This is an automated email from the ASF dual-hosted git repository.
leandron pushed a change to branch ci-docker-staging
in repository https://gitbox.apache.org/repos/asf/tvm.git.
omit 9bb026e cleanup
omit 39d2a03 Add @slow decorator to run tests on `main`
add 5956125 [ONNX] only broadcast matmul if the shape has changed (#10321)
add b5f1dab [TIR] Tir constants integration into compilation pipeline
(#8509)
add 27d9960 Simple workaround for PyTorch symbol crash problem in meta
schedule test (#10342)
add 97648d8 add reading of nRF5340 DK product ID to determine which COM
port to use (#10304)
add 6c6e873 [ARM_CPU] Conv2d int8 intrinsic for cortex-A72 (#10310)
add dcebd4d [CI][Hexagon] Add Hexagon Tests to pipeline (#10302)
add 270c1b3 [TIR] Misc minor updates (#10335)
add 0abb1bd [CUBLAS] Fix cublas batch matmul strategy plevel (#10351)
add 70cc2ab [CI] Re-introduce redirect follow and update hash for Boost
download (#10343)
add faa2e6a Add per channel quantization to QLinearConv and fix related
bugs (#10354)
add c338299 [CI] Fix Flaky Test `test_task_scheduler_gradient` (#10360)
add 8947729 [TOPI] VNNI support for batch matmul (#10332)
add d3d8715 [TIR] TIR Schedule Misc Update (#10341)
add d8f639a [AOT] BugFix of workspace calculation (#10337)
add 1cd42e8 [runtime] Improved log information with function signature
(#10326)
add cc82214 refactored GraphProto.from_onnx into smaller functions
(#10267)
add fbea569 [skip ci] Fix onnx frontend lint (#10363)
add c5baa59 [COMMUNITY] csullivan -> Committer (#10364)
add 0c210fc [BUGFIX][ARITH] Fix FloorMod Simplifier (#10336)
add b329544 [Lint] Fix Pylint Issues (#10358)
add d295ea6 [TIR][Transform] relax LoopPartition restriction that the
intersection of all conditions can not be none. (#10340)
add d110aa1 [ETHOSN] Improved identification of driver library version
(#10285)
add 25cc311 [ETHOSN] Stricter data type conversion checks (#10271)
add f1ff61a [microNPU][4] Add the cascader Proposal generator (#9959)
add 4102ebf [Runtime][Pipeline Executor] multiple threads management and
the data forwarding notification mechanism. (#10234)
add cb7f773 [Hexagon] RPC server/client for simulator (#10361)
add 7e2467a [TIR, Relay] improve bfloat16 support (#10112)
add 35df464 [ci] Check more events before pinging reviewers (#10208)
add dcbdedd Lower cache_read and cache_write to Hexagon DMA via tensorize
(#10365)
add d9fac4f [microNPU] adding more tests with USMP (#10362)
add 308d320a [RELAY] [VIRTUALDEVICE] Change syntax for device planning and
store parameter virtual devices in virtual_device_ field (#10352)
add d62a364 [VirtualMachine] new method allowing to set one input tensor
by its index or name (#10293)
add 07ad977 [Hexagon] Replace strlen in constant initialization with
sizeof (#10381)
add 221329f check to avoid crash in opt_level=0 vm build (#10347)
add 329d8a5 [DOCS] Add how to contribute TVM docs with images. (#10287)
add b31a5cc [MetaSchedule] Update Tuning Interfaces. (#10367)
add 920654c [Bugfix][TVMScript] Convert BufferSlice to BufferLoad when
used as range/loop start and end (#10370)
add 8b30198 [FIX,PROFILING] Add extra precision to numbers when
serializing to json (#10392)
add 37f7e57 Fix plint error. (#10394)
add 4a9db23 meta schedule misc update (#10389)
add 01f306f Fix tvmc run error message when inputs aren't found. (#10017)
add 784618e [Runtime][PipelineExecutor] Polish the name and comments of
variable. (#10395)
add 40f881b Enable groups argument for conv2d_transpose on the cudnn
backend (#10396)
add b7caa12 Fixed a bug in the convert_fully_connected() function (#10371)
add 9ca2139 [TensorIR] Renormalize split pattern (#10401)
add 7127296 [MetaSchedule] Arithmetic analysis (#10403)
add 5b5d969 Add @slow decorator to run tests on `main` (#10057)
add b56770b [microTVM] Zephyr: refactor _find_openocd_serial_port (#10346)
add b55997d [microTVM][RVM] Skip USB device attach if device is already
attached (#8737)
add b72e115 Realize the function op during forward rewrite (#10410)
add 92a80e9 [ci][1/2] Shard `frontend: GPU` job into 2 jobs (#10413)
add 111b2da RelayViz Graphviz renderer (#10400)
add 5e353d5 [Runtime][ThreadPool]Refactor affinity function and support
CPU affinity list setting. (#9802)
add d101c50 [CI][1/2] Update the Python version of pyxir (#10406)
add bd14a4d Modify debug output (#10372)
new 977bdd3 Upgrade tensorflow to version to 2.6.x
This update added new revisions after undoing existing revisions.
