You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2021/12/15 21:40:45 UTC

[GitHub] [tvm] mbs-octoml opened a new pull request #9749: [Relay] Fix invalid shape function for "copy" operator

mbs-octoml opened a new pull request #9749:
URL: https://github.com/apache/tvm/pull/9749


   The 'script' for of the shape function was ill-formed,
   resulting in a TIR shape function which did not assign
   to it's output, which in turn caused either OOM or
   assert fails as uninitialized dimensions worked their
   way downstream. That fix is in python/tvm/relay/op/tensor.py.
   
   Everything else is for testing and debugging as I tracked
   this down.
   
   [This is CORE-112 in OctoML JIRA.]


-- 
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



[GitHub] [tvm] mbs-octoml commented on a change in pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
mbs-octoml commented on a change in pull request #9749:
URL: https://github.com/apache/tvm/pull/9749#discussion_r770992333



##########
File path: src/target/compilation_config.cc
##########
@@ -61,47 +61,47 @@ void CompilationConfigNode::EstablishDefaultSEScopes(const transform::PassContex
   if (host_target.defined()) {
     CHECK(!host_target->host.defined()) << "Host targets are not expected to have hosts";
     host_device_type = static_cast<DLDeviceType>(host_target->kind->device_type);
-    DLOG(INFO) << "Using the given host target " << host_target->ToDebugString()
-               << " of device type " << host_device_type << " for the host target";
+    VLOG(1) << "Using the given host target " << host_target->ToDebugString() << " of device type "
+            << host_device_type << " for the host target";

Review comment:
       Yeah. In this case these were my own DLOGs and they were too verbose even for me.
   
   In general getting folks to think about debugability when writing code in the first place would be a fine thing. Eg I don't know if there's any way to see TE, or Schedules.




-- 
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



[GitHub] [tvm] electriclilies commented on a change in pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
electriclilies commented on a change in pull request #9749:
URL: https://github.com/apache/tvm/pull/9749#discussion_r770956692



##########
File path: include/tvm/runtime/debug.h
##########
@@ -0,0 +1,54 @@
+/*
+ * 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 tvm/runtime/debug.h
+ * \brief Helpers for debugging at runtime.
+ */
+#ifndef TVM_RUNTIME_DEBUG_H_
+#define TVM_RUNTIME_DEBUG_H_
+
+#include <tvm/runtime/container/adt.h>
+#include <tvm/runtime/ndarray.h>
+
+#include <ostream>
+#include <string>
+
+namespace tvm {
+namespace runtime {
+
+/*!
+ * \brief Helpers to describe runtime objects in human-friendly form. For \p nd_arrays we show their
+ * shapes and dtypes, but also their contents if 'small' and on the \p host_device (mostly so that
+ * we can see dynamic shapes as they are computed). For \p adts we show the ADT fields. For
+ * \p objects we dispatch to one of the above as appropriate.
+ */
+void AppendNDArray(std::ostream& os, const NDArray& nd_array, const DLDevice& host_device,
+                   bool show_content = true);
+void AppendADT(std::ostream& os, const ADT& adt, const DLDevice& host_device,
+               bool show_content = true);
+void AppendRuntimeObject(std::ostream& os, const ObjectRef& object, const DLDevice& host_device,
+                         bool show_content = true);

Review comment:
       Nice, this will make debugging shape functions easier!

##########
File path: src/target/compilation_config.cc
##########
@@ -61,47 +61,47 @@ void CompilationConfigNode::EstablishDefaultSEScopes(const transform::PassContex
   if (host_target.defined()) {
     CHECK(!host_target->host.defined()) << "Host targets are not expected to have hosts";
     host_device_type = static_cast<DLDeviceType>(host_target->kind->device_type);
-    DLOG(INFO) << "Using the given host target " << host_target->ToDebugString()
-               << " of device type " << host_device_type << " for the host target";
+    VLOG(1) << "Using the given host target " << host_target->ToDebugString() << " of device type "
+            << host_device_type << " for the host target";

Review comment:
       Have we discussed VLOG vs DLOG? It seems like people are still adding DLOGs. Might be worth discussing unifying around one debugging system at some point in the future.

##########
File path: tests/python/relay/dyn/test_dynamic_op_level3.py
##########
@@ -410,5 +418,75 @@ def verify_sparse_fill_empty_rows(
     )
 
 
+def test_dyn_copy():
+    target = tvm.target.Target("llvm")
+    dev = tvm.cpu()
+    mod = tvm.parser.fromtext(
+        """
+        #[version = "0.0.5"]
+        def @main(%x: Tensor[(?, 3), int64]) -> Tensor[(?, 3), int64] {
+          copy(%x)
+        }
+        """
+    )
+    x_data = np.random.rand(15, 3).astype("int64")
+    expected = x_data
+    check_on_vm(target, dev, [x_data], expected, mod)
+
+
+def test_dyn_copy_scalar():
+    target = tvm.target.Target("llvm")
+    dev = tvm.cpu()
+    mod = tvm.parser.fromtext(
+        """
+        #[version = "0.0.5"]
+        def @main(%x: int32, %y: Tensor[(?), int32]) -> Tensor[(?), int32] {
+          %0 = copy(%x);
+          %1 = expand_dims(%0, axis=0);
+          %2 = (%y, %1);
+          concatenate(%2)
+        }
+        """
+    )
+    x_data = 3
+    y_data = np.random.rand(7).astype("int32")
+    expected = np.concatenate((y_data, np.expand_dims(x_data, axis=0)))
+    check_on_vm(target, dev, [x_data, y_data], expected, mod)
+
+
+def test_dyn_expand_dims():

Review comment:
       There's another test_dyn_expand_dims in this file, possibly bad rebase?

