Julia 1.11 private tests are failing

[vchuravy]

Test Failed at /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:98
  Expression: !(occursin("gcframe", IR))
   Evaluated: !(occursin("gcframe", "; Function Signature: cpu_reduce_private(KernelAbstractions.CompilerMetadata{KernelAbstractions.NDIteration.DynamicSize, KernelAbstractions.NDIteration.NoDynamicCheck, Base.IteratorsMD.CartesianIndex{1}, Base.IteratorsMD.CartesianIndices{1, Tuple{Base.OneTo{Int64}}}, KernelAbstractions.NDIteration.NDRange{1, KernelAbstractions.NDIteration.DynamicSize, KernelAbstractions.NDIteration.StaticSize{(8,)}, Base.IteratorsMD.CartesianIndices{1, Tuple{Base.OneTo{Int64}}}, Nothing}}, Array{Float64, 1}, Array{Float64, 2})\n;  @ none within `cpu_reduce_private`\ndefine void @julia_cpu_reduce_private_15381({ [1 x [1 x i64]], [1 x [1 x [1 x i64]]], { [1 x [1 x [1 x i64]]] } }* nocapture noundef nonnull readonly align 8 dereferenceable(24) %\"__ctx__::CompilerMetadata\", {}* noundef nonnull align 8 dereferenceable(24) %\"out::Array\", {}* noundef nonnull align 8 dereferenceable(32) %\"A::Array\") #0 {\ntop:\n  %gcframe663 = alloca [3 x {}*], align 16\n  %gcframe663.sub = getelementptr inbounds [3 x {}*], [3 x {}*]* %gcframe663, i64 0, i64 0\n  %0 = bitcast [3 x {}*]* %gcframe663 to i8*\n  call void @llvm.memset.p0i8.i64(i8* align 16 %0, i8 0, i64 24, i1 true)\n  %\"new::Tuple116\" = alloca [2 x i64], align 8\n  %\"new::Tuple241\" = alloca [1 x i64], align 8\n  %\"new::Tuple253\" = alloca [1 x i64], align 8\n  %thread_ptr = call i8* asm \"movq %fs:0, \$0\", \"=r\"() #16\n  %tls_ppgcstack = getelementptr i8, i8* %thread_ptr, i64 -8\n  %1 = bitcast i8* %tls_ppgcstack to {}****\n  %tls_pgcstack = load {}***, {}**** %1, align 8\n  %frame.nroots661 = bitcast [3 x {}*]* %gcframe663 to i64*\n  store i64 4, i64* %frame.nroots661, align 16\n  %frame.prev = getelementptr inbounds [3 x {}*], [3 x {}*]* %gcframe663, i64 0, i64 1\n  %2 = bitcast {}** %frame.prev to {}***\n  %task.gcstack = load {}**, {}*** %tls_pgcstack, align 8\n  store {}** %task.gcstack, {}*** %2, align 8\n  %3 = bitcast {}*** %tls_pgcstack to {}***\n  store {}** %gcframe663.sub, {}*** %3, align 8\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:178 within `Scratchpad`\n   %lcnt = load volatile i64, i64* inttoptr (i64 129620768 to i64*), align 32\n   %4 = add i64 %lcnt, 1\n   store volatile i64 %4, i64* inttoptr (i64 129620768 to i64*), align 32\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:179 within `Scratchpad`\n   %lcnt1 = load volatile i64, i64* inttoptr (i64 129620776 to i64*), align 8\n   %5 = add i64 %lcnt1, 1\n   store volatile i64 %5, i64* inttoptr (i64 129620776 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:25 within `__groupsize`\n    %lcnt2 = load volatile i64, i64* inttoptr (i64 57810808 to i64*), align 8\n    %6 = add i64 %lcnt2, 1\n    store volatile i64 %6, i64* inttoptr (i64 57810808 to i64*), align 8\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n     %lcnt3 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n     %7 = add i64 %lcnt3, 1\n     store volatile i64 %7, i64* inttoptr (i64 57810792 to i64*), align 8\n; ││└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n     %lcnt4 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n     %8 = add i64 %lcnt4, 1\n     store volatile i64 %8, i64* inttoptr (i64 81001848 to i64*), align 8\n; │└└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/KernelAbstractions.jl:698 within `__size`\n    %lcnt5 = load volatile i64, i64* inttoptr (i64 41261072 to i64*), align 16\n    %9 = add i64 %lcnt5, 1\n    store volatile i64 %9, i64* inttoptr (i64 41261072 to i64*), align 16\n; │└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:175 within `ScratchArray`\n    %lcnt7 = load volatile i64, i64* inttoptr (i64 129620744 to i64*), align 8\n    %10 = add i64 %lcnt7, 1\n    store volatile i64 %10, i64* inttoptr (i64 129620744 to i64*), align 8\n; └└\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:29 within `__workitems_iterspace`\n   %lcnt8 = load volatile i64, i64* inttoptr (i64 57810840 to i64*), align 8\n   %11 = add i64 %lcnt8, 1\n   store volatile i64 %11, i64* inttoptr (i64 57810840 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n    %lcnt9 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n    %12 = add i64 %lcnt9, 1\n    store volatile i64 %12, i64* inttoptr (i64 57810792 to i64*), align 8\n; │└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n    %lcnt10 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n    %13 = add i64 %lcnt10, 1\n    store volatile i64 %13, i64* inttoptr (i64 81001848 to i64*), align 8\n; └└\n;  @ none within `cpu_reduce_private` @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/macros.jl:284\n  %lcnt11 = load volatile i64, i64* inttoptr (i64 31813520 to i64*), align 16\n  %14 = add i64 %lcnt11, 1\n  store volatile i64 %14, i64* inttoptr (i64 31813520 to i64*), align 16\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:29 within `__workitems_iterspace`\n   %lcnt12 = load volatile i64, i64* inttoptr (i64 57810840 to i64*), align 8\n   %15 = add i64 %lcnt12, 1\n   store volatile i64 %15, i64* inttoptr (i64 57810840 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n    %lcnt13 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n    %16 = add i64 %lcnt13, 1\n    store volatile i64 %16, i64* inttoptr (i64 57810792 to i64*), align 8\n; │└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n    %lcnt14 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n    %17 = add i64 %lcnt14, 1\n    store volatile i64 %17, i64* inttoptr (i64 81001848 to i64*), align 8\n    %\"__ctx__::CompilerMetadata.groupindex_ptr.I_ptr[1]_ptr\" = getelementptr inbounds { [1 x [1 x i64]], [1 x [1 x [1 x i64]]], { [1 x [1 x [1 x i64]]] } }, { [1 x [1 x i64]], [1 x [1 x [1 x i64]]], { [1 x [1 x [1 x i64]]] } }* %\"__ctx__::CompilerMetadata\", i64 0, i32 0, i64 0, i64 0\n    %\"__ctx__::CompilerMetadata.groupindex_ptr.I_ptr[1]_ptr.unbox\" = load i64, i64* %\"__ctx__::CompilerMetadata.groupindex_ptr.I_ptr[1]_ptr\", align 8\n    %18 = shl i64 %\"__ctx__::CompilerMetadata.groupindex_ptr.I_ptr[1]_ptr.unbox\", 3\n    %19 = bitcast {}* %\"A::Array\" to i8*\n    %20 = getelementptr inbounds i8, i8* %19, i64 16\n    %.size118.sroa.0.0..sroa_idx = bitcast i8* %20 to i64*\n    %.size118.sroa.2.0..sroa_idx279 = getelementptr inbounds i8, i8* %19, i64 24\n    %21 = bitcast i8* %.size118.sroa.2.0..sroa_idx279 to i64*\n    %22 = bitcast {}* %\"A::Array\" to { i8*, {}* }*\n    %23 = getelementptr inbounds { i8*, {}* }, { i8*, {}* }* %22, i64 0, i32 1\n    %24 = bitcast {}** %23 to i64**\n    %25 = bitcast {}* %\"A::Array\" to i8**\n    %26 = bitcast {}* %\"out::Array\" to i8*\n    %27 = getelementptr inbounds i8, i8* %26, i64 16\n    %memcpy_refined_src = bitcast i8* %27 to i64*\n    %28 = bitcast {}* %\"out::Array\" to { i8*, {}* }*\n    %29 = bitcast {}* %\"out::Array\" to i8**\n    %30 = getelementptr inbounds { i8*, {}* }, { i8*, {}* }* %28, i64 0, i32 1\n    br label %L129\n\nL129:                                             ; preds = %load224, %top\n    %value_phi = phi i64 [ 1, %top ], [ %166, %load224 ]\n; └└\n;  @ none within `cpu_reduce_private` @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/macros.jl:285\n  %lcnt21 = load volatile i64, i64* inttoptr (i64 31813528 to i64*), align 8\n  %31 = add i64 %lcnt21, 1\n  store volatile i64 %31, i64* inttoptr (i64 31813528 to i64*), align 8\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:151 within `__validindex`\n   %lcnt22 = load volatile i64, i64* inttoptr (i64 62257400 to i64*), align 8\n   %32 = add i64 %lcnt22, 1\n   store volatile i64 %32, i64* inttoptr (i64 62257400 to i64*), align 8\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:153 within `__validindex`\n   %lcnt23 = load volatile i64, i64* inttoptr (i64 62257416 to i64*), align 8\n   %33 = add i64 %lcnt23, 1\n   store volatile i64 %33, i64* inttoptr (i64 62257416 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:26 within `__dynamic_checkbounds`\n    %lcnt24 = load volatile i64, i64* inttoptr (i64 57810816 to i64*), align 128\n    %34 = add i64 %lcnt24, 1\n    store volatile i64 %34, i64* inttoptr (i64 57810816 to i64*), align 128\n; │└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:157 within `__validindex`\n   %lcnt25 = load volatile i64, i64* inttoptr (i64 62257448 to i64*), align 8\n   %35 = add i64 %lcnt25, 1\n   store volatile i64 %35, i64* inttoptr (i64 62257448 to i64*), align 8\n; └\n;  @ none within `cpu_reduce_private` @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/macros.jl:286\n  %lcnt26 = load volatile i64, i64* inttoptr (i64 31813536 to i64*), align 32\n  %36 = add i64 %lcnt26, 1\n  store volatile i64 %36, i64* inttoptr (i64 31813536 to i64*), align 32\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/KernelAbstractions.jl:405 within `__index_Global_NTuple`\n   %lcnt27 = load volatile i64, i64* inttoptr (i64 33976264 to i64*), align 8\n   %37 = add i64 %lcnt27, 1\n   store volatile i64 %37, i64* inttoptr (i64 33976264 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:147 within `__index_Global_Cartesian`\n    %lcnt28 = load volatile i64, i64* inttoptr (i64 62257368 to i64*), align 8\n    %38 = add i64 %lcnt28, 1\n    store volatile i64 %38, i64* inttoptr (i64 62257368 to i64*), align 8\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:148 