Upgrade stack in release/1.3 branch

[peri044]

Description

Please include a summary of the change and which issue is fixed. Please also include relevant motivation and context. List any dependencies that are required for this change.

Fixes # (issue)

Type of change

Please delete options that are not relevant and/or add your own.

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• This change requires a documentation update

Checklist:

• My code follows the style guidelines of this project (You can use the linters)
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas and hacks
• I have made corresponding changes to the documentation
• I have added tests to verify my fix or my feature
• New and existing unit tests pass locally with my changes
• I have added the relevant labels to my PR in so that relevant reviewers are notified

[github-actions]

Code conforms to C++ style guidelines

[github-actions]

Code conforms to Python style guidelines

[github-actions]

Code conforms to C++ style guidelines

[github-actions]

There are some changes that do not conform to Python style guidelines:

--- py/torch_tensorrt/fx/test/passes/test_fuse_permute_linear_trt.py	2022-11-29 06:52:42.337276 +0000
+++ py/torch_tensorrt/fx/test/passes/test_fuse_permute_linear_trt.py	2022-11-29 06:53:06.464407 +0000
@@ -49,10 +49,11 @@
            TestModule(10, 30),
            inputs,
            {acc_ops.permute, trt_transposed_linear},
            apply_passes=[fuse_permute_linear],
        )
+
    # TODO: The following test has been disabled due to a bug in TRT 8.5.1.7
    # with self.linear2. Issue : https://github.com/pytorch/TensorRT/issues/1444
    # def test_multi_fuse_permute_linear(self):
    #     """
    #     Fusion when permute output is shared by multiple linears

[github-actions]

Code conforms to C++ style guidelines

[github-actions]

There are some changes that do not conform to Python style guidelines:

--- py/torch_tensorrt/fx/test/passes/test_fuse_permute_linear_trt.py	2022-11-29 08:13:26.901017 +0000
+++ py/torch_tensorrt/fx/test/passes/test_fuse_permute_linear_trt.py	2022-11-29 08:13:50.867941 +0000
@@ -49,10 +49,11 @@
            TestModule(10, 30),
            inputs,
            {acc_ops.permute, trt_transposed_linear},
            apply_passes=[fuse_permute_linear],
        )
+
    # TODO: The following test has been disabled due to a bug in TRT 8.5.1.7
    # with self.linear2. Issue : https://github.com/pytorch/TensorRT/issues/1444
    # def test_multi_fuse_permute_linear(self):
    #     """
    #     Fusion when permute output is shared by multiple linears

[github-actions]

Code conforms to Python style guidelines

[github-actions]

Code conforms to C++ style guidelines

[narendasan]

Just a note for later, should we be keeping this object around?

[narendasan]

@peri044 why arent we using input_binding_names here. TensorRT does not guarantee binding order. This would be potentially unsafe.

I would expect that the workflow here would be to iterate not through inputs.size() but through the names in input_binding_names

[narendasan]

this should be std::string name = compiled_engine->input_binding_names[i]

[narendasan]

Same thing here, use output_binding_names to pull out the TensorRT binding index / address

[narendasan]

Other parts of the codebase have been using logging for this sort of role.

[yinghai]

Can we just use unittest skip?

[yinghai]

Please address the warning and unittest skip comment. Otherwise LGTM.

[github-actions]

Code conforms to Python style guidelines

[github-actions]

Code conforms to C++ style guidelines

[github-actions]

There are some changes that do not conform to C++ style guidelines:

diff --git a/home/runner/work/TensorRT/TensorRT/core/runtime/execute_engine.cpp b/tmp/changes.txt
index b799201..7fd4b8e 100644
--- a/home/runner/work/TensorRT/TensorRT/core/runtime/execute_engine.cpp
+++ b/tmp/changes.txt
@@ -128,8 +128,7 @@ std::vector<at::Tensor> execute_engine(std::vector<at::Tensor> inputs, c10::intr
          std::make_unique<torch::autograd::profiler::RecordProfile>(compiled_engine->input_profile_path);
    }
    for (size_t i = 0; i < inputs.size(); i++) {
-      std::string name =
-          compiled_engine->in_binding_names[i];
+      std::string name = compiled_engine->in_binding_names[i];
      TORCHTRT_CHECK(
          inputs[i].is_cuda(), "Expected input tensors to have device cuda, found device " << inputs[i].device());
      auto expected_type =
@@ -158,8 +157,7 @@ std::vector<at::Tensor> execute_engine(std::vector<at::Tensor> inputs, c10::intr

    for (size_t o = inputs.size(); o < (compiled_engine->num_io.first + compiled_engine->num_io.second); o++) {
      uint64_t pyt_idx = compiled_engine->out_binding_map[o];
-      std::string name =
-          compiled_engine->out_binding_names[pyt_idx];
+      std::string name = compiled_engine->out_binding_names[pyt_idx];
      auto out_shape = compiled_engine->exec_ctx->getTensorShape(name.c_str());
      LOG_DEBUG("Output Name: " << name << " Shape: " << out_shape);
      auto dims = core::util::toVec(out_shape);
ERROR: Some files do not conform to style guidelines

[github-actions]

Code conforms to Python style guidelines

[github-actions]

Code conforms to C++ style guidelines

[github-actions]

Code conforms to Python style guidelines

[yinghai]

lgtm

[narendasan]

LGTM, Thanks for this