255 lines
10 KiB
ArmAsm
255 lines
10 KiB
ArmAsm
# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py
|
|
# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=bdver2 -iterations=500 < %s | FileCheck %s
|
|
|
|
# LLVM-MCA-BEGIN
|
|
vpinsrb $0, %eax, %xmm0, %xmm0
|
|
vpinsrb $1, %eax, %xmm0, %xmm0
|
|
# LLVM-MCA-END
|
|
|
|
# LLVM-MCA-BEGIN
|
|
vpinsrw $0, %eax, %xmm0, %xmm0
|
|
vpinsrw $1, %eax, %xmm0, %xmm0
|
|
# LLVM-MCA-END
|
|
|
|
# LLVM-MCA-BEGIN
|
|
vpinsrd $0, %eax, %xmm0, %xmm0
|
|
vpinsrd $1, %eax, %xmm0, %xmm0
|
|
# LLVM-MCA-END
|
|
|
|
# LLVM-MCA-BEGIN
|
|
vpinsrq $0, %rax, %xmm0, %xmm0
|
|
vpinsrq $1, %rax, %xmm0, %xmm0
|
|
# LLVM-MCA-END
|
|
|
|
# CHECK: [0] Code Region
|
|
|
|
# CHECK: Iterations: 500
|
|
# CHECK-NEXT: Instructions: 1000
|
|
# CHECK-NEXT: Total Cycles: 2003
|
|
# CHECK-NEXT: Total uOps: 2000
|
|
|
|
# CHECK: Dispatch Width: 4
|
|
# CHECK-NEXT: uOps Per Cycle: 1.00
|
|
# CHECK-NEXT: IPC: 0.50
|
|
# CHECK-NEXT: Block RThroughput: 3.0
|
|
|
|
# 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 12 1.50 vpinsrb $0, %eax, %xmm0, %xmm0
|
|
# CHECK-NEXT: 2 12 1.50 vpinsrb $1, %eax, %xmm0, %xmm0
|
|
|
|
# CHECK: Resources:
|
|
# CHECK-NEXT: [0.0] - PdAGLU01
|
|
# CHECK-NEXT: [0.1] - PdAGLU01
|
|
# CHECK-NEXT: [1] - PdBranch
|
|
# CHECK-NEXT: [2] - PdCount
|
|
# CHECK-NEXT: [3] - PdDiv
|
|
# CHECK-NEXT: [4] - PdEX0
|
|
# CHECK-NEXT: [5] - PdEX1
|
|
# CHECK-NEXT: [6] - PdFPCVT
|
|
# CHECK-NEXT: [7.0] - PdFPFMA
|
|
# CHECK-NEXT: [7.1] - PdFPFMA
|
|
# CHECK-NEXT: [8.0] - PdFPMAL
|
|
# CHECK-NEXT: [8.1] - PdFPMAL
|
|
# CHECK-NEXT: [9] - PdFPMMA
|
|
# CHECK-NEXT: [10] - PdFPSTO
|
|
# CHECK-NEXT: [11] - PdFPU0
|
|
# CHECK-NEXT: [12] - PdFPU1
|
|
# CHECK-NEXT: [13] - PdFPU2
|
|
# CHECK-NEXT: [14] - PdFPU3
|
|
# CHECK-NEXT: [15] - PdFPXBR
|
|
# CHECK-NEXT: [16.0] - PdLoad
|
|
# CHECK-NEXT: [16.1] - PdLoad
|
|
# CHECK-NEXT: [17] - PdMul
|
|
# CHECK-NEXT: [18] - PdStore
|
|
|
|
# CHECK: Resource pressure per iteration:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18]
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 3.00 - - 1.00 1.00 - - - - - - -
|
|
|
|
# CHECK: Resource pressure by instruction:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18] Instructions:
|
|
# CHECK-NEXT: - - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - vpinsrb $0, %eax, %xmm0, %xmm0
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - - vpinsrb $1, %eax, %xmm0, %xmm0
|
|
|
|
# CHECK: [1] Code Region
|
|
|
|
# CHECK: Iterations: 500
|
|
# CHECK-NEXT: Instructions: 1000
|
|
# CHECK-NEXT: Total Cycles: 2003
|
|
# CHECK-NEXT: Total uOps: 2000
|
|
