Skip to content

Fix misaligned addresses while reading tensor attributes from raw data buffers #27312

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
yuslepukhin merged 4 commits into from
Feb 15, 2026
+14 −35
Merged

Fix misaligned addresses while reading tensor attributes from raw data buffers #27312

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
5e90c46
Fix misaligned addresses while reading tensor attributes from raw dat…
n-v-k Feb 11, 2026
8bfaf0c
Code formatting
n-v-k Feb 11, 2026
12a2a75
Move from raw_data to utils::UnpackInitializerData in Tensor attribut…
n-v-k Feb 12, 2026
5be0175
Data type independent tensor attribute hashing and comparison
n-v-k Feb 13, 2026
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
49 changes: 14 additions & 35 deletions onnxruntime/core/optimizer/common_subexpression_elimination.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -172,29 +172,18 @@ bool AreRangesEqual(const Range& lhs, const Range& rhs) {
}

// Check if two tensor attributes are equal scalar tensors, mainly to support ConstantOfShape Op.
// Currently support float, float16 and int64 data types, and requires the data are raw data in TensorProto.
bool AreScalarTensorAttributeEqual(const ONNX_NAMESPACE::TensorProto& lhs_t, const ONNX_NAMESPACE::TensorProto& rhs_t) {
if (!(utils::HasDataType(lhs_t) && utils::HasDataType(rhs_t) && lhs_t.data_type() == rhs_t.data_type() &&
(lhs_t.data_type() == onnx::TensorProto_DataType_FLOAT ||
lhs_t.data_type() == onnx::TensorProto_DataType_FLOAT16 ||
lhs_t.data_type() == onnx::TensorProto_DataType_INT64) &&
lhs_t.dims_size() == 1 && rhs_t.dims_size() == 1 && lhs_t.dims()[0] == 1 && rhs_t.dims()[0] == 1 &&
utils::HasRawData(lhs_t) && utils::HasRawData(rhs_t))) {
lhs_t.data_type() != onnx::TensorProto_DataType_STRING &&
lhs_t.dims_size() == 1 && rhs_t.dims_size() == 1 && lhs_t.dims()[0] == 1 && rhs_t.dims()[0] == 1)) {
return false;
}
const void* lhs_value = lhs_t.raw_data().data();
const void* rhs_value = rhs_t.raw_data().data();
switch (lhs_t.data_type()) {
case onnx::TensorProto_DataType_FLOAT:
return *reinterpret_cast<const float*>(lhs_value) == *reinterpret_cast<const float*>(rhs_value);
case onnx::TensorProto_DataType_FLOAT16:
return *reinterpret_cast<const MLFloat16*>(lhs_value) == *reinterpret_cast<const MLFloat16*>(rhs_value);
case onnx::TensorProto_DataType_INT64:
return *reinterpret_cast<const int64_t*>(lhs_value) == *reinterpret_cast<const int64_t*>(rhs_value);
default:
break;
std::vector<uint8_t> unpacked_lhs_tensor, unpacked_rhs_tensor;
if (!utils::UnpackInitializerData(lhs_t, unpacked_lhs_tensor).IsOK() ||
!utils::UnpackInitializerData(rhs_t, unpacked_rhs_tensor).IsOK()) {
return false;
}
return false;
return unpacked_lhs_tensor == unpacked_rhs_tensor;
}

bool AreEqual(const ONNX_NAMESPACE::AttributeProto& lhs, const ONNX_NAMESPACE::AttributeProto& rhs) {
Expand Down Expand Up @@ -235,26 +224,16 @@ bool AreEqual(const ONNX_NAMESPACE::AttributeProto& lhs, const ONNX_NAMESPACE::A
return false;
}

// Support scalar float/int64/fp16 tensor attribute only for now, and requires data is raw data in TensorProto.
// Support scalar tensor attribute only for now.
std::size_t GetTensorAttributeHash(const ONNX_NAMESPACE::TensorProto& attr_t) {
std::size_t hash = 0;
if (utils::HasDataType(attr_t) && attr_t.dims_size() == 1 && attr_t.dims()[0] == 1 && utils::HasRawData(attr_t)) {
if (utils::HasDataType(attr_t) && attr_t.dims_size() == 1 && attr_t.dims()[0] == 1) {
int data_type = attr_t.data_type();
switch (data_type) {
case onnx::TensorProto_DataType_FLOAT:
UpdateHash(data_type, hash);
UpdateHash(*reinterpret_cast<const float*>(attr_t.raw_data().data()), hash);
break;
case onnx::TensorProto_DataType_FLOAT16:
UpdateHash(data_type, hash);
UpdateHash(static_cast<float>(*reinterpret_cast<const MLFloat16*>(attr_t.raw_data().data())), hash);
break;
case onnx::TensorProto_DataType_INT64:
UpdateHash(data_type, hash);
UpdateHash(*reinterpret_cast<const int64_t*>(attr_t.raw_data().data()), hash);
break;
default:
break;
ORT_ENFORCE(data_type != onnx::TensorProto_DataType_STRING, "Unexpected tensor string type");
std::vector<uint8_t> unpacked_tensor;
if (utils::UnpackInitializerData(attr_t, unpacked_tensor).IsOK()) {
UpdateHash(data_type, hash);
UpdateHashWithContainer(unpacked_tensor, hash);
}
}
return hash;
Expand Down
Loading