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
discard 4542fade2e fixed tag
discard 4b4e5c98e7 Testing a new GPU image
add 209e77c18b [tvmc] add instruments for PassContext (#13136)
add 4fe8e96c77 [Hexagon] Set c++17 standard for launcher (#13140)
add 57b9988922 [Hexagon] Add scripts for e2e MetaSchedule tuning
demonstration (#13135)
add 1dbd1fa4b8 [Hexagon] Support fetching and building Hexagon runtime
with external runtime sources (#13138)
add 9f4c7121b9 [Hexagon] [runtime] Clarify compute resources (#13149)
add 0e21840553 [Hexagon] Fix chunk address table generation in
chunkify_hwio_16b (#13002)
add f28b0ca9c4 [CI] Update PyTorch to v1.12 in GPU image (#13160)
add 9aedb8bdda [Hexagon] refactor HexagonBufferManager class (#13145)
add 1311cac88b Fix typo in test_pipeline_executor.py (#13134)
add 100c050d60 [Relay] fix: trilu check op for i64/i32 (#13123)
add 46aa04f3f1 [ci] Wait longer during retries (#13102)
add eb8f225226 [ci] Ignore JUnit upload failures (#13142)
add 81e6e17ff0 [skip ci][ci] Ignore certain files for triggering docker
builds (#13167)
add e7a72af0c1 [microTVM] Improve code reuse in Corstone300 conv2d tests
(#13051)
add eeb8b7067e [Docker][CI] Remove mounting local download path to docker
(#13108)
add ad117809a1 [Hexagon] Skip
HexagonThreadManagerTest.thread_order_signal_wait unit test (#13171)
add e024b0d207 [Hexagon] Add power manager (#13162)
add 3948944745 [ci] Disable flaky Java RPC tests (#13170)
add e42e4bba28 [Hexagon] Adjust Hexagon pytest port range (#13172)
add f4fdc136b9 [Hexagon] [runtime] Allow creation of thread manager
without hardware resources (#13174)
add e5c7c9371e [TIR] Add the missing nearbyint intrinsic (#13175)
add 24e89befd6 [TVMScript] IRModule parser (#13176)
add 3131cdc56a [Target] Replace utility functions with target.features
(#12455)
add 7827fffe5c [Relay][Frontend][ONNX] Add GridSample operator (#13163)
add 5400b942d2 [ONNX] Add converter for FastGelu from Microsoft
onnxruntime contrib opset (#13119)
add 3e02ac5d2e [Adreno] Fix mem_scope annotations for prim funcs having
several heads (#13153)
add 03d989f441 [Adreno] Adapt reduction schedule for adreno (#13100)
add c302b3dce3 [AOT] Add CreateFunctionMetadata analysis pass (#13095)
add bb8aad160d [ci] Temporarily disable MacOS RPC tests (#13186)
add 7950271ceb [TVMScript] TIR parser (#13190)
add 0d4caa54ff [docs] Add 'Edit on GitHub' button (#13183)
add e41d0ed6eb [Relay] Rewrite division by constant to multiply (#13182)
add 23c2909f29 [Hexagon] Add support for instrumentation based profiling
for Hexagon (#12971)
add 7983ef2faa [Relay][ONNX] Resolve proto ingestion issue for non-tensor
type and add CastLike op (#13184)
add 30c38e0dce [Fix,Auto_scheduler] Default to extent of 1 if extent
cannot be determined (#13196)
add f44ef53a6b [Relay][ONNX] Add Optional, OptionalHasElement and
OptionalGetElement ops based on Sequence op (#13189)
add 62789a6423 [Hexagon] Add fix for vtcm allocation searches (#13197)
add 2a2dd9ac50 [ci] Update Docker images (#13200)
add 7cdb7a71ab [Hexagon]: Add upload function to hexagon session (#13161)
add 9dd0a70a07 [Hexagon] Fix correctness error in conv2d hvx (#13165)
add 3836eb9ca9 [Hexagon] [runtime] Per-thread hardware resource management
(#13181)
add 9b9b7458e9 [docs] Fix empty code blocks in tutorials (#13188)
add 697533e136 [TIR] Add utility for anchor block extraction (#13194)
add 04afd83113 [Hexagon] Add a test to show how to use multi input async
dma pipelin… (#13110)
add 6d2aa09cd9 [Hexagon] Fix port range bug and add random.seed (#13207)
add 645a5eafa3 [Relay][Hexagon] Add per-channel FixedPointMultiply
operation (#13080)
add 5c9066d816 [AMP] refine AMP and the corresponding tests for bfloat16
(#12787)
add 5acf3f90c6 [ci] Protect release branches (#13208)
add 3157cf188c [IR] Remove shadowing in
IRSubstituteWithDataTypeLegalization (#13219)
add 0554a46789 [TIR][Arith] Support negative coeff in ModularSet (#13081)
add 0c10302fd0 [Arith] Add internal NarrowPredicateExpression utility
(#13041)
add de7f762436 typo in 'build' NNPACK (#13215)
add 1c697d7460 [Frontend][Paddle] Add depthwise_conv2d_transpose op
mapping (#13214)
