Strip debug symbols from cuda-core Linux wheels

Add -Wl,--strip-all to extra_link_args for wheel builds on Linux,
matching the existing behavior in cuda_bindings/build_hooks.py.

Without stripping, the 0.7.0 Linux wheel is ~30 MB (103 MB extracted)
because every .so ships with debug_info. After stripping, extracted
size drops from 103 MB to ~11 MB, bringing the wheel in line with
the ~4-5 MB Windows wheels.

Closes #1881

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: leofang

cuda_core/build_hooks.py

@@ -91,7 +91,7 @@ def _determine_cuda_major_version() -> str:
 _extensions = None
 
 
-def _build_cuda_core():
+def _build_cuda_core(strip=False):
     # Customizing the build hooks is needed because we must defer cythonization until cuda-bindings,
     # now a required build-time dependency that's dynamically installed via the other hook below,
     # is installed. Otherwise, cimport any cuda.bindings modules would fail!
@@ -136,6 +136,9 @@ def get_sources(mod_name):
 
     all_include_dirs = [os.path.join(_get_cuda_path(), "include")]
     extra_compile_args = []
+    extra_link_args = []
+    if strip and sys.platform == "linux":
+        extra_link_args += ["-Wl,--strip-all"]
     if COMPILE_FOR_COVERAGE:
         # CYTHON_TRACE_NOGIL indicates to trace nogil functions.  It is not
         # related to free-threading builds.
@@ -152,6 +155,7 @@ def get_sources(mod_name):
             + all_include_dirs,
             language="c++",
             extra_compile_args=extra_compile_args,
+            extra_link_args=extra_link_args,
         )
         for mod in module_names()
     )
@@ -254,7 +258,7 @@ def _add_cython_include_paths_to_pth(wheel_path: str) -> None:
 
 
 def build_editable(wheel_directory, config_settings=None, metadata_directory=None):
-    _build_cuda_core()
+    _build_cuda_core(strip=False)
     wheel_name = _build_meta.build_editable(wheel_directory, config_settings, metadata_directory)
 
     # Patch the .pth file to add Cython include paths
@@ -265,7 +269,7 @@ def build_editable(wheel_directory, config_settings=None, metadata_directory=Non
 
 
 def build_wheel(wheel_directory, config_settings=None, metadata_directory=None):
-    _build_cuda_core()
+    _build_cuda_core(strip=True)
     return _build_meta.build_wheel(wheel_directory, config_settings, metadata_directory)