add conditional _ld_preload.py file extension for TVM EP

setup.py

@@ -145,6 +145,28 @@ def _rewrite_ld_preload_tensorrt(self, to_preload):
                     f.write('    import os\n')
                     f.write('    os.environ["ORT_TENSORRT_UNAVAILABLE"] = "1"\n')
 
+        def _rewrite_ld_preload_tvm(self):
+            with open('onnxruntime/capi/_ld_preload.py', 'a') as f:
+                f.write('import warnings\n\n')
+                f.write('try:\n')
+                f.write('   # This import is necessary in order to delegate the loading of libtvm.so to TVM.\n')
+                f.write('   import tvm\n')
+                f.write('except ImportError as e:\n')
+                f.write('   warnings.warn(\n')
+                f.write('       f"WARNING: Failed to import TVM, libtvm.so was not loaded. More details: {e}"\n')
+                f.write('   )\n\n')
+                f.write('try:\n')
+                f.write('   # Working between the C++ and Python parts in TVM EP is done using\n')
+                f.write('   # the PackedFunc and Registry classes. In order to use a Python function in C++ code,\n')
+                f.write('   # it must be registered in the global table of functions. Registration is carried out\n')
+                f.write('   # through JIT interface, so it is necessary to call special functions for registration.\n')
+                f.write('   # To do this, we need to make the following import.\n')
+                f.write('   import onnxruntime.providers.stvm\n')
+                f.write('except ImportError as e:\n')
+                f.write('   warnings.warn(\n')
+                f.write('       f"WARNING: Failed to register python functions for TVM EP. More details: {e}"\n')
+                f.write('   )\n')
+
         def run(self):
             if is_manylinux:
                 source = 'onnxruntime/capi/onnxruntime_pybind11_state.so'
@@ -207,6 +229,8 @@ def run(self):
                 self._rewrite_ld_preload(to_preload)
                 self._rewrite_ld_preload_cuda(to_preload_cuda)
                 self._rewrite_ld_preload_tensorrt(to_preload_tensorrt)
+            if package_name == 'onnxruntime-tvm':
+                self._rewrite_ld_preload_tvm()
             _bdist_wheel.run(self)
             if is_manylinux and not disable_auditwheel_repair:
                 file = glob(path.join(self.dist_dir, '*linux*.whl'))[0]