backend/cuda: Use NVCC_FLAGS with user-provided NVCC

[raffenet]

Pull Request Description

These flags were ignored when the user specified a compiler other than the nvcc included in the CUDA installation. Make sure to include them for consistency. See pmodels/mpich#6954.

Expected Impact

Make application of NVCC_FLAGS consistent across configurations.

Author Checklist

• [x] Reference appropriate issues (with "Fixes" or "See" as appropriate)
• [x] Commits are self-contained and do not do two things at once
• [x] Commit message is of the form: module: short description and follows good practice
• [x] Add comments such that someone without knowledge of the code could understand
• [x] Have read and agree to the Yaksa CLA terms (https://github.com/pmodels/yaksa/wiki/Yaksa-Contributor-License-Agreement)