Skip to content

Don't launch a kernel to permute empty arrays #798

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 27, 2026
Merged

Don't launch a kernel to permute empty arrays #798

Show file tree
Hide file tree
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
2 changes: 2 additions & 0 deletions src/host/linalg.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1057,6 +1057,8 @@ function _permutedims!(::Type{IT}, dest::AbstractGPUArray,
src::AbstractGPUArray, perm::NTuple{N}) where {IT,N}
@assert length(src) <= typemax(IT)
Base.checkdims_perm(dest, src, perm)
# not every back-end supports launching an empty kernel (JuliaGPU/Metal.jl#980)
isempty(dest) && return dest
dest_strides = ntuple(k->k==1 ? 1 : prod(i->size(dest, i), 1:k-1), N)
dest_strides_perm = ntuple(i->IT(dest_strides[findfirst(==(i), perm)]), N)
size_src = IT.(size(src))
Expand Down
2 changes: 2 additions & 0 deletions test/testsuite/linalg.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,6 +26,8 @@
@test compare(x -> permutedims(x, (2, 1)), AT, rand(Float32, 2, 3))
@test compare(x -> permutedims(x, (2, 1, 3)), AT, rand(Float32, 4, 5, 6))
@test compare(x -> permutedims(x, (3, 1, 2)), AT, rand(Float32, 4, 5, 6))
@test compare(x -> permutedims(x, (2, 1)), AT, rand(Float32, 0, 3))
@test compare(x -> permutedims(x, (3, 1, 2)), AT, rand(Float32, 4, 0, 6))
if ComplexF32 in eltypes
@test compare(x -> permutedims(x, [2,1,4,3]), AT, randn(ComplexF32,3,4,5,1))
# test UInt64 version to make sure it works properly when array length is larger than typemax of UInt32.
Expand Down
Loading