That is to say, some revisions that were in the old version of the
branch are not in the new version. This situation occurs
when a user --force pushes a change and generates a repository
containing something like this:
* -- * -- B -- O -- O -- O (9bb026e)
\
N -- N -- N refs/heads/ci-docker-staging (977bdd3)
You should already have received notification emails for all of the O
revisions, and so the following emails describe only the N revisions
from the common base, B.
Any revisions marked "omit" are not gone; other references still
refer to them. Any revisions marked "discard" are gone forever.
The 1 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails. The revisions
listed as "add" were already present in the repository and have only
been added to this reference.
Summary of changes:
CONTRIBUTORS.md | 1 +
Jenkinsfile | 29 +
apps/hexagon_api/CMakeLists.txt | 29 +-
.../template_project/src/example_project/model.c | 2 +-
apps/microtvm/reference-vm/base-box-tool.py | 29 +-
apps/microtvm/zephyr/template_project/boards.json | 4 +-
.../zephyr/template_project/microtvm_api_server.py | 90 +-
apps/microtvm/zephyr_cmsisnn/src/main.c | 2 +-
cmake/modules/Hexagon.cmake | 34 +-
cmake/utils/FindEthosN.cmake | 11 +-
docker/install/ubuntu1804_install_python.sh | 4 +-
docker/install/ubuntu_install_boost.sh | 4 +-
docker/install/ubuntu_install_tensorflow.sh | 6 +-
.../install/ubuntu_install_vitis_ai_packages_ci.sh | 2 +-
docs/contribute/document.rst | 18 +
docs/reference/api/python/contrib.rst | 4 +-
gallery/how_to/work_with_relay/using_relay_viz.py | 2 +
include/tvm/ir/function.h | 10 -
include/tvm/ir/module.h | 42 +
include/tvm/ir/transform.h | 59 +-
include/tvm/meta_schedule/builder.h | 4 +-
include/tvm/meta_schedule/tune_context.h | 2 +
include/tvm/node/structural_hash.h | 9 +
include/tvm/relay/executor.h | 7 +-
include/tvm/relay/interpreter.h | 4 +-
include/tvm/relay/runtime.h | 2 +
include/tvm/relay/transform.h | 17 +
include/tvm/runtime/module.h | 2 -
include/tvm/runtime/packed_func.h | 238 +++-
include/tvm/runtime/registry.h | 5 +-
include/tvm/runtime/threading_backend.h | 48 +-
include/tvm/runtime/vm/executable.h | 7 +
include/tvm/runtime/vm/vm.h | 55 +-
include/tvm/target/virtual_device.h | 2 +-
include/tvm/tir/builtin.h | 7 +
include/tvm/tir/function.h | 38 +-
include/tvm/tir/op.h | 16 +-
include/tvm/tir/stmt.h | 98 +-
include/tvm/tir/stmt_functor.h | 4 +
include/tvm/tir/transform.h | 16 +
include/tvm/tir/var.h | 2 +-
python/tvm/autotvm/tuner/tuner.py | 18 +-
python/tvm/contrib/cudnn.py | 8 +-
python/tvm/contrib/ethosu/cascader/proposal.py | 106 ++
.../contrib/ethosu/cascader/proposal_generator.py | 58 +
python/tvm/contrib/hexagon/session.py | 2 +
python/tvm/contrib/pipeline_executor.py | 16 +-
python/tvm/contrib/relay_viz/__init__.py | 10 +-
