From 2c9031a0ec6a97e693e85be52fcf90722e71b72f Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Thu, 7 May 2026 19:30:32 -0500
Subject: [PATCH 01/22] allow for unary reductions with arbitrary dims

---
 lib/cunumeric_jl_wrapper/src/ndarray.cpp | 17 +++++++++++++
 src/ndarray/detail/ndarray.jl            |  8 ++++++
 src/ndarray/unary.jl                     |  9 ++++++-
 test/tests/unary_tests.jl                | 31 ++++++++++++++++++++++++
 4 files changed, 64 insertions(+), 1 deletion(-)

diff --git a/lib/cunumeric_jl_wrapper/src/ndarray.cpp b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
index 47cdf3b02..8b24f6e55 100644
--- a/lib/cunumeric_jl_wrapper/src/ndarray.cpp
+++ b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
@@ -237,6 +237,23 @@ void nda_unary_reduction(CN_NDArray* out, CuPyNumericUnaryRedCode op_code,
   out->obj.unary_reduction(op_code, input->obj);
 }
 
+CN_NDArray* nda_unary_reduction_axes(CuPyNumericUnaryRedCode op_code, CN_NDArray* input, const int32_t* axes, int32_t num_axes, bool keepdims) {
+    std::vector<int32_t> axis_vec(axes, axes + num_axes);
+    NDArray result = input->obj.perform_unary_reduction(
+        static_cast<int32_t>(op_code),
+        input->obj,
+        axis_vec,
+        std::nullopt,  // dtype
+        std::nullopt,  // res_dtype
+        std::nullopt,  // out
+        keepdims,
+        {},            // args
+        std::nullopt,  // initial
+        std::nullopt   // where
+    );
+    return new CN_NDArray{NDArray(std::move(result))};
+}
+
 NDArray get_slice(NDArray arr, std::vector<legate::Slice> slices) {
   switch (slices.size()) {
     case 1: {
diff --git a/src/ndarray/detail/ndarray.jl b/src/ndarray/detail/ndarray.jl
index b11c41015..57160b835 100644
--- a/src/ndarray/detail/ndarray.jl
+++ b/src/ndarray/detail/ndarray.jl
@@ -219,6 +219,14 @@ function nda_unary_reduction(out::NDArray, op_code::UnaryRedCode, input::NDArray
     return out
 end
 
+function nda_unary_reduction_axes(op_code::UnaryRedCode, input::NDArray{T,N}, axes::Vector{Int32}, keepdims::Bool) where {T,N}
+    axes_c = collect(Int32, axes)
+    ptr = ccall((:nda_unary_reduction_axes, libnda),
+        NDArray_t, (UnaryRedCode, NDArray_t, Ptr{Int32}, Int32, Cint),
+        op_code, input.ptr, axes_c, Int32(length(axes_c)), keepdims)
+    return NDArray(ptr)
+end
+
 function nda_array_equal(rhs1::NDArray{T,N}, rhs2::NDArray{T,N}) where {T,N}
     ptr = ccall((:nda_array_equal, libnda),
         NDArray_t, (NDArray_t, NDArray_t),
diff --git a/src/ndarray/unary.jl b/src/ndarray/unary.jl
index 6cc7b7491..0c0fbc765 100644
--- a/src/ndarray/unary.jl
+++ b/src/ndarray/unary.jl
@@ -248,9 +248,16 @@ for (base_func, op_code) in unary_reduction_map
         function $(Symbol(base_func))(input::NDArray{T}) where {T}
             T_OUT = Base.promote_op($base_func, Vector{T})
             is_wider_type(T_OUT, T) && assertpromotion($base_func, T, T_OUT)
-            out = cuNumeric.zeros(T_OUT) #0D result (not right if reducing along dims)
+            out = cuNumeric.zeros(T_OUT)
             return nda_unary_reduction(out, $(op_code), unchecked_promote_arr(input, T_OUT))
         end
+
+        function $(Symbol(base_func))(input::NDArray{T,N}; dims) where {T,N}
+            T_OUT = Base.promote_op($base_func, Vector{T})
+            is_wider_type(T_OUT, T) && assertpromotion($base_func, T, T_OUT)
+            axes = collect(Int32, (d - 1 for d in (dims isa Integer ? (dims,) : dims)))
+            return nda_unary_reduction_axes($(op_code), unchecked_promote_arr(input, T_OUT), axes, false)
+        end
     end
 end
 