add cf2a8ea426 [Hexagon] Update tests to use AOT Executor (#13221)
add 187a2570ed [TIR][Schedule] Add cache_index to precompute index of
buffer load (#13192)
add 3cce9738bd [BugFix][TIR] Affine-binding check should not simplify
trivial iterators (#13203)
add e1ac5658ad [Hexagon]Pull and build specific LLVM sha for hexagon
docker CI (#13199)
add f42826eec4 [MetaSchedule] Enable anchor-block tuning (#13206)
add 602c194f25 [TIR][Arith] Implement kApplyConstraintsToBooleanBranches
extension (#13129)
add c0f33dfb97 [Logging][Hexagon] Improve logging on Hexagon (#13072)
add 3149ee5a73 [skip ci] Remove naut-thomas from .asf.yaml (#13231)
add 60e865a6fe [CI] Enable iOS RPC tests (#13229)
add 6cd1bb5e89 [Hexagon] Update search pattern to find .so address for on
device runs (#13230)
add b9e5c02d0b [TOPI][Hexagon] Implement quantized depthwise conv2d
(#12499)
add 20082614be [skip ci] Revert "[skip ci] Remove naut-thomas from
.asf.yaml (#13231)" (#13232)
add da76587505 [runtime] Fix typo in runtime/registry.h (#13224)
add 9dd0225a70 [TIR] Use Optional<Stmt> for IfThenElseNode::else_case
(#13218)
add f07f22fac9 [MetaSchedule][Minor] Fix Memory Database Module Equality
(#13198)
add e971956d76 [Relay][transform][SimplifyExpr] simplify adjacent muls and
adds with constants (#13213)
add 25a0d47d2b [Arith][TIR] Check for constant offsets of known literal
constraints (#13023)
add 76cd298bd9 [Hexagon] Add E2E test demonstrating how to apply blocked
layout schedule to conv2d via metaschedule (#13180)
add c9b10a80fd [ci][java] Use mavenCentral for dependencies (#13239)
add e48dded3b4 [BugFix][LLVM] Add UseInitArray flag in target_options_
(#13115)
add c3bb62e5c4 [skip ci][docs] Disable scipy intersphinx linking (#13245)
add 67fa959f18 [CI] Make additional_flags parameter optional in
tests/scripts/ci.py (#13238)
add 0683ece044 [MetaSchedule] Fix thread bindings of
MultiLevelTilingTensorCore (#13243)
add c69f8ce9c9 [Relay] Add ClipAndConsecutiveCast and CastClip to
SimplifyExpr (#13236)
add 3259580bd5 [Hexagon] Make pytest use a random port if not running in
CI (#13244)
add 9cdc97fe23 [DLPack][runtime] Update DLPack to v0.7 (#13177)
add 2c1fecd097 [TIR][Primitive] Support rolling_buffer schedule primitive
in TensorIR (#13033)
add e9e8c4b263 fix GPU other build (#13235)
add 5d15428994 [Relay] Add set_attrs_type registry to broadcast_to op
(#13096)
add 6551b71588 [COMMUNITY] Jyotsna Verma -> Reviewer (#13251)
add 87f52af48a [skip ci] Revert "fix GPU other build (#13235)" (#13261)
add 7536068e80 [MetaSchedule] Swap the order of RewriteTensorize and
VerifyGPUCode to reduce tuning time (#13259)
add 84fadc45d0 [CI] Skip failing Caffe tests due to broken URL (#13228)
add 4ecf303695 [TVMC] Apply constant folding when converting layout
(#13216)
add c3c1454233 Apply group write permissions to Python virtual environment
(#13252)
add da4bb4a65b Fix a typo in rpc/client.py (#12842)
add d261fa8838 [Hexagon] Add pylint on tests (#13233)
add 404d95f054 [build][relay][te][tir] remove unused vars / args (#13266)
add ff6aaeb12a [Frontend][Tensorflow2] Import graph_def to default graph
before calling function_def_to_graph_def (#13260)
add d998187e16 [Frontend][PaddlePaddle] Fix UnboundLocalError: local
variable 'shape… (#13247)
add e9ba9865a7 [skip ci] Revert "[ci] Protect release branches (#13208)"
(#13274)
add f15afd2251 [Docs] Minimal dependencies for Fedora/CentOS (#13248)
add 9df3a33ff4 [build][doc] Fix clang doxygen warnings (#13270)
add 0d553127e5 [build][tir] fix clang redundant-move warning (#13268)
add 75921fb559 [ETHOSN] Inline non-compute-intensive partitions (#13092)
add 47da418fbf [ETHOSN] Throw error message when inference fails (#13022)
add 1d1db35236 [MetaSchedule] Fix Task Hanging in EvolutionarySearch
(#13246)
add 215f0e2fc2 [Bugfix][TIR] Fix version conflict with `typing` for Python
3.9 (#13269)
add b98b9f92da [MetaSchedule] Improve the script for TorchBench model
tuning & benchmarking (#13255)
add 90ed632280 [Relay] Add tensor rank check for `nn.instance_norm`
(#13280)
add b1a099b213 [Relay] Enhancement for fold_scale_axis and simplify_expr
(#13275)
add de8a79d9ba [skip-ci][COMMUNITY] New committer Ashutosh Parkhi (#13286)
