| 1 |
# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py |
| 2 |
# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=haswell -iterations=1 -resource-pressure=false -instruction-info=false -timeline < %s | FileCheck %s -check-prefixes=ALL,HASWELL |
| 3 |
|
| 4 |
# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=broadwell -iterations=1 -resource-pressure=false -instruction-info=false -timeline < %s | FileCheck %s -check-prefixes=ALL,BDWELL |
| 5 |
|
| 6 |
# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=skylake -iterations=1 -resource-pressure=false -instruction-info=false -timeline < %s | FileCheck %s -check-prefixes=ALL,SKYLAKE |
| 7 |
|
| 8 |
# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=icelake-server -iterations=1 -resource-pressure=false -instruction-info=false -timeline < %s | FileCheck %s -check-prefixes=ALL,SKYLAKE |
| 9 |
|
| 10 |
# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=znver1 -iterations=1 -resource-pressure=false -instruction-info=false -timeline < %s | FileCheck %s -check-prefixes=ALL,ZNVER1 |
| 11 |
|
| 12 |
# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=znver2 -iterations=1 -resource-pressure=false -instruction-info=false -timeline < %s | FileCheck %s -check-prefixes=ALL,ZNVER1 |
| 13 |
|
| 14 |
vaddps %xmm0, %xmm0, %xmm2 |
| 15 |
vfmadd213ps (%rdi), %xmm1, %xmm2 |
| 16 |
|
| 17 |
# ALL: Iterations: 1 |
| 18 |
# ALL-NEXT: Instructions: 2 |
| 19 |
|
| 20 |
# BDWELL-NEXT: Total Cycles: 13 |
| 21 |
# BDWELL-NEXT: Total uOps: 3 |
| 22 |
|
| 23 |
# HASWELL-NEXT: Total Cycles: 14 |
| 24 |
# HASWELL-NEXT: Total uOps: 3 |
| 25 |
|
| 26 |
# SKYLAKE-NEXT: Total Cycles: 13 |
| 27 |
# SKYLAKE-NEXT: Total uOps: 3 |
| 28 |
|
| 29 |
# ZNVER1-NEXT: Total Cycles: 15 |
| 30 |
# ZNVER1-NEXT: Total uOps: 2 |
| 31 |
|
| 32 |
# BDWELL: Dispatch Width: 4 |
| 33 |
# BDWELL-NEXT: uOps Per Cycle: 0.23 |
| 34 |
# BDWELL-NEXT: IPC: 0.15 |
| 35 |
# BDWELL-NEXT: Block RThroughput: 1.0 |
| 36 |
|
| 37 |
# HASWELL: Dispatch Width: 4 |
| 38 |
# HASWELL-NEXT: uOps Per Cycle: 0.21 |
| 39 |
# HASWELL-NEXT: IPC: 0.14 |
| 40 |
# HASWELL-NEXT: Block RThroughput: 1.0 |
| 41 |
|
| 42 |
# SKYLAKE: Dispatch Width: 6 |
| 43 |
# SKYLAKE-NEXT: uOps Per Cycle: 0.23 |
| 44 |
# SKYLAKE-NEXT: IPC: 0.15 |
| 45 |
# SKYLAKE-NEXT: Block RThroughput: 1.0 |
| 46 |
|
| 47 |
# ZNVER1: Dispatch Width: 4 |
| 48 |
# ZNVER1-NEXT: uOps Per Cycle: 0.13 |
| 49 |
# ZNVER1-NEXT: IPC: 0.13 |
| 50 |
# ZNVER1-NEXT: Block RThroughput: 0.5 |
| 51 |
|
| 52 |
# ALL: Timeline view: |
| 53 |
|
| 54 |
# BDWELL-NEXT: 012 |
| 55 |
# HASWELL-NEXT: 0123 |
| 56 |
# SKYLAKE-NEXT: 012 |
| 57 |
# ZNVER1-NEXT: 01234 |
| 58 |
|
| 59 |
# ALL-NEXT: Index 0123456789 |
| 60 |
|
| 61 |
# BDWELL: [0,0] DeeeER . . vaddps %xmm0, %xmm0, %xmm2 |
| 62 |
# BDWELL-NEXT: [0,1] DeeeeeeeeeeER vfmadd213ps (%rdi), %xmm1, %xmm2 |
| 63 |
|
| 64 |
# HASWELL: [0,0] DeeeER . . vaddps %xmm0, %xmm0, %xmm2 |
| 65 |
# HASWELL-NEXT: [0,1] DeeeeeeeeeeeER vfmadd213ps (%rdi), %xmm1, %xmm2 |
| 66 |
|
| 67 |
# SKYLAKE: [0,0] DeeeeER . . vaddps %xmm0, %xmm0, %xmm2 |
| 68 |
# SKYLAKE-NEXT: [0,1] DeeeeeeeeeeER vfmadd213ps (%rdi), %xmm1, %xmm2 |
| 69 |
|
| 70 |
# ZNVER1: [0,0] DeeeER . . vaddps %xmm0, %xmm0, %xmm2 |
| 71 |
# ZNVER1-NEXT: [0,1] DeeeeeeeeeeeeER vfmadd213ps (%rdi), %xmm1, %xmm2 |
| 72 |
|
| 73 |
# ALL: Average Wait times (based on the timeline view): |
| 74 |
# ALL-NEXT: [0]: Executions |
| 75 |
# ALL-NEXT: [1]: Average time spent waiting in a scheduler's queue |
| 76 |
# ALL-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready |
| 77 |
# ALL-NEXT: [3]: Average time elapsed from WB until retire stage |
| 78 |
|
| 79 |
# ALL: [0] [1] [2] [3] |
| 80 |
# ALL-NEXT: 0. 1 1.0 1.0 0.0 vaddps %xmm0, %xmm0, %xmm2 |
| 81 |
# ALL-NEXT: 1. 1 1.0 0.0 0.0 vfmadd213ps (%rdi), %xmm1, %xmm2 |
| 82 |
# ALL-NEXT: 1 1.0 0.5 0.0 <total> |