From: Roman Lebedev Date: Mon, 10 May 2021 20:40:34 +0000 (+0300) Subject: [X86] AMD Zen 3: same-reg AVX XMM VPCMP is dep breaking one-idiom X-Git-Tag: llvmorg-14-init~7116 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=29532453370044a4c2ddeea130a3db1648b42aa9;p=platform%2Fupstream%2Fllvm.git [X86] AMD Zen 3: same-reg AVX XMM VPCMP is dep breaking one-idiom As measured by exegesis, and confirmed by ref docs. Again, it's not zero-cycle. --- diff --git a/llvm/lib/Target/X86/X86ScheduleZnver3.td b/llvm/lib/Target/X86/X86ScheduleZnver3.td index b1e1e4c..3f753d7 100644 --- a/llvm/lib/Target/X86/X86ScheduleZnver3.td +++ b/llvm/lib/Target/X86/X86ScheduleZnver3.td @@ -1556,6 +1556,11 @@ def : IsDepBreakingFunction<[ DepBreakingClass<[ PCMPEQBrr, PCMPEQWrr, PCMPEQDrr, PCMPEQQrr ], ZeroIdiomPredicate>, + + // AVX XMM + DepBreakingClass<[ + VPCMPEQBrr, VPCMPEQWrr, VPCMPEQDrr, VPCMPEQQrr + ], ZeroIdiomPredicate>, ]>; } // SchedModel diff --git a/llvm/test/tools/llvm-mca/X86/Znver3/one-idioms-avx-xmm.s b/llvm/test/tools/llvm-mca/X86/Znver3/one-idioms-avx-xmm.s index 0a4ed38..1431875 100644 --- a/llvm/test/tools/llvm-mca/X86/Znver3/one-idioms-avx-xmm.s +++ b/llvm/test/tools/llvm-mca/X86/Znver3/one-idioms-avx-xmm.s @@ -25,12 +25,12 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Iterations: 1000 # CHECK-NEXT: Instructions: 2000 -# CHECK-NEXT: Total Cycles: 2003 +# CHECK-NEXT: Total Cycles: 504 # CHECK-NEXT: Total uOps: 2000 # CHECK: Dispatch Width: 6 -# CHECK-NEXT: uOps Per Cycle: 1.00 -# CHECK-NEXT: IPC: 1.00 +# CHECK-NEXT: uOps Per Cycle: 3.97 +# CHECK-NEXT: IPC: 3.97 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: @@ -47,12 +47,12 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Register File statistics: # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 74 # CHECK: * Register File #1 -- Zn3FpPRF: # CHECK-NEXT: Number of physical registers: 160 # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 74 # CHECK: * Register File #2 -- Zn3IntegerPRF: # CHECK-NEXT: Number of physical registers: 192 @@ -90,16 +90,16 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Resource pressure by instruction: # 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: -# CHECK-NEXT: - - - - - - - - - 0.50 - 0.50 - - - - - - - - - - - vpcmpeqb %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: - - - - - - - - 0.50 - 0.50 - - - - - - - - - - - - vpaddb %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: - - - - - - - - 0.25 0.25 0.50 - - - - - - - - - - - - vpcmpeqb %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: - - - - - - - - 0.25 0.25 - 0.50 - - - - - - - - - - - vpaddb %xmm0, %xmm0, %xmm0 # CHECK: Timeline view: -# CHECK-NEXT: Index 0123456 +# CHECK-NEXT: Index 01234 -# CHECK: [0,0] DeER .. vpcmpeqb %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [0,1] D=eER.. vpaddb %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,0] D==eER. vpcmpeqb %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,1] D===eER vpaddb %xmm0, %xmm0, %xmm0 +# CHECK: [0,0] DeER. vpcmpeqb %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [0,1] D=eER vpaddb %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,0] DeE-R vpcmpeqb %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,1] D=eER vpaddb %xmm0, %xmm0, %xmm0 # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions @@ -108,20 +108,20 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] -# CHECK-NEXT: 0. 2 2.0 0.5 0.0 vpcmpeqb %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 1. 2 3.0 0.0 0.0 vpaddb %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 2 2.5 0.3 0.0 +# CHECK-NEXT: 0. 2 1.0 1.0 0.5 vpcmpeqb %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 1. 