diff --git a/test/tests/unary_tests.jl b/test/tests/unary_tests.jl
index 88f284ff6..1ad7b34f8 100644
--- a/test/tests/unary_tests.jl
+++ b/test/tests/unary_tests.jl
@@ -88,3 +88,34 @@ function test_unary_function_set(func_dict, T, N)
         test_unary_operation(func, julia_arr_2D, cunumeric_arr_2D, T)
     end
 end
+
+function test_unary_reduction_dims(func, julia_arr::AbstractArray{T,N}, cunumeric_arr::NDArray{T,N}) where {T,N}
+    for d in 1:N
+        julia_res = func(julia_arr, dims=d)
+        cunumeric_res = func(cunumeric_arr, dims=d)
+        allowscalar() do
+            @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
+        end
+    end
+
+    if N >= 2
+        julia_res = func(julia_arr, dims=(1,2))
+        cunumeric_res = func(cunumeric_arr, dims=(1,2))
+        allowscalar() do
+            @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
+        end
+    end
+end
+
+@testset "unary reductions with dims" begin
+    for T in (Float32, Float64, Int32, Int64)
+        julia_arr_1D, julia_arr_2D = make_julia_arrays(T, N, :unit_interval)
+        cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(
+            [julia_arr_1D], [julia_arr_2D], T, N
+        )
+        @testset "$func $T" for (func, _) in unary_reduction_map
+            test_unary_reduction_dims(func, julia_arr_1D, cunumeric_arr_1D)
+            test_unary_reduction_dims(func, julia_arr_2D, cunumeric_arr_2D)
+        end
+    end
+end

From 12fff072e61ee4ac3252024058620568959e7f18 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Tue, 12 May 2026 12:45:53 -0500
Subject: [PATCH 02/22] fix test bug

---
 test/tests/unary_tests.jl | 1 +
 1 file changed, 1 insertion(+)

diff --git a/test/tests/unary_tests.jl b/test/tests/unary_tests.jl
index 1ad7b34f8..f2cd8f68e 100644
--- a/test/tests/unary_tests.jl
+++ b/test/tests/unary_tests.jl
@@ -108,6 +108,7 @@ function test_unary_reduction_dims(func, julia_arr::AbstractArray{T,N}, cunumeri
 end
 
 @testset "unary reductions with dims" begin
+    N = 100
     for T in (Float32, Float64, Int32, Int64)
         julia_arr_1D, julia_arr_2D = make_julia_arrays(T, N, :unit_interval)
         cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(

From 33d8f60255e44c5f04a8663d69d4a514db109544 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Tue, 12 May 2026 12:53:04 -0500
Subject: [PATCH 03/22] more testing fixes

---
 test/tests/unary_tests.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/tests/unary_tests.jl b/test/tests/unary_tests.jl
index f2cd8f68e..3c562ee1f 100644
--- a/test/tests/unary_tests.jl
+++ b/test/tests/unary_tests.jl
@@ -114,7 +114,7 @@ end
         cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(
             [julia_arr_1D], [julia_arr_2D], T, N
         )
-        @testset "$func $T" for (func, _) in unary_reduction_map
+        @testset "$func $T" for (func, _) in cuNumeric.unary_reduction_map
             test_unary_reduction_dims(func, julia_arr_1D, cunumeric_arr_1D)
             test_unary_reduction_dims(func, julia_arr_2D, cunumeric_arr_2D)
         end

From 0a61ee5543f62b390efb19c26362280534b83644 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Tue, 12 May 2026 13:46:10 -0500
Subject: [PATCH 04/22] fix c++

---
 lib/cunumeric_jl_wrapper/src/ndarray.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/cunumeric_jl_wrapper/src/ndarray.cpp b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
index 8b24f6e55..ed157b077 100644
--- a/lib/cunumeric_jl_wrapper/src/ndarray.cpp
+++ b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
@@ -239,7 +239,7 @@ void nda_unary_reduction(CN_NDArray* out, CuPyNumericUnaryRedCode op_code,
 
 CN_NDArray* nda_unary_reduction_axes(CuPyNumericUnaryRedCode op_code, CN_NDArray* input, const int32_t* axes, int32_t num_axes, bool keepdims) {
     std::vector<int32_t> axis_vec(axes, axes + num_axes);
-    NDArray result = input->obj.perform_unary_reduction(
+    NDArray result = input->obj._perform_unary_reduction(
         static_cast<int32_t>(op_code),
         input->obj,
         axis_vec,

From 8381f19ca68c6f6d7429521cfd9cce978c6edef4 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Tue, 12 May 2026 14:03:44 -0500
Subject: [PATCH 05/22] keepdims

---
 src/ndarray/unary.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/ndarray/unary.jl b/src/ndarray/unary.jl
index 0c0fbc765..2bf2036d8 100644
--- a/src/ndarray/unary.jl
+++ b/src/ndarray/unary.jl
@@ -256,7 +256,7 @@ for (base_func, op_code) in unary_reduction_map
             T_OUT = Base.promote_op($base_func, Vector{T})
             is_wider_type(T_OUT, T) && assertpromotion($base_func, T, T_OUT)
             axes = collect(Int32, (d - 1 for d in (dims isa Integer ? (dims,) : dims)))
-            return nda_unary_reduction_axes($(op_code), unchecked_promote_arr(input, T_OUT), axes, false)
+            return nda_unary_reduction_axes($(op_code), unchecked_promote_arr(input, T_OUT), axes, true)
         end
     end
 end

From d7f7635c372357e6e745f0a56944602c75f20cae Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Tue, 12 May 2026 14:09:38 -0500
Subject: [PATCH 06/22] allow promotion on unit test