add ccb7d07159 [TIR][Arith] Use TryCompare to narrow inequalities if
possible (#13024)
add e860884923 [build][hexagon] remove unused variable (#13291)
add 6da298b3a0 [BugFix][Pattern] Fixed a crash when AltPattern and
FunctionPattern are used nested (#13278)
add dec74cb93d [build][tir] suppress -Woverloaded-virtual warning (#13267)
add be44e9c811 [Tensorize] Add logs to comparator to make debugging
tensorize failures easier (#13285)
add 62fadacd12 [Hexagon] Lint tests part 2 (#13271)
add 56878fab7c [TE] Make `elem_offset` of the buffers created by
`te.extern` a variable to avoid crash (#13297)
add 1e793641d8 [TIR] Added unit test for dynamic parameter in layout
transform (#13298)
add 732e34fe3d [TIR] Preserve loop annotation after loop partitioning
(#13292)
add b51c491c00 [FIX] Handle matmul where one inner dimension is unknown
(#13287)
add f2a740331f [DOCS][TVMC] Use correct argument to reuse tuning records
(#13302)
add 60e2c98fdb [Hexagon] Fix Hexagon external libs check (#13257)
add dd257e478e [Relay][Op] Add support for large index fp16 mean and var
(#13289)
add 6b238c4b6e [Bugfix][Runtime] Fix sched_setaffinity in Android (#13158)
add e398d16de8 [Torch] Fix advanced indexing with boolean mask (#13306)
add ce777fde18 [TOPI] Enhance VNNI dot product (#12588)
add b16a64d6ed [MetaSchedule] Refactor ScheduleRule Attributes (#13195)
add 904ae7748f [Frontend][PaddlePaddle] Add test case for interpolate op
convert func… (#13277)
add f8691180e0 [BugFix][Driver] Correctly propogate simple-mode flag in
LowerSchedule (#13311)
add e43841d2ef [microTVM] Fix RPC session close on runtime side (#13310)
add b807613c79 [Hexagon] [runtime] Move lock/unlock to HexagonHtp
temporarily (#13318)
add c898dc6b55 [TIR] Add thread sync if access index doesn't depend on
thread index (#13314)
add 79093a1718 [ETHOSN] Consolidate target string usage (#13159)
add be30238947 [Adreno][Textures] Fix static memory planner (#13253)
add bf77e793ab Fixup libtorch backend build (#13320)
add 15752e4009 [TVMScript] Hide trailing return type if None (#13308)
add 750ba9f742 [OpenCL][unit tests] Fix opencl cpp unit tests (#13254)
add 16bb1a6c2e [microTVM][CRT] Add memory size as project option (#13313)
add 36b1c5c4d9 [TIR] Remove redundant add in vnni/arm intrin (#13319)
add 244bceb457 [TIR] Allow folding cast with broadcast and ramp (#13317)
add 65dbee7f0c [Doc] Add desc of keep_orig_output_dtype for
ToMixedPrecision (#13321)
add 0e395c389c [AOT] Add CreateExecutorMetadata analysis pass (#13250)
add fbe174bd6c [microTVM][CRT][DOCS] Add a PyTorch tutorial for microTVM
with CRT (#13324)
add 999eee8c1a [ci] Update Jenkins readme to match new directory structure
(#13333)
add 8453c9c357 [MetaSchedule] Fix the order of applying `AutoInline` in
`ScheduleUsingAnchorTrace` (#13329)
add 5dc4186338 [MetaSchedule] Add JSON Database Validation Scripts (#12948)
add b4b90d76d9 [QNN, ONNX] Extension of QLinearMatMul in ONNX front-end
for all ranks of input tensors (#13322)
add 6d9d2132c1 [TIR] Check producer predicate in `ReverseComputeInline`
(#13338)
add a16a890483 [TOPI] Fix conv2d transpose for small channel (#13341)
add 1228104726 [Minor][Testing] Consolidate IRs into corresponding
functions (#13339)
add a0dcab2b24 [CPP_RPC][ANDROID] Fix cpp_rpc build failure (#13305)
add 3a30df6701 [Hexagon] Make allocate_hexagon_array a hexagon contrib API
(#13336)
add 54bd5e1f5f [microNPU] Fixed MergeConstants pass on striped networks
(#13281)
add 23ade0c14b [TVMC] Global pass context for compile and tune (#13309)
add 7cd203dc3e [TIR] Update ReductionIterNotIndexOutputBuffer to check
BlockRealizeN… (#13301)
add c66bb00a48 [Docker]Refactor timezone script and NRF installation
(#13342)
add 3a639a41a8 [TIR][Arith] Fix divisor checking in `TryConstFold` (#13348)
add b582cd12ae [MetaSchedule][Minor] Fix Typo in ApplyCustomRule Schedule
Rule (#13353)
add 93fdf83e8f [MetaSchedule] Improve inlining and `VerifyGPUCode` for
quantized model workload (#13334)
add f950b118aa [MetaSchedule][Minor] Allow Zero Run Time In Benchmarking
Result (#13354)
add 6d68aff030 [Bugfix][TIR] Patch for PR#13269 to support Python 3.10
(#13350)
