50 - - - - 0.50 0.50 - vpcmpeqw %xmm3, %xmm3, %xmm5116 117# CHECK: Timeline view:118# CHECK-NEXT: 0119# CHECK-NEXT: Index 0123456789120 121# CHECK: [0,0] DeER . . pcmpeqb %mm2, %mm2122# CHECK-NEXT: [0,1] DeER . . pcmpeqd %mm2, %mm2123# CHECK-NEXT: [0,2] .DeER. . pcmpeqw %mm2, %mm2124# CHECK-NEXT: [0,3] .DeER. . pcmpeqb %xmm2, %xmm2125# CHECK-NEXT: [0,4] . DeER . pcmpeqd %xmm2, %xmm2126# CHECK-NEXT: [0,5] . DeER . pcmpeqq %xmm2, %xmm2127# CHECK-NEXT: [0,6] . DeER . pcmpeqw %xmm2, %xmm2128# CHECK-NEXT: [0,7] . DeER . vpcmpeqb %xmm3, %xmm3, %xmm3129# CHECK-NEXT: [0,8] . DeER . vpcmpeqd %xmm3, %xmm3, %xmm3130# CHECK-NEXT: [0,9] . DeER . vpcmpeqq %xmm3, %xmm3, %xmm3131# CHECK-NEXT: [0,10] . DeER . vpcmpeqw %xmm3, %xmm3, %xmm3132# CHECK-NEXT: [0,11] . DeER . vpcmpeqb %xmm3, %xmm3, %xmm5133# CHECK-NEXT: [0,12] . .DeER. vpcmpeqd %xmm3, %xmm3, %xmm5134# CHECK-NEXT: [0,13] . .DeER. vpcmpeqq %xmm3, %xmm3, %xmm5135# CHECK-NEXT: [0,14] . . DeER vpcmpeqw %xmm3, %xmm3, %xmm5136 137# CHECK: Average Wait times (based on the timeline view):138# CHECK-NEXT: [0]: Executions139# CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue140# CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready141# CHECK-NEXT: [3]: Average time elapsed from WB until retire stage142 143# CHECK: [0] [1] [2] [3]144# CHECK-NEXT: 0. 1 1.0 1.0 0.0 pcmpeqb %mm2, %mm2145# CHECK-NEXT: 1. 1 1.0 1.0 0.0 pcmpeqd %mm2, %mm2146# CHECK-NEXT: 2. 1 1.0 1.0 0.0 pcmpeqw %mm2, %mm2147# CHECK-NEXT: 3. 1 1.0 1.0 0.0 pcmpeqb %xmm2, %xmm2148# CHECK-NEXT: 4. 1 1.0 1.0 0.0 pcmpeqd %xmm2, %xmm2149# CHECK-NEXT: 5. 1 1.0 1.0 0.0 pcmpeqq %xmm2, %xmm2150# CHECK-NEXT: 6. 1 1.0 1.0 0.0 pcmpeqw %xmm2, %xmm2151# CHECK-NEXT: 7. 1 1.0 1.0 0.0 vpcmpeqb %xmm3, %xmm3, %xmm3152# CHECK-NEXT: 8. 1 1.0 1.0 0.0 vpcmpeqd %xmm3, %xmm3, %xmm3153# CHECK-NEXT: 9. 1 1.0 1.0 0.0 vpcmpeqq %xmm3, %xmm3, %xmm3154# CHECK-NEXT: 10. 1 1.0 1.0 0.0 vpcmpeqw %xmm3, %xmm3, %xmm3155# CHECK-NEXT: 11. 1 1.0 1.0 0.0 vpcmpeqb %xmm3, %xmm3, %xmm5156# CHECK-NEXT: 12. 1 1.0 1.0 0.0 vpcmpeqd %xmm3, %xmm3, %xmm5157# CHECK-NEXT: 13. 1 1.0 1.0 0.0 vpcmpeqq %xmm3, %xmm3, %xmm5158# CHECK-NEXT: 14. 1 1.0 1.0 0.0 vpcmpeqw %xmm3, %xmm3, %xmm5159# CHECK-NEXT: 1 1.0 1.0 0.0 <total>160