python/tvm/contrib/relay_viz/dot.py | 221 ++++
python/tvm/contrib/relay_viz/interface.py | 10 +-
python/tvm/contrib/relay_viz/terminal.py | 5 +-
python/tvm/driver/tvmc/runner.py | 2 +-
python/tvm/meta_schedule/builder/builder.py | 1 +
python/tvm/meta_schedule/builder/local_builder.py | 7 +-
python/tvm/meta_schedule/cost_model/xgb_model.py | 4 +-
python/tvm/meta_schedule/integration.py | 6 +-
python/tvm/meta_schedule/tune.py | 229 ++--
.../relay/backend/contrib/ethosu/tir/compiler.py | 4 +-
.../backend/contrib/ethosu/tir_to_cs_translator.py | 5 +-
python/tvm/relay/backend/te_compiler.py | 2 +
python/tvm/relay/frontend/onnx.py | 128 +-
python/tvm/relay/frontend/qnn_torch.py | 1 +
python/tvm/relay/frontend/tflite.py | 4 +-
python/tvm/relay/op/strategy/arm_cpu.py | 36 +-
python/tvm/relay/op/strategy/cuda.py | 14 +-
python/tvm/relay/qnn/op/legalizations.py | 4 +-
python/tvm/runtime/vm.py | 25 +
python/tvm/script/parser.py | 5 +-
python/tvm/script/tir/node.py | 3 +-
python/tvm/script/tir/scope_handler.py | 64 +-
python/tvm/te/__init__.py | 2 +-
python/tvm/te/operation.py | 22 +
python/tvm/tir/__init__.py | 11 +-
python/tvm/tir/stmt.py | 34 +
python/tvm/tir/transform/transform.py | 22 +
python/tvm/topi/arm_cpu/arm_utils.py | 6 +
python/tvm/topi/arm_cpu/conv2d_alter_op.py | 105 +-
python/tvm/topi/arm_cpu/conv2d_int8.py | 135 +-
python/tvm/topi/arm_cpu/injective.py | 2 +-
python/tvm/topi/arm_cpu/tensor_intrin.py | 134 +-
python/tvm/topi/cuda/conv2d_transpose.py | 6 +-
python/tvm/topi/generic/conv2d.py | 236 +++-
python/tvm/topi/nn/dense.py | 23 +-
python/tvm/topi/x86/batch_matmul.py | 73 +-
python/tvm/topi/x86/conv2d_alter_op.py | 118 +-
python/tvm/topi/x86/dense.py | 13 +-
python/tvm/topi/x86/dense_alter_op.py | 56 +-
src/arith/canonical_simplify.cc | 14 +-
src/arith/iter_affine_map.cc | 175 +--
src/arith/modular_set.cc | 12 +
src/arith/rewrite_simplify.cc | 14 +-
src/auto_scheduler/feature.cc | 20 +-
src/autotvm/touch_extractor.h | 20 +-
src/contrib/ethosu/cascader/pareto.cc | 27 +
src/contrib/ethosu/cascader/pareto.h | 4 +
src/contrib/ethosu/cascader/proposal.cc | 82 ++
src/contrib/ethosu/cascader/proposal.h | 105 ++
src/contrib/ethosu/cascader/proposal_generator.cc | 219 ++++
src/contrib/ethosu/cascader/proposal_generator.h | 86 ++
src/contrib/hybrid/codegen_hybrid.cc | 3 +
src/driver/driver_api.cc | 1 +
src/ir/module.cc | 7 +
src/ir/transform.cc | 56 -
src/node/structural_hash.cc | 68 +-
src/parser/parser.cc | 42 +-
src/printer/relay_text_printer.cc | 7 +-
src/printer/text_printer.h | 2 +
src/printer/tir_text_printer.cc | 13 +
src/printer/tvmscript_printer.cc | 72 +-
src/relay/backend/aot_executor_codegen.cc | 103 +-
src/relay/backend/build_module.cc | 29 +-
src/relay/backend/contrib/codegen_c/codegen_c.h | 2 +
src/relay/backend/contrib/ethosn/codegen.cc | 2 +