add a15663654b [MetaSchedule] Unannotate `schedule_rule` if corresponding
schedule func is not found (#13346)
add f3eb239989 [MetaSchedule] Skip empty fx graph in TorchBench tuning
script (#13356)
add 5364e5a39a [MetaSchedule] Fuse loops around shared to global store
block in `MultiLevelTilingTensorCore` (#13357)
add 45327127d5 [TIR][Schedule] Make consistent implementation for
GetProducers() & GetConsumers() (#13344)
add f9ed60aaef [Relay] Refactor constant folding over expr into a utility
function (#13343)
add ce0e9abea0 [TIR] Make syntax of AST nodes different than ops (#13358)
add 5ffcfd9327 [FQ2I] Add cast back to input data type after AvgPool2d
(#13332)
add 8897983484 [IRBuilder][Minor] Add intrinsics like `T.int32x4` (#13361)
add 38771170cc [TIR][Schedule] Fix cache_read loc detecting and
region_cover checking (#13345)
add b20b7c4ad4 [TVMScript] Reorganize the folder structure (#12496)
add b8384d105b [ci] Assert some tests are not skipped in the CI (#12915)
add 5a767d053d [CI] Separate the ci scripts into Github and Jenkins
scripts (#13368)
add b6fae9b35e [TIR][Bugfix] Fix AXIS_SEPARATORS in
tir.Schedule.transform_layout (#13326)
add 68f51e692b [ci] Fix Jenkins quoting (#13380)
add 41a2243260 [CI] Do not merge before running CI on main (#13372)
add 3aa16f72dd [Codegen] Fix CUDA codegen for int64 Ramp (#13382)
add 647be2b425 [MetaSchedule] Add `from-target` Defaults for LLVM VNNI
Targets (#13383)
add 034dc67d03 [TFLite] Enable int64 biases for int16 quantized operators
(#12042)
add 4fd34b1bad [Fix][Warning] tvm.target.create() deprecated (#13391)
add 41b7a45083 feat: fix pylint issues for conftest (#13386)
add 24790d1d56 [RUNTIME][ALIGNMENT] Configurable kAllocAlignment if needed
(#13307)
add 72c60ad78a [AOT][FIX] Handle device contexts properly in
CreateFunctionMetadata (#13392)
add 8c30bda738 [MetaSchedule] Support schedules with cache read in
RewriteLayout (#13384)
add 42f9a766ff [TOPI] Add padding for dense/batch matmul for x86 vnni
(#13385)
add bac450a645 [Hexagon] Use VTCM while scheduling conv2d op (#13388)
add aa0c0db71b [tvm4j] Fix tvm4j build on MacOS (#13370)
add 557f1a9ded [docs] Add some more release process docs (#13351)
add 4f4b4edafd [TOPI][Hexagon] Implement quantized adaptive_avg_pool1d for
hexagon (#13282)
add 2bb3382c89 [OpenCL] Introduce OpenCL wrapper to TVM (#13362)
add 86a5ceec27 [TVMScript] Use tir::Evaluate if expression is in statement
context (#13396)
add 49c5d2f173 [Hexagon] Change single dma alloc buffers to be vtcm
(#13374)
add e030b146d9 [TECompiler] Replace static constant index with NameSupply
(#13397)
add edc1bfd105 Add support for cast double to fp16 (#13395)
add a042a93cd9 [TIR] Add test to cover specific case of reducer match
buffer checking (#13373)
add 6401d0ef62 [TIR] Fix extern_primfunc buffer order bug (#13347)
add 0d9b1850af [ci] fix GPU other build (#13366)
add 78b53221f8 [TIR] Remove PrimFuncNode::preflattened_buffer_map (#10940)
add 44ed06ac9f [ETHOSN] Relax concatenate offloading requirements (#13405)
add 52739ef8cd [ci][tvmbot] Fix spelling error, metionable -> mentionable
(#13276)
add 271ad43029 [TOPI] Update names for pooling ops (#13401)
add b4d4b82dbb [Hexagon] Fix TIR vrmpy tensorization (#13404)
add a80cdc26e2 [TIR][Analysis][Arith] Implement basic data-flow analysis
(#13130)
add 14342a37f5 [Hexagon] Enable Hexagon User DMA bypass mode (#13381)
add eca361de4f [MetaSchedule] Fix segfault in gradient based scheduler
(#13399)
add 59abd9e106 [Build] Fix MSVC compile option /bigobj (#13411)
add ad5c811411 [TIR] Unify index data type when creating prim func (#13327)
add c98f3cd6f8 [ACL] Enable int8 data type in QNN ADD (#13407)
add 5b1d2cc3e8 [usmp] Hill Climb greedy layout size check relaxed (#13369)
add 25ad54058a [CI] Update minor git options (#13398)
add 01a4725b6b [MetaSchedule][Fix] Fix Empty Run Time Issue when
Benchmarking Result (#13406)
add b29ab5c6ba [Hexagon] Add test to show scheduling of resnet50 with
async dma pipe… (#13352)
add 9cb36b1aad [COMMUNITY] Egor Churaev -> Committer (#13422)
add 53824d697a [Hexagon][QNN] Add TOPI strategies for qnn ops
mul/tanh/subtract (#13416)
new ddf6f35f11 Apply PR 13394.
