167 lines · plain
1# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py2# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=znver3 -timeline -iterations=1 < %s | FileCheck %s3 4# PR51495: WriteIMulH reports an incorrect latency for the RM variants of MULX.5 6# LLVM-MCA-BEGIN7mulxl (%rdi), %eax, %ecx8add %eax, %eax9# LLVM-MCA-END10 11# LLVM-MCA-BEGIN12mulxq (%rdi), %rax, %rcx13add %rax, %rax14# LLVM-MCA-END15 16# CHECK: [0] Code Region17 18# CHECK: Iterations: 119# CHECK-NEXT: Instructions: 220# CHECK-NEXT: Total Cycles: 1121# CHECK-NEXT: Total uOps: 322 23# CHECK: Dispatch Width: 624# CHECK-NEXT: uOps Per Cycle: 0.2725# CHECK-NEXT: IPC: 0.1826# CHECK-NEXT: Block RThroughput: 1.027 28# CHECK: Instruction Info:29# CHECK-NEXT: [1]: #uOps30# CHECK-NEXT: [2]: Latency31# CHECK-NEXT: [3]: RThroughput32# CHECK-NEXT: [4]: MayLoad33# CHECK-NEXT: [5]: MayStore34# CHECK-NEXT: [6]: HasSideEffects (U)35 36# CHECK: [1] [2] [3] [4] [5] [6] Instructions:37# CHECK-NEXT: 2 8 1.00 * mulxl (%rdi), %eax, %ecx38# CHECK-NEXT: 1 1 0.25 addl %eax, %eax39 40# CHECK: Resources:41# CHECK-NEXT: [0] - Zn3AGU042# CHECK-NEXT: [1] - Zn3AGU143# CHECK-NEXT: [2] - Zn3AGU244# CHECK-NEXT: [3] - Zn3ALU045# CHECK-NEXT: [4] - Zn3ALU146# CHECK-NEXT: [5] - Zn3ALU247# CHECK-NEXT: [6] - Zn3ALU348# CHECK-NEXT: [7] - Zn3BRU149# CHECK-NEXT: [8] - Zn3FP050# CHECK-NEXT: [9] - Zn3FP151# CHECK-NEXT: [10] - Zn3FP252# CHECK-NEXT: [11] - Zn3FP353# CHECK-NEXT: [12.0] - Zn3FP4554# CHECK-NEXT: [12.1] - Zn3FP4555# CHECK-NEXT: [13] - Zn3FPSt56# CHECK-NEXT: [14.0] - Zn3LSU57# CHECK-NEXT: [14.1] - Zn3LSU58# CHECK-NEXT: [14.2] - Zn3LSU59# CHECK-NEXT: [15.0] - Zn3Load60# CHECK-NEXT: [15.1] - Zn3Load61# CHECK-NEXT: [15.2] - Zn3Load62# CHECK-NEXT: [16.0] - Zn3Store63# CHECK-NEXT: [16.1] - Zn3Store64 65# CHECK: Resource pressure per iteration:66# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12.0] [12.1] [13] [14.0] [14.1] [14.2] [15.0] [15.1] [15.2] [16.0] [16.1]67# CHECK-NEXT: - - 1.00 - 1.00 - 1.00 - - - - - - - - - - 1.00 - - 1.00 - -68 69# CHECK: Resource pressure by instruction:70# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12.0] [12.1] [13] [14.0] [14.1] [14.2] [15.0] [15.1] [15.2] [16.0] [16.1] Instructions:71# CHECK-NEXT: - - 1.00 - 1.00 - - - - - - - - - - - - 1.00 - - 1.00 - - mulxl (%rdi), %eax, %ecx72# CHECK-NEXT: - - - - - - 1.00 - - - - - - - - - - - - - - - - addl %eax, %eax73 74# CHECK: Timeline view:75# CHECK-NEXT: 076# CHECK-NEXT: Index 012345678977 78# CHECK: [0,0] DeeeeeeeeER mulxl (%rdi), %eax, %ecx79# CHECK-NEXT: [0,1] D=======eER addl %eax, %eax80 81# CHECK: Average Wait times (based on the timeline view):82# CHECK-NEXT: [0]: Executions83# CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue84# CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready85# CHECK-NEXT: [3]: Average time elapsed from WB until retire stage86 87# CHECK: [0] [1] [2] [3]88# CHECK-NEXT: 0. 