2 2.0 0.0 0.0 vpaddb %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 2 1.5 0.5 0.3 # CHECK: [1] Code Region # CHECK: Iterations: 1000 # CHECK-NEXT: Instructions: 2000 -# CHECK-NEXT: Total Cycles: 2003 +# CHECK-NEXT: Total Cycles: 504 # CHECK-NEXT: Total uOps: 2000 # CHECK: Dispatch Width: 6 -# CHECK-NEXT: uOps Per Cycle: 1.00 -# CHECK-NEXT: IPC: 1.00 +# CHECK-NEXT: uOps Per Cycle: 3.97 +# CHECK-NEXT: IPC: 3.97 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: @@ -138,12 +138,12 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Register File statistics: # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 74 # CHECK: * Register File #1 -- Zn3FpPRF: # CHECK-NEXT: Number of physical registers: 160 # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 74 # CHECK: * Register File #2 -- Zn3IntegerPRF: # CHECK-NEXT: Number of physical registers: 192 @@ -181,16 +181,16 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Resource pressure by instruction: # 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: -# CHECK-NEXT: - - - - - - - - - 0.50 - 0.50 - - - - - - - - - - - vpcmpeqw %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: - - - - - - - - 0.50 - 0.50 - - - - - - - - - - - - vpaddw %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: - - - - - - - - 0.25 0.25 0.50 - - - - - - - - - - - - vpcmpeqw %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: - - - - - - - - 0.25 0.25 - 0.50 - - - - - - - - - - - vpaddw %xmm0, %xmm0, %xmm0 # CHECK: Timeline view: -# CHECK-NEXT: Index 0123456 +# CHECK-NEXT: Index 01234 -# CHECK: [0,0] DeER .. vpcmpeqw %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [0,1] D=eER.. vpaddw %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,0] D==eER. vpcmpeqw %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,1] D===eER vpaddw %xmm0, %xmm0, %xmm0 +# CHECK: [0,0] DeER. vpcmpeqw %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [0,1] D=eER vpaddw %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,0] DeE-R vpcmpeqw %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,1] D=eER vpaddw %xmm0, %xmm0, %xmm0 # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions @@ -199,20 +199,20 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] -# CHECK-NEXT: 0. 2 2.0 0.5 0.0 vpcmpeqw %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 1. 2 3.0 0.0 0.0 vpaddw %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 2 2.5 0.3 0.0 +# CHECK-NEXT: 0. 2 1.0 1.0 0.5 vpcmpeqw %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 1. 2 2.0 0.0 0.0 vpaddw %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 2 1.5 0.5 0.3 # CHECK: [2] Code Region # CHECK: Iterations: 1000 # CHECK-NEXT: Instructions: 2000 -# CHECK-NEXT: Total Cycles: 2003 +# CHECK-NEXT: Total Cycles: 504 # CHECK-NEXT: Total uOps: 2000 # CHECK: Dispatch Width: 6 -# CHECK-NEXT: uOps Per Cycle: 1.00 -# CHECK-NEXT: IPC: 1.00 +# CHECK-NEXT: uOps Per Cycle: 3.97 +# CHECK-NEXT: IPC: 3.97 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: @@ -229,12 +229,12 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Register File statistics: # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 74 # CHECK: * Register File #1 -- Zn3FpPRF: # CHECK-NEXT: Number of physical registers: 160 # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 74 # CHECK: * Register File #2 -- Zn3IntegerPRF: # CHECK-NEXT: Number of physical registers: 192 @@ -272,16 +272,16 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Resource pressure by instruction: # 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: -# CHECK-NEXT: - - - - - - - - - 0.50 - 0.50 - - - - - - - - - - - vpcmpeqd %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: - - - - - - - - 0.50 - 0.50 - - - - - - - - - - - - vpaddd %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: - - - - - - - - 0.25 0.25 0.50 - - - - - - - - - - - - vpcmpeqd %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: - - - - - - - - 0.25 0.25 - 0.50 - - - - - - - - - - - vpaddd %xmm0, %xmm0, %xmm0 # CHECK: Timeline view: -# CHECK-NEXT: Index 0123456 +# CHECK-NEXT: Index 01234 -# CHECK: [0,0] DeER .. vpcmpeqd %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [0,1] D=eER.. vpaddd %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,0] D==eER. vpcmpeqd %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,1] D===eER vpaddd %xmm0, %xmm0, %xmm0 +# CHECK: [0,0] DeER. vpcmpeqd %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [0,1] D=eER vpaddd %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,0] DeE-R vpcmpeqd %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,1] D=eER vpaddd %xmm0, %xmm0, %xmm0 # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions @@ -290,20 +290,20 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] -# CHECK-NEXT: 0. 2 2.0 0.5 0.0 vpcmpeqd %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 1. 2 3.0 0.0 0.0 vpaddd %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 2 2.5 0.3 0.0 +# CHECK-NEXT: 0. 2 1.0 1.0 0.5 vpcmpeqd %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 1. 2 2.0 0.0 0.0 vpaddd %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 2 1.5 0.5 0.3 # CHECK: [3] Code Region # CHECK: Iterations: 1000 # CHECK-NEXT: Instructions: 2000 -# CHECK-NEXT: Total Cycles: 2003 +# CHECK-NEXT: Total Cycles: 504 # CHECK-NEXT: Total uOps: 2000 # CHECK: Dispatch Width: 6 -# CHECK-NEXT: uOps Per Cycle: 1.00 -# CHECK-NEXT: IPC: 1.00 +# CHECK-NEXT: uOps Per Cycle: 3.97 +# CHECK-NEXT: IPC: 3.97 # CHECK-NEXT: Block RThroughput: 0.5 # CHECK: Instruction Info: @@ -320,12 +320,12 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Register File statistics: # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 73 # CHECK: * Register File #1 -- Zn3FpPRF: # CHECK-NEXT: Number of physical registers: 160 # CHECK-NEXT: Total number of mappings created: 2000 -# CHECK-NEXT: Max number of mappings used: 66 +# CHECK-NEXT: Max number of mappings used: 73 # CHECK: * Register File #2 -- Zn3IntegerPRF: # CHECK-NEXT: Number of physical registers: 192 @@ -367,12 +367,12 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK-NEXT: - - - - - - - - - - 0.50 0.50 - - - - - - - - - - - vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Timeline view: -# CHECK-NEXT: Index 0123456 +# CHECK-NEXT: Index 01234 -# CHECK: [0,0] DeER .. vpcmpeqq %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [0,1] D=eER.. vpaddq %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,0] D==eER. vpcmpeqq %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: [1,1] D===eER vpaddq %xmm0, %xmm0, %xmm0 +# CHECK: [0,0] DeER. vpcmpeqq %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [0,1] D=eER vpaddq %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,0] DeE-R vpcmpeqq %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: [1,1] D=eER vpaddq %xmm0, %xmm0, %xmm0 # CHECK: Average Wait times (based on the timeline view): # CHECK-NEXT: [0]: Executions @@ -381,6 +381,6 @@ vpaddq %xmm0, %xmm0, %xmm0 # CHECK-NEXT: [3]: Average time elapsed from WB until retire stage # CHECK: [0] [1] [2] [3] -# CHECK-NEXT: 0. 2 2.0 0.5 0.0 vpcmpeqq %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 1. 2 3.0 0.0 0.0 vpaddq %xmm0, %xmm0, %xmm0 -# CHECK-NEXT: 2 2.5 0.3 0.0 +# CHECK-NEXT: 0. 2 1.0 1.0 0.5 vpcmpeqq %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 1. 2 2.0 0.0 0.0 vpaddq %xmm0, %xmm0, %xmm0 +# CHECK-NEXT: 2 1.5 0.5 0.3