Revision bc7ba3d5c8b2dab1c0e19537739b67c2da902d11 authored by Keno Fischer on 20 March 2024, 06:35:46 UTC, committed by GitHub on 20 March 2024, 06:35:46 UTC
This passes slightly more information into this function (the full
`inst` rather than just the `stmt`) in order to allow external absint to
access additional fields (the flags and the info) if necessary to make
concrete evaluation decisions. It also splits out the actual concrete
evaluation from the part that just maps the `inst` to a CodeInstance.
1 parent e0bb95a
Raw File
simdloop.jl
# This file is a part of Julia. License is MIT: https://julialang.org/license

function simd_loop_example_from_manual(x, y, z)
    s = zero(eltype(z))
    n = min(length(x),length(y),length(z))
    @simd for i in 1:n
        @inbounds begin
            z[i] = x[i]-y[i]
            s += z[i]*z[i]
        end
    end
    s
end

function simd_loop_axpy!(a, X, Y)
    @simd ivdep for i in eachindex(X)
        @inbounds Y[i] += a*X[i]
    end
    return Y
end

function simd_loop_with_multiple_reductions(x, y, z)
    # Use non-zero initial value to make sure reduction values include it.
    (s,t) = (one(eltype(x)),one(eltype(y)))
    @simd for i in 1:length(z)
        @inbounds begin
            s += x[i]
            t += 2*y[i]
            s += z[i]   # Two reductions go into s
        end
    end
    (s,t)
end

for T in [Int32,Int64,Float32,Float64]
    # Try various lengths to make sure "remainder loop" works
    for n in [0,1,2,3,4,255,256,257]
        local n, a, b, c, s, t
        # Dataset chosen so that results will be exact with only 24 bits of mantissa
        a = convert(Array{T},[2*j+1 for j in 1:n])
        b = convert(Array{T},[3*j+2 for j in 1:n])
        c = convert(Array{T},[5*j+3 for j in 1:n])
        s = simd_loop_example_from_manual(a,b,c)

        @test a==[2*j+1 for j in 1:n]
        @test b==[3*j+2 for j in 1:n]
        @test c==[-j-1 for j in 1:n]
        @test s==sum(c.*c)
        (s,t) = simd_loop_with_multiple_reductions(a,b,c)
        @test s==sum(a)+sum(c)+1
        @test t==2*sum(b)+1

        X = ones(T, n)
        Y = zeros(T, n)
        simd_loop_axpy!(T(2), X, Y)
        @test all(y->y==T(2), Y)
    end
end

# Test that scope rules match regular for
let j=4, k=4
    # Use existing local variable.
    @simd for j=1:0 end
          for k=1:0 end
    @test j==k
    @simd for j=1:3 end
          for k=1:3 end
    @test j==k

    # Use global variable
    global simd_glob = 4
    global glob = 4
    @simd for simd_glob=1:0 end
          for      glob=1:0 end
    @test simd_glob==glob
    @simd for simd_glob=1:3 end
          for      glob=1:3 end
    @test simd_glob==glob

    # Index that is local to loop
    @simd for simd_loop_local=1:0 end
    simd_loop_local_present = true
    try
        simd_loop_local += 1
    catch
        simd_loop_local_present = false
    end
    @test !simd_loop_local_present
end

import Base.SimdLoop.SimdError

# Test that @simd rejects inner loop body with invalid control flow statements
# issue #8613
@test_throws SimdError("break is not allowed inside a @simd loop body") @macroexpand begin
    @simd for x = 1:10
        x == 1 && break
    end
end

@test_throws SimdError("continue is not allowed inside a @simd loop body") @macroexpand begin
    @simd for x = 1:10
        x < 5 && continue
    end
end

@test_throws SimdError("@goto is not allowed inside a @simd loop body") @macroexpand begin
    @simd for x = 1:10
        x == 1 || @goto exit_loop
    end
    @label exit_loop
end

# @simd with cartesian iteration
function simd_cartesian_range!(indices, crng)
    @simd for I in crng
        push!(indices, I)
    end
    indices
end

crng = CartesianIndices(map(Base.Slice, (2:4, 0:1, 1:1, 3:5)))
indices = simd_cartesian_range!(Vector{eltype(crng)}(), crng)
@test indices == vec(collect(crng))

crng = CartesianIndices(map(Base.Slice, (-1:1, 1:3)))
indices = simd_cartesian_range!(Vector{eltype(crng)}(), crng)
@test indices == vec(collect(crng))

crng = CartesianIndices(map(Base.Slice, (-1:-1, 1:3)))
indices = simd_cartesian_range!(Vector{eltype(crng)}(), crng)
@test indices == vec(collect(crng))

crng = CartesianIndices(map(Base.Slice, (2:4,)))
indices = simd_cartesian_range!(Vector{eltype(crng)}(), crng)
@test indices == collect(crng)

crng = CartesianIndices(())
indices = simd_cartesian_range!(Vector{eltype(crng)}(), crng)
@test indices == vec(collect(crng))

# @simd with array as "range"
# issue #13869
function simd_sum_over_array(a)
    s = zero(eltype(a))
    @inbounds @simd for x in a
        s += x
    end
    s
end
@test 2001000 == simd_sum_over_array(Vector(1:2000))
@test 2001000 == simd_sum_over_array(Float32[i+j*500 for i=1:500, j=0:3])

#Opt out of simd
struct iter31113{T}
    parent::T
end
Base.iterate(it::iter31113, args...) = iterate(it.parent, args...)
Base.eltype(it::iter31113) = eltype(it.parent)
Base.SimdLoop.simd_index(v::iter31113, j, i) = j
Base.SimdLoop.simd_inner_length(v::iter31113, j) = 1
Base.SimdLoop.simd_outer_range(v::iter31113) = v
@test 2001000 == simd_sum_over_array(iter31113(Vector(1:2000)))
back to top