# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py # RUN: llvm-mca -mtriple=aarch64 -mcpu=neoverse-v3ae --timeline --timeline-max-iterations=4 < %s | FileCheck %s # LLVM-MCA-BEGIN FPR8-bit ldr b0, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN FPR16-bit ldr h0, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN FPR32-bit ldr s0, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN FPR64-bit ldr d0, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN FPR128-bit ldr q0, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN SIMD64-bit-b ld1 {v0.8b}, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN SIMD64-bit-h ld1 {v0.4h}, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN SIMD64-bit-s ld1 {v0.2s}, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN SIMD64-bit-d ld1 {v0.1d}, [sp] add z0.d, z0.d, z0.d # LLVM-MCA-END # LLVM-MCA-BEGIN insr insr z0.s, w0 add z0.s, z0.s, z0.s # LLVM-MCA-END # CHECK: [0] Code Region - FPR8-bit # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ldr b0, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ldr b0, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ldr b0, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ldr b0, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ldr b0, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ldr b0, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ldr b0, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [1] Code Region - FPR16-bit # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ldr h0, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ldr h0, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ldr h0, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ldr h0, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ldr h0, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ldr h0, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ldr h0, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [2] Code Region - FPR32-bit # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ldr s0, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ldr s0, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ldr s0, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ldr s0, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ldr s0, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ldr s0, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ldr s0, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [3] Code Region - FPR64-bit # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ldr d0, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ldr d0, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ldr d0, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ldr d0, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ldr d0, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ldr d0, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ldr d0, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [4] Code Region - FPR128-bit # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ldr q0, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ldr q0, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ldr q0, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ldr q0, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ldr q0, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ldr q0, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ldr q0, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [5] Code Region - SIMD64-bit-b # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ld1 { v0.8b }, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ld1 { v0.8b }, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ld1 { v0.8b }, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ld1 { v0.8b }, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ld1 { v0.8b }, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ld1 { v0.8b }, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ld1 { v0.8b }, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [6] Code Region - SIMD64-bit-h # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ld1 { v0.4h }, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ld1 { v0.4h }, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ld1 { v0.4h }, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ld1 { v0.4h }, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ld1 { v0.4h }, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ld1 { v0.4h }, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ld1 { v0.4h }, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [7] Code Region - SIMD64-bit-s # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ld1 { v0.2s }, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ld1 { v0.2s }, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ld1 { v0.2s }, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ld1 { v0.2s }, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ld1 { v0.2s }, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ld1 { v0.2s }, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ld1 { v0.2s }, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [8] Code Region - SIMD64-bit-d # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 60 # CHECK-NEXT: Total uOps: 200 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 3.33 # CHECK-NEXT: IPC: 3.33 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 6 0.33 * ld1 { v0.1d }, [sp] # CHECK-NEXT: 1 2 0.50 add z0.d, z0.d, z0.d # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - 0.50 0.50 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - 0.33 0.33 0.34 - - - - - - - - - - - ld1 { v0.1d }, [sp] # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 0.50 0.50 add z0.d, z0.d, z0.d # CHECK: Timeline view: # CHECK-NEXT: 01 # CHECK-NEXT: Index 0123456789 # CHECK: [0,0] DeeeeeeER .. ld1 { v0.1d }, [sp] # CHECK-NEXT: [0,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [1,0] DeeeeeeE--R. ld1 { v0.1d }, [sp] # CHECK-NEXT: [1,1] D======eeER. add z0.d, z0.d, z0.d # CHECK-NEXT: [2,0] DeeeeeeE--R. ld1 { v0.1d }, [sp] # CHECK-NEXT: [2,1] D=======eeER add z0.d, z0.d, z0.d # CHECK-NEXT: [3,0] D=eeeeeeE--R ld1 { v0.1d }, [sp] # CHECK-NEXT: [3,1] D=======eeER add z0.d, z0.d, z0.d # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 1.3 1.3 1.5 ld1 { v0.1d }, [sp] # CHECK-NEXT: 1. 4 7.5 0.3 0.0 add z0.d, z0.d, z0.d # CHECK-NEXT: 8 4.4 0.8 0.8 # CHECK: [9] Code Region - insr # CHECK: Iterations: 100 # CHECK-NEXT: Instructions: 200 # CHECK-NEXT: Total Cycles: 703 # CHECK-NEXT: Total uOps: 300 # CHECK: Dispatch Width: 10 # CHECK-NEXT: uOps Per Cycle: 0.43 # CHECK-NEXT: IPC: 0.28 # CHECK-NEXT: Block RThroughput: 1.0 # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps # CHECK-NEXT: [2]: Latency # CHECK-NEXT: [3]: RThroughput # CHECK-NEXT: [4]: MayLoad # CHECK-NEXT: [5]: MayStore # CHECK-NEXT: [6]: HasSideEffects (U) # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 2 5 1.00 insr z0.s, w0 # CHECK-NEXT: 1 2 0.50 add z0.s, z0.s, z0.s # CHECK: Resources: # CHECK-NEXT: [0.0] - V3AEUnitB # CHECK-NEXT: [0.1] - V3AEUnitB # CHECK-NEXT: [0.2] - V3AEUnitB # CHECK-NEXT: [1.0] - V3AEUnitD # CHECK-NEXT: [1.1] - V3AEUnitD # CHECK-NEXT: [2.0] - V3AEUnitFlg # CHECK-NEXT: [2.1] - V3AEUnitFlg # CHECK-NEXT: [2.2] - V3AEUnitFlg # CHECK-NEXT: [2.3] - V3AEUnitFlg # CHECK-NEXT: [3.0] - V3AEUnitL12 # CHECK-NEXT: [3.1] - V3AEUnitL12 # CHECK-NEXT: [4] - V3AEUnitLS0 # CHECK-NEXT: [5] - V3AEUnitM0 # CHECK-NEXT: [6] - V3AEUnitM1 # CHECK-NEXT: [7] - V3AEUnitS0 # CHECK-NEXT: [8] - V3AEUnitS1 # CHECK-NEXT: [9] - V3AEUnitS2 # CHECK-NEXT: [10] - V3AEUnitS3 # CHECK-NEXT: [11] - V3AEUnitS4 # CHECK-NEXT: [12] - V3AEUnitS5 # CHECK-NEXT: [13] - V3AEUnitST1 # CHECK-NEXT: [14] - V3AEUnitV0 # CHECK-NEXT: [15] - V3AEUnitV1 # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] # CHECK-NEXT: - - - - - - - - - - - - 1.00 - - - - - - - - 1.00 1.00 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0.0] [0.1] [0.2] [1.0] [1.1] [2.0] [2.1] [2.2] [2.3] [3.0] [3.1] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] Instructions: # CHECK-NEXT: - - - - - - - - - - - - 1.00 - - - - - - - - - 1.00 insr z0.s, w0 # CHECK-NEXT: - - - - - - - - - - - - - - - - - - - - - 1.00 - add z0.s, z0.s, z0.s # CHECK: Timeline view: # CHECK-NEXT: 0123456789 0 # CHECK-NEXT: Index 0123456789 0123456789 # CHECK: [0,0] DeeeeeER . . . . . insr z0.s, w0 # CHECK-NEXT: [0,1] D=====eeER. . . . . add z0.s, z0.s, z0.s # CHECK-NEXT: [1,0] D=======eeeeeER. . . . insr z0.s, w0 # CHECK-NEXT: [1,1] D============eeER . . . add z0.s, z0.s, z0.s # CHECK-NEXT: [2,0] D==============eeeeeER . . insr z0.s, w0 # CHECK-NEXT: [2,1] D===================eeER . . add z0.s, z0.s, z0.s # CHECK-NEXT: [3,0] .D====================eeeeeER . insr z0.s, w0 # CHECK-NEXT: [3,1] .D=========================eeER add z0.s, z0.s, z0.s # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions # CHECK-NEXT: [1]: Average time spent waiting in a scheduler's queue # CHECK-NEXT: [2]: Average time spent waiting in a scheduler's queue while ready # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] # CHECK-NEXT: 0. 4 11.3 0.3 0.0 insr z0.s, w0 # CHECK-NEXT: 1. 4 16.3 0.0 0.0 add z0.s, z0.s, z0.s # CHECK-NEXT: 8 13.8 0.1 0.0