perf: add fast-math and vectorization nvcc/cxx build flags

- centralize CXX_FLAGS/NVCC_FLAGS in csrc/build.py as single source
- add --use_fast_math, --ptxas-options=-O3,-v, --extra-device-vectorization
- add -march=native -funroll-loops host flags
- setup.py reads shared cxx_flags/nvcc_flags from registry
- sync pure-C test build commands with new flags

setup.py

@@ -41,7 +41,10 @@ if _should_build():
             CUDAExtension(
                 f"astrai.extension.{name}",
                 info["sources"],
-                extra_compile_args={"cxx": ["-O3"], "nvcc": info["nvcc_flags"]},
+                extra_compile_args={
+                    "cxx": info["cxx_flags"],
+                    "nvcc": info["nvcc_flags"],
+                },
                 extra_link_args=[f"-Wl,-rpath,{_torch_lib}"],
             )
         )