##########
File path: src/relay/backend/te_compiler_cache.cc
##########
@@ -635,20 +640,16 @@ class MakeShapeFunc : public backend::MemoizedExprTranslator<Array<te::Tensor>>
     // Get output ndims
     auto ret_type = call_node->checked_type();
     Array<IndexExpr> out_ndims;
-    if (const auto* ttype = ret_type.as<TensorTypeNode>()) {
+    for (const auto& ttype : FlattenTupleType(ret_type)) {
       out_ndims.push_back(IntImm(DataType::Int(32), ttype->shape.size()));

Review comment:
       Is this change related to fixing the copy operator or just a general improvement?




-- 
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



[GitHub] [tvm] mbs-octoml commented on a change in pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
mbs-octoml commented on a change in pull request #9749:
URL: https://github.com/apache/tvm/pull/9749#discussion_r770992589



##########
File path: tests/python/relay/dyn/test_dynamic_op_level3.py
##########
@@ -410,5 +418,75 @@ def verify_sparse_fill_empty_rows(
     )
 
 
+def test_dyn_copy():
+    target = tvm.target.Target("llvm")
+    dev = tvm.cpu()
+    mod = tvm.parser.fromtext(
+        """
+        #[version = "0.0.5"]
+        def @main(%x: Tensor[(?, 3), int64]) -> Tensor[(?, 3), int64] {
+          copy(%x)
+        }
+        """
+    )
+    x_data = np.random.rand(15, 3).astype("int64")
+    expected = x_data
+    check_on_vm(target, dev, [x_data], expected, mod)
+
+
+def test_dyn_copy_scalar():
+    target = tvm.target.Target("llvm")
+    dev = tvm.cpu()
+    mod = tvm.parser.fromtext(
+        """
+        #[version = "0.0.5"]
+        def @main(%x: int32, %y: Tensor[(?), int32]) -> Tensor[(?), int32] {
+          %0 = copy(%x);
+          %1 = expand_dims(%0, axis=0);
+          %2 = (%y, %1);
+          concatenate(%2)
+        }
+        """
+    )
+    x_data = 3
+    y_data = np.random.rand(7).astype("int32")
+    expected = np.concatenate((y_data, np.expand_dims(x_data, axis=0)))
+    check_on_vm(target, dev, [x_data, y_data], expected, mod)
+
+
+def test_dyn_expand_dims():

Review comment:
       Oh, no just ignorance on my part -- no need to test it again! thx.




-- 
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



[GitHub] [tvm] mbs-octoml commented on a change in pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
mbs-octoml commented on a change in pull request #9749:
URL: https://github.com/apache/tvm/pull/9749#discussion_r770991209



##########
File path: include/tvm/runtime/debug.h
##########
@@ -0,0 +1,54 @@
+/*
+ * 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 tvm/runtime/debug.h
+ * \brief Helpers for debugging at runtime.
+ */
+#ifndef TVM_RUNTIME_DEBUG_H_
+#define TVM_RUNTIME_DEBUG_H_
+
+#include <tvm/runtime/container/adt.h>
+#include <tvm/runtime/ndarray.h>
+
+#include <ostream>
+#include <string>
+
+namespace tvm {
+namespace runtime {
+
+/*!
+ * \brief Helpers to describe runtime objects in human-friendly form. For \p nd_arrays we show their
+ * shapes and dtypes, but also their contents if 'small' and on the \p host_device (mostly so that
+ * we can see dynamic shapes as they are computed). For \p adts we show the ADT fields. For
+ * \p objects we dispatch to one of the above as appropriate.
+ */
+void AppendNDArray(std::ostream& os, const NDArray& nd_array, const DLDevice& host_device,
+                   bool show_content = true);
+void AppendADT(std::ostream& os, const ADT& adt, const DLDevice& host_device,
+               bool show_content = true);
+void AppendRuntimeObject(std::ostream& os, const ObjectRef& object, const DLDevice& host_device,
+                         bool show_content = true);

Review comment:
       Yeah, eventually I'd like to have spans in the VM instruction stream too.




-- 
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



[GitHub] [tvm] mbs-octoml commented on pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
mbs-octoml commented on pull request #9749:
URL: https://github.com/apache/tvm/pull/9749#issuecomment-996276674


   Thanks @electriclilies . Here goes another 6 hours of CI!


-- 
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



[GitHub] [tvm] mbs-octoml commented on a change in pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
mbs-octoml commented on a change in pull request #9749:
URL: https://github.com/apache/tvm/pull/9749#discussion_r770991464



##########
File path: src/relay/backend/te_compiler_cache.cc
##########
@@ -635,20 +640,16 @@ class MakeShapeFunc : public backend::MemoizedExprTranslator<Array<te::Tensor>>
     // Get output ndims
     auto ret_type = call_node->checked_type();
     Array<IndexExpr> out_ndims;
-    if (const auto* ttype = ret_type.as<TensorTypeNode>()) {
+    for (const auto& ttype : FlattenTupleType(ret_type)) {
       out_ndims.push_back(IntImm(DataType::Int(32), ttype->shape.size()));

Review comment:
       Sorry, unrelated. Cleaning up flattening as I go.




-- 
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



[GitHub] [tvm] mbs-octoml commented on pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
mbs-octoml commented on pull request #9749:
URL: https://github.com/apache/tvm/pull/9749#issuecomment-996869014


   Dang, stepped on my own toes in #9759. Here we go again.


-- 
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



[GitHub] [tvm] masahi merged pull request #9749: [Relay] Fix invalid shape function for "copy" operator

Posted by GitBox <gi...@apache.org>.
masahi merged pull request #9749:
URL: https://github.com/apache/tvm/pull/9749


   


-- 
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