Skip to content

Elide bounds checks when kernels contains manual ones. #486

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
Jan 16, 2025
Merged

Elide bounds checks when kernels contains manual ones. #486

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
17 changes: 14 additions & 3 deletions src/device/array.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,6 +48,7 @@ Base.size(g::oneDeviceArray) = g.dims
Base.sizeof(x::oneDeviceArray) = Base.elsize(x) * length(x)

# we store the array length too; computing prod(size) is expensive
Base.size(g::oneDeviceArray{<:Any, 1}) = (g.len,)
Base.length(g::oneDeviceArray) = g.len

Base.pointer(x::oneDeviceArray{T,<:Any,A}) where {T,A} = Base.unsafe_convert(LLVMPtr{T,A}, x)
Expand Down Expand Up @@ -81,7 +82,11 @@ Base.unsafe_convert(::Type{LLVMPtr{T,A}}, x::oneDeviceArray{T,<:Any,A}) where {T
end

@device_function @inline function arrayref(A::oneDeviceArray{T}, index::Integer) where {T}
@boundscheck checkbounds(A, index)
# simplified bounds check to avoid the OneTo construction, which calls `max`
# and breaks elimination of redundant bounds checks in the generated code.
#@boundscheck checkbounds(A, index)
@boundscheck index <= length(A) || Base.throw_boundserror(A, index)

if isbitstype(T)
arrayref_bits(A, index)
else #if isbitsunion(T)
Expand Down Expand Up @@ -123,7 +128,10 @@ end
end

@device_function @inline function arrayset(A::oneDeviceArray{T}, x::T, index::Integer) where {T}
@boundscheck checkbounds(A, index)
# simplified bounds check (see `arrayref`)
#@boundscheck checkbounds(A, index)
@boundscheck index <= length(A) || Base.throw_boundserror(A, index)

if isbitstype(T)
arrayset_bits(A, x, index)
else #if isbitsunion(T)
Expand Down Expand Up @@ -154,7 +162,10 @@ end
end

@device_function @inline function const_arrayref(A::oneDeviceArray{T}, index::Integer) where {T}
@boundscheck checkbounds(A, index)
# simplified bounds check (see `arrayset`)
#@boundscheck checkbounds(A, index)
@boundscheck index <= length(A) || Base.throw_boundserror(A, index)

align = alignment(A)
unsafe_cached_load(pointer(A), index, Val(align))
end
Expand Down
Loading