---
 test/tests/unary_tests.jl | 24 +++++++++++++-----------
 1 file changed, 13 insertions(+), 11 deletions(-)

diff --git a/test/tests/unary_tests.jl b/test/tests/unary_tests.jl
index 3c562ee1f..84392cbfe 100644
--- a/test/tests/unary_tests.jl
+++ b/test/tests/unary_tests.jl
@@ -90,19 +90,21 @@ function test_unary_function_set(func_dict, T, N)
 end
 
 function test_unary_reduction_dims(func, julia_arr::AbstractArray{T,N}, cunumeric_arr::NDArray{T,N}) where {T,N}
-    for d in 1:N
-        julia_res = func(julia_arr, dims=d)
-        cunumeric_res = func(cunumeric_arr, dims=d)
-        allowscalar() do
-            @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
+    allowpromotion(true) do
+        for d in 1:N
+            julia_res = func(julia_arr, dims=d)
+            cunumeric_res = func(cunumeric_arr, dims=d)
+            allowscalar() do
+                @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
+            end
         end
-    end
 
-    if N >= 2
-        julia_res = func(julia_arr, dims=(1,2))
-        cunumeric_res = func(cunumeric_arr, dims=(1,2))
-        allowscalar() do
-            @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
+        if N >= 2
+            julia_res = func(julia_arr, dims=(1,2))
+            cunumeric_res = func(cunumeric_arr, dims=(1,2))
+            allowscalar() do
+                @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
+            end
         end
     end
 end

From 9808d8c2d1e33fb3ef209244663e6da63ec83a9f Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Tue, 12 May 2026 17:58:56 -0500
Subject: [PATCH 07/22] support multi-axis reductions

---
 lib/cunumeric_jl_wrapper/src/ndarray.cpp | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/lib/cunumeric_jl_wrapper/src/ndarray.cpp b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
index ed157b077..c0520270b 100644
--- a/lib/cunumeric_jl_wrapper/src/ndarray.cpp
+++ b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
@@ -239,7 +239,15 @@ void nda_unary_reduction(CN_NDArray* out, CuPyNumericUnaryRedCode op_code,
 
 CN_NDArray* nda_unary_reduction_axes(CuPyNumericUnaryRedCode op_code, CN_NDArray* input, const int32_t* axes, int32_t num_axes, bool keepdims) {
     std::vector<int32_t> axis_vec(axes, axes + num_axes);
-    NDArray result = input->obj._perform_unary_reduction(
+    
+    auto input_shape = input->obj.shape();
+    std::vector<uint64_t> out_shape = input_shape;
+    for (int i = 0; i < num_axes; i++) {
+        out_shape[axes[i]] = 1;
+    }
+    
+    NDArray out = zeros(out_shape, input->obj.type());
+    out._perform_unary_reduction(
         static_cast<int32_t>(op_code),
         input->obj,
         axis_vec,
@@ -251,7 +259,7 @@ CN_NDArray* nda_unary_reduction_axes(CuPyNumericUnaryRedCode op_code, CN_NDArray
         std::nullopt,  // initial
         std::nullopt   // where
     );
-    return new CN_NDArray{NDArray(std::move(result))};
+    return new CN_NDArray{NDArray(std::move(out))};
 }
 
 NDArray get_slice(NDArray arr, std::vector<legate::Slice> slices) {

From 21891b9b6e786cb3c81bd6b46e53712b771c5164 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Tue, 12 May 2026 18:07:29 -0500
Subject: [PATCH 08/22] ok that didnt work

---
 lib/cunumeric_jl_wrapper/src/ndarray.cpp | 12 ++----------
 1 file changed, 2 insertions(+), 10 deletions(-)

diff --git a/lib/cunumeric_jl_wrapper/src/ndarray.cpp b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
index c0520270b..ed157b077 100644
--- a/lib/cunumeric_jl_wrapper/src/ndarray.cpp
+++ b/lib/cunumeric_jl_wrapper/src/ndarray.cpp
@@ -239,15 +239,7 @@ void nda_unary_reduction(CN_NDArray* out, CuPyNumericUnaryRedCode op_code,
 
 CN_NDArray* nda_unary_reduction_axes(CuPyNumericUnaryRedCode op_code, CN_NDArray* input, const int32_t* axes, int32_t num_axes, bool keepdims) {
     std::vector<int32_t> axis_vec(axes, axes + num_axes);
-    
-    auto input_shape = input->obj.shape();
-    std::vector<uint64_t> out_shape = input_shape;
-    for (int i = 0; i < num_axes; i++) {
-        out_shape[axes[i]] = 1;
-    }
-    
-    NDArray out = zeros(out_shape, input->obj.type());
-    out._perform_unary_reduction(
+    NDArray result = input->obj._perform_unary_reduction(
         static_cast<int32_t>(op_code),
         input->obj,
         axis_vec,
@@ -259,7 +251,7 @@ CN_NDArray* nda_unary_reduction_axes(CuPyNumericUnaryRedCode op_code, CN_NDArray
         std::nullopt,  // initial
         std::nullopt   // where
     );
-    return new CN_NDArray{NDArray(std::move(out))};
+    return new CN_NDArray{NDArray(std::move(result))};
 }
 