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 (4542fade2e)
\
N -- N -- N refs/heads/ci-docker-staging (ddf6f35f11)
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:
.github/disabled_workflows/pr_comment_bot.yml | 4 +-
.github/workflows/cc_bot.yml | 2 +-
.github/workflows/nightly_docker_update.yml | 2 +-
.github/workflows/ping_reviewers.yml | 2 +-
.github/workflows/tag_teams.yml | 2 +-
.github/workflows/tvmbot.yml | 2 +-
.../workflows/update_last_successful_branch.yml | 2 +-
.gitmodules | 3 +
3rdparty/OpenCL-Headers | 1 +
3rdparty/dlpack | 2 +-
CMakeLists.txt | 11 +-
CONTRIBUTORS.md | 7 +-
Jenkinsfile | 643 ++--
LICENSE | 1 +
apps/android_camera/app/src/main/jni/Android.mk | 1 +
.../android_camera/app/src/main/jni/make/config.mk | 2 +-
apps/android_camera/app/src/main/jni/tvm_runtime.h | 8 +-
apps/android_deploy/README.md | 34 +-
apps/android_deploy/app/download-models.gradle | 2 +-
apps/android_deploy/app/src/main/jni/Android.mk | 3 +-
.../android_deploy/app/src/main/jni/make/config.mk | 2 +-
apps/android_deploy/app/src/main/jni/tvm_runtime.h | 3 +
apps/android_rpc/README.md | 32 +-
apps/android_rpc/app/src/main/jni/Android.mk | 3 +-
apps/android_rpc/app/src/main/jni/make/config.mk | 2 +-
apps/android_rpc/app/src/main/jni/tvm_runtime.h | 7 +-
apps/cpp_rpc/CMakeLists.txt | 10 +
apps/cpp_rpc/README.md | 10 +-
apps/hexagon_api/CMakeLists.txt | 2 +
apps/hexagon_launcher/CMakeLists.txt | 2 +
apps/hexagon_launcher/README.md | 40 +
apps/hexagon_launcher/cmake/hexagon/CMakeLists.txt | 5 +-
apps/hexagon_launcher/launcher_android.cc | 9 +-
apps/hexagon_launcher/launcher_core.h | 3 +
apps/hexagon_launcher/launcher_hexagon.cc | 10 +-
apps/hexagon_launcher/launcher_main.cc | 14 +-
apps/hexagon_launcher/launcher_rpc.idl | 2 +-
apps/howto_deploy/prepare_test_libs.py | 2 +-
apps/ios_rpc/tvmrpc/TVMRuntime.mm | 2 +-
apps/microtvm/cmsisnn/requirements.txt | 79 +-
apps/microtvm/ethosu/requirements.txt | 79 +-
apps/microtvm/reference-vm/base-box-tool.py | 1 +
.../reference-vm/base-box/base_box_setup.sh | 18 +-
ci/jenkins/Deploy.groovy.j2 | 4 +-
ci/jenkins/DockerBuild.groovy.j2 | 4 +-
ci/jenkins/Jenkinsfile.j2 | 22 +-
ci/jenkins/Prepare.groovy.j2 | 65 +-
ci/jenkins/README.md | 10 +-
ci/jenkins/macros.j2 | 24 +-
ci/scripts/{ => github}/__init__.py | 2 +-
ci/scripts/{ => github}/github_cc_reviewers.py | 4 +
ci/scripts/{ => github}/github_commenter.py | 6 +
ci/scripts/{ => github}/github_docs_comment.py | 0
ci/scripts/{ => github}/github_pr_comment.py | 6 +
.../{ => github}/github_skipped_tests_comment.py | 77 +-
ci/scripts/{ => github}/github_tag_teams.py | 5 +
ci/scripts/{ => github}/github_tvmbot.py | 9 +-
ci/scripts/{ => github}/ping_reviewers.py | 11 +-
ci/scripts/{ => github}/update_branch.py | 7 +-
ci/scripts/{ => jenkins}/check_pr.py | 0
ci/scripts/{ => jenkins}/cmd_utils.py | 3 +-
.../{ => jenkins}/determine_docker_images.py | 3 +-
ci/scripts/{ => jenkins}/git_change_docker.sh | 14 +
ci/scripts/{ => jenkins}/git_change_docs.sh | 0
ci/scripts/{ => jenkins}/git_skip_ci.py | 0
ci/scripts/{ => jenkins}/git_skip_ci_globs.py | 0
ci/scripts/{ => jenkins}/git_utils.py | 0
ci/scripts/{ => jenkins}/http_utils.py | 0
