Skip to content

Fix segmentation fault in grad_fn #9292

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

Closed
vishwakftw wants to merge 11 commits into from
Closed

Fix segmentation fault in grad_fn #9292

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
4751a2c
Fix segmentation fault in grad_fn
vishwakftw Jul 9, 2018
5ddc3cb
Fix one more source of segmentation fault
vishwakftw Jul 10, 2018
7fa71cb
Revert "Fix one more source of segmentation fault"
vishwakftw Jul 10, 2018
315035f
Assert if correct number of inputs have been passed
vishwakftw Jul 10, 2018
9e5c36a
Fix bug in DelayedError
vishwakftw Jul 11, 2018
f26f344
Make operator() virtual for Function
vishwakftw Jul 11, 2018
2486b41
Move error check to THPCppFunction_call
vishwakftw Jul 11, 2018
3b81072
Exclude check for Error and DelayedError
vishwakftw Jul 12, 2018
d00cd68
Add number of inputs to constructor of DelayedError, modify Function …
vishwakftw Jul 12, 2018
de06ce6
Remove unnecessary code
vishwakftw Jul 12, 2018
33a56d7
Update basic_ops.h
vishwakftw Jul 13, 2018
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
8 changes: 4 additions & 4 deletions torch/autograd/function.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -204,13 +204,13 @@ def wrapper(ctx, *args):
if not requires_grad:
return outputs

err_fn = torch._C._functions.DelayedError(
b"trying to differentiate twice a function that was marked"
b"with @once_differentiable")

if not isinstance(outputs, tuple):
outputs = (outputs,)

err_fn = torch._C._functions.DelayedError(
b"trying to differentiate twice a function that was marked"
b"with @once_differentiable", len(outputs))

# Create aliases of each output that has requires_grad=True. We need
# at least one of the inputs to err_fn to require grad so that the
# output will have a grad_fn.
Expand Down
7 changes: 5 additions & 2 deletions torch/csrc/autograd/functions/basic_ops.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,8 +25,11 @@ struct Error : public Function {

// Identity in forward, Error in backward. Used to implement @once_differentiable
struct DelayedError : public Function {
DelayedError(std::string msg)
: msg(std::move(msg)) {};
DelayedError(std::string msg, int num_inputs)
: msg(std::move(msg)) {
for (int i = 0; i < num_inputs; i++)
add_input_metadata(Function::undefined_input());
}

virtual variable_list apply(const variable_list& inputs) override;

Expand Down
6 changes: 4 additions & 2 deletions torch/csrc/autograd/functions/init.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,11 +16,13 @@ using torch::TupleParser;
struct DelayedErrorCtor {
DelayedError* operator()(PyObject* args) {
std::string msg;
int num_inputs;

TupleParser parser(args, 1);
TupleParser parser(args, 2);
parser.parse(msg, "msg");
parser.parse(num_inputs, "num_inputs");

return new DelayedError(msg);
return new DelayedError(msg, num_inputs);
}
};

Expand Down
5 changes: 5 additions & 0 deletions torch/csrc/autograd/python_cpp_function.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,11 @@ PyObject* THPCppFunction_call(PyObject* self, PyObject* args, PyObject *kwargs)
}

int num_inputs = PyTuple_GET_SIZE(args);
int num_inputs_required = ((THPCppFunction*)self)->cdata->num_inputs();
if (num_inputs != num_inputs_required) {
return PyErr_Format(PyExc_TypeError, "expected %d arguments, got %d instead",
num_inputs_required, num_inputs);
}
variable_list vars(num_inputs);
for (int i = 0; i != num_inputs; ++i) {
PyObject* arg = PyTuple_GET_ITEM(args, i);
Expand Down