 NDArray get_slice(NDArray arr, std::vector<legate::Slice> slices) {

From 51d7a44b10742fa6090587c0060d57dece530dea Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Wed, 13 May 2026 11:03:56 -0500
Subject: [PATCH 09/22] refactor impl to add dims keyword

---
 src/ndarray/unary.jl | 47 ++++++++++++++++++++++++++++----------------
 1 file changed, 30 insertions(+), 17 deletions(-)

diff --git a/src/ndarray/unary.jl b/src/ndarray/unary.jl
index 2bf2036d8..b7244f522 100644
--- a/src/ndarray/unary.jl
+++ b/src/ndarray/unary.jl
@@ -242,33 +242,46 @@ global const unary_reduction_map = Dict{Function,UnaryRedCode}(
 
 #! IT WOULD BE NICE IF THESE JUST RETURNED SCALARS WHEN APPROPRIATE
 # #*TODO HOW TO GET THESE ACTING ON CERTAIN DIMS
+
+function _unary_reduction_impl(base_func, op_code, input::NDArray{T}, ::Colon) where {T}
+    T_OUT = Base.promote_op(base_func, Vector{T})
+    is_wider_type(T_OUT, T) && assertpromotion(base_func, T, T_OUT)
+    out = cuNumeric.zeros(T_OUT)
+    return nda_unary_reduction(out, op_code, unchecked_promote_arr(input, T_OUT))
+end
+
+function _unary_reduction_impl(base_func, op_code, input::NDArray{T,N}, dims) where {T,N}
+    T_OUT = Base.promote_op(base_func, Vector{T})
+    is_wider_type(T_OUT, T) && assertpromotion(base_func, T, T_OUT)
+    axes = collect(Int32, (d - 1 for d in (dims isa Integer ? (dims,) : dims)))
+    return nda_unary_reduction_axes(op_code, unchecked_promote_arr(input, T_OUT), axes, true)
+end
+
 # Generate code for all unary reductions.
 for (base_func, op_code) in unary_reduction_map
     @eval begin
-        function $(Symbol(base_func))(input::NDArray{T}) where {T}
-            T_OUT = Base.promote_op($base_func, Vector{T})
-            is_wider_type(T_OUT, T) && assertpromotion($base_func, T, T_OUT)
-            out = cuNumeric.zeros(T_OUT)
-            return nda_unary_reduction(out, $(op_code), unchecked_promote_arr(input, T_OUT))
-        end
-
-        function $(Symbol(base_func))(input::NDArray{T,N}; dims) where {T,N}
-            T_OUT = Base.promote_op($base_func, Vector{T})
-            is_wider_type(T_OUT, T) && assertpromotion($base_func, T, T_OUT)
-            axes = collect(Int32, (d - 1 for d in (dims isa Integer ? (dims,) : dims)))
-            return nda_unary_reduction_axes($(op_code), unchecked_promote_arr(input, T_OUT), axes, true)
+        function $(Symbol(base_func))(input::NDArray{T,N}; dims=Colon()) where {T,N}
+            return _unary_reduction_impl($base_func, $(op_code), input, dims)
         end
     end
 end
 
-function Base.all(input::NDArray{Bool})
+function _bool_reduction_impl(op_code, input::NDArray{Bool}, ::Colon)
     out = cuNumeric.zeros(Bool)
-    return nda_unary_reduction(out, cuNumeric.ALL, input)
+    return nda_unary_reduction(out, op_code, input)
 end
 
-function Base.any(input::NDArray{Bool})
-    out = cuNumeric.zeros(Bool)
-    return nda_unary_reduction(out, cuNumeric.ANY, input)
+function _bool_reduction_impl(op_code, input::NDArray{Bool}, dims)
+    axes = collect(Int32, (d - 1 for d in (dims isa Integer ? (dims,) : dims)))
+    return nda_unary_reduction_axes(op_code, input, axes, true)
+end
+
+function Base.all(input::NDArray{Bool}; dims=Colon())
+    return _bool_reduction_impl(cuNumeric.ALL, input, dims)
+end
+
+function Base.any(input::NDArray{Bool}; dims=Colon())
+    return _bool_reduction_impl(cuNumeric.ANY, input, dims)
 end
 
 #! ONLY ADD ONCE REDUCTIONS RETURN A SCALAR

From 82b04753a9b33c3dd5f7cd4653c97c0feaa57952 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Wed, 13 May 2026 13:34:22 -0500
Subject: [PATCH 10/22] add docs

---
 src/ndarray/unary.jl | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/src/ndarray/unary.jl b/src/ndarray/unary.jl
index b7244f522..922826734 100644
--- a/src/ndarray/unary.jl
+++ b/src/ndarray/unary.jl
@@ -209,9 +209,11 @@ The following unary reduction operations are supported and can be applied direct
   • `prod`
   • `sum`
 
-
 These operations follow standard Julia semantics.
 
