wrongtest commented on a change in pull request #4459: [RUNTIME] Implement
TVMDSOOp(TensorFlow custom op) for TVM runtime
URL: https://github.com/apache/incubator-tvm/pull/4459#discussion_r367282927
##########
File path: src/contrib/tf_op/index_seq.h
##########
@@ -0,0 +1,44 @@
+/**
+ * Refer to std::index_sequence (since c++14)
+ * Utilities to invoke variadic function with template <size_t N>
+ */
+#ifndef __TFTVM_INDEX_SEQ__
+#define __TFTVM_INDEX_SEQ__
+
+template <std::size_t ...>
Review comment:
> Do we need these helpers? I think we can probably directly invoke using
PackedFunc, right?
PackedFunc is invoked variadic but DLPack input tensors is array. Thus
direct implementation maybe like:
```c++
switch(NUM_INPUTS) {
case 1 :
f(args[0]); break;
case 2:
f(args[0], args[1]); break;
case 3:
f(args[0], args[1], args[2]); break;
// update to 8 ...
}
```
This seems quite strange so we use some template utility to call the
PackedFunc.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