within `__index_Global_Cartesian`\n    %lcnt29 = load volatile i64, i64* inttoptr (i64 62257376 to i64*), align 32\n    %39 = add i64 %lcnt29, 1\n    store volatile i64 %39, i64* inttoptr (i64 62257376 to i64*), align 32\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n     %lcnt30 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n     %40 = add i64 %lcnt30, 1\n     store volatile i64 %40, i64* inttoptr (i64 57810792 to i64*), align 8\n; ││└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:24 within `__groupindex`\n     %lcnt31 = load volatile i64, i64* inttoptr (i64 57810800 to i64*), align 16\n     %41 = add i64 %lcnt31, 1\n     store volatile i64 %41, i64* inttoptr (i64 57810800 to i64*), align 16\n; ││└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:73 within `expand`\n     %lcnt32 = load volatile i64, i64* inttoptr (i64 81001944 to i64*), align 8\n     %42 = add i64 %lcnt32, 1\n     store volatile i64 %42, i64* inttoptr (i64 81001944 to i64*), align 8\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:74 within `expand`\n     %lcnt33 = load volatile i64, i64* inttoptr (i64 81001952 to i64*), align 32\n     %43 = add i64 %lcnt33, 1\n     store volatile i64 %43, i64* inttoptr (i64 81001952 to i64*), align 32\n; │││┌ @ ntuple.jl:48 within `ntuple`\n; ││││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:75 within `#1`\n       %lcnt34 = load volatile i64, i64* inttoptr (i64 81001960 to i64*), align 8\n       %44 = add i64 %lcnt34, 1\n       store volatile i64 %44, i64* inttoptr (i64 81001960 to i64*), align 8\n; │││││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:76 within `#1`\n       %lcnt35 = load volatile i64, i64* inttoptr (i64 81001968 to i64*), align 16\n       %45 = add i64 %lcnt35, 1\n       store volatile i64 %45, i64* inttoptr (i64 81001968 to i64*), align 16\n; │││││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n        %lcnt36 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n        %46 = add i64 %lcnt36, 1\n        store volatile i64 %46, i64* inttoptr (i64 81001848 to i64*), align 8\n; │││││└\n; │││││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:77 within `#1`\n       %lcnt37 = load volatile i64, i64* inttoptr (i64 81001976 to i64*), align 8\n       %47 = add i64 %lcnt37, 1\n       store volatile i64 %47, i64* inttoptr (i64 81001976 to i64*), align 8\n; │││││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:78 within `#1`\n       %lcnt40 = load volatile i64, i64* inttoptr (i64 81001984 to i64*), align 512\n       %48 = add i64 %lcnt40, 1\n       store volatile i64 %48, i64* inttoptr (i64 81001984 to i64*), align 512\n; │││││┌ @ int.jl:88 within `*`\n        %49 = add nsw i64 %value_phi, -8\n; │││││└\n; │││││┌ @ int.jl:87 within `+`\n        %50 = add i64 %49, %18\n; │││└└└\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:80 within `expand`\n     %lcnt41 = load volatile i64, i64* inttoptr (i64 81002000 to i64*), align 16\n     %51 = add i64 %lcnt41, 1\n     store volatile i64 %51, i64* inttoptr (i64 81002000 to i64*), align 16\n; └└└\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:124 within `__index_Local_Linear`\n   %lcnt42 = load volatile i64, i64* inttoptr (i64 57810560 to i64*), align 128\n   %52 = add i64 %lcnt42, 1\n   store volatile i64 %52, i64* inttoptr (i64 57810560 to i64*), align 128\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:125 within `__index_Local_Linear`\n   %lcnt43 = load volatile i64, i64* inttoptr (i64 57810568 to i64*), align 8\n   %53 = add i64 %lcnt43, 1\n   store volatile i64 %53, i64* inttoptr (i64 57810568 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n    %lcnt44 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n    %54 = add i64 %lcnt44, 1\n    store volatile i64 %54, i64* inttoptr (i64 57810792 to i64*), align 8\n; │└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n    %lcnt45 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n    %55 = add i64 %lcnt45, 1\n    store volatile i64 %55, i64* inttoptr (i64 81001848 to i64*), align 8\n; │└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:126 within `__index_Local_Linear`\n   %lcnt46 = load volatile i64, i64* inttoptr (i64 57810576 to i64*), align 16\n   %56 = add i64 %lcnt46, 1\n   store volatile i64 %56, i64* inttoptr (i64 57810576 to i64*), align 16\n; └\n;  @ none within `cpu_reduce_private` @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/macros.jl:287\n  %lcnt48 = load volatile i64, i64* inttoptr (i64 31813544 to i64*), align 8\n  %57 = add i64 %lcnt48, 1\n  store volatile i64 %57, i64* inttoptr (i64 31813544 to i64*), align 8\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:53 within `macro expansion`\n   %lcnt49 = load volatile i64, i64* inttoptr (i64 61927352 to i64*), align 8\n   %58 = add i64 %lcnt49, 1\n   store volatile i64 %58, i64* inttoptr (i64 61927352 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:124 within `__index_Local_Linear`\n    %lcnt50 = load volatile i64, i64* inttoptr (i64 57810560 to i64*), align 128\n    %59 = add i64 %lcnt50, 1\n    store volatile i64 %59, i64* inttoptr (i64 57810560 to i64*), align 128\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:125 within `__index_Local_Linear`\n    %lcnt51 = load volatile i64, i64* inttoptr (i64 57810568 to i64*), align 8\n    %60 = add i64 %lcnt51, 1\n    store volatile i64 %60, i64* inttoptr (i64 57810568 to i64*), align 8\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n     %lcnt52 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n     %61 = add i64 %lcnt52, 1\n     store volatile i64 %61, i64* inttoptr (i64 57810792 to i64*), align 8\n; ││└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n     %lcnt53 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n     %62 = add i64 %lcnt53, 1\n     store volatile i64 %62, i64* inttoptr (i64 81001848 to i64*), align 8\n; ││└\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:126 within `__index_Local_Linear`\n    %lcnt54 = load volatile i64, i64* inttoptr (i64 57810576 to i64*), align 16\n    %63 = add i64 %lcnt54, 1\n    store volatile i64 %63, i64* inttoptr (i64 57810576 to i64*), align 16\n; │└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:189 within `getindex`\n    %lcnt60 = load volatile i64, i64* inttoptr (i64 129620856 to i64*), align 8\n    %64 = add i64 %lcnt60, 1\n    store volatile i64 %64, i64* inttoptr (i64 129620856 to i64*), align 8\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:190 within `getindex`\n    %lcnt61 = load volatile i64, i64* inttoptr (i64 129620864 to i64*), align 128\n    %65 = add i64 %lcnt61, 1\n    store volatile i64 %65, i64* inttoptr (i64 129620864 to i64*), align 128\n; ││┌ @ ntuple.jl:48 within `ntuple`\n; │││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:190 within `#22`\n      %lcnt62 = load volatile i64, i64* inttoptr (i64 129620864 to i64*), align 128\n      %66 = add i64 %lcnt62, 1\n      store volatile i64 %66, i64* inttoptr (i64 129620864 to i64*), align 128\n; ││└└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:184 within `aview`\n     %lcnt63 = load volatile i64, i64* inttoptr (i64 129620816 to i64*), align 16\n     %67 = add i64 %lcnt63, 1\n     store volatile i64 %67, i64* inttoptr (i64 129620816 to i64*), align 16\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:185 within `aview`\n     %lcnt64 = load volatile i64, i64* inttoptr (i64 129620824 to i64*), align 8\n     %68 = add i64 %lcnt64, 1\n     store volatile i64 %68, i64* inttoptr (i64 129620824 to i64*), align 8\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:186 within `aview`\n     %lcnt65 = load volatile i64, i64* inttoptr (i64 129620832 to i64*), align 32\n     %69 = add i64 %lcnt65, 1\n     store volatile i64 %69, i64* inttoptr (i64 129620832 to i64*), align 32\n; │└└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:54 within `macro expansion`\n   %lcnt75 = load volatile i64, i64* inttoptr (i64 61927360 to i64*), align 64\n   %70 = add i64 %lcnt75, 1\n   store volatile i64 %70, i64* inttoptr (i64 61927360 to i64*), align 64\n; │┌ @ array.jl:191 within `size`\n    %.size.sroa.1.0.copyload = load i64, i64* %21, align 8\n; │└\n; │┌ @ range.jl:5 within `Colon`\n; ││┌ @ range.jl:404 within `UnitRange`\n; │││┌ @ range.jl:415 within `unitrange_last`\n      %71 = call i64 @llvm.smax.i64(i64 %.size.sroa.1.0.copyload, i64 0)\n; │└└└\n; │┌ @ range.jl:898 within `iterate`\n; ││┌ @ range.jl:675 within `isempty`\n; │││┌ @ operators.jl:378 within `>`\n; ││││┌ @ int.jl:83 within `<`\n       %72 = icmp sgt i64 %.size.sroa.1.0.copyload, 0\n; ││└└└\n    br i1 %72, label %L150.split, label %L129.L361_crit_edge\n\nL129.L361_crit_edge:                              ; preds = %L129\n; │└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:57 within `macro expansion`\n; │┌ @ array.jl:965 within `setindex!