areusch commented on a change in pull request #6917: URL: https://github.com/apache/incubator-tvm/pull/6917#discussion_r528405903
########## File path: src/target/llvm/codegen_params.cc ########## @@ -0,0 +1,151 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one + * or more contributor license agreements. See the NOTICE file + * distributed with this work for additional information + * regarding copyright ownership. The ASF licenses this file + * to you under the Apache License, Version 2.0 (the + * "License"); you may not use this file except in compliance + * with the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, + * software distributed under the License is distributed on an + * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY + * KIND, either express or implied. See the License for the + * specific language governing permissions and limitations + * under the License. + */ + +/*! + * \file codegen_params.cc + */ +#ifdef TVM_LLVM_VERSION + +#include "codegen_params.h" + +#include <memory> +#include <vector> + +namespace tvm { +namespace codegen { + +namespace { +class DLManagedTensorDeleter { + public: + void operator()(DLManagedTensor* ptr) { ptr->deleter(ptr); } +}; +} // namespace + +llvm::ConstantArray* NDArrayToLLVMArray(llvm::LLVMContext* ctx, ::tvm::runtime::NDArray arr) { + llvm::Type* element_type = nullptr; + Review comment: the thing is, because it will be used directly by instructions (I.e. load half-word), I prefer to let LLVM deal with alignment and whatever else is necessary to make that work. my understanding was we tried an experiment compiling large constants and the bulk of the time was spent in parsing, so this shouldn't actually load LLVM too much? also, we just do this a few times for the whole model ---------------------------------------------------------------- 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: [email protected]