src/relay/backend/contrib/ethosn/ethosn_api.cc | 32 +-
src/relay/backend/contrib/ethosn/ethosn_api.h | 7 +-
src/relay/backend/executor.cc | 1 +
src/relay/backend/interpreter.cc | 5 +-
src/relay/backend/te_compiler.cc | 10 +-
src/relay/backend/te_compiler_cache.cc | 67 +-
src/relay/backend/te_compiler_cache.h | 4 +-
src/relay/backend/utils.h | 9 +-
src/relay/backend/vm/compiler.cc | 18 +-
src/relay/backend/vm/lambda_lift.cc | 9 +-
src/relay/ir/expr_functor.cc | 46 +-
src/relay/ir/transform.cc | 2 +-
src/relay/op/memory/on_device.cc | 44 +-
src/relay/op/memory/on_device.h | 27 -
src/relay/op/nn/nn.cc | 3 +-
src/relay/transforms/device_aware_visitors.cc | 16 +-
src/relay/transforms/device_aware_visitors.h | 6 +-
src/relay/transforms/device_planner.cc | 48 +-
src/relay/transforms/fold_constant.cc | 7 +-
src/relay/transforms/forward_rewrite.cc | 3 +
src/relay/transforms/fuse_ops.cc | 41 +-
src/relay/transforms/memory_alloc.cc | 3 +
src/relay/transforms/pattern_utils.h | 26 +
src/relay/transforms/split_args.cc | 3 +-
src/relay/transforms/to_a_normal_form.cc | 6 +-
src/runtime/crt/common/packed_func.c | 3 +
src/runtime/hexagon/hexagon/hexagon_buffer.cc | 2 +-
.../hexagon/hexagon/hexagon_device_api_v2.cc | 12 +
.../hexagon/rpc/simulator/hexagon_sim_proto.h | 75 ++
src/runtime/hexagon/rpc/simulator/rpc_server.cc | 319 +++++
src/runtime/hexagon/rpc/simulator/session.cc | 1322 ++++++++++++++++++++
src/runtime/pipeline/pipeline_executor.cc | 16 +-
src/runtime/pipeline/pipeline_executor.h | 8 +-
src/runtime/pipeline/pipeline_scheduler.cc | 18 +-
src/runtime/pipeline/pipeline_scheduler.h | 4 -
src/runtime/pipeline/pipeline_struct.h | 299 ++++-
src/runtime/profiling.cc | 6 +-
src/runtime/thread_pool.cc | 38 +-
src/runtime/threading_backend.cc | 251 ++--
src/runtime/vm/bytecode.cc | 3 +
src/runtime/vm/executable.cc | 25 +-
src/runtime/vm/vm.cc | 122 +-
src/target/llvm/codegen_amdgpu.cc | 2 +-
src/target/llvm/codegen_llvm.cc | 13 +-
src/target/llvm/codegen_llvm.h | 1 +
src/target/llvm/codegen_nvptx.cc | 2 +-
src/target/source/codegen_c.cc | 34 +-
src/target/source/codegen_c.h | 6 +
src/target/source/codegen_c_host.cc | 75 +-
src/target/source/codegen_c_host.h | 4 -
src/target/source/codegen_cuda.cc | 2 +-
src/target/source/codegen_opencl.cc | 3 +-
src/target/source/codegen_opencl.h | 2 +-
src/target/source/codegen_params.cc | 16 +-
src/target/source/intrin_rule_cuda.cc | 9 +-
src/target/source/source_module.cc | 2 +-
src/target/spirv/codegen_spirv.cc | 2 +-
src/tir/analysis/verify_gpu_code.cc | 4 +-
src/tir/ir/function.cc | 8 -
src/tir/ir/stmt.cc | 79 +-
src/tir/ir/stmt_functor.cc | 19 +
src/tir/op/builtin.cc | 3 +
src/tir/op/op.cc | 14 +
src/tir/schedule/instruction_traits.h | 13 +
src/tir/schedule/primitive/loop_transformation.cc | 16 +-
src/tir/transforms/arg_binder.cc | 3 +-