+Reduction over specific dimensions is supported via the `dims` keyword argument,
+following the same semantics as Julia's base reduction functions.
+
 Examples
 --------
 
@@ -220,6 +222,14 @@ A = cuNumeric.ones(5)
 
 maximum(A)
 sum(A)
+
+# Reduce over a specific dimension
+B = cuNumeric.ones(3, 4)
+sum(B, dims=1)    # 1×4 result
+sum(B, dims=2)    # 3×1 result
+
+# Reduce over multiple dimensions
+sum(B, dims=(1,2))  # 1×1 result
 ```
 """
 global const unary_reduction_map = Dict{Function,UnaryRedCode}(

From 416450df9cfe7d8fa5616bc35ac5620d7c13b937 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Wed, 13 May 2026 16:22:19 -0500
Subject: [PATCH 11/22] add runtime check

---
 src/ndarray/unary.jl | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/src/ndarray/unary.jl b/src/ndarray/unary.jl
index 922826734..15db3a959 100644
--- a/src/ndarray/unary.jl
+++ b/src/ndarray/unary.jl
@@ -264,6 +264,9 @@ function _unary_reduction_impl(base_func, op_code, input::NDArray{T,N}, dims) wh
     T_OUT = Base.promote_op(base_func, Vector{T})
     is_wider_type(T_OUT, T) && assertpromotion(base_func, T, T_OUT)
     axes = collect(Int32, (d - 1 for d in (dims isa Integer ? (dims,) : dims)))
+    if length(axes) > 1
+        error("$(base_func): reducing over multiple dimensions is not yet supported. Got dims=$dims")
+    end
     return nda_unary_reduction_axes(op_code, unchecked_promote_arr(input, T_OUT), axes, true)
 end
 

From b727d215ae46040a3dc37c37297ba825f9931f86 Mon Sep 17 00:00:00 2001
From: Nader Rahhal <107228500+Nader-Rahhal@users.noreply.github.com>
Date: Wed, 13 May 2026 16:56:57 -0500
Subject: [PATCH 12/22] add dispatch

---
 src/ndarray/unary.jl | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/src/ndarray/unary.jl b/src/ndarray/unary.jl
index 15db3a959..96e69f981 100644
--- a/src/ndarray/unary.jl
+++ b/src/ndarray/unary.jl
@@ -260,13 +260,21 @@ function _unary_reduction_impl(base_func, op_code, input::NDArray{T}, ::Colon) w
     return nda_unary_reduction(out, op_code, unchecked_promote_arr(input, T_OUT))
 end
 
-function _unary_reduction_impl(base_func, op_code, input::NDArray{T,N}, dims) where {T,N}
+function _unary_reduction_impl(base_func, op_code, input::NDArray{T,N}, dims::Integer) where {T,N}
     T_OUT = Base.promote_op(base_func, Vector{T})
     is_wider_type(T_OUT, T) && assertpromotion(base_func, T, T_OUT)
-    axes = collect(Int32, (d - 1 for d in (dims isa Integer ? (dims,) : dims)))
-    if length(axes) > 1
+    axes = Int32[dims - 1]
+    return nda_unary_reduction_axes(op_code, unchecked_promote_arr(input, T_OUT), axes, true)
+end
+
+function _unary_reduction_impl(base_func, op_code, input::NDArray{T,N}, dims::Tuple) where {T,N}
+    if length(dims) > 1
         error("$(base_func): reducing over multiple dimensions is not yet supported. Got dims=$dims")
     end
+    # single element tuple
+    T_OUT = Base.promote_op(base_func, Vector{T})
+    is_wider_type(T_OUT, T) && assertpromotion(base_func, T, T_OUT)
+    axes = Int32[dims[1] - 1]
     return nda_unary_reduction_axes(op_code, unchecked_promote_arr(input, T_OUT), axes, true)
 end
 

From 53c0bdbede824e1b27b008b15d0cdbf6f635be3c Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 08:29:39 -0500
Subject: [PATCH 13/22] expected failures on multi-axis unary reductions

---
 test/runtests.jl          | 18 +++++++++++++++++-
 test/tests/unary_tests.jl | 32 +++++++++-----------------------
 2 files changed, 26 insertions(+), 24 deletions(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index d2396b3f8..31852281b 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -1,4 +1,4 @@
-#= Copyright 2026 Northwestern University, 
+#= Copyright 2026 Northwestern University,
  *                   Carnegie Mellon University University
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
@@ -167,6 +167,22 @@ end
     end
 end
 
+@testset "Unary Reductions with Dims" begin
+    N = 100
+    for T in (Float32, Float64, Int32, Int64)
+        @testset "T = $T" begin
+            julia_arr_1D, julia_arr_2D = make_julia_arrays(T, N, :unit_interval)
+            cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(
+                [julia_arr_1D], [julia_arr_2D], T, N
+            )
+            for (func, _) in cuNumeric.unary_reduction_map
+                test_unary_reduction_dims(func, julia_arr_1D, cunumeric_arr_1D)
+                test_unary_reduction_dims(func, julia_arr_2D, cunumeric_arr_2D)
+            end
+        end
+    end
+end
+
 @testset verbose = true "Binary Ops" begin
     N = 100
 
diff --git a/test/tests/unary_tests.jl b/test/tests/unary_tests.jl
index 84392cbfe..8c1e4f527 100644