`\n; ││┌ @ int.jl:86 within `-`\n     %.pre548 = add i64 %50, -1\n; │└└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:54 within `macro expansion`\n; │┌ @ range.jl:898 within `iterate`\n    br label %L361\n\nL150.split:                                       ; preds = %L129\n    %lcnt263 = load volatile i64, i64* inttoptr (i64 61927360 to i64*), align 64\n    %73 = add i64 %lcnt263, 1\n    store volatile i64 %73, i64* inttoptr (i64 61927360 to i64*), align 64\n    %74 = add i64 %50, -1\n; │└\n   br label %L155\n\nL155:                                             ; preds = %L300, %L150.split\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:55 within `macro expansion`\n; │┌ @ subarray.jl:321 within `getindex` @ /home/runner/.julia/packages/StaticArrays/cZ1ET/src/MArray.jl:25\n; ││┌ @ pointer.jl:152 within `unsafe_load`\n     %pointerref = phi double [ 0.000000e+00, %L150.split ], [ %115, %L300 ]\n     %value_phi83 = phi i64 [ 1, %L150.split ], [ %118, %L300 ]\n; │└└\n   %lcnt85 = load volatile i64, i64* inttoptr (i64 61927368 to i64*), align 8\n   %75 = add i64 %lcnt85, 1\n   store volatile i64 %75, i64* inttoptr (i64 61927368 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:124 within `__index_Local_Linear`\n    %lcnt86 = load volatile i64, i64* inttoptr (i64 57810560 to i64*), align 128\n    %76 = add i64 %lcnt86, 1\n    store volatile i64 %76, i64* inttoptr (i64 57810560 to i64*), align 128\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:125 within `__index_Local_Linear`\n    %lcnt87 = load volatile i64, i64* inttoptr (i64 57810568 to i64*), align 8\n    %77 = add i64 %lcnt87, 1\n    store volatile i64 %77, i64* inttoptr (i64 57810568 to i64*), align 8\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n     %lcnt88 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n     %78 = add i64 %lcnt88, 1\n     store volatile i64 %78, i64* inttoptr (i64 57810792 to i64*), align 8\n; ││└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n     %lcnt89 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n     %79 = add i64 %lcnt89, 1\n     store volatile i64 %79, i64* inttoptr (i64 81001848 to i64*), align 8\n; ││└\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:126 within `__index_Local_Linear`\n    %lcnt90 = load volatile i64, i64* inttoptr (i64 57810576 to i64*), align 16\n    %80 = add i64 %lcnt90, 1\n    store volatile i64 %80, i64* inttoptr (i64 57810576 to i64*), align 16\n; │└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:189 within `getindex`\n    %lcnt96 = load volatile i64, i64* inttoptr (i64 129620856 to i64*), align 8\n    %81 = add i64 %lcnt96, 1\n    store volatile i64 %81, i64* inttoptr (i64 129620856 to i64*), align 8\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:190 within `getindex`\n    %lcnt97 = load volatile i64, i64* inttoptr (i64 129620864 to i64*), align 128\n    %82 = add i64 %lcnt97, 1\n    store volatile i64 %82, i64* inttoptr (i64 129620864 to i64*), align 128\n; ││┌ @ ntuple.jl:48 within `ntuple`\n; │││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:190 within `#22`\n      %lcnt98 = load volatile i64, i64* inttoptr (i64 129620864 to i64*), align 128\n      %83 = add i64 %lcnt98, 1\n      store volatile i64 %83, i64* inttoptr (i64 129620864 to i64*), align 128\n; ││└└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:184 within `aview`\n     %lcnt99 = load volatile i64, i64* inttoptr (i64 129620816 to i64*), align 16\n     %84 = add i64 %lcnt99, 1\n     store volatile i64 %84, i64* inttoptr (i64 129620816 to i64*), align 16\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:185 within `aview`\n     %lcnt100 = load volatile i64, i64* inttoptr (i64 129620824 to i64*), align 8\n     %85 = add i64 %lcnt100, 1\n     store volatile i64 %85, i64* inttoptr (i64 129620824 to i64*), align 8\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:186 within `aview`\n     %lcnt101 = load volatile i64, i64* inttoptr (i64 129620832 to i64*), align 32\n     %86 = add i64 %lcnt101, 1\n     store volatile i64 %86, i64* inttoptr (i64 129620832 to i64*), align 32\n; ││└\n; ││ @ array.jl:909 within `getindex`\n; ││┌ @ abstractarray.jl:699 within `checkbounds` @ abstractarray.jl:678\n; │││┌ @ abstractarray.jl:98 within `axes`\n; ││││┌ @ array.jl:194 within `size`\n       %.size118.sroa.0.0.copyload = load i64, i64* %.size118.sroa.0.0..sroa_idx, align 8\n       %.size118.sroa.2.0.copyload = load i64, i64* %21, align 8\n; │││└└\n; │││┌ @ abstractarray.jl:725 within `checkbounds_indices`\n; ││││┌ @ abstractarray.jl:758 within `checkindex`\n; │││││┌ @ int.jl:513 within `<`\n        %87 = icmp uge i64 %74, %.size118.sroa.0.0.copyload\n; ││││└└\n; ││││ @ abstractarray.jl:725 within `checkbounds_indices` @ abstractarray.jl:725\n; ││││┌ @ abstractarray.jl:758 within `checkindex`\n; │││││┌ @ int.jl:86 within `-`\n        %88 = add nsw i64 %value_phi83, -1\n; │││││└\n; │││││┌ @ int.jl:513 within `<`\n        %89 = icmp uge i64 %88, %.size118.sroa.2.0.copyload\n; │││└└└\n; │││ @ abstractarray.jl:699 within `checkbounds`\n     %90 = or i1 %87, %89\n     br i1 %90, label %L239, label %L242\n\nL239:                                             ; preds = %L155\n     %91 = getelementptr inbounds [2 x i64], [2 