ci/scripts/{ => jenkins}/open_docker_update_pr.py | 3 +-
ci/scripts/{ => jenkins}/pytest_ids.py | 0
ci/scripts/{ => jenkins}/pytest_wrapper.py | 3 +-
ci/scripts/{ => jenkins}/retry.sh | 2 +-
ci/scripts/{ => jenkins}/should_rebuild_docker.py | 0
ci/scripts/{ => jenkins}/should_run_slow_tests.py | 6 +-
cmake/config.cmake | 6 +-
cmake/modules/Hexagon.cmake | 43 +
cmake/modules/HexagonSDK.cmake | 6 +
cmake/modules/LibInfo.cmake | 1 +
cmake/modules/OpenCL.cmake | 30 +-
cmake/modules/StandaloneCrt.cmake | 2 +-
cmake/utils/FindOpenCL.cmake | 2 +-
docker/Dockerfile.ci_adreno | 1 +
docker/Dockerfile.ci_arm | 3 +
docker/Dockerfile.ci_cortexm | 7 +
docker/Dockerfile.ci_cpu | 4 +
docker/Dockerfile.ci_gpu | 3 +
docker/Dockerfile.ci_hexagon | 4 +
docker/Dockerfile.ci_i386 | 3 +
docker/Dockerfile.ci_minimal | 3 +
docker/Dockerfile.ci_riscv | 3 +
docker/Dockerfile.ci_wasm | 3 +
docker/Dockerfile.demo_android | 4 +
docker/Dockerfile.demo_rocm | 3 +
docker/Dockerfile.demo_vitis_ai | 3 +
docker/Dockerfile.docs | 3 +
docker/bash.sh | 13 +-
docker/install/ubuntu_install_core.sh | 5 -
.../install/ubuntu_install_ethosn_driver_stack.sh | 7 +-
.../install/ubuntu_install_ethosu_driver_stack.sh | 10 +-
docker/install/ubuntu_install_hexagon.sh | 31 +-
..._install_core.sh => ubuntu_install_nrfjprog.sh} | 44 +-
docker/install/ubuntu_install_papi.sh | 6 +-
docker/install/ubuntu_install_python.sh | 1 +
docker/install/ubuntu_install_python_package.sh | 2 +-
docker/install/ubuntu_install_vela.sh | 2 +-
.../{ubuntu_install_vela.sh => ubuntu_setup_tz.sh} | 4 +-
docs/conf.py | 31 +-
docs/contribute/release_process.rst | 27 +-
docs/install/from_source.rst | 9 +
docs/install/nnpack.rst | 2 +-
gallery/how_to/compile_models/from_tflite.py | 8 +-
.../deploy_models/deploy_model_on_android.py | 5 +-
.../optimize_operators/opt_conv_tensorcore.py | 12 +-
.../how_to/tune_with_autotvm/tune_conv2d_cuda.py | 12 +-
gallery/how_to/tune_with_autotvm/tune_relay_arm.py | 12 +-
.../how_to/tune_with_autotvm/tune_relay_cuda.py | 12 +-
.../tune_with_autotvm/tune_relay_mobile_gpu.py | 12 +-
gallery/how_to/work_with_microtvm/micro_aot.py | 6 +-
gallery/how_to/work_with_microtvm/micro_ethosu.py | 4 +-
gallery/how_to/work_with_microtvm/micro_pytorch.py | 206 ++
gallery/how_to/work_with_microtvm/micro_tflite.py | 12 +-
gallery/tutorial/autotvm_matmul_x86.py | 12 +-
gallery/tutorial/cross_compilation_and_rpc.py | 12 +-
gallery/tutorial/install.py | 12 +-
gallery/tutorial/tvmc_python.py | 4 +-
include/tvm/arith/analyzer.h | 24 +-
include/tvm/arith/int_set.h | 9 +
include/tvm/ir/expr.h | 4 +
include/tvm/meta_schedule/database.h | 8 +
include/tvm/meta_schedule/mutator.h | 2 +
include/tvm/meta_schedule/postproc.h | 2 +
include/tvm/meta_schedule/schedule/cpu/.gitignore | 0
.../tvm/meta_schedule/schedule/cuda/thread_bind.h | 50 +-
.../tvm/meta_schedule/schedule/generic/winograd.h | 30 +-
include/tvm/meta_schedule/schedule/x86/.gitignore | 0
include/tvm/meta_schedule/schedule_rule.h | 25 +-
include/tvm/meta_schedule/search_strategy.h | 2 +
include/tvm/relay/attrs/nn.h | 6 +-
include/tvm/relay/attrs/transform.h | 17 +
include/tvm/runtime/builtin_fp16.h | 1 +
