Skip to content

fix: prevent unsigned wraparound in erase_tensor_id when shrinking tensor_list #1447

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
wants to merge 1 commit into
base: tengine-lite
Choose a base branch
from
Open

fix: prevent unsigned wraparound in erase_tensor_id when shrinking tensor_list #1447

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
20 changes: 13 additions & 7 deletions tools/convert_tool/utils/graph_optimizer/graph_opt.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,12 +26,16 @@

static int erase_tensor_id(ir_graph_t* graph, int16_t id)
{
if (!graph || graph->tensor_num == 0 || id < 0 || id >= graph->tensor_num) {
return -1;
}

ir_tensor_t* tensor_del = get_ir_graph_tensor(graph, id);
std::map<int16_t, int16_t> old_new_id;
int16_t j = 0;
for (size_t i = 0; i < graph->tensor_num; i++)
{
if (i == id) continue;
if (i == (size_t)id) continue;

ir_tensor_t* tensor = get_ir_graph_tensor(graph, i);
tensor->index = j;
Expand All @@ -41,18 +45,20 @@ static int erase_tensor_id(ir_graph_t* graph, int16_t id)
for (size_t i = 0; i < graph->node_num; i++)
{
ir_node_t* node = get_ir_graph_node(graph, i);
for (size_t j = 0; j < node->input_num; j++)
for (size_t k = 0; k < node->input_num; k++)
{
node->input_tensors[j] = old_new_id[node->input_tensors[j]];
node->input_tensors[k] = old_new_id[node->input_tensors[k]];
}
for (size_t j = 0; j < node->output_num; j++)
for (size_t k = 0; k < node->output_num; k++)
{
node->output_tensors[j] = old_new_id[node->output_tensors[j]];
node->output_tensors[k] = old_new_id[node->output_tensors[k]];
}
}

ir_tensor_t** new_tensor_list = (ir_tensor_t**)sys_realloc(graph->tensor_list, sizeof(ir_tensor_t*) * (graph->tensor_num - 1));
graph->tensor_list = new_tensor_list;
size_t new_size = (size_t)(graph->tensor_num - 1) * sizeof(ir_tensor_t*);
ir_tensor_t** new_tensor_list =
(ir_tensor_t**)sys_realloc(graph->tensor_list, new_size);
if (new_tensor_list) graph->tensor_list = new_tensor_list;
graph->tensor_num--;

destroy_ir_tensor(graph, tensor_del);
Expand Down
Loading