From 0579a5a2a97c5dfffc37d95d9e3011fd45b86f51 Mon Sep 17 00:00:00 2001 From: Simon Pilgrim Date: Tue, 22 Jan 2019 17:52:15 +0000 Subject: [PATCH] [llvm-mca][X86] Tidyup avx512 placeholder tests Ensure we keep avx512f/bw/dq + vl versions separate, add example broadcast tests - this should allow us to better the test coverage of test\CodeGen\X86\avx512-schedule.ll git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@351848 91177308-0d34-0410-b5e6-96231b3b80d8 --- .../llvm-mca/X86/Generic/resources-avx512.s | 204 ++++++++------ .../llvm-mca/X86/Generic/resources-avx512bw.s | 143 ++++++++++ .../X86/Generic/resources-avx512bwvl.s | 257 +++++++++++++++++ .../llvm-mca/X86/Generic/resources-avx512vl.s | 146 +++++++++- .../X86/SkylakeServer/resources-avx512.s | 204 ++++++++------ .../X86/SkylakeServer/resources-avx512bw.s | 145 ++++++++++ .../X86/SkylakeServer/resources-avx512bwvl.s | 259 ++++++++++++++++++ .../X86/SkylakeServer/resources-avx512vl.s | 146 +++++++++- 8 files changed, 1348 insertions(+), 156 deletions(-) create mode 100644 test/tools/llvm-mca/X86/Generic/resources-avx512bw.s create mode 100644 test/tools/llvm-mca/X86/Generic/resources-avx512bwvl.s create mode 100644 test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bw.s create mode 100644 test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bwvl.s diff --git a/test/tools/llvm-mca/X86/Generic/resources-avx512.s b/test/tools/llvm-mca/X86/Generic/resources-avx512.s index b5d40ed766e..c9fcee86914 100644 --- a/test/tools/llvm-mca/X86/Generic/resources-avx512.s +++ b/test/tools/llvm-mca/X86/Generic/resources-avx512.s @@ -3,129 +3,143 @@ vaddpd %zmm16, %zmm17, %zmm19 vaddpd (%rax), %zmm17, %zmm19 +vaddpd (%rax){1to8}, %zmm17, %zmm19 vaddpd %zmm16, %zmm17, %zmm19 {k1} vaddpd (%rax), %zmm17, %zmm19 {k1} +vaddpd (%rax){1to8}, %zmm17, %zmm19 {k1} vaddpd %zmm16, %zmm17, %zmm19 {z}{k1} vaddpd (%rax), %zmm17, %zmm19 {z}{k1} +vaddpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vaddps %zmm16, %zmm17, %zmm19 vaddps (%rax), %zmm17, %zmm19 +vaddps (%rax){1to16}, %zmm17, %zmm19 vaddps %zmm16, %zmm17, %zmm19 {k1} vaddps (%rax), %zmm17, %zmm19 {k1} +vaddps (%rax){1to16}, %zmm17, %zmm19 {k1} vaddps %zmm16, %zmm17, %zmm19 {z}{k1} vaddps (%rax), %zmm17, %zmm19 {z}{k1} +vaddps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vdivpd %zmm16, %zmm17, %zmm19 vdivpd (%rax), %zmm17, %zmm19 +vdivpd (%rax){1to8}, %zmm17, %zmm19 vdivpd %zmm16, %zmm17, %zmm19 {k1} vdivpd (%rax), %zmm17, %zmm19 {k1} +vdivpd (%rax){1to8}, %zmm17, %zmm19 {k1} vdivpd %zmm16, %zmm17, %zmm19 {z}{k1} vdivpd (%rax), %zmm17, %zmm19 {z}{k1} +vdivpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vdivps %zmm16, %zmm17, %zmm19 vdivps (%rax), %zmm17, %zmm19 +vdivps (%rax){1to16}, %zmm17, %zmm19 vdivps %zmm16, %zmm17, %zmm19 {k1} vdivps (%rax), %zmm17, %zmm19 {k1} +vdivps (%rax){1to16}, %zmm17, %zmm19 {k1} vdivps %zmm16, %zmm17, %zmm19 {z}{k1} vdivps (%rax), %zmm17, %zmm19 {z}{k1} +vdivps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vmulpd %zmm16, %zmm17, %zmm19 vmulpd (%rax), %zmm17, %zmm19 +vmulpd (%rax){1to8}, %zmm17, %zmm19 vmulpd %zmm16, %zmm17, %zmm19 {k1} vmulpd (%rax), %zmm17, %zmm19 {k1} +vmulpd (%rax){1to8}, %zmm17, %zmm19 {k1} vmulpd %zmm16, %zmm17, %zmm19 {z}{k1} vmulpd (%rax), %zmm17, %zmm19 {z}{k1} +vmulpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vmulps %zmm16, %zmm17, %zmm19 vmulps (%rax), %zmm17, %zmm19 +vmulps (%rax){1to16}, %zmm17, %zmm19 vmulps %zmm16, %zmm17, %zmm19 {k1} vmulps (%rax), %zmm17, %zmm19 {k1} +vmulps (%rax){1to16}, %zmm17, %zmm19 {k1} vmulps %zmm16, %zmm17, %zmm19 {z}{k1} vmulps (%rax), %zmm17, %zmm19 {z}{k1} +vmulps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vpabsd %zmm16, %zmm19 vpabsd (%rax), %zmm19 +vpabsd (%rax){1to16}, %zmm19 vpabsd %zmm16, %zmm19 {k1} vpabsd (%rax), %zmm19 {k1} +vpabsd (%rax){1to16}, %zmm19 {k1} vpabsd %zmm16, %zmm19 {z}{k1} vpabsd (%rax), %zmm19 {z}{k1} +vpabsd (%rax){1to16}, %zmm19 {z}{k1} vpabsq %zmm16, %zmm19 vpabsq (%rax), %zmm19 +vpabsq (%rax){1to8}, %zmm19 vpabsq %zmm16, %zmm19 {k1} vpabsq (%rax), %zmm19 {k1} +vpabsq (%rax){1to8}, %zmm19 {k1} vpabsq %zmm16, %zmm19 {z}{k1} vpabsq (%rax), %zmm19 {z}{k1} - -vpaddb %zmm16, %zmm17, %zmm19 -vpaddb (%rax), %zmm17, %zmm19 -vpaddb %zmm16, %zmm17, %zmm19 {k1} -vpaddb (%rax), %zmm17, %zmm19 {k1} -vpaddb %zmm16, %zmm17, %zmm19 {z}{k1} -vpaddb (%rax), %zmm17, %zmm19 {z}{k1} +vpabsq (%rax){1to8}, %zmm19 {z}{k1} vpaddd %zmm16, %zmm17, %zmm19 vpaddd (%rax), %zmm17, %zmm19 +vpaddd (%rax){1to16}, %zmm17, %zmm19 vpaddd %zmm16, %zmm17, %zmm19 {k1} vpaddd (%rax), %zmm17, %zmm19 {k1} +vpaddd (%rax){1to16}, %zmm17, %zmm19 {k1} vpaddd %zmm16, %zmm17, %zmm19 {z}{k1} vpaddd (%rax), %zmm17, %zmm19 {z}{k1} +vpaddd (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vpaddq %zmm16, %zmm17, %zmm19 vpaddq (%rax), %zmm17, %zmm19 +vpaddq (%rax){1to8}, %zmm17, %zmm19 vpaddq %zmm16, %zmm17, %zmm19 {k1} vpaddq (%rax), %zmm17, %zmm19 {k1} +vpaddq (%rax){1to8}, %zmm17, %zmm19 {k1} vpaddq %zmm16, %zmm17, %zmm19 {z}{k1} vpaddq (%rax), %zmm17, %zmm19 {z}{k1} - -vpaddw %zmm16, %zmm17, %zmm19 -vpaddw (%rax), %zmm17, %zmm19 -vpaddw %zmm16, %zmm17, %zmm19 {k1} -vpaddw (%rax), %zmm17, %zmm19 {k1} -vpaddw %zmm16, %zmm17, %zmm19 {z}{k1} -vpaddw (%rax), %zmm17, %zmm19 {z}{k1} - -vpsubb %zmm16, %zmm17, %zmm19 -vpsubb (%rax), %zmm17, %zmm19 -vpsubb %zmm16, %zmm17, %zmm19 {k1} -vpsubb (%rax), %zmm17, %zmm19 {k1} -vpsubb %zmm16, %zmm17, %zmm19 {z}{k1} -vpsubb (%rax), %zmm17, %zmm19 {z}{k1} +vpaddq (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vpsubd %zmm16, %zmm17, %zmm19 vpsubd (%rax), %zmm17, %zmm19 +vpsubd (%rax){1to16}, %zmm17, %zmm19 vpsubd %zmm16, %zmm17, %zmm19 {k1} vpsubd (%rax), %zmm17, %zmm19 {k1} +vpsubd (%rax){1to16}, %zmm17, %zmm19 {k1} vpsubd %zmm16, %zmm17, %zmm19 {z}{k1} vpsubd (%rax), %zmm17, %zmm19 {z}{k1} +vpsubd (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vpsubq %zmm16, %zmm17, %zmm19 vpsubq (%rax), %zmm17, %zmm19 +vpsubq (%rax){1to8}, %zmm17, %zmm19 vpsubq %zmm16, %zmm17, %zmm19 {k1} vpsubq (%rax), %zmm17, %zmm19 {k1} +vpsubq (%rax){1to8}, %zmm17, %zmm19 {k1} vpsubq %zmm16, %zmm17, %zmm19 {z}{k1} vpsubq (%rax), %zmm17, %zmm19 {z}{k1} - -vpsubw %zmm16, %zmm17, %zmm19 -vpsubw (%rax), %zmm17, %zmm19 -vpsubw %zmm16, %zmm17, %zmm19 {k1} -vpsubw (%rax), %zmm17, %zmm19 {k1} -vpsubw %zmm16, %zmm17, %zmm19 {z}{k1} -vpsubw (%rax), %zmm17, %zmm19 {z}{k1} +vpsubq (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vsubpd %zmm16, %zmm17, %zmm19 vsubpd (%rax), %zmm17, %zmm19 +vsubpd (%rax){1to8}, %zmm17, %zmm19 vsubpd %zmm16, %zmm17, %zmm19 {k1} vsubpd (%rax), %zmm17, %zmm19 {k1} +vsubpd (%rax){1to8}, %zmm17, %zmm19 {k1} vsubpd %zmm16, %zmm17, %zmm19 {z}{k1} vsubpd (%rax), %zmm17, %zmm19 {z}{k1} +vsubpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vsubps %zmm16, %zmm17, %zmm19 vsubps (%rax), %zmm17, %zmm19 +vsubps (%rax){1to16}, %zmm17, %zmm19 vsubps %zmm16, %zmm17, %zmm19 {k1} vsubps (%rax), %zmm17, %zmm19 {k1} +vsubps (%rax){1to16}, %zmm17, %zmm19 {k1} vsubps %zmm16, %zmm17, %zmm19 {z}{k1} vsubps (%rax), %zmm17, %zmm19 {z}{k1} +vsubps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps @@ -138,112 +152,130 @@ vsubps (%rax), %zmm17, %zmm19 {z}{k1} # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 3 1.00 vaddpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 10 1.00 * vaddpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 10 1.00 * vaddpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 3 1.00 vaddpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vaddpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vaddpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vaddpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vaddps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 10 1.00 * vaddps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 10 1.00 * vaddps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 3 1.00 vaddps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vaddps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vaddps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vaddps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 3 45 44.00 vdivpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 4 52 44.00 * vdivpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 4 52 44.00 * vdivpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 3 45 44.00 vdivpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 4 52 44.00 * vdivpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 4 52 44.00 * vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 3 45 44.00 vdivpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 4 52 44.00 * vdivpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 4 52 44.00 * vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 3 29 28.00 vdivps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 4 36 28.00 * vdivps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 4 36 28.00 * vdivps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 3 29 28.00 vdivps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 4 36 28.00 * vdivps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 4 36 28.00 * vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 3 29 28.00 vdivps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 4 36 28.00 * vdivps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 4 36 28.00 * vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 5 1.00 vmulpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 12 1.00 * vmulpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 12 1.00 * vmulpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 5 1.00 vmulpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 12 1.00 * vmulpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 12 1.00 * vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 5 1.00 vmulpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 12 1.00 * vmulpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 12 1.00 * vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 5 1.00 vmulps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 12 1.00 * vmulps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 12 1.00 * vmulps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 5 1.00 vmulps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 12 1.00 * vmulps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 12 1.00 * vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 5 1.00 vmulps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 12 1.00 * vmulps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 12 1.00 * vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.50 vpabsd %zmm16, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpabsd (%rax), %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpabsd (%rax){1to16}, %zmm19 # CHECK-NEXT: 1 1 0.50 vpabsd %zmm16, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpabsd (%rax), %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsd (%rax){1to16}, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.50 vpabsd %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpabsd (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsd (%rax){1to16}, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.50 vpabsq %zmm16, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpabsq (%rax), %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpabsq (%rax){1to8}, %zmm19 # CHECK-NEXT: 1 1 0.50 vpabsq %zmm16, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpabsq (%rax), %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsq (%rax){1to8}, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.50 vpabsq %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpabsq (%rax), %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.50 vpaddb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.50 vpaddb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.50 vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsq (%rax){1to8}, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.50 vpaddd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpaddd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.50 vpaddd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpaddd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.50 vpaddd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpaddd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.50 vpaddq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpaddq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.50 vpaddq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpaddq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.50 