151 lines · plain
1# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py2# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=skylake -bottleneck-analysis < %s | FileCheck %s3 4.LBB0_4:5 vmovups (%rsi,%rax,2), %xmm06 vpermilps $255, %xmm0, %xmm77 vmulps -24(%rsp), %xmm7, %xmm88 vpermilps $170, %xmm0, %xmm69 vpermilps $85, %xmm0, %xmm510 vbroadcastss %xmm0, %xmm011 vfmadd231ps %xmm9, %xmm6, %xmm812 vfmadd213ps %xmm8, %xmm10, %xmm513 vfmadd213ps %xmm5, %xmm11, %xmm014 vfmadd213ps %xmm0, %xmm12, %xmm415 vfmadd213ps %xmm4, %xmm13, %xmm116 vmovaps %xmm7, %xmm417 vfmadd213ps %xmm1, %xmm14, %xmm218 vmovaps %xmm6, %xmm119 vfmadd213ps %xmm2, %xmm15, %xmm320 vpermilps $170, %xmm3, %xmm021 vmovups %xmm3, (%rdx,%rax)22 vpermilps $255, %xmm3, %xmm223 addq $16, %rax24 decl %ecx25 vmovaps %xmm0, %xmm326 jne .LBB0_427 28# CHECK: Iterations: 10029# CHECK-NEXT: Instructions: 220030# CHECK-NEXT: Total Cycles: 103931# CHECK-NEXT: Total uOps: 240032 33# CHECK: Dispatch Width: 634# CHECK-NEXT: uOps Per Cycle: 2.3135# CHECK-NEXT: IPC: 2.1236# CHECK-NEXT: Block RThroughput: 6.037 38# CHECK: Cycles with backend pressure increase [ 92.69% ]39# CHECK-NEXT: Throughput Bottlenecks:40# CHECK-NEXT: Resource Pressure [ 46.78% ]41# CHECK-NEXT: - SKLPort0 [ 14.24% ]42# CHECK-NEXT: - SKLPort1 [ 14.24% ]43# CHECK-NEXT: - SKLPort5 [ 46.49% ]44# CHECK-NEXT: - SKLPort6 [ 8.66% ]45# CHECK-NEXT: Data Dependencies: [ 64.97% ]46# CHECK-NEXT: - Register Dependencies [ 64.97% ]47# CHECK-NEXT: - Memory Dependencies [ 0.00% ]48 49# CHECK: Critical sequence based on the simulation:50 51# CHECK: Instruction Dependency Information52# CHECK-NEXT: +----< 18. addq $16, %rax53# CHECK-NEXT: |54# CHECK-NEXT: | < loop carried >55# CHECK-NEXT: |56# CHECK-NEXT: +----> 0. vmovups (%rsi,%rax,2), %xmm0 ## REGISTER dependency: %rax57# CHECK-NEXT: | 1. vpermilps $255, %xmm0, %xmm758# CHECK-NEXT: | 2. vmulps -24(%rsp), %xmm7, %xmm859# CHECK-NEXT: +----> 3. vpermilps $170, %xmm0, %xmm6 ## REGISTER dependency: %xmm060# CHECK-NEXT: | 4. vpermilps $85, %xmm0, %xmm561# CHECK-NEXT: | 5. vbroadcastss %xmm0, %xmm062# CHECK-NEXT: +----> 6. vfmadd231ps %xmm9, %xmm6, %xmm8 ## REGISTER dependency: %xmm663# CHECK-NEXT: +----> 7. vfmadd213ps %xmm8, %xmm10, %xmm5 ## REGISTER dependency: %xmm864# CHECK-NEXT: +----> 8. vfmadd213ps %xmm5, %xmm11, %xmm0 ## REGISTER dependency: %xmm565# CHECK-NEXT: +----> 9. vfmadd213ps %xmm0, %xmm12, %xmm4 ## REGISTER dependency: %xmm066# CHECK-NEXT: +----> 10. vfmadd213ps %xmm4, %xmm13, %xmm1 ## REGISTER dependency: %xmm467# CHECK-NEXT: | 11. vmovaps %xmm7, %xmm468# CHECK-NEXT: +----> 12. vfmadd213ps %xmm1, %xmm14, %xmm2 ## REGISTER dependency: %xmm169# CHECK-NEXT: | 13. vmovaps %xmm6, %xmm170# CHECK-NEXT: +----> 14. vfmadd213ps %xmm2, %xmm15, %xmm3 ## REGISTER dependency: %xmm271# CHECK-NEXT: +----> 15. vpermilps $170, %xmm3, %xmm0 ## REGISTER dependency: %xmm372# CHECK-NEXT: | 16. vmovups %xmm3, (%rdx,%rax)73# CHECK-NEXT: | 17. vpermilps $255, %xmm3, %xmm274# CHECK-NEXT: | 18. addq $16, %rax75# CHECK-NEXT: | 19. decl %ecx76# CHECK-NEXT: +----> 20. vmovaps %xmm0, %xmm3 ## REGISTER dependency: %xmm077# CHECK-NEXT: 21. jne .LBB0_478 79# CHECK: Instruction Info:80# CHECK-NEXT: [1]: #uOps81# CHECK-NEXT: [2]: Latency82# CHECK-NEXT: [3]: RThroughput83# CHECK-NEXT: [4]: MayLoad84# CHECK-NEXT: [5]: MayStore85# CHECK-NEXT: [6]: HasSideEffects (U)86 87# CHECK: [1] [2] [3] [4] [5] [6] Instructions:88# CHECK-NEXT: 1 6 0.50 * vmovups (%rsi,%rax,2), %xmm089# CHECK-NEXT: 1 1 1.00 vpermilps $255, %xmm0, %xmm790# CHECK-NEXT: 2 10 0.50 * vmulps -24(%rsp), %xmm7, %xmm891# CHECK-NEXT: 1 1 1.00 vpermilps $170, %xmm0, %xmm692# CHECK-NEXT: 1 1 1.00 vpermilps $85, %xmm0, %xmm593# CHECK-NEXT: 1 1 1.00 vbroadcastss %xmm0, %xmm094# CHECK-NEXT: 1 4 0.50 vfmadd231ps %xmm9, %xmm6, %xmm895# CHECK-NEXT: 1 4 0.50 vfmadd213ps %xmm8, %xmm10, %xmm596# CHECK-NEXT: 1 4 0.50 vfmadd213ps %xmm5, %xmm11, %xmm097# CHECK-NEXT: 1 4 0.50 vfmadd213ps %xmm0, %xmm12, %xmm498# CHECK-NEXT: 1 4 0.50 vfmadd213ps %xmm4, %xmm13, %xmm199# CHECK-NEXT: 1 1 0.33 vmovaps %xmm7, %xmm4100# CHECK-NEXT: 1 4 0.50 vfmadd213ps %xmm1, %xmm14, %xmm2101# CHECK-NEXT: 1 1 0.33 vmovaps %xmm6, %xmm1102# CHECK-NEXT: 1 4 0.50 vfmadd213ps %xmm2, %xmm15, %xmm3103# CHECK-NEXT: 1 1 1.00 vpermilps $170, %xmm3, %xmm0104# CHECK-NEXT: 2 1 1.00 * vmovups %xmm3, (%rdx,%rax)105# CHECK-NEXT: 1 1 1.00 vpermilps $255, %xmm3, %xmm2106# CHECK-NEXT: 1 1 0.25 addq $16, %rax107# CHECK-NEXT: 1 1 0.25 decl %ecx108# CHECK-NEXT: 1 1 0.33 vmovaps %xmm0, %xmm3109# CHECK-NEXT: 1 1 0.50 jne .LBB0_4110 111# CHECK: Resources:112# CHECK-NEXT: [0] - SKLDivider113# CHECK-NEXT: [1] - SKLFPDivider114# CHECK-NEXT: [2] - SKLPort0115# CHECK-NEXT: [3] - SKLPort1116# CHECK-NEXT: [4] - SKLPort2117# CHECK-NEXT: [5] - SKLPort3118# CHECK-NEXT: [6] - SKLPort4119# CHECK-NEXT: [7] - SKLPort5120# CHECK-NEXT: [8] - SKLPort6121# CHECK-NEXT: [9] - SKLPort7122 123# CHECK: Resource pressure per iteration:124# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9]125# CHECK-NEXT: - - 5.52 5.53 1.01 1.03 1.00 6.02 2.93 0.96126 127# CHECK: Resource pressure by instruction:128# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Instructions:129# CHECK-NEXT: - - - - 0.04 0.96 - - - - vmovups (%rsi,%rax,2), %xmm0130# CHECK-NEXT: - - - - - - - 1.00 - - vpermilps $255, %xmm0, %xmm7131# CHECK-NEXT: - - 0.03 0.97 0.96 0.04 - - - - vmulps -24(%rsp), %xmm7, %xmm8132# CHECK-NEXT: - - - - - - - 1.00 - - vpermilps $170, %xmm0, %xmm6133# CHECK-NEXT: - - - - - - - 1.00 - - vpermilps $85, %xmm0, %xmm5134# CHECK-NEXT: - - - - - - - 1.00 - - vbroadcastss %xmm0, %xmm0135# CHECK-NEXT: - - 0.95 0.05 - - - - - - vfmadd231ps %xmm9, %xmm6, %xmm8136# CHECK-NEXT: - - 0.50 0.50 - - - - - - vfmadd213ps %xmm8, %xmm10, %xmm5137# CHECK-NEXT: - - 0.92 0.08 - - - - - - vfmadd213ps %xmm5, %xmm11, %xmm0138# CHECK-NEXT: - - 0.95 0.05 - - - - - - vfmadd213ps %xmm0, %xmm12, %xmm4139# CHECK-NEXT: - - 0.51 0.49 - - - - - - vfmadd213ps %xmm4, %xmm13, %xmm1140# CHECK-NEXT: - - 0.52 0.48 - - - - - - vmovaps %xmm7, %xmm4141# CHECK-NEXT: - - 0.49 0.51 - - - - - - vfmadd213ps %xmm1, %xmm14, %xmm2142# CHECK-NEXT: - - 0.04 0.95 - - - 0.01 - - vmovaps %xmm6, %xmm1143# CHECK-NEXT: - - 0.51 0.49 - - - - - - vfmadd213ps %xmm2, %xmm15, %xmm3144# CHECK-NEXT: - - - - - - - 1.00 - - vpermilps $170, %xmm3, %xmm0145# CHECK-NEXT: - - - - 0.01 0.03 1.00 - - 0.96 vmovups %xmm3, (%rdx,%rax)146# CHECK-NEXT: - - - - - - - 1.00 - - vpermilps $255, %xmm3, %xmm2147# CHECK-NEXT: - - - - - - - - 1.00 - addq $16, %rax148# CHECK-NEXT: - - 0.04 0.01 - - - 0.01 0.94 - decl %ecx149# CHECK-NEXT: - - 0.05 0.95 - - - - - - vmovaps %xmm0, %xmm3150# CHECK-NEXT: - - 0.01 - - - - - 0.99 - jne .LBB0_4151