--- a/test/tests/unary_tests.jl
+++ b/test/tests/unary_tests.jl
@@ -1,4 +1,4 @@
-#= Copyright 2026 Northwestern University, 
+#= Copyright 2026 Northwestern University,
  *                   Carnegie Mellon University University
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
@@ -89,36 +89,22 @@ function test_unary_function_set(func_dict, T, N)
     end
 end
 
-function test_unary_reduction_dims(func, julia_arr::AbstractArray{T,N}, cunumeric_arr::NDArray{T,N}) where {T,N}
+function test_unary_reduction_dims(
+    func, julia_arr::AbstractArray{T,N}, cunumeric_arr::NDArray{T,N}
+) where {T,N}
     allowpromotion(true) do
         for d in 1:N
-            julia_res = func(julia_arr, dims=d)
-            cunumeric_res = func(cunumeric_arr, dims=d)
+            julia_res = func(julia_arr; dims=d)
+            cunumeric_res = func(cunumeric_arr; dims=d)
             allowscalar() do
                 @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
             end
         end
 
+        # we are testing a multi axis reduction. This will throw a runtime error.
+        # https://github.com/nv-legate/cupynumeric/blob/main/src/cupynumeric/ndarray.cc#L1132
         if N >= 2
-            julia_res = func(julia_arr, dims=(1,2))
-            cunumeric_res = func(cunumeric_arr, dims=(1,2))
-            allowscalar() do
-                @test cuNumeric.compare(julia_res, cunumeric_res, atol(T), rtol(T))
-            end
-        end
-    end
-end
-
-@testset "unary reductions with dims" begin
-    N = 100
-    for T in (Float32, Float64, Int32, Int64)
-        julia_arr_1D, julia_arr_2D = make_julia_arrays(T, N, :unit_interval)
-        cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(
-            [julia_arr_1D], [julia_arr_2D], T, N
-        )
-        @testset "$func $T" for (func, _) in cuNumeric.unary_reduction_map
-            test_unary_reduction_dims(func, julia_arr_1D, cunumeric_arr_1D)
-            test_unary_reduction_dims(func, julia_arr_2D, cunumeric_arr_2D)
+            @test_throws Exception func(cunumeric_arr, dims=(1, 2))
         end
     end
 end

From 6ca427521be8d20af0e393050dc550d645f746b5 Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 08:51:09 -0500
Subject: [PATCH 14/22] add verbose=true

---
 .githash         | 2 +-
 test/runtests.jl | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/.githash b/.githash
index cee7a1140..0b20e71f0 100644
--- a/.githash
+++ b/.githash
@@ -1 +1 @@
-ff81537a0c8e23806869eef5c28c235b0dc3fbbe
+53c0bdbede824e1b27b008b15d0cdbf6f635be3c
diff --git a/test/runtests.jl b/test/runtests.jl
index 31852281b..01a9dfee4 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -167,7 +167,7 @@ end
     end
 end
 
-@testset "Unary Reductions with Dims" begin
+@testset verbose=true "Unary Reductions with Dims" begin
     N = 100
     for T in (Float32, Float64, Int32, Int64)
         @testset "T = $T" begin

From 840b61f97570142e010f8d85e96409ec1230cc18 Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 09:11:00 -0500
Subject: [PATCH 15/22] model after Unary Reductions

---
 test/runtests.jl | 25 +++++++++++++++++--------
 1 file changed, 17 insertions(+), 8 deletions(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index 01a9dfee4..657c6e879 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -169,16 +169,25 @@ end
 
 @testset verbose=true "Unary Reductions with Dims" begin
     N = 100
-    for T in (Float32, Float64, Int32, Int64)
-        @testset "T = $T" begin
-            julia_arr_1D, julia_arr_2D = make_julia_arrays(T, N, :unit_interval)
-            cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(
-                [julia_arr_1D], [julia_arr_2D], T, N
+
+    @testset for T in Base.uniontypes(cuNumeric.SUPPORTED_ARRAY_TYPES)
+        julia_arr_1D, julia_arr_2D = make_julia_arrays(T, N, :unit_interval)
+        cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(
+            [julia_arr_1D], [julia_arr_2D], T, N
+        )
+
+        @testset "$(func)" for (func, _) in cuNumeric.unary_reduction_map
+            # Skip reductions not supported by the cuNumeric backend for complex types
+            if T <: Complex && (
+                func == Base.maximum ||
+                func == Base.minimum ||
+                func == Base.prod
             )
-            for (func, _) in cuNumeric.unary_reduction_map
-                test_unary_reduction_dims(func, julia_arr_1D, cunumeric_arr_1D)
-                test_unary_reduction_dims(func, julia_arr_2D, cunumeric_arr_2D)
+                continue
             end
+
+            test_unary_reduction_dims(func, julia_arr_1D, cunumeric_arr_1D)