vpaddq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpaddq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.50 vpaddw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.50 vpaddw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.50 vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.50 vpsubb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.50 vpsubb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.50 vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.50 vpsubd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpsubd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.50 vpsubd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpsubd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.50 vpsubd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpsubd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.50 vpsubq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpsubq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.50 vpsubq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpsubq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.50 vpsubq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpsubq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.50 vpsubw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.50 vpsubw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.50 vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vsubpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 10 1.00 * vsubpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 10 1.00 * vsubpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 3 1.00 vsubpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vsubpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vsubpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vsubpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vsubps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 10 1.00 * vsubps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 10 1.00 * vsubps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 3 1.00 vsubps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vsubps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vsubps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vsubps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK: Resources: # CHECK-NEXT: [0] - SBDivider @@ -257,115 +289,133 @@ vsubps (%rax), %zmm17, %zmm19 {z}{k1} # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] -# CHECK-NEXT: - 432.00 42.00 54.00 - 36.00 27.00 27.00 +# CHECK-NEXT: - 648.00 63.00 63.00 - 36.00 42.00 42.00 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] Instructions: # CHECK-NEXT: - - - 1.00 - - - - vaddpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - - - vaddpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vaddpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vaddps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - - - vaddps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vaddps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 44.00 2.50 - - 0.50 - - vdivpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - 44.00 2.50 - - 0.50 - - vdivpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 44.00 2.50 - - 0.50 - - vdivpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 28.00 2.50 - - 0.50 - - vdivps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - 28.00 2.50 - - 0.50 - - vdivps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 28.00 2.50 - - 0.50 - - vdivps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - vmulpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - 1.00 - - - - - vmulpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - vmulpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - vmulps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - 1.00 - - - - - vmulps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - vmulps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsd %zmm16, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsd (%rax), %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsd (%rax){1to16}, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsd %zmm16, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsd (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsd (%rax){1to16}, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsd %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsd (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsd (%rax){1to16}, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsq %zmm16, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsq (%rax), %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsq (%rax){1to8}, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsq %zmm16, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsq (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsq (%rax){1to8}, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsq %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsq (%rax), %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsq (%rax){1to8}, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vsubpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - - - vsubpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vsubpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vsubps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - - 1.00 - - - - vsubps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vsubps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} diff --git a/test/tools/llvm-mca/X86/Generic/resources-avx512bw.s b/test/tools/llvm-mca/X86/Generic/resources-avx512bw.s new file mode 100644 index 00000000000..8d4091eb8c2 --- /dev/null +++ b/test/tools/llvm-mca/X86/Generic/resources-avx512bw.s @@ -0,0 +1,143 @@ +# 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 + +vpabsb %zmm16, %zmm19 +vpabsb (%rax), %zmm19 +vpabsb %zmm16, %zmm19 {k1} +vpabsb (%rax), %zmm19 {k1} +vpabsb %zmm16, %zmm19 {z}{k1} +vpabsb (%rax), %zmm19 {z}{k1} + +vpabsw %zmm16, %zmm19 +vpabsw (%rax), %zmm19 +vpabsw %zmm16, %zmm19 {k1} +vpabsw (%rax), %zmm19 {k1} +vpabsw %zmm16, %zmm19 {z}{k1} +vpabsw (%rax), %zmm19 {z}{k1} + +vpaddb %zmm16, %zmm17, %zmm19 +vpaddb (%rax), %zmm17, %zmm19 +vpaddb %zmm16, %zmm17, %zmm19 {k1} +vpaddb (%rax), %zmm17, %zmm19 {k1} +vpaddb %zmm16, %zmm17, %zmm19 {z}{k1} +vpaddb (%rax), %zmm17, %zmm19 {z}{k1} + +vpaddw %zmm16, %zmm17, %zmm19 +vpaddw (%rax), %zmm17, %zmm19 +vpaddw %zmm16, %zmm17, %zmm19 {k1} +vpaddw (%rax), %zmm17, %zmm19 {k1} +vpaddw %zmm16, %zmm17, %zmm19 {z}{k1} +vpaddw (%rax), %zmm17, %zmm19 {z}{k1} + +vpsubb %zmm16, %zmm17, %zmm19 +vpsubb (%rax), %zmm17, %zmm19 +vpsubb %zmm16, %zmm17, %zmm19 {k1} +vpsubb (%rax), %zmm17, %zmm19 {k1} +vpsubb %zmm16, %zmm17, %zmm19 {z}{k1} +vpsubb (%rax), %zmm17, %zmm19 {z}{k1} + +vpsubw %zmm16, %zmm17, %zmm19 +vpsubw (%rax), %zmm17, %zmm19 +vpsubw %zmm16, %zmm17, %zmm19 {k1} +vpsubw (%rax), %zmm17, %zmm19 {k1} +vpsubw %zmm16, %zmm17, %zmm19 {z}{k1} +vpsubw (%rax), %zmm17, %zmm19 {z}{k1} + +# 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 vpabsb %zmm16, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %zmm19 +# CHECK-NEXT: 1 1 0.50 vpabsb %zmm16, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsb %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpabsw %zmm16, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %zmm19 +# CHECK-NEXT: 1 1 0.50 vpabsw %zmm16, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsw %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpaddb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.50 vpaddb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpaddw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.50 vpaddw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpsubb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.50 vpsubb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpsubw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.50 vpsubw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} + +# 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: - - - 18.00 - 18.00 9.00 9.00 + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] Instructions: +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %zmm16, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %zmm16, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %zmm16, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %zmm16, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} diff --git a/test/tools/llvm-mca/X86/Generic/resources-avx512bwvl.s b/test/tools/llvm-mca/X86/Generic/resources-avx512bwvl.s new file mode 100644 index 00000000000..5efa5a3cf1e --- /dev/null +++ b/test/tools/llvm-mca/X86/Generic/resources-avx512bwvl.s @@ -0,0 +1,257 @@ +# 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 + +vpabsb %xmm16, %xmm19 +vpabsb (%rax), %xmm19 +vpabsb %xmm16, %xmm19 {k1} +vpabsb (%rax), %xmm19 {k1} +vpabsb %xmm16, %xmm19 {z}{k1} +vpabsb (%rax), %xmm19 {z}{k1} + +vpabsb %ymm16, %ymm19 +vpabsb (%rax), %ymm19 +vpabsb %ymm16, %ymm19 {k1} +vpabsb (%rax), %ymm19 {k1} +vpabsb %ymm16, %ymm19 {z}{k1} +vpabsb (%rax), %ymm19 {z}{k1} + +vpabsw %xmm16, %xmm19 +vpabsw (%rax), %xmm19 +vpabsw %xmm16, %xmm19 {k1} +vpabsw (%rax), %xmm19 {k1} +vpabsw %xmm16, %xmm19 {z}{k1} +vpabsw (%rax), %xmm19 {z}{k1} + +vpabsw %ymm16, %ymm19 +vpabsw (%rax), %ymm19 +vpabsw %ymm16, %ymm19 {k1} +vpabsw (%rax), %ymm19 {k1} +vpabsw %ymm16, %ymm19 {z}{k1} +vpabsw (%rax), %ymm19 {z}{k1} + +vpaddb %xmm16, %xmm17, %xmm19 +vpaddb (%rax), %xmm17, %xmm19 +vpaddb %xmm16, %xmm17, %xmm19 {k1} +vpaddb (%rax), %xmm17, %xmm19 {k1} +vpaddb %xmm16, %xmm17, %xmm19 {z}{k1} +vpaddb (%rax), %xmm17, %xmm19 {z}{k1} + +vpaddb %ymm16, %ymm17, %ymm19 +vpaddb (%rax), %ymm17, %ymm19 +vpaddb %ymm16, %ymm17, %ymm19 {k1} +vpaddb (%rax), %ymm17, %ymm19 {k1} +vpaddb %ymm16, %ymm17, %ymm19 {z}{k1} +vpaddb (%rax), %ymm17, %ymm19 {z}{k1} + +vpaddw %xmm16, %xmm17, %xmm19 +vpaddw (%rax), %xmm17, %xmm19 +vpaddw %xmm16, %xmm17, %xmm19 {k1} +vpaddw (%rax), %xmm17, %xmm19 {k1} +vpaddw %xmm16, %xmm17, %xmm19 {z}{k1} +vpaddw (%rax), %xmm17, %xmm19 {z}{k1} + +vpaddw %ymm16, %ymm17, %ymm19 +vpaddw (%rax), %ymm17, %ymm19 +vpaddw %ymm16, %ymm17, %ymm19 {k1} +vpaddw (%rax), %ymm17, %ymm19 {k1} +vpaddw %ymm16, %ymm17, %ymm19 {z}{k1} +vpaddw (%rax), %ymm17, %ymm19 {z}{k1} + +vpsubb %xmm16, %xmm17, %xmm19 +vpsubb (%rax), %xmm17, %xmm19 +vpsubb %xmm16, %xmm17, %xmm19 {k1} +vpsubb (%rax), %xmm17, %xmm19 {k1} +vpsubb %xmm16, %xmm17, %xmm19 {z}{k1} +vpsubb (%rax), %xmm17, %xmm19 {z}{k1} + +vpsubb %ymm16, %ymm17, %ymm19 +vpsubb (%rax), %ymm17, %ymm19 +vpsubb %ymm16, %ymm17, %ymm19 {k1} +vpsubb (%rax), %ymm17, %ymm19 {k1} +vpsubb %ymm16, %ymm17, %ymm19 {z}{k1} +vpsubb (%rax), %ymm17, %ymm19 {z}{k1} + +vpsubw %xmm16, %xmm17, %xmm19 +vpsubw (%rax), %xmm17, %xmm19 +vpsubw %xmm16, %xmm17, %xmm19 {k1} +vpsubw (%rax), %xmm17, %xmm19 {k1} +vpsubw %xmm16, %xmm17, %xmm19 {z}{k1} +vpsubw (%rax), %xmm17, %xmm19 {z}{k1} + +vpsubw %ymm16, %ymm17, %ymm19 +vpsubw (%rax), %ymm17, %ymm19 +vpsubw %ymm16, %ymm17, %ymm19 {k1} +vpsubw (%rax), %ymm17, %ymm19 {k1} +vpsubw %ymm16, %ymm17, %ymm19 {z}{k1} +vpsubw (%rax), %ymm17, %ymm19 {z}{k1} + +# 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 vpabsb %xmm16, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpabsb (%rax), %xmm19 +# CHECK-NEXT: 1 1 0.50 vpabsb %xmm16, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpabsb (%rax), %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsb %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpabsb (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpabsb %ymm16, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %ymm19 +# CHECK-NEXT: 1 1 0.50 vpabsb %ymm16, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsb %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpabsw %xmm16, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpabsw (%rax), %xmm19 +# CHECK-NEXT: 1 1 0.50 vpabsw %xmm16, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpabsw (%rax), %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsw %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpabsw (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpabsw %ymm16, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %ymm19 +# CHECK-NEXT: 1 1 0.50 vpabsw %ymm16, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsw %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpaddb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpaddb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.50 vpaddb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpaddb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpaddb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpaddb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpaddb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.50 