x i64]* %\"new::Tuple116\", i64 0, i64 1\n     %92 = getelementptr inbounds [2 x i64], [2 x i64]* %\"new::Tuple116\", i64 0, i64 0\n; │││ @ abstractarray.jl:697 within `checkbounds`\n     store i64 %50, i64* %92, align 8\n     store i64 %value_phi83, i64* %91, align 8\n; │││ @ abstractarray.jl:699 within `checkbounds`\n     call void @j_throw_boundserror_15410({}* nonnull %\"A::Array\", [2 x i64]* nocapture readonly %\"new::Tuple116\") #4\n     unreachable\n\nL242:                                             ; preds = %L155\n; ││└\n; ││ @ array.jl:910 within `getindex`\n; ││┌ @ abstractarray.jl:1349 within `_to_linear_index`\n; │││┌ @ abstractarray.jl:2979 within `_sub2ind` @ abstractarray.jl:2995\n; ││││┌ @ abstractarray.jl:3011 within `_sub2ind_recurse` @ abstractarray.jl:3011\n; │││││┌ @ int.jl:88 within `*`\n        %93 = mul i64 %.size118.sroa.0.0.copyload, %88\n; │││││└\n; │││││┌ @ int.jl:87 within `+`\n        %94 = add i64 %93, %50\n; ││└└└└\n; ││ @ array.jl:910 within `getindex` @ essentials.jl:790\n    %95 = add i64 %94, -1\n; ││┌ @ essentials.jl:12 within `length`\n     %96 = load i64*, i64** %24, align 8\n; ││└\n    %bitcast149 = load i64, i64* %96, align 8\n    %.not = icmp ult i64 %95, %bitcast149\n    br i1 %.not, label %L266, label %L263\n\nL263:                                             ; preds = %L242\n    %97 = getelementptr inbounds [1 x i64], [1 x i64]* %\"new::Tuple253\", i64 0, i64 0\n    store i64 %94, i64* %97, align 8\n    call void @j_throw_boundserror_15409({}* nonnull %\"A::Array\", [1 x i64]* nocapture readonly %\"new::Tuple253\") #4\n    unreachable\n\nL266:                                             ; preds = %L242\n; ││ @ array.jl:910 within `getindex` @ essentials.jl:791\n    %98 = load i8*, i8** %25, align 8\n    %99 = shl nuw nsw i64 %bitcast149, 1\n    %100 = add i64 %bitcast149, %95\n    %.not289 = icmp ult i64 %100, %99\n    %101 = bitcast i8* %98 to double*\n    %102 = getelementptr inbounds double, double* %101, i64 %95\n    %.data_ptr = getelementptr inbounds i64, i64* %96, i64 1\n    %103 = bitcast i64* %.data_ptr to i8**\n    %104 = load i8*, i8** %103, align 8\n    %105 = ptrtoint i8* %104 to i64\n    %106 = ptrtoint double* %102 to i64\n    %107 = sub i64 %106, %105\n    %108 = shl nuw nsw i64 %bitcast149, 3\n    %109 = icmp ult i64 %107, %108\n    %110 = and i1 %.not289, %109\n    br i1 %110, label %L287, label %oob\n\nL287:                                             ; preds = %L266\n    %111 = bitcast i8* %98 to {}**\n    %112 = getelementptr inbounds {}*, {}** %111, i64 %95\n    %113 = bitcast {}** %112 to double*\n    %114 = load double, double* %113, align 8\n; │└\n; │┌ @ float.jl:460 within `+`\n    %115 = fadd double %pointerref, %114\n; │└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:56 within `macro expansion`\n   %lcnt166 = load volatile i64, i64* inttoptr (i64 61927376 to i64*), align 16\n   %116 = add i64 %lcnt166, 1\n   store volatile i64 %116, i64* inttoptr (i64 61927376 to i64*), align 16\n; │┌ @ range.jl:902 within `iterate`\n; ││┌ @ promotion.jl:620 within `==`\n     %.not293 = icmp eq i64 %value_phi83, %71\n; ││└\n    br i1 %.not293, label %L361, label %L300\n\nL300:                                             ; preds = %L287\n; ││ @ range.jl:903 within `iterate`\n; ││┌ @ int.jl:87 within `+`\n     %lcnt247 = load volatile i64, i64* inttoptr (i64 61927376 to i64*), align 16\n     %117 = add i64 %lcnt247, 1\n     store volatile i64 %117, i64* inttoptr (i64 61927376 to i64*), align 16\n     %118 = add nuw i64 %value_phi83, 1\n     br label %L155\n\nL361:                                             ; preds = %L287, %L129.L361_crit_edge\n; │└└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:57 within `macro expansion`\n; │┌ @ array.jl:965 within `setindex!`\n; ││┌ @ int.jl:86 within `-`\n     %.pre-phi = phi i64 [ %.pre548, %L129.L361_crit_edge ], [ %74, %L287 ]\n; │└└\n; │┌ @ subarray.jl:321 within `getindex` @ /home/runner/.julia/packages/StaticArrays/cZ1ET/src/MArray.jl:25\n; ││┌ @ pointer.jl:152 within `unsafe_load`\n     %pointerref203 = phi double [ 0.000000e+00, %L129.L361_crit_edge ], [ %115, %L287 ]\n; │└└\n   %lcnt171 = load volatile i64, i64* inttoptr (i64 61927384 to i64*), align 8\n   %119 = add i64 %lcnt171, 1\n   store volatile i64 %119, i64* inttoptr (i64 61927384 to i64*), align 8\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:124 within `__index_Local_Linear`\n    %lcnt172 = load volatile i64, i64* inttoptr (i64 57810560 to i64*), align 128\n    %120 = add i64 %lcnt172, 1\n    store volatile i64 %120, i64* inttoptr (i64 57810560 to i64*), align 128\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:125 within `__index_Local_Linear`\n    %lcnt173 = load volatile i64, i64* inttoptr (i64 57810568 to i64*), align 8\n    %121 = add i64 %lcnt173, 1\n    store volatile i64 %121, i64* inttoptr (i64 57810568 to i64*), align 8\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/compiler.jl:23 within `__iterspace`\n     %lcnt174 = load volatile i64, i64* inttoptr (i64 57810792 to i64*), align 8\n     %122 = add i64 %lcnt174, 1\n     store volatile i64 %122, i64* inttoptr (i64 57810792 to i64*), align 8\n; ││└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/nditeration.jl:63 within `workitems`\n     %lcnt175 = load volatile i64, i64* inttoptr (i64 81001848 to i64*), align 8\n     %123 = add i64 %lcnt175, 1\n     store volatile i64 %123, i64* inttoptr (i64 81001848 to i64*), align 8\n; ││└\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:126 within `__index_Local_Linear`\n    %lcnt176 = load volatile i64, i64* inttoptr (i64 57810576 to i64*), align 16\n    %124 = add i64 %lcnt176, 1\n    store volatile i64 %124, i64* inttoptr (i64 57810576 to i64*), align 16\n; │└\n; │┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:189 within `getindex`\n    %lcnt182 = load volatile i64, i64* inttoptr (i64 129620856 to i64*), align 8\n    %125 = add i64 %lcnt182, 1\n    store volatile i64 %125, i64* inttoptr (i64 129620856 to i64*), align 8\n; ││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:190 within `getindex`\n    %lcnt183 = load volatile i64, i64* inttoptr (i64 129620864 to i64*), align 128\n    %126 = add i64 %lcnt183, 1\n    store volatile i64 %126, i64* inttoptr (i64 129620864 to i64*), align 128\n; ││┌ @ ntuple.jl:48 within `ntuple`\n; │││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:190 within `#22`\n      %lcnt184 = load volatile i64, i64* inttoptr (i64 129620864 to i64*), align 128\n      %127 = add i64 %lcnt184, 1\n      store volatile i64 %127, i64* inttoptr (i64 129620864 to i64*), align 128\n; ││└└\n; ││┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:184 within `aview`\n     %lcnt185 = load volatile i64, i64* inttoptr (i64 129620816 to i64*), align 16\n     %128 = add i64 %lcnt185, 1\n     store volatile i64 %128, i64* inttoptr (i64 129620816 to i64*), align 16\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:185 within `aview`\n     %lcnt186 = load volatile i64, i64* inttoptr (i64 129620824 to i64*), align 8\n     %129 = add i64 %lcnt186, 1\n     store volatile i64 %129, i64* inttoptr (i64 129620824 to i64*), align 8\n; │││ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/cpu.jl:186 within `aview`\n     %lcnt187 = load volatile i64, i64* inttoptr (i64 129620832 to i64*), align 32\n     %130 = add i64 %lcnt187, 1\n     store volatile i64 %130, i64* inttoptr (i64 129620832 to i64*), align 32\n; │└└\n; │┌ @ array.jl:965 within `setindex!`\n; ││┌ @ essentials.jl:11 within `length`\n     %131 = load i64, i64* %memcpy_refined_src, align 8\n; ││└\n; ││┌ @ int.jl:513 within `<`\n     %.not294 = icmp ult i64 %.pre-phi, %131\n; ││└\n    br i1 %.not294, label %L382, label %L379\n\nL379:                                             ; preds = %L361\n    %lcnt238 = load volatile i64, i64* inttoptr (i64 31813544 to i64*), align 8\n    %132 = add i64 %lcnt238, 1\n    store volatile i64 %132, i64* inttoptr (i64 31813544 to i64*), align 8\n    %lcnt239 = load volatile i64, i64* inttoptr (i64 61927384 to i64*), align 8\n    %133 = add i64 %lcnt239, 1\n    store volatile i64 %133, i64* inttoptr (i64 61927384 to i64*), align 8\n    %134 = getelementptr inbounds [1 x i64], [1 x i64]* %\"new::Tuple241\", i64 0, i64 0\n    store i64 %50, i64* %134, align 8\n    call void @j_throw_boundserror_15406({}* nonnull %\"out::Array\", [1 x i64]* nocapture readonly %\"new::Tuple241\") #4\n    unreachable\n\nL382:                                             ; preds = %L361\n; ││ @ array.jl:966 within `setindex!`\n; ││┌ @ Base.jl:47 within `getproperty`\n     %135 = load i8*, i8** %29, align 8\n     %136 = load {}*, {}** %30, align 8\n; ││└\n    %137 = bitcast {}* %136 to i64*\n    %138 = load i64, i64* %137, align 8\n    %139 = shl nuw nsw i64 %138, 1\n    %140 = add i64 %138, %.pre-phi\n    %.not295 = icmp ult i64 %140, %139\n    %141 = bitcast i8* %135 to double*\n    %142 = getelementptr inbounds double, double* %141, i64 %.pre-phi\n    %143 = bitcast {}* %136 to { i64, {}** }*\n    %.data_ptr216 = getelementptr inbounds { i64, {}** }, { i64, {}** }* %143, i64 0, i32 1\n    %144 = bitcast {}*** %.data_ptr216 to i8**\n    %145 = load i8*, i8** %144, align 8\n    %146 = ptrtoint i8* %145 to i64\n    %147 = ptrtoint double* %142 to i64\n    %148 = sub i64 %147, %146\n    %149 = shl nuw nsw i64 %138, 3\n    %150 = icmp