+            test_unary_reduction_dims(func, julia_arr_2D, cunumeric_arr_2D)
         end
     end
 end

From fd8603035a07b18e4b4c072501e9824e2b97ad7c Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 09:32:06 -0500
Subject: [PATCH 16/22] not sure why developer ci turned on jll mode in the
 last run

---
 .github/workflows/developer.yml | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/developer.yml b/.github/workflows/developer.yml
index 10c5094bd..1244c9847 100644
--- a/.github/workflows/developer.yml
+++ b/.github/workflows/developer.yml
@@ -89,14 +89,14 @@ jobs:
 
       - name: Setup cuNumeric.jl with build from src wrappers
         run: |
-          julia --color=yes -e '
+          julia --color=yes --project -e '
               using Pkg;
               Pkg.develop(PackageSpec(path = "lib/CNPreferences"))
               using CNPreferences; CNPreferences.use_developer_mode();
               Pkg.develop(PackageSpec(path = "."))
           '
-          julia --color=yes -e 'using Pkg; Pkg.build("cuNumeric")'
+          julia --color=yes --project -e 'using Pkg; Pkg.build()'
 
       - name: Perform Test
         run: |
-          GPUTESTS=0 julia --color=yes -e 'using Pkg; Pkg.test("cuNumeric")'
+          GPUTESTS=0 julia --color=yes --project -e 'using Pkg; Pkg.test()'

From f30a619171b96a0b72e354b616ac5f58911dfa01 Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 09:52:32 -0500
Subject: [PATCH 17/22] fix uuid

---
 .github/workflows/developer.yml | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/developer.yml b/.github/workflows/developer.yml
index 1244c9847..ea5d7acad 100644
--- a/.github/workflows/developer.yml
+++ b/.github/workflows/developer.yml
@@ -93,10 +93,14 @@ jobs:
               using Pkg;
               Pkg.develop(PackageSpec(path = "lib/CNPreferences"))
               using CNPreferences; CNPreferences.use_developer_mode();
+          '
+          julia --color=yes -e '
+              using Pkg;
+              Pkg.develop(PackageSpec(path = "lib/CNPreferences"))
               Pkg.develop(PackageSpec(path = "."))
+              Pkg.build("cuNumeric")
           '
-          julia --color=yes --project -e 'using Pkg; Pkg.build()'
 
       - name: Perform Test
         run: |
-          GPUTESTS=0 julia --color=yes --project -e 'using Pkg; Pkg.test()'
+          GPUTESTS=0 julia --color=yes -e 'using Pkg; Pkg.test("cuNumeric")'

From 787d502a50ef74bc346e89bae63ad276a94892c9 Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 10:08:48 -0500
Subject: [PATCH 18/22] try this

---
 .github/workflows/developer.yml | 9 +++++----
 test/Project.toml               | 1 +
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/.github/workflows/developer.yml b/.github/workflows/developer.yml
index ea5d7acad..f2bef3a89 100644
--- a/.github/workflows/developer.yml
+++ b/.github/workflows/developer.yml
@@ -89,17 +89,18 @@ jobs:
 
       - name: Setup cuNumeric.jl with build from src wrappers
         run: |
-          julia --color=yes --project -e '
+          julia --color=yes -e '
               using Pkg;
               Pkg.develop(PackageSpec(path = "lib/CNPreferences"))
               using CNPreferences; CNPreferences.use_developer_mode();
+              Pkg.develop(PackageSpec(path = "."))
           '
-          julia --color=yes -e '
+          julia --color=yes --project=test -e '
               using Pkg;
               Pkg.develop(PackageSpec(path = "lib/CNPreferences"))
-              Pkg.develop(PackageSpec(path = "."))
-              Pkg.build("cuNumeric")
+              using CNPreferences; CNPreferences.use_developer_mode();
           '
+          julia --color=yes -e 'using Pkg; Pkg.build("cuNumeric")'
 
       - name: Perform Test
         run: |
diff --git a/test/Project.toml b/test/Project.toml
index 7dd0521e1..8e41dd68b 100644
--- a/test/Project.toml
+++ b/test/Project.toml
@@ -1,4 +1,5 @@
 [deps]
+CNPreferences = "3e078157-ea10-49d5-bf32-908f777cd46f"
 CUDA = "052768ef-5323-5732-b1bb-66c8b64840ba"
 CUDA_Driver_jll = "4ee394cb-3365-5eb0-8335-949819d2adfc"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"

From 38a48c841ac20c0ea3a01fe4336176d6981eb43b Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 10:22:17 -0500
Subject: [PATCH 19/22] fix array init

---
 test/runtests.jl | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index 657c6e879..c33e794dc 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -171,10 +171,11 @@ end
     N = 100
 
     @testset for T in Base.uniontypes(cuNumeric.SUPPORTED_ARRAY_TYPES)
-        julia_arr_1D, julia_arr_2D = make_julia_arrays(T, N, :unit_interval)
