Skip to content

GH-54570: [C++][Compute] Fix segfault casting null list with empty child to fixed_size_list #54574

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

Open
ShivanshhhG wants to merge 1 commit into
base: main
Choose a base branch
from
+87 −14
Open

GH-54570: [C++][Compute] Fix segfault casting null list with empty child to fixed_size_list #54574

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
35 changes: 21 additions & 14 deletions cpp/src/arrow/compute/kernels/scalar_cast_nested.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -236,9 +236,9 @@ struct CastVarToFixedList {
RETURN_NOT_OK(builder.Reserve(in_array.length * list_size));
for (int64_t offset_i = 0; offset_i < in_array.length; ++offset_i) {
if (in_array.IsNull(offset_i)) {
// If element is null, just fill in the null slots with first value.
// If element is null, fill in the null slots with null values.
for (int64_t j = 0; j < list_size; ++j) {
builder.UnsafeAppend(0);
builder.UnsafeAppendNull();
}
} else {
int64_t value_i = offsets[offset_i];
Expand Down Expand Up @@ -275,18 +275,25 @@ struct CastVarToFixedList {
DCHECK(cast_values_datum.is_array());
std::shared_ptr<ArrayData> cast_values = cast_values_datum.array();

const auto* offsets = in_array.GetValues<src_offset_type>(1);
if (in_array.GetNullCount() > 0 && !nulls_have_correct_length) {
// We need to fill in the null slots, so we'll use Take on the values.
ARROW_ASSIGN_OR_RAISE(auto indices, BuildTakeIndices(in_array, list_size, ctx));
ARROW_ASSIGN_OR_RAISE(auto take_result,
Take(cast_values, Datum(indices),
TakeOptions::NoBoundsCheck(), ctx->exec_context()));
DCHECK(take_result.is_array());
cast_values = take_result.array();
} else if (offsets[0] != 0) {
// No nulls, but we need to slice the values
cast_values = cast_values->Slice(offsets[0], in_array.length * list_size);
if (cast_values->length == 0) {

@dmitry-chirkov-dremio dmitry-chirkov-dremio Oct 21, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The list-view inputs mentioned in the description still fail before reaching this branch. ValidateLengths reads offsets[length], but list views have only length offsets and a separate sizes buffer.

Please add regression tests for all-null ListView and LargeListView inputs with empty children. Either handle these layouts here or narrow the claimed scope and track the remaining bug separately.

ARROW_ASSIGN_OR_RAISE(
auto null_values,
MakeArrayOfNull(child_type, in_array.length * list_size, ctx->memory_pool()));
cast_values = null_values->data();
} else {
const auto* offsets = in_array.GetValues<src_offset_type>(1);
if (in_array.GetNullCount() > 0 && !nulls_have_correct_length) {
// We need to fill in the null slots, so we'll use Take on the values.
ARROW_ASSIGN_OR_RAISE(auto indices, BuildTakeIndices(in_array, list_size, ctx));
ARROW_ASSIGN_OR_RAISE(auto take_result,
Take(cast_values, Datum(indices),
TakeOptions::NoBoundsCheck(), ctx->exec_context()));
DCHECK(take_result.is_array());
cast_values = take_result.array();
} else if (offsets[0] != 0 || cast_values->length > in_array.length * list_size) {
// No nulls or nulls have correct length, but we need to slice the values
cast_values = cast_values->Slice(offsets[0], in_array.length * list_size);
}
}

out_array->child_data.emplace_back(std::move(cast_values));
Expand Down
66 changes: 66 additions & 0 deletions cpp/src/arrow/compute/kernels/scalar_cast_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3787,6 +3787,72 @@ TEST(Cast, ListToFSL) {
// Invalid fixed_size_list cast if inconsistent size
ASSERT_RAISES(Invalid, Cast(ArrayFromJSON(list(int32()), "[[0, 1, 2], null, [3, 4]]"),
CastOptions::Safe(fixed_size_list(int32(), 3))));

// GH-54570: Null list array with empty child array cast to FixedSizeList
{
auto child = ArrayFromJSON(int64(), "[]");
ASSERT_OK_AND_ASSIGN(
auto from,
ListArray::FromArrays(*ArrayFromJSON(int32(), "[0, 0]"), *child,

@dmitry-chirkov-dremio dmitry-chirkov-dremio Oct 21, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The C++ build has not run yet however compiling this locally reports an error because the third argument to FromArrays expects MemoryPool*, but receives an Array.

Did you compile/run this before submitting?

*ArrayFromJSON(boolean(), "[false]")));
auto to = ArrayFromJSON(fixed_size_list(int64(), 3), "[null]");
CheckCast(from, to);
}

// GH-54570: LargeList with empty child cast to FixedSizeList
{
auto child = ArrayFromJSON(int64(), "[]");
ASSERT_OK_AND_ASSIGN(
auto from,
LargeListArray::FromArrays(*ArrayFromJSON(int64(), "[0, 0]"), *child,
*ArrayFromJSON(boolean(), "[false]")));
auto to = ArrayFromJSON(fixed_size_list(int64(), 3), "[null]");
CheckCast(from, to);
}

// GH-54570: Multiple null lists with empty child
{
auto child = ArrayFromJSON(int64(), "[]");
ASSERT_OK_AND_ASSIGN(
auto from,
ListArray::FromArrays(*ArrayFromJSON(int32(), "[0, 0, 0]"), *child,
*ArrayFromJSON(boolean(), "[false, false]")));
auto to = ArrayFromJSON(fixed_size_list(int64(), 2), "[null, null]");
CheckCast(from, to);
}

// GH-54570: Null list with unallocated/null child data buffers
{
auto child_data = ArrayData::Make(int64(), 0, {nullptr, nullptr});
auto list_bitmap = ArrayFromJSON(boolean(), "[false]")->data()->buffers[1];
auto list_offsets = ArrayFromJSON(int32(), "[0, 0]")->data()->buffers[1];
auto list_data = ArrayData::Make(list(int64()), 1, {list_bitmap, list_offsets},
{child_data}, 1);
auto from = MakeArray(list_data);
auto to = ArrayFromJSON(fixed_size_list(int64(), 3), "[null]");
CheckCast(from, to);
}

// GH-54570: Null slots with length 0 alongside valid slots
{
auto child = ArrayFromJSON(int32(), "[10, 20]");
ASSERT_OK_AND_ASSIGN(
auto from,
ListArray::FromArrays(*ArrayFromJSON(int32(), "[0, 0, 2]"), *child,
*ArrayFromJSON(boolean(), "[false, true]")));
auto to = ArrayFromJSON(fixed_size_list(int32(), 2), "[null, [10, 20]]");
CheckCast(from, to);
}

{
auto child = ArrayFromJSON(int32(), "[10, 20]");
ASSERT_OK_AND_ASSIGN(
auto from,
ListArray::FromArrays(*ArrayFromJSON(int32(), "[0, 2, 2]"), *child,
*ArrayFromJSON(boolean(), "[true, false]")));
auto to = ArrayFromJSON(fixed_size_list(int32(), 2), "[[10, 20], null]");
CheckCast(from, to);
}
}

TEST(Cast, CastMap) {
Expand Down
Loading