vpaddb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpaddb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpaddw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpaddw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.50 vpaddw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpaddw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpaddw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpaddw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpaddw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.50 vpaddw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpaddw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpsubb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpsubb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.50 vpsubb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpsubb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpsubb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpsubb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpsubb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.50 vpsubb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpsubb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpsubw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpsubw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.50 vpsubw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpsubw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpsubw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpsubw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpsubw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.50 vpsubw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpsubw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %ymm17, %ymm19 {%k1} {z} + +# 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: - - - 36.00 - 36.00 18.00 18.00 + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] Instructions: +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %xmm16, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %xmm16, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %ymm16, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %ymm16, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsb %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsb (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %xmm16, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %xmm16, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %ymm16, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %ymm16, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpabsw %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpabsw (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpaddw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpaddw (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 0.50 - 0.50 - - vpsubw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 0.50 - 0.50 0.50 0.50 vpsubw (%rax), %ymm17, %ymm19 {%k1} {z} diff --git a/test/tools/llvm-mca/X86/Generic/resources-avx512vl.s b/test/tools/llvm-mca/X86/Generic/resources-avx512vl.s index 7bfcd8a5db2..81358314d4a 100644 --- a/test/tools/llvm-mca/X86/Generic/resources-avx512vl.s +++ b/test/tools/llvm-mca/X86/Generic/resources-avx512vl.s @@ -3,115 +3,163 @@ vaddpd %xmm16, %xmm17, %xmm19 vaddpd (%rax), %xmm17, %xmm19 +vaddpd (%rax){1to2}, %xmm17, %xmm19 vaddpd %xmm16, %xmm17, %xmm19 {k1} vaddpd (%rax), %xmm17, %xmm19 {k1} +vaddpd (%rax){1to2}, %xmm17, %xmm19 {k1} vaddpd %xmm16, %xmm17, %xmm19 {z}{k1} vaddpd (%rax), %xmm17, %xmm19 {z}{k1} +vaddpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vaddpd %ymm16, %ymm17, %ymm19 vaddpd (%rax), %ymm17, %ymm19 +vaddpd (%rax){1to4}, %ymm17, %ymm19 vaddpd %ymm16, %ymm17, %ymm19 {k1} vaddpd (%rax), %ymm17, %ymm19 {k1} +vaddpd (%rax){1to4}, %ymm17, %ymm19 {k1} vaddpd %ymm16, %ymm17, %ymm19 {z}{k1} vaddpd (%rax), %ymm17, %ymm19 {z}{k1} +vaddpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vaddps %xmm16, %xmm17, %xmm19 vaddps (%rax), %xmm17, %xmm19 +vaddps (%rax){1to4}, %xmm17, %xmm19 vaddps %xmm16, %xmm17, %xmm19 {k1} vaddps (%rax), %xmm17, %xmm19 {k1} +vaddps (%rax){1to4}, %xmm17, %xmm19 {k1} vaddps %xmm16, %xmm17, %xmm19 {z}{k1} vaddps (%rax), %xmm17, %xmm19 {z}{k1} +vaddps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vaddps %ymm16, %ymm17, %ymm19 vaddps (%rax), %ymm17, %ymm19 +vaddps (%rax){1to8}, %ymm17, %ymm19 vaddps %ymm16, %ymm17, %ymm19 {k1} vaddps (%rax), %ymm17, %ymm19 {k1} +vaddps (%rax){1to8}, %ymm17, %ymm19 {k1} vaddps %ymm16, %ymm17, %ymm19 {z}{k1} vaddps (%rax), %ymm17, %ymm19 {z}{k1} +vaddps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} vdivpd %xmm16, %xmm17, %xmm19 vdivpd (%rax), %xmm17, %xmm19 +vdivpd (%rax){1to2}, %xmm17, %xmm19 vdivpd %xmm16, %xmm17, %xmm19 {k1} vdivpd (%rax), %xmm17, %xmm19 {k1} +vdivpd (%rax){1to2}, %xmm17, %xmm19 {k1} vdivpd %xmm16, %xmm17, %xmm19 {z}{k1} vdivpd (%rax), %xmm17, %xmm19 {z}{k1} +vdivpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vdivpd %ymm16, %ymm17, %ymm19 vdivpd (%rax), %ymm17, %ymm19 +vdivpd (%rax){1to4}, %ymm17, %ymm19 vdivpd %ymm16, %ymm17, %ymm19 {k1} vdivpd (%rax), %ymm17, %ymm19 {k1} +vdivpd (%rax){1to4}, %ymm17, %ymm19 {k1} vdivpd %ymm16, %ymm17, %ymm19 {z}{k1} vdivpd (%rax), %ymm17, %ymm19 {z}{k1} +vdivpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vdivps %xmm16, %xmm17, %xmm19 vdivps (%rax), %xmm17, %xmm19 +vdivps (%rax){1to4}, %xmm17, %xmm19 vdivps %xmm16, %xmm17, %xmm19 {k1} vdivps (%rax), %xmm17, %xmm19 {k1} +vdivps (%rax){1to4}, %xmm17, %xmm19 {k1} vdivps %xmm16, %xmm17, %xmm19 {z}{k1} vdivps (%rax), %xmm17, %xmm19 {z}{k1} +vdivps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vdivps %ymm16, %ymm17, %ymm19 vdivps (%rax), %ymm17, %ymm19 +vdivps (%rax){1to8}, %ymm17, %ymm19 vdivps %ymm16, %ymm17, %ymm19 {k1} vdivps (%rax), %ymm17, %ymm19 {k1} +vdivps (%rax){1to8}, %ymm17, %ymm19 {k1} vdivps %ymm16, %ymm17, %ymm19 {z}{k1} vdivps (%rax), %ymm17, %ymm19 {z}{k1} +vdivps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} vmulpd %xmm16, %xmm17, %xmm19 vmulpd (%rax), %xmm17, %xmm19 +vmulpd (%rax){1to2}, %xmm17, %xmm19 vmulpd %xmm16, %xmm17, %xmm19 {k1} vmulpd (%rax), %xmm17, %xmm19 {k1} +vmulpd (%rax){1to2}, %xmm17, %xmm19 {k1} vmulpd %xmm16, %xmm17, %xmm19 {z}{k1} vmulpd (%rax), %xmm17, %xmm19 {z}{k1} +vmulpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vmulpd %ymm16, %ymm17, %ymm19 vmulpd (%rax), %ymm17, %ymm19 +vmulpd (%rax){1to4}, %ymm17, %ymm19 vmulpd %ymm16, %ymm17, %ymm19 {k1} vmulpd (%rax), %ymm17, %ymm19 {k1} +vmulpd (%rax){1to4}, %ymm17, %ymm19 {k1} vmulpd %ymm16, %ymm17, %ymm19 {z}{k1} vmulpd (%rax), %ymm17, %ymm19 {z}{k1} +vmulpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vmulps %xmm16, %xmm17, %xmm19 vmulps (%rax), %xmm17, %xmm19 +vmulps (%rax){1to4}, %xmm17, %xmm19 vmulps %xmm16, %xmm17, %xmm19 {k1} vmulps (%rax), %xmm17, %xmm19 {k1} +vmulps (%rax){1to4}, %xmm17, %xmm19 {k1} vmulps %xmm16, %xmm17, %xmm19 {z}{k1} vmulps (%rax), %xmm17, %xmm19 {z}{k1} +vmulps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vmulps %ymm16, %ymm17, %ymm19 vmulps (%rax), %ymm17, %ymm19 +vmulps (%rax){1to8}, %ymm17, %ymm19 vmulps %ymm16, %ymm17, %ymm19 {k1} vmulps (%rax), %ymm17, %ymm19 {k1} +vmulps (%rax){1to8}, %ymm17, %ymm19 {k1} vmulps %ymm16, %ymm17, %ymm19 {z}{k1} vmulps (%rax), %ymm17, %ymm19 {z}{k1} +vmulps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} vsubpd %xmm16, %xmm17, %xmm19 vsubpd (%rax), %xmm17, %xmm19 +vsubpd (%rax){1to2}, %xmm17, %xmm19 vsubpd %xmm16, %xmm17, %xmm19 {k1} vsubpd (%rax), %xmm17, %xmm19 {k1} +vsubpd (%rax){1to2}, %xmm17, %xmm19 {k1} vsubpd %xmm16, %xmm17, %xmm19 {z}{k1} vsubpd (%rax), %xmm17, %xmm19 {z}{k1} +vsubpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vsubpd %ymm16, %ymm17, %ymm19 vsubpd (%rax), %ymm17, %ymm19 +vsubpd (%rax){1to4}, %ymm17, %ymm19 vsubpd %ymm16, %ymm17, %ymm19 {k1} vsubpd (%rax), %ymm17, %ymm19 {k1} +vsubpd (%rax){1to4}, %ymm17, %ymm19 {k1} vsubpd %ymm16, %ymm17, %ymm19 {z}{k1} vsubpd (%rax), %ymm17, %ymm19 {z}{k1} +vsubpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vsubps %xmm16, %xmm17, %xmm19 vsubps (%rax), %xmm17, %xmm19 +vsubps (%rax){1to4}, %xmm17, %xmm19 vsubps %xmm16, %xmm17, %xmm19 {k1} vsubps (%rax), %xmm17, %xmm19 {k1} +vsubps (%rax){1to4}, %xmm17, %xmm19 {k1} vsubps %xmm16, %xmm17, %xmm19 {z}{k1} vsubps (%rax), %xmm17, %xmm19 {z}{k1} +vsubps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vsubps %ymm16, %ymm17, %ymm19 vsubps (%rax), %ymm17, %ymm19 +vsubps (%rax){1to8}, %ymm17, %ymm19 vsubps %ymm16, %ymm17, %ymm19 {k1} vsubps (%rax), %ymm17, %ymm19 {k1} +vsubps (%rax){1to8}, %ymm17, %ymm19 {k1} vsubps %ymm16, %ymm17, %ymm19 {z}{k1} vsubps (%rax), %ymm17, %ymm19 {z}{k1} +vsubps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps @@ -124,100 +172,148 @@ vsubps (%rax), %ymm17, %ymm19 {z}{k1} # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 3 1.00 vaddpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 9 1.00 * vaddpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 9 1.00 * vaddpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 3 1.00 vaddpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 9 1.00 * vaddpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 9 1.00 * vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vaddpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 9 1.00 * vaddpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 9 1.00 * vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vaddpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 10 1.00 * vaddpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 10 1.00 * vaddpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 1 3 1.00 vaddpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vaddpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 3 1.00 vaddpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vaddpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vaddps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 9 1.00 * vaddps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 9 1.00 * vaddps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 3 1.00 vaddps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 9 1.00 * vaddps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 9 1.00 * vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vaddps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 9 1.00 * vaddps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 9 1.00 * vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vaddps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 10 1.00 * vaddps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 10 1.00 * vaddps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 1 3 1.00 vaddps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vaddps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 3 1.00 vaddps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vaddps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 22 22.00 vdivpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 28 22.00 * vdivpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 28 22.00 * vdivpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 22 22.00 vdivpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 28 22.00 * vdivpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 28 22.00 * vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 22 22.00 vdivpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 28 22.00 * vdivpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 28 22.00 * vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 3 45 44.00 vdivpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 4 52 44.00 * vdivpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 4 52 44.00 * vdivpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 3 45 44.00 vdivpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 4 52 44.00 * vdivpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 4 52 44.00 * vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 3 45 44.00 vdivpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 4 52 44.00 * vdivpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 4 52 44.00 * vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 14 14.00 vdivps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 20 14.00 * vdivps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 20 14.00 * vdivps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 14 14.00 vdivps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 20 14.00 * vdivps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 20 14.00 * vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 14 14.00 vdivps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 20 14.00 * vdivps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 20 14.00 * vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 3 29 28.00 vdivps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 4 36 28.00 * vdivps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 4 36 28.00 * vdivps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 3 29 28.00 vdivps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 4 36 28.00 * vdivps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 4 36 28.00 * vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 3 29 28.00 vdivps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 4 36 28.00 * vdivps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 4 36 28.00 * vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 5 1.00 vmulpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 11 1.00 * vmulpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 11 1.00 * vmulpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 5 1.00 vmulpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 11 1.00 * vmulpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 11 1.00 * vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 5 1.00 vmulpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 11 1.00 * vmulpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 11 1.00 * vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 5 1.00 vmulpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 12 1.00 * vmulpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 12 1.00 * vmulpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 1 5 1.00 vmulpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 12 1.00 * vmulpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 12 1.00 * vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 5 1.00 vmulpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 12 1.00 * vmulpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 12 1.00 * vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 5 1.00 vmulps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 11 1.00 * vmulps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 11 1.00 * vmulps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 5 1.00 vmulps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 11 1.00 * vmulps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 11 1.00 * vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 5 1.00 vmulps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 11 1.00 * vmulps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 11 1.00 * vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 5 1.00 vmulps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 12 1.00 * vmulps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 12 1.00 * vmulps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 1 5 1.00 vmulps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 12 1.00 * vmulps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 12 1.00 * vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 5 1.00 vmulps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 12 1.00 * vmulps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 12 1.00 * vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vsubpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 9 1.00 * vsubpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 9 1.00 * vsubpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 3 1.00 vsubpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 9 1.00 * vsubpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 9 1.00 * vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vsubpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 9 1.00 * vsubpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 9 1.00 * vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vsubpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 10 1.00 * vsubpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 10 1.00 * vsubpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 1 3 1.00 vsubpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vsubpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 3 1.00 vsubpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vsubpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vsubps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 9 1.00 * vsubps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 9 1.00 * vsubps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 3 1.00 vsubps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 9 1.00 * vsubps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 9 1.00 * vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 3 1.00 vsubps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 9 1.00 * vsubps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 9 1.00 * vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 3 1.00 vsubps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 10 1.00 * vsubps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 10 1.00 * vsubps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 1 3 1.00 vsubps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 10 1.00 * vsubps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 10 1.00 * vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 3 1.00 vsubps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 10 1.00 * vsubps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 10 1.00 * vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK: Resources: # CHECK-NEXT: [0] - SBDivider @@ -231,103 +327,151 @@ vsubps (%rax), %ymm17, %ymm19 {z}{k1} # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] -# CHECK-NEXT: - 648.00 66.00 48.00 - 6.00 24.00 24.00 +# CHECK-NEXT: - 972.00 99.00 72.00 - 9.00 48.00 48.00 # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6.0] [6.1] Instructions: # CHECK-NEXT: - - - 1.00 - - - - vaddpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - - - vaddpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vaddpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vaddpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - - - vaddpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vaddpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vaddps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - - - vaddps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vaddps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vaddps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - - - vaddps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vaddps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 22.00 1.00 - - - - - vdivpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - 22.00 1.00 - - - 0.50 0.50 vdivpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - 22.00 1.00 - - - 0.50 0.50 vdivpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - 22.00 1.00 - - - - - vdivpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 22.00 1.00 - - - 0.50 0.50 vdivpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - 22.00 1.00 - - - 0.50 0.50 vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 22.00 1.00 - - - - - vdivpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 22.00 1.00 - - - 0.50 0.50 vdivpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - 22.00 1.00 - - - 0.50 0.50 vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 44.00 2.50 - - 0.50 - - vdivpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - 44.00 2.50 - - 0.50 - - vdivpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 44.00 2.50 - - 0.50 - - vdivpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - 44.00 2.50 - - 0.50 0.50 0.50 vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 14.00 1.00 - - - - - vdivps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - 14.00 1.00 - - - 0.50 0.50 vdivps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - 14.00 1.00 - - - 0.50 0.50 vdivps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - 14.00 1.00 - - - - - vdivps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 14.00 1.00 - - - 0.50 0.50 vdivps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - 14.00 1.00 - - - 0.50 0.50 vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 14.00 1.00 - - - - - vdivps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 14.00 1.00 - - - 0.50 0.50 vdivps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - 14.00 1.00 - - - 0.50 0.50 vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 28.00 2.50 - - 0.50 - - vdivps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - 28.00 2.50 - - 0.50 - - vdivps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 28.00 2.50 - - 0.50 - - vdivps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - 28.00 2.50 - - 0.50 0.50 0.50 vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - vmulpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - - 1.00 - - - - - vmulpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - vmulpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - vmulpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - - 1.00 - - - - - vmulpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - vmulpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - vmulps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - - 1.00 - - - - - vmulps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - vmulps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - vmulps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - - 1.00 - - - - - vmulps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - vmulps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - - - 0.50 0.50 vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vsubpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - - - vsubpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vsubpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vsubpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - - - vsubpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vsubpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vsubps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - - - 1.00 - - - - vsubps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vsubps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - - - vsubps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - - - 1.00 - - - - vsubps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - - 1.00 - - - - vsubps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - - 1.00 - - 0.50 0.50 vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} diff --git a/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512.s b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512.s index a941c0a3a66..c2e955a6157 100644 --- a/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512.s +++ b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512.s @@ -3,129 +3,143 @@ vaddpd %zmm16, %zmm17, %zmm19 vaddpd (%rax), %zmm17, %zmm19 +vaddpd (%rax){1to8}, %zmm17, %zmm19 vaddpd %zmm16, %zmm17, %zmm19 {k1} vaddpd (%rax), %zmm17, %zmm19 {k1} +vaddpd (%rax){1to8}, %zmm17, %zmm19 {k1} vaddpd %zmm16, %zmm17, %zmm19 {z}{k1} vaddpd (%rax), %zmm17, %zmm19 {z}{k1} +vaddpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vaddps %zmm16, %zmm17, %zmm19 vaddps (%rax), %zmm17, %zmm19 +vaddps (%rax){1to16}, %zmm17, %zmm19 vaddps %zmm16, %zmm17, %zmm19 {k1} vaddps (%rax), %zmm17, %zmm19 {k1} +vaddps (%rax){1to16}, %zmm17, %zmm19 {k1} vaddps %zmm16, %zmm17, %zmm19 {z}{k1} vaddps (%rax), %zmm17, %zmm19 {z}{k1} +vaddps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vdivpd %zmm16, %zmm17, %zmm19 vdivpd (%rax), %zmm17, %zmm19 +vdivpd (%rax){1to8}, %zmm17, %zmm19 vdivpd %zmm16, %zmm17, %zmm19 {k1} vdivpd (%rax), %zmm17, %zmm19 {k1} +vdivpd (%rax){1to8}, %zmm17, %zmm19 {k1} vdivpd %zmm16, %zmm17, %zmm19 {z}{k1} vdivpd (%rax), %zmm17, %zmm19 {z}{k1} +vdivpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vdivps %zmm16, %zmm17, %zmm19 vdivps (%rax), %zmm17, %zmm19 +vdivps (%rax){1to16}, %zmm17, %zmm19 vdivps %zmm16, %zmm17, %zmm19 {k1} vdivps (%rax), %zmm17, %zmm19 {k1} +vdivps (%rax){1to16}, %zmm17, %zmm19 {k1} vdivps %zmm16, %zmm17, %zmm19 {z}{k1} vdivps (%rax), %zmm17, %zmm19 {z}{k1} +vdivps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vmulpd %zmm16, %zmm17, %zmm19 vmulpd (%rax), %zmm17, %zmm19 +vmulpd (%rax){1to8}, %zmm17, %zmm19 vmulpd %zmm16, %zmm17, %zmm19 {k1} vmulpd (%rax), %zmm17, %zmm19 {k1} +vmulpd (%rax){1to8}, %zmm17, %zmm19 {k1} vmulpd %zmm16, %zmm17, %zmm19 {z}{k1} vmulpd (%rax), %zmm17, %zmm19 {z}{k1} +vmulpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vmulps %zmm16, %zmm17, %zmm19 vmulps (%rax), %zmm17, %zmm19 +vmulps (%rax){1to16}, %zmm17, %zmm19 vmulps %zmm16, %zmm17, %zmm19 {k1} vmulps (%rax), %zmm17, %zmm19 {k1} +vmulps (%rax){1to16}, %zmm17, %zmm19 {k1} vmulps %zmm16, %zmm17, %zmm19 {z}{k1} vmulps (%rax), %zmm17, %zmm19 {z}{k1} +vmulps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vpabsd %zmm16, %zmm19 vpabsd (%rax), %zmm19 +vpabsd (%rax){1to16}, %zmm19 vpabsd %zmm16, %zmm19 {k1} vpabsd (%rax), %zmm19 {k1} +vpabsd (%rax){1to16}, %zmm19 {k1} vpabsd %zmm16, %zmm19 {z}{k1} vpabsd (%rax), %zmm19 {z}{k1} +vpabsd (%rax){1to16}, %zmm19 {z}{k1} vpabsq %zmm16, %zmm19 vpabsq (%rax), %zmm19 +vpabsq (%rax){1to8}, %zmm19 vpabsq %zmm16, %zmm19 {k1} vpabsq (%rax), %zmm19 {k1} +vpabsq (%rax){1to8}, %zmm19 {k1} vpabsq %zmm16, %zmm19 {z}{k1} vpabsq (%rax), %zmm19 {z}{k1} - -vpaddb %zmm16, %zmm17, %zmm19 -vpaddb (%rax), %zmm17, %zmm19 -vpaddb %zmm16, %zmm17, %zmm19 {k1} -vpaddb (%rax), %zmm17, %zmm19 {k1} -vpaddb %zmm16, %zmm17, %zmm19 {z}{k1} -vpaddb (%rax), %zmm17, %zmm19 {z}{k1} +vpabsq (%rax){1to8}, %zmm19 {z}{k1} vpaddd %zmm16, %zmm17, %zmm19 vpaddd (%rax), %zmm17, %zmm19 +vpaddd (%rax){1to16}, %zmm17, %zmm19 vpaddd %zmm16, %zmm17, %zmm19 {k1} vpaddd (%rax), %zmm17, %zmm19 {k1} +vpaddd (%rax){1to16}, %zmm17, %zmm19 {k1} vpaddd %zmm16, %zmm17, %zmm19 {z}{k1} vpaddd (%rax), %zmm17, %zmm19 {z}{k1} +vpaddd (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vpaddq %zmm16, %zmm17, %zmm19 vpaddq (%rax), %zmm17, %zmm19 +vpaddq (%rax){1to8}, %zmm17, %zmm19 vpaddq %zmm16, %zmm17, %zmm19 {k1} vpaddq (%rax), %zmm17, %zmm19 {k1} +vpaddq (%rax){1to8}, %zmm17, %zmm19 {k1} vpaddq %zmm16, %zmm17, %zmm19 {z}{k1} vpaddq (%rax), %zmm17, %zmm19 {z}{k1} - -vpaddw %zmm16, %zmm17, %zmm19 -vpaddw (%rax), %zmm17, %zmm19 -vpaddw %zmm16, %zmm17, %zmm19 {k1} -vpaddw (%rax), %zmm17, %zmm19 {k1} -vpaddw %zmm16, %zmm17, %zmm19 {z}{k1} -vpaddw (%rax), %zmm17, %zmm19 {z}{k1} - -vpsubb %zmm16, %zmm17, %zmm19 -vpsubb (%rax), %zmm17, %zmm19 -vpsubb %zmm16, %zmm17, %zmm19 {k1} -vpsubb (%rax), %zmm17, %zmm19 {k1} -vpsubb %zmm16, %zmm17, %zmm19 {z}{k1} -vpsubb (%rax), %zmm17, %zmm19 {z}{k1} +vpaddq (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vpsubd %zmm16, %zmm17, %zmm19 vpsubd (%rax), %zmm17, %zmm19 +vpsubd (%rax){1to16}, %zmm17, %zmm19 vpsubd %zmm16, %zmm17, %zmm19 {k1} vpsubd (%rax), %zmm17, %zmm19 {k1} +vpsubd (%rax){1to16}, %zmm17, %zmm19 {k1} vpsubd %zmm16, %zmm17, %zmm19 {z}{k1} vpsubd (%rax), %zmm17, %zmm19 {z}{k1} +vpsubd (%rax){1to16}, %zmm17, %zmm19 {z}{k1} vpsubq %zmm16, %zmm17, %zmm19 vpsubq (%rax), %zmm17, %zmm19 +vpsubq (%rax){1to8}, %zmm17, %zmm19 vpsubq %zmm16, %zmm17, %zmm19 {k1} vpsubq (%rax), %zmm17, %zmm19 {k1} +vpsubq (%rax){1to8}, %zmm17, %zmm19 {k1} vpsubq %zmm16, %zmm17, %zmm19 {z}{k1} vpsubq (%rax), %zmm17, %zmm19 {z}{k1} - -vpsubw %zmm16, %zmm17, %zmm19 -vpsubw (%rax), %zmm17, %zmm19 -vpsubw %zmm16, %zmm17, %zmm19 {k1} -vpsubw (%rax), %zmm17, %zmm19 {k1} -vpsubw %zmm16, %zmm17, %zmm19 {z}{k1} -vpsubw (%rax), %zmm17, %zmm19 {z}{k1} +vpsubq (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vsubpd %zmm16, %zmm17, %zmm19 vsubpd (%rax), %zmm17, %zmm19 +vsubpd (%rax){1to8}, %zmm17, %zmm19 vsubpd %zmm16, %zmm17, %zmm19 {k1} vsubpd (%rax), %zmm17, %zmm19 {k1} +vsubpd (%rax){1to8}, %zmm17, %zmm19 {k1} vsubpd %zmm16, %zmm17, %zmm19 {z}{k1} vsubpd (%rax), %zmm17, %zmm19 {z}{k1} +vsubpd (%rax){1to8}, %zmm17, %zmm19 {z}{k1} vsubps %zmm16, %zmm17, %zmm19 vsubps (%rax), %zmm17, %zmm19 +vsubps (%rax){1to16}, %zmm17, %zmm19 vsubps %zmm16, %zmm17, %zmm19 {k1} vsubps (%rax), %zmm17, %zmm19 {k1} +vsubps (%rax){1to16}, %zmm17, %zmm19 {k1} vsubps %zmm16, %zmm17, %zmm19 {z}{k1} vsubps (%rax), %zmm17, %zmm19 {z}{k1} +vsubps (%rax){1to16}, %zmm17, %zmm19 {z}{k1} # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps @@ -138,112 +152,130 @@ vsubps (%rax), %zmm17, %zmm19 {z}{k1} # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 4 0.50 vaddpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 11 0.50 * vaddpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 11 0.50 * vaddpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 4 0.50 vaddpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vaddpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vaddpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vaddpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vaddps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 11 0.50 * vaddps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 11 0.50 * vaddps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 4 0.50 vaddps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vaddps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vaddps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vaddps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 3 23 16.00 vdivpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 4 30 16.00 * vdivpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 4 30 16.00 * vdivpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 3 23 16.00 vdivpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 4 30 16.00 * vdivpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 4 30 16.00 * vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 3 23 16.00 vdivpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 4 30 16.00 * vdivpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 4 30 16.00 * vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 3 18 10.00 vdivps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 4 25 10.00 * vdivps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 4 25 10.00 * vdivps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 3 18 10.00 vdivps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 4 25 10.00 * vdivps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 4 25 10.00 * vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 3 18 10.00 vdivps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 4 25 10.00 * vdivps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 4 25 10.00 * vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vmulpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 11 0.50 * vmulpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 11 0.50 * vmulpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 4 0.50 vmulpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vmulpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vmulpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vmulpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vmulps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 11 0.50 * vmulps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 11 0.50 * vmulps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 4 0.50 vmulps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vmulps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vmulps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vmulps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 1.00 vpabsd %zmm16, %zmm19 # CHECK-NEXT: 2 8 1.00 * vpabsd (%rax), %zmm19 +# CHECK-NEXT: 2 8 1.00 * vpabsd (%rax){1to16}, %zmm19 # CHECK-NEXT: 1 1 1.00 vpabsd %zmm16, %zmm19 {%k1} # CHECK-NEXT: 2 8 1.00 * vpabsd (%rax), %zmm19 {%k1} +# CHECK-NEXT: 2 8 1.00 * vpabsd (%rax){1to16}, %zmm19 {%k1} # CHECK-NEXT: 1 1 1.00 vpabsd %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 1.00 * vpabsd (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 1.00 * vpabsd (%rax){1to16}, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 1.00 vpabsq %zmm16, %zmm19 # CHECK-NEXT: 2 8 1.00 * vpabsq (%rax), %zmm19 +# CHECK-NEXT: 2 8 1.00 * vpabsq (%rax){1to8}, %zmm19 # CHECK-NEXT: 1 1 1.00 vpabsq %zmm16, %zmm19 {%k1} # CHECK-NEXT: 2 8 1.00 * vpabsq (%rax), %zmm19 {%k1} +# CHECK-NEXT: 2 8 1.00 * vpabsq (%rax){1to8}, %zmm19 {%k1} # CHECK-NEXT: 1 1 1.00 vpabsq %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 1.00 * vpabsq (%rax), %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.33 vpaddb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.33 vpaddb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.33 vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 1.00 * vpabsq (%rax){1to8}, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.33 vpaddd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpaddd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.33 vpaddd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpaddd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.33 vpaddd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpaddd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.33 vpaddq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpaddq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.33 vpaddq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpaddq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.33 vpaddq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpaddq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.33 vpaddw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.33 vpaddw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.33 vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.33 vpsubb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.33 vpsubb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.33 vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.33 vpsubd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpsubd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.33 vpsubd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpsubd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.33 vpsubd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpsubd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 1 0.33 vpsubq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 8 0.50 * vpsubq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 1 0.33 vpsubq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 8 0.50 * vpsubq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 1 0.33 vpsubq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 8 0.50 * vpsubq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 1 1 0.33 vpsubw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: 1 1 0.33 vpsubw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: 1 1 0.33 vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vsubpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 11 0.50 * vsubpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 11 0.50 * vsubpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: 1 4 0.50 vsubpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vsubpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vsubpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vsubpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vsubps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: 2 11 0.50 * vsubps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 2 11 0.50 * vsubps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: 1 4 0.50 vsubps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vsubps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vsubps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vsubps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK: Resources: # CHECK-NEXT: [0] - SKXDivider @@ -259,115 +291,133 @@ vsubps (%rax), %zmm17, %zmm19 {z}{k1} # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] -# CHECK-NEXT: - 156.00 70.00 16.00 27.00 27.00 - 46.00 - - +# CHECK-NEXT: - 234.00 93.00 12.00 42.00 42.00 - 57.00 - - # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Instructions: # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vaddpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vaddpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vaddpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vaddps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vaddps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vaddps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vaddps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 16.00 2.00 - - - - 1.00 - - vdivpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - 16.00 2.00 - 0.50 0.50 - 1.00 - - vdivpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - 16.00 2.00 - 0.50 0.50 - 1.00 - - vdivpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - 16.00 2.00 - - - - 1.00 - - vdivpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 16.00 2.00 - 0.50 0.50 - 1.00 - - vdivpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - 16.00 2.00 - 0.50 0.50 - 1.00 - - vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 16.00 2.00 - - - - 1.00 - - vdivpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 16.00 2.00 - 0.50 0.50 - 1.00 - - vdivpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - 16.00 2.00 - 0.50 0.50 - 1.00 - - vdivpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 10.00 2.00 - - - - 1.00 - - vdivps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - 10.00 2.00 - 0.50 0.50 - 1.00 - - vdivps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - 10.00 2.00 - 0.50 0.50 - 1.00 - - vdivps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - 10.00 2.00 - - - - 1.00 - - vdivps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 10.00 2.00 - 0.50 0.50 - 1.00 - - vdivps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - 10.00 2.00 - 0.50 0.50 - 1.00 - - vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - 10.00 2.00 - - - - 1.00 - - vdivps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - 10.00 2.00 - 0.50 0.50 - 1.00 - - vdivps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - 10.00 2.00 - 0.50 0.50 - 1.00 - - vdivps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vmulpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vmulpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vmulpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vmulps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vmulps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vmulps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vmulps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - - - vpabsd %zmm16, %zmm19 # CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsd (%rax), %zmm19 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsd (%rax){1to16}, %zmm19 # CHECK-NEXT: - - 1.00 - - - - - - - vpabsd %zmm16, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsd (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsd (%rax){1to16}, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - - - vpabsd %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsd (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsd (%rax){1to16}, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - - - - - - - vpabsq %zmm16, %zmm19 # CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsq (%rax), %zmm19 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsq (%rax){1to8}, %zmm19 # CHECK-NEXT: - - 1.00 - - - - - - - vpabsq %zmm16, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsq (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsq (%rax){1to8}, %zmm19 {%k1} # CHECK-NEXT: - - 1.00 - - - - - - - vpabsq %zmm16, %zmm19 {%k1} {z} # CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsq (%rax), %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsq (%rax){1to8}, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubd (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubd (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubq %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubq (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubq (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubq %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubq (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubq %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubq (%rax), %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %zmm16, %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %zmm17, %zmm19 -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %zmm17, %zmm19 {%k1} -# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} -# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubq (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vsubpd %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubpd (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubpd (%rax){1to8}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vsubpd %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubpd (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vsubpd %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubpd (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubpd (%rax){1to8}, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vsubps %zmm16, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubps (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubps (%rax){1to16}, %zmm17, %zmm19 # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vsubps %zmm16, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubps (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} # CHECK-NEXT: - - 0.50 - - - - 0.50 - - vsubps %zmm16, %zmm17, %zmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubps (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 - 0.50 0.50 - 0.50 - - vsubps (%rax){1to16}, %zmm17, %zmm19 {%k1} {z} diff --git a/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bw.s b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bw.s new file mode 100644 index 00000000000..e9b502a6090 --- /dev/null +++ b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bw.s @@ -0,0 +1,145 @@ +# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py +# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=skylake-avx512 -instruction-tables < %s | FileCheck %s + +vpabsb %zmm16, %zmm19 +vpabsb (%rax), %zmm19 +vpabsb %zmm16, %zmm19 {k1} +vpabsb (%rax), %zmm19 {k1} +vpabsb %zmm16, %zmm19 {z}{k1} +vpabsb (%rax), %zmm19 {z}{k1} + +vpabsw %zmm16, %zmm19 +vpabsw (%rax), %zmm19 +vpabsw %zmm16, %zmm19 {k1} +vpabsw (%rax), %zmm19 {k1} +vpabsw %zmm16, %zmm19 {z}{k1} +vpabsw (%rax), %zmm19 {z}{k1} + +vpaddb %zmm16, %zmm17, %zmm19 +vpaddb (%rax), %zmm17, %zmm19 +vpaddb %zmm16, %zmm17, %zmm19 {k1} +vpaddb (%rax), %zmm17, %zmm19 {k1} +vpaddb %zmm16, %zmm17, %zmm19 {z}{k1} +vpaddb (%rax), %zmm17, %zmm19 {z}{k1} + +vpaddw %zmm16, %zmm17, %zmm19 +vpaddw (%rax), %zmm17, %zmm19 +vpaddw %zmm16, %zmm17, %zmm19 {k1} +vpaddw (%rax), %zmm17, %zmm19 {k1} +vpaddw %zmm16, %zmm17, %zmm19 {z}{k1} +vpaddw (%rax), %zmm17, %zmm19 {z}{k1} + +vpsubb %zmm16, %zmm17, %zmm19 +vpsubb (%rax), %zmm17, %zmm19 +vpsubb %zmm16, %zmm17, %zmm19 {k1} +vpsubb (%rax), %zmm17, %zmm19 {k1} +vpsubb %zmm16, %zmm17, %zmm19 {z}{k1} +vpsubb (%rax), %zmm17, %zmm19 {z}{k1} + +vpsubw %zmm16, %zmm17, %zmm19 +vpsubw (%rax), %zmm17, %zmm19 +vpsubw %zmm16, %zmm17, %zmm19 {k1} +vpsubw (%rax), %zmm17, %zmm19 {k1} +vpsubw %zmm16, %zmm17, %zmm19 {z}{k1} +vpsubw (%rax), %zmm17, %zmm19 {z}{k1} + +# 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 vpabsb %zmm16, %zmm19 +# CHECK-NEXT: 2 8 1.00 * vpabsb (%rax), %zmm19 +# CHECK-NEXT: 1 1 1.00 vpabsb %zmm16, %zmm19 {%k1} +# CHECK-NEXT: 2 8 1.00 * vpabsb (%rax), %zmm19 {%k1} +# CHECK-NEXT: 1 1 1.00 vpabsb %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 1.00 * vpabsb (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 1.00 vpabsw %zmm16, %zmm19 +# CHECK-NEXT: 2 8 1.00 * vpabsw (%rax), %zmm19 +# CHECK-NEXT: 1 1 1.00 vpabsw %zmm16, %zmm19 {%k1} +# CHECK-NEXT: 2 8 1.00 * vpabsw (%rax), %zmm19 {%k1} +# CHECK-NEXT: 1 1 1.00 vpabsw %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 1.00 * vpabsw (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpaddb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.33 vpaddb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpaddw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.33 vpaddw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpsubb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.33 vpsubb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpsubw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: 1 1 0.33 vpsubw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} + +# CHECK: Resources: +# CHECK-NEXT: [0] - SKXDivider +# CHECK-NEXT: [1] - SKXFPDivider +# CHECK-NEXT: [2] - SKXPort0 +# CHECK-NEXT: [3] - SKXPort1 +# CHECK-NEXT: [4] - SKXPort2 +# CHECK-NEXT: [5] - SKXPort3 +# CHECK-NEXT: [6] - SKXPort4 +# CHECK-NEXT: [7] - SKXPort5 +# CHECK-NEXT: [8] - SKXPort6 +# CHECK-NEXT: [9] - SKXPort7 + +# CHECK: Resource pressure per iteration: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] +# CHECK-NEXT: - - 20.00 8.00 9.00 9.00 - 8.00 - - + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Instructions: +# CHECK-NEXT: - - 1.00 - - - - - - - vpabsb %zmm16, %zmm19 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsb (%rax), %zmm19 +# CHECK-NEXT: - - 1.00 - - - - - - - vpabsb %zmm16, %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsb (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - - - - - vpabsb %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsb (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - - - - - - - vpabsw %zmm16, %zmm19 +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsw (%rax), %zmm19 +# CHECK-NEXT: - - 1.00 - - - - - - - vpabsw %zmm16, %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsw (%rax), %zmm19 {%k1} +# CHECK-NEXT: - - 1.00 - - - - - - - vpabsw %zmm16, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 1.00 - 0.50 0.50 - - - - vpabsw (%rax), %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %zmm16, %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %zmm17, %zmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %zmm17, %zmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %zmm16, %zmm17, %zmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %zmm17, %zmm19 {%k1} {z} diff --git a/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bwvl.s b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bwvl.s new file mode 100644 index 00000000000..0eedb6eaf2b --- /dev/null +++ b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512bwvl.s @@ -0,0 +1,259 @@ +# NOTE: Assertions have been autogenerated by utils/update_mca_test_checks.py +# RUN: llvm-mca -mtriple=x86_64-unknown-unknown -mcpu=skylake-avx512 -instruction-tables < %s | FileCheck %s + +vpabsb %xmm16, %xmm19 +vpabsb (%rax), %xmm19 +vpabsb %xmm16, %xmm19 {k1} +vpabsb (%rax), %xmm19 {k1} +vpabsb %xmm16, %xmm19 {z}{k1} +vpabsb (%rax), %xmm19 {z}{k1} + +vpabsb %ymm16, %ymm19 +vpabsb (%rax), %ymm19 +vpabsb %ymm16, %ymm19 {k1} +vpabsb (%rax), %ymm19 {k1} +vpabsb %ymm16, %ymm19 {z}{k1} +vpabsb (%rax), %ymm19 {z}{k1} + +vpabsw %xmm16, %xmm19 +vpabsw (%rax), %xmm19 +vpabsw %xmm16, %xmm19 {k1} +vpabsw (%rax), %xmm19 {k1} +vpabsw %xmm16, %xmm19 {z}{k1} +vpabsw (%rax), %xmm19 {z}{k1} + +vpabsw %ymm16, %ymm19 +vpabsw (%rax), %ymm19 +vpabsw %ymm16, %ymm19 {k1} +vpabsw (%rax), %ymm19 {k1} +vpabsw %ymm16, %ymm19 {z}{k1} +vpabsw (%rax), %ymm19 {z}{k1} + +vpaddb %xmm16, %xmm17, %xmm19 +vpaddb (%rax), %xmm17, %xmm19 +vpaddb %xmm16, %xmm17, %xmm19 {k1} +vpaddb (%rax), %xmm17, %xmm19 {k1} +vpaddb %xmm16, %xmm17, %xmm19 {z}{k1} +vpaddb (%rax), %xmm17, %xmm19 {z}{k1} + +vpaddb %ymm16, %ymm17, %ymm19 +vpaddb (%rax), %ymm17, %ymm19 +vpaddb %ymm16, %ymm17, %ymm19 {k1} +vpaddb (%rax), %ymm17, %ymm19 {k1} +vpaddb %ymm16, %ymm17, %ymm19 {z}{k1} +vpaddb (%rax), %ymm17, %ymm19 {z}{k1} + +vpaddw %xmm16, %xmm17, %xmm19 +vpaddw (%rax), %xmm17, %xmm19 +vpaddw %xmm16, %xmm17, %xmm19 {k1} +vpaddw (%rax), %xmm17, %xmm19 {k1} +vpaddw %xmm16, %xmm17, %xmm19 {z}{k1} +vpaddw (%rax), %xmm17, %xmm19 {z}{k1} + +vpaddw %ymm16, %ymm17, %ymm19 +vpaddw (%rax), %ymm17, %ymm19 +vpaddw %ymm16, %ymm17, %ymm19 {k1} +vpaddw (%rax), %ymm17, %ymm19 {k1} +vpaddw %ymm16, %ymm17, %ymm19 {z}{k1} +vpaddw (%rax), %ymm17, %ymm19 {z}{k1} + +vpsubb %xmm16, %xmm17, %xmm19 +vpsubb (%rax), %xmm17, %xmm19 +vpsubb %xmm16, %xmm17, %xmm19 {k1} +vpsubb (%rax), %xmm17, %xmm19 {k1} +vpsubb %xmm16, %xmm17, %xmm19 {z}{k1} +vpsubb (%rax), %xmm17, %xmm19 {z}{k1} + +vpsubb %ymm16, %ymm17, %ymm19 +vpsubb (%rax), %ymm17, %ymm19 +vpsubb %ymm16, %ymm17, %ymm19 {k1} +vpsubb (%rax), %ymm17, %ymm19 {k1} +vpsubb %ymm16, %ymm17, %ymm19 {z}{k1} +vpsubb (%rax), %ymm17, %ymm19 {z}{k1} + +vpsubw %xmm16, %xmm17, %xmm19 +vpsubw (%rax), %xmm17, %xmm19 +vpsubw %xmm16, %xmm17, %xmm19 {k1} +vpsubw (%rax), %xmm17, %xmm19 {k1} +vpsubw %xmm16, %xmm17, %xmm19 {z}{k1} +vpsubw (%rax), %xmm17, %xmm19 {z}{k1} + +vpsubw %ymm16, %ymm17, %ymm19 +vpsubw (%rax), %ymm17, %ymm19 +vpsubw %ymm16, %ymm17, %ymm19 {k1} +vpsubw (%rax), %ymm17, %ymm19 {k1} +vpsubw %ymm16, %ymm17, %ymm19 {z}{k1} +vpsubw (%rax), %ymm17, %ymm19 {z}{k1} + +# 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 vpabsb %xmm16, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpabsb (%rax), %xmm19 +# CHECK-NEXT: 1 1 0.50 vpabsb %xmm16, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpabsb (%rax), %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsb %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpabsb (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpabsb %ymm16, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %ymm19 +# CHECK-NEXT: 1 1 0.50 vpabsb %ymm16, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsb %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsb (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpabsw %xmm16, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpabsw (%rax), %xmm19 +# CHECK-NEXT: 1 1 0.50 vpabsw %xmm16, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpabsw (%rax), %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsw %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpabsw (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.50 vpabsw %ymm16, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %ymm19 +# CHECK-NEXT: 1 1 0.50 vpabsw %ymm16, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.50 vpabsw %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpabsw (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpaddb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpaddb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.33 vpaddb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpaddb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpaddb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpaddb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpaddb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.33 vpaddb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpaddb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpaddw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpaddw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.33 vpaddw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpaddw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpaddw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpaddw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpaddw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.33 vpaddw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpaddw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpaddw (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpsubb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpsubb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.33 vpsubb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpsubb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpsubb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpsubb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpsubb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.33 vpsubb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpsubb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpsubw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: 2 7 0.50 * vpsubw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 1 1 0.33 vpsubw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 7 0.50 * vpsubw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpsubw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 7 0.50 * vpsubw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 1 1 0.33 vpsubw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 1 1 0.33 vpsubw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 1 1 0.33 vpsubw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 8 0.50 * vpsubw (%rax), %ymm17, %ymm19 {%k1} {z} + +# CHECK: Resources: +# CHECK-NEXT: [0] - SKXDivider +# CHECK-NEXT: [1] - SKXFPDivider +# CHECK-NEXT: [2] - SKXPort0 +# CHECK-NEXT: [3] - SKXPort1 +# CHECK-NEXT: [4] - SKXPort2 +# CHECK-NEXT: [5] - SKXPort3 +# CHECK-NEXT: [6] - SKXPort4 +# CHECK-NEXT: [7] - SKXPort5 +# CHECK-NEXT: [8] - SKXPort6 +# CHECK-NEXT: [9] - SKXPort7 + +# CHECK: Resource pressure per iteration: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] +# CHECK-NEXT: - - 28.00 28.00 18.00 18.00 - 16.00 - - + +# CHECK: Resource pressure by instruction: +# CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Instructions: +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsb %xmm16, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsb (%rax), %xmm19 +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsb %xmm16, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsb (%rax), %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsb %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsb (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsb %ymm16, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsb (%rax), %ymm19 +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsb %ymm16, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsb (%rax), %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsb %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsb (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsw %xmm16, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsw (%rax), %xmm19 +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsw %xmm16, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsw (%rax), %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsw %xmm16, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsw (%rax), %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsw %ymm16, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsw (%rax), %ymm19 +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsw %ymm16, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsw (%rax), %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 - - - - - - vpabsw %ymm16, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vpabsw (%rax), %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpaddw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpaddw (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubb %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubb (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %xmm16, %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %xmm16, %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %xmm16, %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %ymm16, %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %ymm16, %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.33 0.33 - - - 0.33 - - vpsubw %ymm16, %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.33 0.33 0.50 0.50 - 0.33 - - vpsubw (%rax), %ymm17, %ymm19 {%k1} {z} diff --git a/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512vl.s b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512vl.s index 888601a04c0..3ec9efc028a 100644 --- a/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512vl.s +++ b/test/tools/llvm-mca/X86/SkylakeServer/resources-avx512vl.s @@ -3,115 +3,163 @@ vaddpd %xmm16, %xmm17, %xmm19 vaddpd (%rax), %xmm17, %xmm19 +vaddpd (%rax){1to2}, %xmm17, %xmm19 vaddpd %xmm16, %xmm17, %xmm19 {k1} vaddpd (%rax), %xmm17, %xmm19 {k1} +vaddpd (%rax){1to2}, %xmm17, %xmm19 {k1} vaddpd %xmm16, %xmm17, %xmm19 {z}{k1} vaddpd (%rax), %xmm17, %xmm19 {z}{k1} +vaddpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vaddpd %ymm16, %ymm17, %ymm19 vaddpd (%rax), %ymm17, %ymm19 +vaddpd (%rax){1to4}, %ymm17, %ymm19 vaddpd %ymm16, %ymm17, %ymm19 {k1} vaddpd (%rax), %ymm17, %ymm19 {k1} +vaddpd (%rax){1to4}, %ymm17, %ymm19 {k1} vaddpd %ymm16, %ymm17, %ymm19 {z}{k1} vaddpd (%rax), %ymm17, %ymm19 {z}{k1} +vaddpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vaddps %xmm16, %xmm17, %xmm19 vaddps (%rax), %xmm17, %xmm19 +vaddps (%rax){1to4}, %xmm17, %xmm19 vaddps %xmm16, %xmm17, %xmm19 {k1} vaddps (%rax), %xmm17, %xmm19 {k1} +vaddps (%rax){1to4}, %xmm17, %xmm19 {k1} vaddps %xmm16, %xmm17, %xmm19 {z}{k1} vaddps (%rax), %xmm17, %xmm19 {z}{k1} +vaddps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vaddps %ymm16, %ymm17, %ymm19 vaddps (%rax), %ymm17, %ymm19 +vaddps (%rax){1to8}, %ymm17, %ymm19 vaddps %ymm16, %ymm17, %ymm19 {k1} vaddps (%rax), %ymm17, %ymm19 {k1} +vaddps (%rax){1to8}, %ymm17, %ymm19 {k1} vaddps %ymm16, %ymm17, %ymm19 {z}{k1} vaddps (%rax), %ymm17, %ymm19 {z}{k1} +vaddps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} vdivpd %xmm16, %xmm17, %xmm19 vdivpd (%rax), %xmm17, %xmm19 +vdivpd (%rax){1to2}, %xmm17, %xmm19 vdivpd %xmm16, %xmm17, %xmm19 {k1} vdivpd (%rax), %xmm17, %xmm19 {k1} +vdivpd (%rax){1to2}, %xmm17, %xmm19 {k1} vdivpd %xmm16, %xmm17, %xmm19 {z}{k1} vdivpd (%rax), %xmm17, %xmm19 {z}{k1} +vdivpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vdivpd %ymm16, %ymm17, %ymm19 vdivpd (%rax), %ymm17, %ymm19 +vdivpd (%rax){1to4}, %ymm17, %ymm19 vdivpd %ymm16, %ymm17, %ymm19 {k1} vdivpd (%rax), %ymm17, %ymm19 {k1} +vdivpd (%rax){1to4}, %ymm17, %ymm19 {k1} vdivpd %ymm16, %ymm17, %ymm19 {z}{k1} vdivpd (%rax), %ymm17, %ymm19 {z}{k1} +vdivpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vdivps %xmm16, %xmm17, %xmm19 vdivps (%rax), %xmm17, %xmm19 +vdivps (%rax){1to4}, %xmm17, %xmm19 vdivps %xmm16, %xmm17, %xmm19 {k1} vdivps (%rax), %xmm17, %xmm19 {k1} +vdivps (%rax){1to4}, %xmm17, %xmm19 {k1} vdivps %xmm16, %xmm17, %xmm19 {z}{k1} vdivps (%rax), %xmm17, %xmm19 {z}{k1} +vdivps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vdivps %ymm16, %ymm17, %ymm19 vdivps (%rax), %ymm17, %ymm19 +vdivps (%rax){1to8}, %ymm17, %ymm19 vdivps %ymm16, %ymm17, %ymm19 {k1} vdivps (%rax), %ymm17, %ymm19 {k1} +vdivps (%rax){1to8}, %ymm17, %ymm19 {k1} vdivps %ymm16, %ymm17, %ymm19 {z}{k1} vdivps (%rax), %ymm17, %ymm19 {z}{k1} +vdivps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} vmulpd %xmm16, %xmm17, %xmm19 vmulpd (%rax), %xmm17, %xmm19 +vmulpd (%rax){1to2}, %xmm17, %xmm19 vmulpd %xmm16, %xmm17, %xmm19 {k1} vmulpd (%rax), %xmm17, %xmm19 {k1} +vmulpd (%rax){1to2}, %xmm17, %xmm19 {k1} vmulpd %xmm16, %xmm17, %xmm19 {z}{k1} vmulpd (%rax), %xmm17, %xmm19 {z}{k1} +vmulpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vmulpd %ymm16, %ymm17, %ymm19 vmulpd (%rax), %ymm17, %ymm19 +vmulpd (%rax){1to4}, %ymm17, %ymm19 vmulpd %ymm16, %ymm17, %ymm19 {k1} vmulpd (%rax), %ymm17, %ymm19 {k1} +vmulpd (%rax){1to4}, %ymm17, %ymm19 {k1} vmulpd %ymm16, %ymm17, %ymm19 {z}{k1} vmulpd (%rax), %ymm17, %ymm19 {z}{k1} +vmulpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vmulps %xmm16, %xmm17, %xmm19 vmulps (%rax), %xmm17, %xmm19 +vmulps (%rax){1to4}, %xmm17, %xmm19 vmulps %xmm16, %xmm17, %xmm19 {k1} vmulps (%rax), %xmm17, %xmm19 {k1} +vmulps (%rax){1to4}, %xmm17, %xmm19 {k1} vmulps %xmm16, %xmm17, %xmm19 {z}{k1} vmulps (%rax), %xmm17, %xmm19 {z}{k1} +vmulps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vmulps %ymm16, %ymm17, %ymm19 vmulps (%rax), %ymm17, %ymm19 +vmulps (%rax){1to8}, %ymm17, %ymm19 vmulps %ymm16, %ymm17, %ymm19 {k1} vmulps (%rax), %ymm17, %ymm19 {k1} +vmulps (%rax){1to8}, %ymm17, %ymm19 {k1} vmulps %ymm16, %ymm17, %ymm19 {z}{k1} vmulps (%rax), %ymm17, %ymm19 {z}{k1} +vmulps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} vsubpd %xmm16, %xmm17, %xmm19 vsubpd (%rax), %xmm17, %xmm19 +vsubpd (%rax){1to2}, %xmm17, %xmm19 vsubpd %xmm16, %xmm17, %xmm19 {k1} vsubpd (%rax), %xmm17, %xmm19 {k1} +vsubpd (%rax){1to2}, %xmm17, %xmm19 {k1} vsubpd %xmm16, %xmm17, %xmm19 {z}{k1} vsubpd (%rax), %xmm17, %xmm19 {z}{k1} +vsubpd (%rax){1to2}, %xmm17, %xmm19 {z}{k1} vsubpd %ymm16, %ymm17, %ymm19 vsubpd (%rax), %ymm17, %ymm19 +vsubpd (%rax){1to4}, %ymm17, %ymm19 vsubpd %ymm16, %ymm17, %ymm19 {k1} vsubpd (%rax), %ymm17, %ymm19 {k1} +vsubpd (%rax){1to4}, %ymm17, %ymm19 {k1} vsubpd %ymm16, %ymm17, %ymm19 {z}{k1} vsubpd (%rax), %ymm17, %ymm19 {z}{k1} +vsubpd (%rax){1to4}, %ymm17, %ymm19 {z}{k1} vsubps %xmm16, %xmm17, %xmm19 vsubps (%rax), %xmm17, %xmm19 +vsubps (%rax){1to4}, %xmm17, %xmm19 vsubps %xmm16, %xmm17, %xmm19 {k1} vsubps (%rax), %xmm17, %xmm19 {k1} +vsubps (%rax){1to4}, %xmm17, %xmm19 {k1} vsubps %xmm16, %xmm17, %xmm19 {z}{k1} vsubps (%rax), %xmm17, %xmm19 {z}{k1} +vsubps (%rax){1to4}, %xmm17, %xmm19 {z}{k1} vsubps %ymm16, %ymm17, %ymm19 vsubps (%rax), %ymm17, %ymm19 +vsubps (%rax){1to8}, %ymm17, %ymm19 vsubps %ymm16, %ymm17, %ymm19 {k1} vsubps (%rax), %ymm17, %ymm19 {k1} +vsubps (%rax){1to8}, %ymm17, %ymm19 {k1} vsubps %ymm16, %ymm17, %ymm19 {z}{k1} vsubps (%rax), %ymm17, %ymm19 {z}{k1} +vsubps (%rax){1to8}, %ymm17, %ymm19 {z}{k1} # CHECK: Instruction Info: # CHECK-NEXT: [1]: #uOps @@ -124,100 +172,148 @@ vsubps (%rax), %ymm17, %ymm19 {z}{k1} # CHECK: [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: 1 4 0.50 vaddpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 10 0.50 * vaddpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 10 0.50 * vaddpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 4 0.50 vaddpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 10 0.50 * vaddpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 10 0.50 * vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vaddpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 10 0.50 * vaddpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 10 0.50 * vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vaddpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 