src/tir/transforms/bf16_legalize.cc | 115 +-
src/tir/transforms/bind_params.cc | 133 ++
src/tir/transforms/extract_constants.cc | 115 ++
src/tir/transforms/loop_partition.cc | 35 +-
src/tir/transforms/lower_tvm_builtin.cc | 18 +-
src/tir/transforms/lower_warp_memory.cc | 2 +-
.../merge_dynamic_shared_memory_allocations.cc | 2 +-
src/tir/transforms/renormalize_split_pattern.cc | 212 ++++
src/tir/transforms/split_host_device.cc | 5 +
src/tir/transforms/storage_flatten.cc | 2 +-
src/tir/transforms/storage_rewrite.cc | 33 +-
src/tir/usmp/transform/assign_pool_info.cc | 4 +-
.../convert_pool_allocations_to_offsets.cc | 2 +-
src/tir/usmp/unified_static_memory_planner.cc | 4 +-
src/tir/usmp/utils.cc | 2 +-
tests/cpp/ir_functor_test.cc | 5 +-
tests/cpp/threading_backend_test.cc | 130 +-
tests/lint/python_format.sh | 4 +-
.../contrib/test_ethosn/test_depth_to_space.py | 2 +-
tests/python/contrib/test_ethosn/test_networks.py | 3 +-
tests/python/contrib/test_ethosn/test_split.py | 3 +-
.../contrib/test_ethosu/cascader/conftest.py | 397 ++++++
.../cascader/test_proposal_generator.py | 139 ++
tests/python/contrib/test_ethosu/infra.py | 5 +-
tests/python/contrib/test_ethosu/test_networks.py | 81 +-
tests/python/contrib/test_hexagon/conftest.py | 75 +-
tests/python/contrib/test_hexagon/rpc/__init__.py | 18 -
tests/python/contrib/test_hexagon/rpc/conftest.py | 31 -
.../contrib/test_hexagon/test_cache_read_write.py | 135 ++
.../test_hexagon/{rpc => }/test_launcher.md | 25 +-
.../test_hexagon/{rpc => }/test_launcher.py | 28 +-
tests/python/frontend/onnx/test_forward.py | 36 +-
tests/python/frontend/pytorch/test_forward.py | 13 +-
tests/python/relay/aot/aot_test_utils.py | 99 +-
tests/python/relay/aot/test_c_device_api.py | 4 +-
tests/python/relay/aot/test_crt_aot.py | 80 +-
tests/python/relay/aot/test_crt_aot_usmp.py | 28 +-
.../python/relay/op/annotation/test_annotation.py | 12 -
tests/python/relay/test_cpp_build_module.py | 40 +
tests/python/relay/test_op_level1.py | 46 +-
tests/python/relay/test_op_level10.py | 45 +-
tests/python/relay/test_pass_alter_op_layout.py | 13 +
.../test_pass_fake_quantization_to_integer.py | 4 +-
tests/python/relay/test_pass_fold_constant.py | 21 +-
tests/python/relay/test_pass_fuse_ops.py | 41 +-
tests/python/relay/test_pass_plan_devices.py | 207 ++-
tests/python/relay/test_pipeline_executor.py | 15 +-
tests/python/relay/test_vm.py | 149 +++
tests/python/topi/python/test_topi_conv2d_int8.py | 125 +-
.../unittest/test_arith_canonical_simplify.py | 1 +
tests/python/unittest/test_arith_intset.py | 5 +-
.../python/unittest/test_arith_iter_affine_map.py | 2 +-
tests/python/unittest/test_arith_modular_set.py | 9 +
.../python/unittest/test_arith_rewrite_simplify.py | 5 +
.../unittest/test_auto_scheduler_task_scheduler.py | 2 +-
tests/python/unittest/test_ci.py | 32 +-
