jroesch commented on a change in pull request #9326:
URL: https://github.com/apache/tvm/pull/9326#discussion_r746297164
##########
File path: src/relay/backend/te_compiler.h
##########
@@ -173,31 +173,28 @@ Map<Target, IRModule> GetPerTargetModules(IRModule mod);
* to TE expressions, schedules them, and then to TIR.
*
* \param module The IRModule.
- * \param targets The mapping for devices to targets.
* \param memory_plan The memory plan used during lowering
* \param module_name The name of this module
* \param process_fn Callback allowing one-level up code generators to process
* each function that we lower
* \return The lowered module, see above.
*/
IRModule LowerTE(
- const IRModule& module, TargetMap targets, backend::StaticMemoryPlan
memory_plan,
- const String& module_name, ProcessFn process_fn = [](Function f) {});
+ const IRModule& module, backend::StaticMemoryPlan memory_plan, const
String& module_name,
Review comment:
Do we have a tracking issue for pulling out the mangling from the
lowering? probably makes sense after your most recent draft change?
##########
File path: tests/python/unittest/test_micro_model_library_format.py
##########
@@ -411,4 +410,7 @@ def test_export_byoc_c_module():
if __name__ == "__main__":
- sys.exit(pytest.main([__file__] + sys.argv[1:]))
+ import sys
+
+ # sys.exit(pytest.main([__file__] + sys.argv[1:]))
+ test_export_operator_model_library_format()
Review comment:
We should probably fix this, feel free to do in follow up.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
