fix(ops/recv): use std::vector to store shape to support scalar

GitOrigin-RevId: e1dac3c9199539b03609838908626f15b70de555

imperative/src/impl/ops/io_remote.cpp

@@ -47,8 +47,13 @@ cg::OperatorNodeBase* apply_on_var_node_remote_recv(
     auto group_client = std::make_shared<opr::GroupClientProxy>(
             ssprintf("%s:%d", recv.addr.data(), recv.port));
     auto&& graph = inputs[0]->owner_graph();
+    mgb_assert(!recv.shape.empty());
+    TensorShape shape;
+    for (auto&& dim : recv.shape) {
+        shape[shape.ndim++] = dim;
+    }
     return graph->insert_opr(std::make_unique<mgb::opr::RemoteRecv>(
-            recv.key, inputs[0], *graph, group_client, config, recv.shape, recv.dtype,
+            recv.key, inputs[0], *graph, group_client, config, shape, recv.dtype,
             recv.backend));
 }
 

imperative/src/test/io_remote.cpp

@@ -42,7 +42,7 @@ TEST(TestImperative, IORemote) {
     auto run_recv = [&](std::shared_ptr<HostTensorND> hnd) {
         auto def = imperative::RemoteRecv::make(
                 "io_remote_test", server_addr, port, 0, CompNode::load("gpu1"),
-                TensorShape{vector_size}, dtype::Float32(), "nccl");
+                std::vector<int32_t>{(int32_t)vector_size}, dtype::Float32(), "nccl");
         auto inp = Tensor::make(*hnd);
         auto oup = OpDef::apply_on_physical_tensor(*def, {inp});
         HostTensorND host_v;

src/core/include/megbrain/ir/ops.td

@@ -284,7 +284,7 @@ def RemoteRecv : MgbHashableOp<"RemoteRecv"> {
     MgbUI32Attr:$port,
     MgbUI32Attr:$rank_from,
     MgbCompNodeAttr:$cn,
-    MgbTensorShapeAttr:$shape,
+    MgbArrayAttr<MgbI32Attr>:$shape,
     MgbDTypeAttr:$dtype,
     MgbStringAttr:$backend
   );