-        cunumeric_arr_1D, cunumeric_arr_2D = make_cunumeric_arrays(
-            [julia_arr_1D], [julia_arr_2D], T, N
-        )
+        julia_arr_1D = my_rand(T, N)
+        julia_arr_2D = my_rand(T, (isqrt(N), isqrt(N)))
+
+        cunumeric_arr_1D = @allowscalar NDArray(julia_arr_1D)
+        cunumeric_arr_2D = @allowscalar NDArray(julia_arr_2D)
 
         @testset "$(func)" for (func, _) in cuNumeric.unary_reduction_map
             # Skip reductions not supported by the cuNumeric backend for complex types

From e4cc4b6a778aa828e5647803f594222053797f54 Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 10:41:46 -0500
Subject: [PATCH 20/22] try this?

---
 test/runtests.jl | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/test/runtests.jl b/test/runtests.jl
index c33e794dc..7cf26acfb 100644
--- a/test/runtests.jl
+++ b/test/runtests.jl
@@ -172,7 +172,7 @@ end
 
     @testset for T in Base.uniontypes(cuNumeric.SUPPORTED_ARRAY_TYPES)
         julia_arr_1D = my_rand(T, N)
-        julia_arr_2D = my_rand(T, (isqrt(N), isqrt(N)))
+        julia_arr_2D = my_rand(T, isqrt(N), isqrt(N))
 
         cunumeric_arr_1D = @allowscalar NDArray(julia_arr_1D)
         cunumeric_arr_2D = @allowscalar NDArray(julia_arr_2D)

From 3713a2acb6579891706b08f970e99f9ae5e6472f Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 11:08:12 -0500
Subject: [PATCH 21/22] skip CPU and GPU tests on PRs that aren't into main

---
 .buildkite/pipeline.yml         | 2 +-
 .github/workflows/ci.yml        | 2 ++
 .github/workflows/container.yml | 1 +
 3 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/.buildkite/pipeline.yml b/.buildkite/pipeline.yml
index f7bcfd4f8..e9abf561e 100644
--- a/.buildkite/pipeline.yml
+++ b/.buildkite/pipeline.yml
@@ -28,7 +28,7 @@ steps:
         agents:
           queue: "juliagpu"
           cuda: "*"
-        if: build.message !~ /\[skip tests\]/
+        if: build.message !~ /\[skip tests\]/ && (build.branch == "main" || build.pull_request.base_branch == "main")
         timeout_in_minutes: 44
         env:
           LD_LIBRARY_PATH: ""
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index 0daab8c40..cb77a387c 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -26,6 +26,8 @@ on:
     branches:
       - main
   pull_request:
+    branches:
+      - main
     paths:
       - 'src/**'
       - 'scripts/**'
diff --git a/.github/workflows/container.yml b/.github/workflows/container.yml
index 00d5d995e..07a68c927 100644
--- a/.github/workflows/container.yml
+++ b/.github/workflows/container.yml
@@ -18,6 +18,7 @@ on:
       - main
 jobs:
   push_to_registry:
+    if: ${{ github.event_name == 'workflow_dispatch' || github.event.workflow_run.conclusion == 'success' }}
     name: Container for ${{ matrix.platform }} - Julia ${{ matrix.julia }} - CUDA ${{ matrix.cuda }}
     permissions:
       contents: read

From 1a2d299743a74c10be82be97f5e1526f13595144 Mon Sep 17 00:00:00 2001
From: krasow <krasow@u.northwestern.edu>
Date: Thu, 14 May 2026 11:35:49 -0500
Subject: [PATCH 22/22] filtering ci.yml

---
 .githash                 |  2 +-
 .github/workflows/ci.yml | 22 ++++++++++++++++------
 2 files changed, 17 insertions(+), 7 deletions(-)

diff --git a/.githash b/.githash
index 6cdfeea5d..f2d1e6df2 100644
--- a/.githash
+++ b/.githash
@@ -1 +1 @@
-dec047f1bd1c8287513c6c437f946982e516ccd4
+3713a2acb6579891706b08f970e99f9ae5e6472f
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index cb77a387c..e979e8968 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -19,26 +19,36 @@ on:
       - 'deps/build.jl'
       - 'Project.toml'
       - 'lib/CNPreferences/src/**'
-      - '.github/workflows/ci.yml'
-      - 'Dockerfile' # container.yml depends on this
     tags:
       - 'v*'
     branches:
       - main
   pull_request:
-    branches:
-      - main
     paths:
       - 'src/**'
       - 'scripts/**'
       - 'deps/build.jl'
       - 'Project.toml'
       - 'lib/CNPreferences/src/**'
-      - '.github/workflows/ci.yml'
-      - 'Dockerfile' # container.yml depends on this
 jobs:
+  check_changes:
+    name: Check for wrapper changes
+    runs-on: ubuntu-latest
+    outputs:
+      wrapper_changed: ${{ steps.filter.outputs.wrapper }}
+    steps:
+      - uses: actions/checkout@v4
+      - uses: dorny/paths-filter@v3
+        id: filter
+        with:
+          filters: |
+            wrapper:
+              - 'lib/cunumeric_jl_wrapper/**'
+
   test:
     name: Julia ${{ matrix.julia }} - ${{ matrix.os }}
+    needs: check_changes
+    if: ${{ needs.check_changes.outputs.wrapper_changed != 'true' }}
     runs-on: ${{ matrix.os }}
     strategy:
       fail-fast: false