include/tvm/runtime/c_runtime_api.h | 73 +-
include/tvm/runtime/device_api.h | 19 +
include/tvm/runtime/logging.h | 30 +-
include/tvm/runtime/registry.h | 4 +-
include/tvm/script/ir_builder/tir/frame.h | 19 +-
include/tvm/script/ir_builder/tir/ir.h | 66 +-
include/tvm/tir/analysis.h | 32 +
include/tvm/tir/builtin.h | 17 +-
include/tvm/tir/data_type_rewriter.h | 155 +
include/tvm/tir/function.h | 43 +-
include/tvm/tir/op_attr_types.h | 31 +
include/tvm/tir/schedule/schedule.h | 25 +
include/tvm/tir/stmt.h | 5 +-
include/tvm/tir/stmt_functor.h | 51 -
include/tvm/tir/transform.h | 6 +
include/tvm/topi/nn/pooling.h | 18 +-
jvm/core/src/main/java/org/apache/tvm/Device.java | 79 +-
jvm/core/src/main/java/org/apache/tvm/NDArray.java | 31 +-
.../main/java/org/apache/tvm/rpc/RPCSession.java | 25 +-
jvm/native/osx-x86_64/pom.xml | 4 +-
python/gen_requirements.py | 2 +-
python/tvm/_ffi/runtime_ctypes.py | 93 +-
python/tvm/contrib/hexagon/build.py | 272 +-
python/tvm/contrib/hexagon/hexagon_profiler.py | 128 +
python/tvm/contrib/hexagon/meta_schedule.py | 25 +-
.../contrib/hexagon/profiling/process_lwp_data.py | 387 +++
python/tvm/contrib/hexagon/pytest_plugin.py | 47 +-
python/tvm/contrib/hexagon/session.py | 55 +-
python/tvm/contrib/hexagon/tools.py | 41 +
python/tvm/contrib/xcode.py | 4 +-
python/tvm/driver/tvmc/autotuner.py | 197 +-
python/tvm/driver/tvmc/compiler.py | 118 +-
python/tvm/driver/tvmc/composite_target.py | 12 -
python/tvm/driver/tvmc/transform.py | 12 +-
python/tvm/ir/memory_pools.py | 52 +
python/tvm/meta_schedule/__init__.py | 7 +-
python/tvm/meta_schedule/database/json_database.py | 4 +
.../tvm/meta_schedule/database/memory_database.py | 4 +
.../meta_schedule/database/schedule_fn_database.py | 4 +
python/tvm/meta_schedule/profiler.py | 2 +-
python/tvm/meta_schedule/relay_integration.py | 24 +-
.../tvm/meta_schedule/schedule}/__init__.py | 4 +-
.../tvm/meta_schedule/schedule/cpu}/__init__.py | 3 +-
.../tvm/meta_schedule/schedule/cuda}/__init__.py | 3 +-
.../meta_schedule/schedule/generic}/__init__.py | 3 +-
.../tvm/meta_schedule/schedule/x86}/__init__.py | 3 +-
python/tvm/meta_schedule/schedule_rule/__init__.py | 7 +-
.../schedule_rule/apply_custom_rule.py | 32 +-
.../tvm/meta_schedule/schedule_rule/auto_inline.py | 17 +
.../search_strategy/evolutionary_search.py | 6 +-
.../meta_schedule/testing/conv2d_winograd_cpu.py | 172 -
.../meta_schedule/testing/conv2d_winograd_cuda.py | 173 -
.../meta_schedule/testing/custom_builder_runner.py | 4 +-
python/tvm/meta_schedule/testing/relay_workload.py | 1 -
.../tvm/meta_schedule/testing/space_generation.py | 2 +-
python/tvm/meta_schedule/testing/te_workload.py | 150 +-
python/tvm/meta_schedule/testing/torchbench/run.py | 292 +-
.../tvm/meta_schedule/testing/torchbench/utils.py | 8 +-
python/tvm/meta_schedule/testing/tune_utils.py | 55 +-
.../tvm/meta_schedule/testing/validate_database.py | 282 ++
python/tvm/meta_schedule/trace_apply.py | 39 +
python/tvm/meta_schedule/tune.py | 4 +
python/tvm/micro/project_api/client.py | 6 +-
python/tvm/micro/session.py | 2 +
python/tvm/relay/backend/aot.py | 60 +
.../relay/backend/contrib/ethosu/tir/compiler.py | 5 +-
