Skip to content

Lowering refactor take 2 #72

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Jun 12, 2020
Merged

Lowering refactor take 2 #72

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e744a3c
Start commenting lowering better, break indexing pass into its own cl…
csarofeen Jun 9, 2020
1d6a5d9
Refactor lowering to break up passes and make logic more incremental.
csarofeen Jun 10, 2020
ac85a9f
removing commented code
jjsjann123 Jun 12, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions caffe2/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -439,8 +439,11 @@ if(NOT INTERN_BUILD_MOBILE OR NOT BUILD_CAFFE2_MOBILE)
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/manager.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/shape_inference.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/mutator.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/lower_index.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/lower_loops.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/lower_unroll.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/lower_utils.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/lower_validation.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/lower2device.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/parser.cpp
${TORCH_SRC_DIR}/csrc/jit/codegen/cuda/partition.cpp
Expand Down
4 changes: 0 additions & 4 deletions test/cpp/jit/test_gpu.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2564,10 +2564,6 @@ void testGPU_FusionReduction4() {
tv2->axis(-2)->parallelize(ParallelType::TIDy);
tv1->axis(-2)->parallelize(ParallelType::TIDy);

fusion.printMath();
GPULower lower(&fusion);
lower.printKernel(std::cout);

prog.device_ = 0;
prog.grid(1, bidy);
prog.block(tidx, tidy);
Expand Down
3 changes: 3 additions & 0 deletions tools/build_variables.bzl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -317,8 +317,11 @@ libtorch_cuda_sources = [
"torch/csrc/jit/codegen/cuda/iter_visitor.cpp",
"torch/csrc/jit/codegen/cuda/kernel.cpp",
"torch/csrc/jit/codegen/cuda/kernel_cache.cpp",
"torch/csrc/jit/codegen/cuda/lower_index.cpp",
"torch/csrc/jit/codegen/cuda/lower_loops.cpp",
"torch/csrc/jit/codegen/cuda/lower_unroll.cpp",
"torch/csrc/jit/codegen/cuda/lower_utils.cpp",
"torch/csrc/jit/codegen/cuda/lower_validation.cpp",
"torch/csrc/jit/codegen/cuda/lower2device.cpp",
"torch/csrc/jit/codegen/cuda/manager.cpp",
"torch/csrc/jit/codegen/cuda/shape_inference.cpp",
Expand Down
4 changes: 2 additions & 2 deletions torch/csrc/jit/codegen/cuda/ir_interface_nodes.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -275,10 +275,10 @@ struct TORCH_CUDA_API TensorView : public Val {
}

friend TORCH_CUDA_API TransformReplay;
// friend TORCH_CUDA_API TransformIter;
friend TORCH_CUDA_API OptOutMutator;
friend TORCH_CUDA_API GPULower;
friend TORCH_CUDA_API LoopNestGenerator;
friend void IRFixComputeAt(Fusion*);
friend void IRReplaceSizes();

protected:
// Make an exact copy of this tensor (similar to clone()), however, also grabs
Expand Down
Loading