[GitHub] [tvm] mbs-octoml commented on a change in pull request #8788: [Relay] Remove DeviceMap from LowerTE

[GitBox <gi...@apache.org>]

mbs-octoml commented on a change in pull request #8788:
URL: https://github.com/apache/tvm/pull/8788#discussion_r720592409



##########
File path: src/relay/backend/graph_executor_codegen.cc
##########
@@ -197,32 +200,30 @@ class GraphExecutorCodegen : public backend::MemoizedExprTranslator<std::vector<
   LoweredOutput Codegen(relay::Function func, String mod_name) {
     mod_name_ = mod_name;
 
-    // TODO(@jroesch): we need to split device planning and memory planning
-    // first we run device assignment, then we perform lowering, and then
-    // storage planning in ideal world.
-
-    memory_plan_ = GraphPlanMemory(func);
-
     // This first phase moves from implicit use of compile engine,
     // to instead explicitly lowering the incoming IRModule, and then
     // performing the preexisting graph executor code generation phase.
     IRModule mod = IRModule::FromExpr(func);
 
-    // Build a map from each operation to device.
-    tec::DeviceMap device_context_map;
-    for (const auto& it : memory_plan_->expr_to_storage_info) {
-      auto expr = it.first;
-      auto storage_info = it.second;
-      auto device_types = storage_info->device_types;
-      // CHECK_EQ(device_types.size(), 1);
-      tvm::Device dev;
-      dev.device_id = 0;
-      dev.device_type = device_types[0];
-      device_context_map.insert({expr, dev});
+    // Make explicit which device each expression should be executed on.

Review comment:
       Most of this device map hacking has gone away now.




-- 
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: commits-unsubscribe@tvm.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org