Skip to content

Test branching reductions with loop-carried booleans #679

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
maleadt merged 1 commit into from
Oct 13, 2026
+22 −0
Merged

Test branching reductions with loop-carried booleans #679

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 22 additions & 0 deletions test/array.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -237,6 +237,28 @@ end
end
end

@testset "branching reduction with loop-carried booleans" begin
# Before libigc_jll 2.34.4+3, IGC promoted its own replacements, leaving undefined
# operands in loop-carried i1 values (JuliaGPU/oneAPI.jl#678). This 2-D reduction
# triggers the bug on Julia 1.12 with --check-bounds=yes.
A = rand(Int32(1):Int32(4), 64, 64)
dA = oneArray(A)
dB = oneArray(A)
init = (; is_missing=false, is_equal=true)
mapper(a, b) = (; is_missing=false, is_equal=a == b)
branching(a, b) = (a.is_missing || b.is_missing) ?
(; is_missing=true, is_equal=false) :
(; is_missing=false, is_equal=a.is_equal & b.is_equal)

@test mapreduce(mapper, branching, dA, dB; init) == init

B = copy(A)
B[end] += Int32(1)
copyto!(dB, B)
@test mapreduce(mapper, branching, dA, dB; init) ==
(; is_missing=false, is_equal=false)
end

@testset "narrow integer reduction with overflowing map" begin
# IGC < 2.34.4+2 kept the result of a 16-bit `mad` in the accumulator at full width, so
# the signed `max` compared the unwrapped value (JuliaGPU/oneAPI.jl#670, TGL and DG2).
Expand Down
Loading