Update on "[wip] quantization: store input_qrange_le_128 flag on quan…

…tized conv"


Summary:

This is a start of fixing the problems surfaced in #46749.
This particular PR only fixes a small part of this:
1. if a conv module is unsafe to run in fbgemm, we now persist this
information with a `input_qrange_le_128` boolean flag stored on `ConvPackedParams{n}d` set to False.
2. if we are in an fbgemm kernel and we detect that the current conv
packed params are tagged as unsafe, we throw an error.

For now, this PR is a WIP to get some early feedback if this is the
right direction, since iteration cost on this is high. In particular,
missing things here are:
* testing serialization of saving v3 and loading it back
* getting all the conv callsites (currently just module + conv2d is handled)

Note: there were some potential improvements discussed on dynamically
dispatching to qnnpack if it is available and the flag is set.  This PR
does not attempt to solve this issue - it can be solved by future PRs.

Test Plan:

```
# test that the error gets thrown when we are trying to run an operation which could
# saturate, and does not get thrown otherwise
python test/test_quantization.py TestQuantizedOps.test_conv_reduce_range

# test that loading older versions of conv packed params works as expected
# TODO(before land): extend these tests with the v3 files
python test/test_quantization.py TestSerialization
```

Reviewers:

Subscribers:

Tasks:

Tags:

Differential Revision: [D29175285](https://our.internmc.facebook.com/intern/diff/D29175285)

[ghstack-poisoned]

.circleci/docker/common/install_conda.sh

@@ -116,6 +116,7 @@ if [ -n "$ANACONDA_PYTHON_VERSION" ]; then
     boto3==1.16.34 \
     coverage==5.5 \
     hypothesis==4.53.2 \
+    expecttest==0.1.3 \
     mypy==0.812 \
     tb-nightly
 

.circleci/scripts/binary_populate_env.sh

@@ -62,7 +62,7 @@ if [[ -z "$DOCKER_IMAGE" ]]; then
   if [[ "$PACKAGE_TYPE" == conda ]]; then
     export DOCKER_IMAGE="pytorch/conda-cuda"
   elif [[ "$DESIRED_CUDA" == cpu ]]; then
-    export DOCKER_IMAGE="pytorch/manylinux-cuda100"
+    export DOCKER_IMAGE="pytorch/manylinux-cpu"
   else
     export DOCKER_IMAGE="pytorch/manylinux-cuda${DESIRED_CUDA:2}"
   fi

.github/scale-config.yml

@@ -27,6 +27,11 @@ runner_types:
     os: linux
     max_available: 50
     disk_size: 150
+  linux.16xlarge.nvidia.gpu:
+    instance_type: g3.16xlarge
+    os: linux
+    max_available: 10
+    disk_size: 150
   windows.4xlarge:
     instance_type: c5d.4xlarge
     os: windows

.github/scripts/generate_ci_workflows.py

@@ -137,15 +137,23 @@ def generate_workflow_file(
     #     docker_image_base=f"{DOCKER_REGISTRY}/pytorch/pytorch-linux-xenial-py3-clang7-onnx",
     #     test_runner_type=LINUX_CPU_TEST_RUNNER,
     # ),
+    PyTorchLinuxWorkflow(
+        build_environment="pytorch-linux-bionic-cuda10.2-cudnn7-py3.9-gcc7",
+        docker_image_base=f"{DOCKER_REGISTRY}/pytorch/pytorch-linux-bionic-cuda10.2-cudnn7-py3.9-gcc7",
+        test_runner_type=LINUX_CUDA_TEST_RUNNER,
+        num_test_shards=2,
+    ),
     PyTorchLinuxWorkflow(
         build_environment="pytorch-linux-xenial-cuda10.2-cudnn7-py3.6-gcc7",
         docker_image_base=f"{DOCKER_REGISTRY}/pytorch/pytorch-linux-xenial-cuda10.2-cudnn7-py3-gcc7",
         test_runner_type=LINUX_CUDA_TEST_RUNNER,
+        num_test_shards=2,
     ),
     PyTorchLinuxWorkflow(
         build_environment="pytorch-linux-xenial-cuda11.1-cudnn8-py3.6-gcc7",
         docker_image_base=f"{DOCKER_REGISTRY}/pytorch/pytorch-linux-xenial-cuda11.1-cudnn8-py3-gcc7",
         test_runner_type=LINUX_CUDA_TEST_RUNNER,
+        num_test_shards=2,
     ),
     # PyTorchLinuxWorkflow(
     #     build_environment="pytorch-libtorch-linux-xenial-cuda11.1-cudnn8-py3.6-gcc7",

.github/templates/linux_ci_workflow.yml.j2

@@ -461,7 +461,7 @@ jobs:
         env:
           PR_NUMBER: ${{ github.event.pull_request.number }}
         run: |
-          echo "See rendered docs at https://d28slxzaq48q8t.cloudfront.net/$PR_NUMBER/"
+          echo "See rendered docs at https://docs-preview.pytorch.org/$PR_NUMBER/"
       - name: Archive artifacts into zip
         run: |
           zip -r pytorch_github_io.zip "${GITHUB_WORKSPACE}/pytorch.github.io"

.github/workflows/lint.yml

@@ -103,6 +103,8 @@ jobs:
             echo 'Running setup.py with Python 2 did not give the expected error message.'
             false
           fi
+      - name: Keep torch.utils.collect_env python2 compliant
+        run: python2 -m py_compile torch/utils/collect_env.py
 
   shellcheck:
     runs-on: ubuntu-18.04
@@ -267,10 +269,10 @@ jobs:
 
   clang-tidy:
     if: github.event_name == 'pull_request'
-    runs-on: ubuntu-18.04
+    runs-on: linux.2xlarge
     container:
       # ubuntu18.04-cuda10.2-py3.6-tidy11
-      image: ghcr.io/pytorch/cilint-clang-tidy:7f0b4616100071a4813318bfdbd5b06ae36c5272
+      image: ghcr.io/pytorch/cilint-clang-tidy:b5a795a1165938adc6ccdab36bfa59bb3829ad47
     steps:
       - name: Checkout PyTorch
         uses: actions/checkout@v2
@@ -288,8 +290,6 @@ jobs:
         run: |
           cd "${GITHUB_WORKSPACE}"
           set -eux
-          git remote add upstream https://github.com/pytorch/pytorch
-          git fetch upstream "$GITHUB_BASE_REF"
 
           if [ ! -d build ]; then
             git submodule update --init --recursive
@@ -329,9 +329,11 @@ jobs:
           # deploy/interpreter files are excluded due to using macros and other techniquies
           # that are not easily converted to accepted c++
           python3 tools/linter/clang_tidy.py \
+            --parallel \
             --verbose \
             --paths torch/csrc/ \
             --diff-file pr.diff \
+            --include-dir /usr/lib/llvm-11/include/openmp \
             -g"-torch/csrc/jit/passes/onnx/helper.cpp" \
             -g"-torch/csrc/jit/passes/onnx/shape_type_inference.cpp" \
             -g"-torch/csrc/jit/serialization/onnx.cpp" \
@@ -401,7 +403,7 @@ jobs:
           set -eux
           pip install -r requirements.txt
           pip install numpy==1.20 # https://github.com/pytorch/pytorch/pull/60472
-          pip install mypy==0.812
+          pip install expecttest==0.1.3 mypy==0.812
           # Needed to check tools/render_junit.py
           pip install junitparser rich
       - name: Run autogen