[tvm] branch v0.10 updated (79c48f3887 -> b051cad9f4)

[an...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

andrewzhaoluo pushed a change to branch v0.10
in repository https://gitbox.apache.org/repos/asf/tvm.git


    from 79c48f3887 [TIR][Bugfix] Correct handling of buffer argument when scheduling (#12816)
     add f9b692765a [BugFix][LLVM] Fix the bug that the generated systemlib cannot register ```__tvm_module_ctx``` symbol sometimes (#12817)
     add a75dcabd3f [MetaSchedule] PyDatabase Complete Function Reload Support (#12838)
     add e18b48bed8 [Fix] naming outputs of graph nodes by op_name:output_index (#12809)
     add 18909a4c13 [ONNX] Upgrade onnx and onnxruntime (#12729)
     add ecd003c742 [ci] Lint PR title/body for @ tags (#12840)
     add d9f7cf3539 Revert "[ci] Lint PR title/body for @ tags (#12840)" (#12848)
     add 5dfa8da00e [Hexagon] 2-Stage Pipeline; Lower Async TIR primitives to Hexagon User DMA (#12785)
     add 534378b935 [Containers] Add Array::Map (#12692)
     add 52dbf102cd Fix caffe, boost install in Python venvs by creating python3.X link (#12828)
     add fa5045bf69 [Metaschedule] MultiLevelTiling for wide vector architectures (#12845)
     add d4e3207cca [TIR] Enhance RemoveNoOp pass to remove negative loop (#12836)
     add b051cad9f4 [FIX,PROFILING] Fix gpu timer name and lookup (#12849)

No new revisions were added by this update.

Summary of changes:
 docker/install/ubuntu_install_boost.sh             |   3 +
 docker/install/ubuntu_install_caffe.sh             |   4 +-
 docker/install/ubuntu_install_onnx.sh              |   4 +-
 docker/install/ubuntu_install_python.sh            |   7 +
 include/tvm/meta_schedule/database.h               |  70 +++++++
 include/tvm/meta_schedule/schedule_rule.h          |  15 ++
 include/tvm/runtime/container/array.h              | 198 +++++++++++++++----
 include/tvm/tir/builtin.h                          |  10 +
 include/tvm/tir/transform.h                        |   5 +
 python/tvm/meta_schedule/database/database.py      |  81 ++++++++
 python/tvm/meta_schedule/schedule_rule/__init__.py |   1 +
 .../schedule_rule/multi_level_tiling.py            |  37 ++++
 src/driver/driver_api.cc                           |  12 +-
 src/ir/type_functor.cc                             |   9 +-
 src/meta_schedule/database/database.cc             |   6 +
 .../schedule_rule/multi_level_tiling.cc            |  35 ++--
 .../schedule_rule/multi_level_tiling.h             |   3 +
 .../multi_level_tiling_wide_vector.cc              | 120 ++++++++++++
 src/runtime/cuda/cuda_device_api.cc                |  18 +-
 src/runtime/graph_executor/graph_executor.cc       |   4 +-
 src/runtime/hexagon/hexagon_device_api.cc          |  25 +++
 src/runtime/profiling.cc                           |  13 ++
 src/target/llvm/codegen_cpu.cc                     |   2 +-
 src/te/operation/create_primfunc.cc                |   2 +-
 src/tir/analysis/device_constraint_utils.cc        |   5 +-
 src/tir/ir/buffer.cc                               |   4 +-
 src/tir/ir/expr.cc                                 |   3 +-
 src/tir/ir/expr_functor.cc                         |  14 +-
 src/tir/ir/functor_common.h                        |   3 +-
 src/tir/ir/index_map.cc                            |   5 +-
 src/tir/ir/specialize.cc                           |  19 +-
 src/tir/ir/stmt_functor.cc                         |   3 +-
 src/tir/op/builtin.cc                              |   6 +
 src/tir/schedule/primitive/decompose_padding.cc    |  15 +-
 src/tir/schedule/transform.cc                      |   8 +-
 src/tir/transforms/inject_virtual_thread.cc        |   4 +-
 src/tir/transforms/lower_async_dma.cc              | 194 +++++++++++++++++++
 src/tir/transforms/lower_match_buffer.cc           |   8 +-
 src/tir/transforms/lower_tvm_builtin.cc            |  30 +++
 src/tir/transforms/remove_no_op.cc                 |  11 ++
 src/tir/transforms/renew_defs.cc                   |  37 ++--
 src/tir/transforms/vectorize_loop.cc               |   6 +-
 tests/cpp/container_test.cc                        | 135 +++++++++++++
 .../test_hexagon/test_software_pipeline_async.py   |  86 +++++++++
 tests/python/frontend/onnx/test_forward.py         |  64 +++++++
 .../python/unittest/test_meta_schedule_database.py | 211 ++++++++++++++++++++-
 .../test_meta_schedule_schedule_rule_mlt.py        | 108 ++++++++++-
 .../test_tir_transform_inject_ptx_async_copy.py    |   4 +-
 .../test_tir_transform_inject_software_pipeline.py |   2 +-
 .../unittest/test_tir_transform_remove_no_op.py    |  16 +-
 50 files changed, 1523 insertions(+), 162 deletions(-)
 create mode 100644 src/meta_schedule/schedule_rule/multi_level_tiling_wide_vector.cc
 create mode 100644 src/tir/transforms/lower_async_dma.cc
 create mode 100644 tests/python/contrib/test_hexagon/test_software_pipeline_async.py