reject overflowing output dimensions in constant pad reshape

[aizu-m]

Each padded output dimension is computed as input_dim + pre_padding + post_padding in size_t with no overflow check (constant-pad-nd.c:188 and static-constant-pad.c:120), so a large padding value wraps the sum to a small dimension that slips past the multiply-only overflow guard in xnn_runtime_tensor_get_size, producing an undersized output tensor. Use xnn_safe_add and reject the reshape when the addition overflows.

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.