11 0.50 * vaddpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 11 0.50 * vaddpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 1 4 0.50 vaddpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vaddpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 4 0.50 vaddpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vaddpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vaddps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 10 0.50 * vaddps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 10 0.50 * vaddps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 4 0.50 vaddps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 10 0.50 * vaddps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 10 0.50 * vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vaddps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 10 0.50 * vaddps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 10 0.50 * vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vaddps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 11 0.50 * vaddps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 11 0.50 * vaddps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 1 4 0.50 vaddps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vaddps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 4 0.50 vaddps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vaddps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 14 3.00 vdivpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 20 4.00 * vdivpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 20 4.00 * vdivpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 14 3.00 vdivpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 20 4.00 * vdivpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 20 4.00 * vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 14 3.00 vdivpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 20 4.00 * vdivpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 20 4.00 * vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 14 5.00 vdivpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 21 8.00 * vdivpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 21 8.00 * vdivpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 1 14 5.00 vdivpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 21 8.00 * vdivpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 21 8.00 * vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 14 5.00 vdivpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 21 8.00 * vdivpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 21 8.00 * vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 11 3.00 vdivps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 17 5.00 * vdivps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 17 5.00 * vdivps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 11 3.00 vdivps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 17 5.00 * vdivps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 17 5.00 * vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 11 3.00 vdivps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 17 5.00 * vdivps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 17 5.00 * vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 11 5.00 vdivps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 18 5.00 * vdivps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 18 5.00 * vdivps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 1 11 5.00 vdivps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 18 5.00 * vdivps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 18 5.00 * vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 11 5.00 vdivps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 18 5.00 * vdivps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 18 5.00 * vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vmulpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 10 0.50 * vmulpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 10 0.50 * vmulpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 4 0.50 vmulpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 10 0.50 * vmulpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 10 0.50 * vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vmulpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 10 0.50 * vmulpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 10 0.50 * vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vmulpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 11 0.50 * vmulpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 11 0.50 * vmulpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 1 4 0.50 vmulpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vmulpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 4 0.50 vmulpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vmulpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vmulps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 10 0.50 * vmulps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 10 0.50 * vmulps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 4 0.50 vmulps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 10 0.50 * vmulps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 10 0.50 * vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vmulps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 10 0.50 * vmulps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 10 0.50 * vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vmulps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 11 0.50 * vmulps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 11 0.50 * vmulps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 1 4 0.50 vmulps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vmulps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 4 0.50 vmulps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vmulps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vsubpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 10 0.50 * vsubpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 10 0.50 * vsubpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: 1 4 0.50 vsubpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 10 0.50 * vsubpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 10 0.50 * vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vsubpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 10 0.50 * vsubpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 10 0.50 * vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vsubpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 11 0.50 * vsubpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 11 0.50 * vsubpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: 1 4 0.50 vsubpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vsubpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 4 0.50 vsubpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vsubpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vsubps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: 2 10 0.50 * vsubps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: 2 10 0.50 * vsubps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: 1 4 0.50 vsubps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 2 10 0.50 * vsubps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: 2 10 0.50 * vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: 1 4 0.50 vsubps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 2 10 0.50 * vsubps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: 2 10 0.50 * vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: 1 4 0.50 vsubps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: 2 11 0.50 * vsubps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: 2 11 0.50 * vsubps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: 1 4 0.50 vsubps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 2 11 0.50 * vsubps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: 2 11 0.50 * vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: 1 4 0.50 vsubps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: 2 11 0.50 * vsubps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: 2 11 0.50 * vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK: Resources: # CHECK-NEXT: [0] - SKXDivider @@ -233,103 +329,151 @@ vsubps (%rax), %ymm17, %ymm19 {z}{k1} # CHECK: Resource pressure per iteration: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] -# CHECK-NEXT: - 114.00 60.00 36.00 24.00 24.00 - - - - +# CHECK-NEXT: - 180.00 90.00 54.00 48.00 48.00 - - - - # CHECK: Resource pressure by instruction: # CHECK-NEXT: [0] [1] [2] [3] [4] [5] [6] [7] [8] [9] Instructions: # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vaddps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vaddps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 3.00 1.00 - - - - - - - vdivpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - 4.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - 4.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - 3.00 1.00 - - - - - - - vdivpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 4.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - 4.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 3.00 1.00 - - - - - - - vdivpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 4.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - 4.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 5.00 1.00 - - - - - - - vdivpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - 8.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - 8.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - 5.00 1.00 - - - - - - - vdivpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 8.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - 8.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 5.00 1.00 - - - - - - - vdivpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 8.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - 8.00 1.00 - 0.50 0.50 - - - - vdivpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 3.00 1.00 - - - - - - - vdivps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - 3.00 1.00 - - - - - - - vdivps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - 3.00 1.00 - - - - - - - vdivps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - 5.00 1.00 - - - - - - - vdivps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - 5.00 1.00 - - - - - - - vdivps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - 5.00 1.00 - - - - - - - vdivps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - 5.00 1.00 - 0.50 0.50 - - - - vdivps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vmulps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vmulps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubpd %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax){1to2}, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubpd %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubpd %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax){1to2}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubpd %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax){1to4}, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubpd %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubpd %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubpd (%rax){1to4}, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubps %xmm16, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax), %xmm17, %xmm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax){1to4}, %xmm17, %xmm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubps %xmm16, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax), %xmm17, %xmm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubps %xmm16, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax), %xmm17, %xmm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax){1to4}, %xmm17, %xmm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubps %ymm16, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax), %ymm17, %ymm19 +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax){1to8}, %ymm17, %ymm19 # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubps %ymm16, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax), %ymm17, %ymm19 {%k1} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} # CHECK-NEXT: - - 0.50 0.50 - - - - - - vsubps %ymm16, %ymm17, %ymm19 {%k1} {z} # CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax), %ymm17, %ymm19 {%k1} {z} +# CHECK-NEXT: - - 0.50 0.50 0.50 0.50 - - - - vsubps (%rax){1to8}, %ymm17, %ymm19 {%k1} {z} -- 2.50.1