tests/python/unittest/test_link_params.py | 37 +-
...meta_schedule_postproc_rewrite_unbound_block.py | 4 +-
...chedule_schedule_rule_cross_thread_reduction.py | 16 +-
...ta_schedule_schedule_rule_multi_level_tiling.py | 20 +-
...hedule_schedule_rule_random_compute_location.py | 2 +-
.../unittest/test_meta_schedule_tune_relay.py | 5 +
.../python/unittest/test_meta_schedule_tune_tir.py | 6 +-
.../unittest/test_micro_model_library_format.py | 24 +-
tests/python/unittest/test_te_schedule_ops.py | 6 +
tests/python/unittest/test_tir_nodes.py | 7 +-
.../unittest/test_tir_schedule_split_fuse.py | 18 +-
tests/python/unittest/test_tir_schedule_trace.py | 31 +
.../test_tir_transform_extract_constants.py | 64 +
.../unittest/test_tir_transform_loop_partition.py | 74 ++
...test_tir_transform_renormalize_split_pattern.py | 119 ++
...ransform_convert_pool_allocations_to_offsets.py | 16 +-
tests/python/unittest/test_tvmscript_roundtrip.py | 98 ++
tests/scripts/ping_reviewers.py | 74 +-
tests/scripts/setup-pytest-env.sh | 4 +-
tests/scripts/task_config_build_hexagon.sh | 7 +-
tests/scripts/task_demo_microtvm.sh | 7 +-
tests/scripts/task_python_frontend.sh | 69 +-
...sk_python_nightly.sh => task_python_hexagon.sh} | 8 +-
258 files changed, 9005 insertions(+), 1954 deletions(-)
create mode 100644 python/tvm/contrib/ethosu/cascader/proposal.py
create mode 100644 python/tvm/contrib/ethosu/cascader/proposal_generator.py
create mode 100644 python/tvm/contrib/relay_viz/dot.py
create mode 100644 src/contrib/ethosu/cascader/proposal.cc
create mode 100644 src/contrib/ethosu/cascader/proposal.h
create mode 100644 src/contrib/ethosu/cascader/proposal_generator.cc
create mode 100644 src/contrib/ethosu/cascader/proposal_generator.h
create mode 100644 src/runtime/hexagon/rpc/simulator/hexagon_sim_proto.h
create mode 100644 src/runtime/hexagon/rpc/simulator/rpc_server.cc
create mode 100644 src/runtime/hexagon/rpc/simulator/session.cc
create mode 100644 src/tir/transforms/bind_params.cc
create mode 100644 src/tir/transforms/extract_constants.cc
create mode 100644 src/tir/transforms/renormalize_split_pattern.cc
create mode 100644
tests/python/contrib/test_ethosu/cascader/test_proposal_generator.py
delete mode 100644 tests/python/contrib/test_hexagon/rpc/__init__.py
delete mode 100644 tests/python/contrib/test_hexagon/rpc/conftest.py
create mode 100644 tests/python/contrib/test_hexagon/test_cache_read_write.py
rename tests/python/contrib/test_hexagon/{rpc => }/test_launcher.md (87%)
rename tests/python/contrib/test_hexagon/{rpc => }/test_launcher.py (91%)
create mode 100644
tests/python/unittest/test_tir_transform_extract_constants.py
create mode 100644
tests/python/unittest/test_tir_transform_renormalize_split_pattern.py
copy tests/scripts/{task_python_nightly.sh => task_python_hexagon.sh} (82%)