ult i64 %148, %149\n    %151 = and i1 %.not295, %150\n    br i1 %151, label %idxend222, label %oob217\n\nL405:                                             ; preds = %load224\n    %frame.prev662 = load {}*, {}** %frame.prev, align 8\n    %152 = bitcast {}*** %tls_pgcstack to {}**\n    store {}* %frame.prev662, {}** %152, align 8\n; └└\n;  @ none within `cpu_reduce_private` @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/macros.jl:288\n  ret void\n\noob:                                              ; preds = %L266\n  %gc_slot_addr_0 = getelementptr inbounds [3 x {}*], [3 x {}*]* %gcframe663, i64 0, i64 2\n  %153 = bitcast {}** %gc_slot_addr_0 to i64**\n  store i64* %96, i64** %153, align 16\n;  @ none within `cpu_reduce_private` @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/macros.jl:287\n; ┌ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:55 within `macro expansion`\n; │┌ @ array.jl:910 within `getindex` @ essentials.jl:791\n    %ptls_field667 = getelementptr inbounds {}**, {}*** %tls_pgcstack, i64 2\n    %154 = bitcast {}*** %ptls_field667 to i8**\n    %ptls_load668669 = load i8*, i8** %154, align 8\n    %\"box::GenericMemoryRef\" = call noalias nonnull dereferenceable(32) {}* @ijl_gc_pool_alloc_instrumented(i8* %ptls_load668669, i32 800, i32 32, i64 139965152542256) #14\n    %155 = bitcast {}* %\"box::GenericMemoryRef\" to i64*\n    %\"box::GenericMemoryRef.tag_addr\" = getelementptr inbounds i64, i64* %155, i64 -1\n    store atomic i64 139965152542256, i64* %\"box::GenericMemoryRef.tag_addr\" unordered, align 8\n    %156 = bitcast {}* %\"box::GenericMemoryRef\" to { i8*, {}* }*\n    %.repack = bitcast {}* %\"box::GenericMemoryRef\" to i8**\n    store i8* %98, i8** %.repack, align 8\n    %.repack291 = getelementptr inbounds { i8*, {}* }, { i8*, {}* }* %156, i64 0, i32 1\n    %157 = bitcast {}** %.repack291 to i64**\n    store i64* %96, i64** %157, align 8\n    call void @ijl_bounds_error_int({}* %\"box::GenericMemoryRef\", i64 %94)\n    unreachable\n\noob217:                                           ; preds = %L382\n    %gc_slot_addr_0653 = getelementptr inbounds [3 x {}*], [3 x {}*]* %gcframe663, i64 0, i64 2\n    store {}* %136, {}** %gc_slot_addr_0653, align 16\n; │└\n; │ @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/test/private.jl:57 within `macro expansion`\n; │┌ @ array.jl:966 within `setindex!`\n    %ptls_field658664 = getelementptr inbounds {}**, {}*** %tls_pgcstack, i64 2\n    %158 = bitcast {}*** %ptls_field658664 to i8**\n    %ptls_load659665666 = load i8*, i8** %158, align 8\n    %\"box::GenericMemoryRef221\" = call noalias nonnull dereferenceable(32) {}* @ijl_gc_pool_alloc_instrumented(i8* %ptls_load659665666, i32 800, i32 32, i64 139965152542256) #14\n    %159 = bitcast {}* %\"box::GenericMemoryRef221\" to i64*\n    %\"box::GenericMemoryRef221.tag_addr\" = getelementptr inbounds i64, i64* %159, i64 -1\n    store atomic i64 139965152542256, i64* %\"box::GenericMemoryRef221.tag_addr\" unordered, align 8\n    %160 = bitcast {}* %\"box::GenericMemoryRef221\" to { i8*, {}* }*\n    %.repack297 = bitcast {}* %\"box::GenericMemoryRef221\" to i8**\n    store i8* %135, i8** %.repack297, align 8\n    %.repack298 = getelementptr inbounds { i8*, {}* }, { i8*, {}* }* %160, i64 0, i32 1\n    store {}* %136, {}** %.repack298, align 8\n    call void @ijl_bounds_error_int({}* %\"box::GenericMemoryRef221\", i64 %50)\n    unreachable\n\nidxend222:                                        ; preds = %L382\n    %161 = icmp eq i64 %138, 0\n    br i1 %161, label %oob223, label %load224\n\noob223:                                           ; preds = %idxend222\n    call void @ijl_bounds_error_int({}* %136, i64 1)\n    unreachable\n\nload224:                                          ; preds = %idxend222\n    %162 = bitcast i8* %135 to {}**\n    %163 = getelementptr inbounds {}*, {}** %162, i64 %.pre-phi\n    %164 = bitcast {}** %163 to double*\n    store double %pointerref203, double* %164, align 8\n; └└\n;  @ none within `cpu_reduce_private` @ /home/runner/work/KernelAbstractions.jl/KernelAbstractions.jl/src/macros.jl:288\n  %lcnt225 = load volatile i64, i64* inttoptr (i64 31813568 to i64*), align 64\n  %165 = add i64 %lcnt225, 1\n  store volatile i64 %165, i64* inttoptr (i64 31813568 to i64*), align 64\n; ┌ @ multidimensional.jl:408 within `iterate`\n; │┌ @ multidimensional.jl:426 within `__inc`\n; ││┌ @ int.jl:87 within `+`\n     %166 = add nuw nsw i64 %value_phi, 1\n; ││└\n; ││ @ multidimensional.jl:427 within `__inc`\n; ││┌ @ operators.jl:276 within `!=`\n; │││┌ @ promotion.jl:620 within `==`\n      %.not300 = icmp eq i64 %value_phi, 8\n; └└└└\n  br i1 %.not300, label %L405, label %L129\n}\n"))