|
|
# CHECK: Dispatch Width: 4
|
|
# CHECK-NEXT: uOps Per Cycle: 1.00
|
|
# CHECK-NEXT: IPC: 0.50
|
|
# CHECK-NEXT: Block RThroughput: 3.0
|
|
|
|
# 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 12 1.50 vpinsrw $0, %eax, %xmm0, %xmm0
|
|
# CHECK-NEXT: 2 12 1.50 vpinsrw $1, %eax, %xmm0, %xmm0
|
|
|
|
# CHECK: Resources:
|
|
# CHECK-NEXT: [0.0] - PdAGLU01
|
|
# CHECK-NEXT: [0.1] - PdAGLU01
|
|
# CHECK-NEXT: [1] - PdBranch
|
|
# CHECK-NEXT: [2] - PdCount
|
|
# CHECK-NEXT: [3] - PdDiv
|
|
# CHECK-NEXT: [4] - PdEX0
|
|
# CHECK-NEXT: [5] - PdEX1
|
|
# CHECK-NEXT: [6] - PdFPCVT
|
|
# CHECK-NEXT: [7.0] - PdFPFMA
|
|
# CHECK-NEXT: [7.1] - PdFPFMA
|
|
# CHECK-NEXT: [8.0] - PdFPMAL
|
|
# CHECK-NEXT: [8.1] - PdFPMAL
|
|
# CHECK-NEXT: [9] - PdFPMMA
|
|
# CHECK-NEXT: [10] - PdFPSTO
|
|
# CHECK-NEXT: [11] - PdFPU0
|
|
# CHECK-NEXT: [12] - PdFPU1
|
|
# CHECK-NEXT: [13] - PdFPU2
|
|
# CHECK-NEXT: [14] - PdFPU3
|
|
# CHECK-NEXT: [15] - PdFPXBR
|
|
# CHECK-NEXT: [16.0] - PdLoad
|
|
# CHECK-NEXT: [16.1] - PdLoad
|
|
# CHECK-NEXT: [17] - PdMul
|
|
# CHECK-NEXT: [18] - PdStore
|
|
|
|
# CHECK: Resource pressure per iteration:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18]
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 3.00 - - 1.00 1.00 - - - - - - -
|
|
|
|
# CHECK: Resource pressure by instruction:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18] Instructions:
|
|
# CHECK-NEXT: - - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - vpinsrw $0, %eax, %xmm0, %xmm0
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - - vpinsrw $1, %eax, %xmm0, %xmm0
|
|
|
|
# CHECK: [2] Code Region
|
|
|
|
# CHECK: Iterations: 500
|
|
# CHECK-NEXT: Instructions: 1000
|
|
# CHECK-NEXT: Total Cycles: 2003
|
|
# CHECK-NEXT: Total uOps: 2000
|
|
|
|
# CHECK: Dispatch Width: 4
|
|
# CHECK-NEXT: uOps Per Cycle: 1.00
|
|
# CHECK-NEXT: IPC: 0.50
|
|
# CHECK-NEXT: Block RThroughput: 3.0
|
|
|
|
# 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 12 1.50 vpinsrd $0, %eax, %xmm0, %xmm0
|
|
# CHECK-NEXT: 2 12 1.50 vpinsrd $1, %eax, %xmm0, %xmm0
|
|
|
|
# CHECK: Resources:
|
|
# CHECK-NEXT: [0.0] - PdAGLU01
|
|
# CHECK-NEXT: [0.1] - PdAGLU01
|
|
# CHECK-NEXT: [1] - PdBranch
|
|
# CHECK-NEXT: [2] - PdCount
|
|
# CHECK-NEXT: [3] - PdDiv
|
|
# CHECK-NEXT: [4] - PdEX0
|
|
# CHECK-NEXT: [5] - PdEX1
|
|
# CHECK-NEXT: [6] - PdFPCVT
|
|
# CHECK-NEXT: [7.0] - PdFPFMA
|
|
# CHECK-NEXT: [7.1] - PdFPFMA
|
|
# CHECK-NEXT: [8.0] - PdFPMAL
|
|
# CHECK-NEXT: [8.1] - PdFPMAL
|
|
# CHECK-NEXT: [9] - PdFPMMA
|
|
# CHECK-NEXT: [10] - PdFPSTO
|
|
# CHECK-NEXT: [11] - PdFPU0
|
|
# CHECK-NEXT: [12] - PdFPU1
|
