Compilation issue with modern AMD + optimization enabled

[webmaster128]

Current master at fc81a55 does not compile (as discussed in https://community.amd.com/thread/234809)

1. Get a single-use Ubuntu 16.04 machine (no GPU required; 18.04 does not work due to kernel version compatibility of rocm-dkms)
2. Login as root

apt update && apt upgrade -y && apt autoremove -y && apt install -y htop libnuma-dev && reboot
wget -qO - http://repo.radeon.com/rocm/apt/debian/rocm.gpg.key | sudo apt-key add -
echo 'deb [arch=amd64] http://repo.radeon.com/rocm/apt/debian/ xenial main' | sudo tee /etc/apt/sources.list.d/rocm.list
apt update && apt install -y rocm-dkms
git clone https://github.com/PlasmaPower/nano-vanity.git && cd nano-vanity
./merge-kernel.py > kernel.cl

/opt/rocm/opencl/bin/x86_64/clang -include/opt/rocm/opencl/include/opencl-c.h -cl-std=CL1.2 -c -O0 kernel.cl
# good: creates kernel.o

/opt/rocm/opencl/bin/x86_64/clang -include/opt/rocm/opencl/include/opencl-c.h -cl-std=CL1.2 -c kernel.cl
# bad: does not terminate

Maybe there is some error which clang does not report. However, since CPU is at 100 % as long as the compiler runs it is more likely to be an infinite loop in the optimization process.

This may or may not be an issue in this project's code.

[webmaster128]

The kernel compiles fine using the CodeXL toolchain on Windows and was now reported as a ROCm compiler issue.