1 1.0 1.0 0.0 mulxl (%rdi), %eax, %ecx89# CHECK-NEXT: 1. 1 8.0 0.0 0.0 addl %eax, %eax90# CHECK-NEXT: 1 4.5 0.5 0.0 <total>91 92# CHECK: [1] Code Region93 94# CHECK: Iterations: 195# CHECK-NEXT: Instructions: 296# CHECK-NEXT: Total Cycles: 1197# CHECK-NEXT: Total uOps: 398 99# CHECK: Dispatch Width: 6100# CHECK-NEXT: uOps Per Cycle: 0.27101# CHECK-NEXT: IPC: 0.18102# CHECK-NEXT: Block RThroughput: 1.0103 104# CHECK: Instruction Info:105# CHECK-NEXT: [1]: #uOps106# CHECK-NEXT: [2]: Latency107# CHECK-NEXT: [3]: RThroughput108# CHECK-NEXT: [4]: MayLoad109# CHECK-NEXT: [5]: MayStore110# CHECK-NEXT: [6]: HasSideEffects (U)111 112# CHECK: [1] [2] [3] [4] [5] [6] Instructions:113# CHECK-NEXT: 2 8 1.00 * mulxq (%rdi), %rax, %rcx114# CHECK-NEXT: 1 1 0.25 addq %rax, %rax115 116# CHECK: Resources:117# CHECK-NEXT: [0] - Zn3AGU0118# CHECK-NEXT: [1] - Zn3AGU1119# CHECK-NEXT: [2] - Zn3AGU2120# CHECK-NEXT: [3] - Zn3ALU0121# CHECK-NEXT: [4] - Zn3ALU1122# CHECK-NEXT: [5] - Zn3ALU2123# CHECK-NEXT: [6] - Zn3ALU3124# CHECK-NEXT: [7] - Zn3BRU1125# CHECK-NEXT: [8] - Zn3FP0126# CHECK-NEXT: [9] - Zn3FP1127# CHECK-NEXT: [10] - Zn3FP2128# CHECK-NEXT: [11] - Zn3FP3129# CHECK-NEXT: [12.0] - Zn3FP45130# CHECK-NEXT: [12.1] - Zn3FP45131# CHECK-NEXT: [13] - Zn3FPSt132# CHECK-NEXT: [14.0] - Zn3LSU133# CHECK-NEXT: [14.1] - Zn3LSU134# CHECK-NEXT: [14.2] - Zn3LSU135# CHECK-NEXT: [15.0] - Zn3Load136# CHECK-NEXT: [15.1] - Zn3Load137# CHECK-NEXT: [15.2] - Zn3Load138# CHECK-NEXT: [16.0] - Zn3Store139# CHECK-NEXT: [16.1] - Zn3Store140 141# CHECK: Resource pressure per iteration:142# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12.0] [12.1] [13] [14.0] [14.1] [14.2] [15.0] [15.1] [15.2] [16.0] [16.1]143# CHECK-NEXT: - - 1.00 - 1.00 - 1.00 - - - - - - - - - - 1.00 - - 1.00 - -144 145# CHECK: Resource pressure by instruction:146# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12.0] [12.1] [13] [14.0] [14.1] [14.2] [15.0] [15.1] [15.2] [16.0] [16.1] Instructions:147# CHECK-NEXT: - - 1.00 - 1.00 - - - - - - - - - - - - 1.00 - - 1.00 - - mulxq (%rdi), %rax, %rcx148# CHECK-NEXT: - - - - - - 1.00 - - - - - - - - - - - - - - - - addq %rax, %rax149 150# CHECK: Timeline view:151# CHECK-NEXT: 0152# CHECK-NEXT: Index 0123456789153 154# CHECK: [0,0] DeeeeeeeeER mulxq (%rdi), %rax, %rcx155# CHECK-NEXT: [0,1] D=======eER addq %rax, %rax156 157# CHECK: Average Wait times (based on the timeline view):158# CHECK-NEXT: [0]: Executions159# CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue160# CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready161# CHECK-NEXT: [3]: Average time elapsed from WB until retire stage162 163# CHECK: [0] [1] [2] [3]164# CHECK-NEXT: 0. 1 1.0 1.0 0.0 mulxq (%rdi), %rax, %rcx165# CHECK-NEXT: 1. 1 8.0 0.0 0.0 addq %rax, %rax166# CHECK-NEXT: 1 4.5 0.5 0.0 <total>167