From ba5714d77352306fbe8eb4da940b76554c05ecf6 Mon Sep 17 00:00:00 2001 From: Simon Pilgrim Date: Sat, 19 Nov 2022 21:20:31 +0000 Subject: [PATCH] [MCA][X86] Add test coverage for VP2INTERSECT instructions NOTE: For IceLakeServer we actually test TigerLake as that's the only target that supports it (we do something similar for F16C on IvyBridge in the SandyBridge tests). --- .../X86/Generic/resources-avx512vp2intersect.s | 49 +++++++++++++++ .../X86/Generic/resources-avx512vp2intersectvl.s | 69 ++++++++++++++++++++ .../IceLakeServer/resources-avx512vp2intersect.s | 53 ++++++++++++++++ .../IceLakeServer/resources-avx512vp2intersectvl.s | 73 ++++++++++++++++++++++ 4 files changed, 244 insertions(+) create mode 100644 llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersect.s create mode 100644 llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersectvl.s create mode 100644 llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersect.s create mode 100644 llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersectvl.s diff --git a/llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersect.s b/llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersect.s new file mode 100644 index 0000000..a05fe32 --- /dev/null +++ b/llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersect.s @@ -0,0 +1,49 @@ +# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py +# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=x86-64 -instruction-tables < %s | FileCheck %s + +vp2intersectd %zmm16, %zmm19, %k0 +vp2intersectd (%rax), %zmm19, %k0 +vp2intersectd (%rax){1to16}, %zmm19, %k0 + +vp2intersectq %zmm16, %zmm19, %k0 +vp2intersectq (%rax), %zmm19, %k0 +vp2intersectq (%rax){1to8}, %zmm19, %k0 + +# 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 1 0.50 vp2intersectd %zmm16, %zmm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectd (%rax), %zmm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectd (%rax){1to16}, %zmm19, %k0 +# CHECK-NEXT: 1 1 0.50 vp2intersectq %zmm16, %zmm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectq (%rax), %zmm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectq (%rax){1to8}, %zmm19, %k0 + +# CHECK: Resources: +# CHECK-NEXT: [0] - SBDivider +# CHECK-NEXT: [1] - SBFPDivider +# CHECK-NEXT: [2] - SBPort0 +# CHECK-NEXT: [3] - SBPort1 +# CHECK-NEXT: [4] - SBPort4 +# CHECK-NEXT: [5] - SBPort5 +# CHECK-NEXT: [6.0] - SBPort23 +# CHECK-NEXT: [6.1] - SBPort23 + +# CHECK: Resource pressure per iteration: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] +# CHECK-NEXT: - - - 3.00 - 3.00 2.00 2.00 + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] Instructions: +# CHECK-NEXT: - - - 0.50 - 0.50 - - vp2intersectd %zmm16, %zmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectd (%rax), %zmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectd (%rax){1to16}, %zmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vp2intersectq %zmm16, %zmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectq (%rax), %zmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectq (%rax){1to8}, %zmm19, %k0 diff --git a/llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersectvl.s b/llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersectvl.s new file mode 100644 index 0000000..73acb88 --- /dev/null +++ b/llvm/test/tools/llvm-mca/X86/Generic/resources-avx512vp2intersectvl.s @@ -0,0 +1,69 @@ +# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py +# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=x86-64 -instruction-tables < %s | FileCheck %s + +vp2intersectd %xmm16, %xmm19, %k0 +vp2intersectd (%rax), %xmm19, %k0 +vp2intersectd (%rax){1to4}, %xmm19, %k0 + +vp2intersectd %ymm16, %ymm19, %k0 +vp2intersectd (%rax), %ymm19, %k0 +vp2intersectd (%rax){1to8}, %ymm19, %k0 + +vp2intersectq %xmm16, %xmm19, %k0 +vp2intersectq (%rax), %xmm19, %k0 +vp2intersectq (%rax){1to2}, %xmm19, %k0 + +vp2intersectq %ymm16, %ymm19, %k0 +vp2intersectq (%rax), %ymm19, %k0 +vp2intersectq (%rax){1to4}, %ymm19, %k0 + +# 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 1 0.50 vp2intersectd %xmm16, %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectd (%rax), %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectd (%rax){1to4}, %xmm19, %k0 +# CHECK-NEXT: 1 1 0.50 vp2intersectd %ymm16, %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectd (%rax), %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectd (%rax){1to8}, %ymm19, %k0 +# CHECK-NEXT: 1 1 0.50 vp2intersectq %xmm16, %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectq (%rax), %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectq (%rax){1to2}, %xmm19, %k0 +# CHECK-NEXT: 1 1 0.50 vp2intersectq %ymm16, %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectq (%rax), %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectq (%rax){1to4}, %ymm19, %k0 + +# CHECK: Resources: +# CHECK-NEXT: [0] - SBDivider +# CHECK-NEXT: [1] - SBFPDivider +# CHECK-NEXT: [2] - SBPort0 +# CHECK-NEXT: [3] - SBPort1 +# CHECK-NEXT: [4] - SBPort4 +# CHECK-NEXT: [5] - SBPort5 +# CHECK-NEXT: [6.0] - SBPort23 +# CHECK-NEXT: [6.1] - SBPort23 + +# CHECK: Resource pressure per iteration: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] +# CHECK-NEXT: - - - 6.00 - 6.00 4.00 4.00 + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] Instructions: +# CHECK-NEXT: - - - 0.50 - 0.50 - - vp2intersectd %xmm16, %xmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectd (%rax), %xmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectd (%rax){1to4}, %xmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vp2intersectd %ymm16, %ymm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectd (%rax), %ymm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectd (%rax){1to8}, %ymm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vp2intersectq %xmm16, %xmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectq (%rax), %xmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectq (%rax){1to2}, %xmm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vp2intersectq %ymm16, %ymm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectq (%rax), %ymm19, %k0 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vp2intersectq (%rax){1to4}, %ymm19, %k0 diff --git a/llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersect.s b/llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersect.s new file mode 100644 index 0000000..3230ae9 --- /dev/null +++ b/llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersect.s @@ -0,0 +1,53 @@ +# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py +# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=tigerlake -instruction-tables < %s | FileCheck %s + +vp2intersectd %zmm16, %zmm19, %k0 +vp2intersectd (%rax), %zmm19, %k0 +vp2intersectd (%rax){1to16}, %zmm19, %k0 + +vp2intersectq %zmm16, %zmm19, %k0 +vp2intersectq (%rax), %zmm19, %k0 +vp2intersectq (%rax){1to8}, %zmm19, %k0 + +# 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 1 1.00 vp2intersectd %zmm16, %zmm19, %k0 +# CHECK-NEXT: 2 8 1.00 * vp2intersectd (%rax), %zmm19, %k0 +# CHECK-NEXT: 2 8 1.00 * vp2intersectd (%rax){1to16}, %zmm19, %k0 +# CHECK-NEXT: 1 1 1.00 vp2intersectq %zmm16, %zmm19, %k0 +# CHECK-NEXT: 2 8 1.00 * vp2intersectq (%rax), %zmm19, %k0 +# CHECK-NEXT: 2 8 1.00 * vp2intersectq (%rax){1to8}, %zmm19, %k0 + +# CHECK: Resources: +# CHECK-NEXT: [0] - ICXDivider +# CHECK-NEXT: [1] - ICXFPDivider +# CHECK-NEXT: [2] - ICXPort0 +# CHECK-NEXT: [3] - ICXPort1 +# CHECK-NEXT: [4] - ICXPort2 +# CHECK-NEXT: [5] - ICXPort3 +# CHECK-NEXT: [6] - ICXPort4 +# CHECK-NEXT: [7] - ICXPort5 +# CHECK-NEXT: [8] - ICXPort6 +# CHECK-NEXT: [9] - ICXPort7 +# CHECK-NEXT: [10] - ICXPort8 +# CHECK-NEXT: [11] - ICXPort9 + +# CHECK: Resource pressure per iteration: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] +# CHECK-NEXT: - - 6.00 - 2.00 2.00 - - - - - - + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] Instructions: +# CHECK-NEXT: - - 1.00 - - - - - - - - - vp2intersectd %zmm16, %zmm19, %k0 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - - - vp2intersectd (%rax), %zmm19, %k0 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - - - vp2intersectd (%rax){1to16}, %zmm19, %k0 +# CHECK-NEXT: - - 1.00 - - - - - - - - - vp2intersectq %zmm16, %zmm19, %k0 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - - - vp2intersectq (%rax), %zmm19, %k0 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - - - vp2intersectq (%rax){1to8}, %zmm19, %k0 diff --git a/llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersectvl.s b/llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersectvl.s new file mode 100644 index 0000000..b7e81f5 --- /dev/null +++ b/llvm/test/tools/llvm-mca/X86/IceLakeServer/resources-avx512vp2intersectvl.s @@ -0,0 +1,73 @@ +# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py +# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=tigerlake -instruction-tables < %s | FileCheck %s + +vp2intersectd %xmm16, %xmm19, %k0 +vp2intersectd (%rax), %xmm19, %k0 +vp2intersectd (%rax){1to4}, %xmm19, %k0 + +vp2intersectd %ymm16, %ymm19, %k0 +vp2intersectd (%rax), %ymm19, %k0 +vp2intersectd (%rax){1to8}, %ymm19, %k0 + +vp2intersectq %xmm16, %xmm19, %k0 +vp2intersectq (%rax), %xmm19, %k0 +vp2intersectq (%rax){1to2}, %xmm19, %k0 + +vp2intersectq %ymm16, %ymm19, %k0 +vp2intersectq (%rax), %ymm19, %k0 +vp2intersectq (%rax){1to4}, %ymm19, %k0 + +# 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 1 0.50 vp2intersectd %xmm16, %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectd (%rax), %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectd (%rax){1to4}, %xmm19, %k0 +# CHECK-NEXT: 1 1 0.50 vp2intersectd %ymm16, %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectd (%rax), %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectd (%rax){1to8}, %ymm19, %k0 +# CHECK-NEXT: 1 1 0.50 vp2intersectq %xmm16, %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectq (%rax), %xmm19, %k0 +# CHECK-NEXT: 2 7 0.50 * vp2intersectq (%rax){1to2}, %xmm19, %k0 +# CHECK-NEXT: 1 1 0.50 vp2intersectq %ymm16, %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectq (%rax), %ymm19, %k0 +# CHECK-NEXT: 2 8 0.50 * vp2intersectq (%rax){1to4}, %ymm19, %k0 + +# CHECK: Resources: +# CHECK-NEXT: [0] - ICXDivider +# CHECK-NEXT: [1] - ICXFPDivider +# CHECK-NEXT: [2] - ICXPort0 +# CHECK-NEXT: [3] - ICXPort1 +# CHECK-NEXT: [4] - ICXPort2 +# CHECK-NEXT: [5] - ICXPort3 +# CHECK-NEXT: [6] - ICXPort4 +# CHECK-NEXT: [7] - ICXPort5 +# CHECK-NEXT: [8] - ICXPort6 +# CHECK-NEXT: [9] - ICXPort7 +# CHECK-NEXT: [10] - ICXPort8 +# CHECK-NEXT: [11] - ICXPort9 + +# CHECK: Resource pressure per iteration: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] +# CHECK-NEXT: - - 6.00 6.00 4.00 4.00 - - - - - - + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] Instructions: +# CHECK-NEXT: - - 0.50 0.50 - - - - - - - - vp2intersectd %xmm16, %xmm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectd (%rax), %xmm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectd (%rax){1to4}, %xmm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 - - - - - - - - vp2intersectd %ymm16, %ymm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectd (%rax), %ymm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectd (%rax){1to8}, %ymm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 - - - - - - - - vp2intersectq %xmm16, %xmm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectq (%rax), %xmm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectq (%rax){1to2}, %xmm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 - - - - - - - - vp2intersectq %ymm16, %ymm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectq (%rax), %ymm19, %k0 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - - - vp2intersectq (%rax){1to4}, %ymm19, %k0 -- 2.7.4