|
# CHECK-NEXT: [13] - PdFPU2
|
|
# CHECK-NEXT: [14] - PdFPU3
|
|
# CHECK-NEXT: [15] - PdFPXBR
|
|
# CHECK-NEXT: [16.0] - PdLoad
|
|
# CHECK-NEXT: [16.1] - PdLoad
|
|
# CHECK-NEXT: [17] - PdMul
|
|
# CHECK-NEXT: [18] - PdStore
|
|
|
|
# CHECK: Resource pressure per iteration:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18]
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 3.00 - - 1.00 1.00 - - - - - - -
|
|
|
|
# CHECK: Resource pressure by instruction:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18] Instructions:
|
|
# CHECK-NEXT: - - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - vpinsrd $0, %eax, %xmm0, %xmm0
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - - vpinsrd $1, %eax, %xmm0, %xmm0
|
|
|
|
# CHECK: [3] Code Region
|
|
|
|
# CHECK: Iterations: 500
|
|
# CHECK-NEXT: Instructions: 1000
|
|
# CHECK-NEXT: Total Cycles: 2003
|
|
# CHECK-NEXT: Total uOps: 2000
|
|
|
|
# CHECK: Dispatch Width: 4
|
|
# CHECK-NEXT: uOps Per Cycle: 1.00
|
|
# CHECK-NEXT: IPC: 0.50
|
|
# CHECK-NEXT: Block RThroughput: 3.0
|
|
|
|
# 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 12 1.50 vpinsrq $0, %rax, %xmm0, %xmm0
|
|
# CHECK-NEXT: 2 12 1.50 vpinsrq $1, %rax, %xmm0, %xmm0
|
|
|
|
# CHECK: Resources:
|
|
# CHECK-NEXT: [0.0] - PdAGLU01
|
|
# CHECK-NEXT: [0.1] - PdAGLU01
|
|
# CHECK-NEXT: [1] - PdBranch
|
|
# CHECK-NEXT: [2] - PdCount
|
|
# CHECK-NEXT: [3] - PdDiv
|
|
# CHECK-NEXT: [4] - PdEX0
|
|
# CHECK-NEXT: [5] - PdEX1
|
|
# CHECK-NEXT: [6] - PdFPCVT
|
|
# CHECK-NEXT: [7.0] - PdFPFMA
|
|
# CHECK-NEXT: [7.1] - PdFPFMA
|
|
# CHECK-NEXT: [8.0] - PdFPMAL
|
|
# CHECK-NEXT: [8.1] - PdFPMAL
|
|
# CHECK-NEXT: [9] - PdFPMMA
|
|
# CHECK-NEXT: [10] - PdFPSTO
|
|
# CHECK-NEXT: [11] - PdFPU0
|
|
# CHECK-NEXT: [12] - PdFPU1
|
|
# CHECK-NEXT: [13] - PdFPU2
|
|
# CHECK-NEXT: [14] - PdFPU3
|
|
# CHECK-NEXT: [15] - PdFPXBR
|
|
# CHECK-NEXT: [16.0] - PdLoad
|
|
# CHECK-NEXT: [16.1] - PdLoad
|
|
# CHECK-NEXT: [17] - PdMul
|
|
# CHECK-NEXT: [18] - PdStore
|
|
|
|
# CHECK: Resource pressure per iteration:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18]
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 3.00 - - 1.00 1.00 - - - - - - -
|
|
|
|
# CHECK: Resource pressure by instruction:
|
|
# CHECK-NEXT: [0.0] [0.1] [1] [2] [3] [4] [5] [6] [7.0] [7.1] [8.0] [8.1] [9] [10] [11] [12] [13] [14] [15] [16.0] [16.1] [17] [18] Instructions:
|
|
# CHECK-NEXT: - - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - vpinsrq $0, %rax, %xmm0, %xmm0
|
|
# CHECK-NEXT: - - - - - - - - - - 3.00 - - - 1.00 - - - - - - - - vpinsrq $1, %rax, %xmm0, %xmm0
|