.../tvm/relay/backend/contrib/ethosu/tir/passes.py | 77 +-
python/tvm/relay/backend/te_compiler.py | 4 +-
python/tvm/relay/frontend/onnx.py | 310 +-
python/tvm/relay/frontend/paddlepaddle.py | 3 +-
python/tvm/relay/frontend/pytorch.py | 15 +-
python/tvm/relay/frontend/tensorflow2.py | 30 +-
python/tvm/relay/frontend/tflite.py | 6 +-
python/tvm/relay/op/_tensor.py | 13 +
python/tvm/relay/op/contrib/arm_compute_lib.py | 2 +-
python/tvm/relay/op/contrib/ethosn.py | 91 +-
python/tvm/relay/op/image/image.py | 2 +-
python/tvm/relay/op/nn/_nn.py | 4 +-
python/tvm/relay/op/strategy/adreno.py | 17 +-
python/tvm/relay/op/strategy/arm_cpu.py | 29 +-
python/tvm/relay/op/strategy/bifrost.py | 10 +-
python/tvm/relay/op/strategy/cuda.py | 72 +-
python/tvm/relay/op/strategy/generic.py | 10 +-
python/tvm/relay/op/strategy/mali.py | 10 +-
python/tvm/relay/op/strategy/x86.py | 10 +-
python/tvm/relay/qnn/op/_qnn.py | 11 +-
python/tvm/relay/qnn/op/legalizations.py | 21 +-
python/tvm/relay/qnn/strategy/generic.py | 27 +
python/tvm/relay/qnn/strategy/hexagon.py | 36 +
.../transform/fake_quantization_to_integer.py | 7 +-
python/tvm/relay/transform/transform.py | 14 +-
python/tvm/rpc/client.py | 21 +-
python/tvm/runtime/ndarray.py | 26 +-
python/tvm/runtime/vm.py | 2 +
python/tvm/script/__init__.py | 7 +-
python/tvm/script/ir_builder/tir/__init__.py | 1 +
python/tvm/script/ir_builder/tir/frame.py | 4 +-
python/tvm/script/ir_builder/tir/ir.py | 553 ++--
.../{_parser/_core.py => parser/__init__.py} | 8 +-
python/tvm/script/{_parser => parser}/_core.py | 5 +-
.../script/{_parser => parser}/core/__init__.py | 0
.../script/{_parser => parser}/core/diagnostics.py | 39 +-
.../script/{_parser => parser}/core/dispatch.py | 0
python/tvm/script/{_parser => parser}/core/doc.py | 0
.../script/{_parser => parser}/core/doc_core.py | 0
.../tvm/script/{_parser => parser}/core/entry.py | 10 +
.../script/{_parser => parser}/core/evaluator.py | 0
.../tvm/script/{_parser => parser}/core/parser.py | 15 +
.../tvm/script/{_parser => parser}/core/utils.py | 36 +-
.../tvm/script/parser/ir/__init__.py | 9 +-
.../backend/aot.py => script/parser/ir/entry.py} | 37 +-
python/tvm/script/parser/ir/parser.py | 66 +
.../tvm/script/parser/tir/__init__.py | 18 +-
python/tvm/script/parser/tir/entry.py | 108 +
python/tvm/script/parser/tir/operation.py | 85 +
python/tvm/script/parser/tir/parser.py | 473 +++
python/tvm/script/{ => parser_v1}/__init__.py | 0
python/tvm/script/{ => parser_v1}/_ffi_api.py | 0
.../script/{ => parser_v1}/context_maintainer.py | 3 -
python/tvm/script/{ => parser_v1}/diagnostics.py | 0
python/tvm/script/{ => parser_v1}/meta_unparser.py | 0
python/tvm/script/{ => parser_v1}/parser.py | 1 -
python/tvm/script/{ => parser_v1}/registry.py | 0
python/tvm/script/{ => parser_v1}/tir/__init__.py | 5 +-
python/tvm/script/{ => parser_v1}/tir/__init__.pyi | 12 -
python/tvm/script/{ => parser_v1}/tir/intrin.py | 80 +-
python/tvm/script/{ => parser_v1}/tir/node.py | 0
python/tvm/script/{ => parser_v1}/tir/prim_func.py | 0
.../script/{ => parser_v1}/tir/scope_handler.py | 0
.../tvm/script/{ => parser_v1}/tir/special_stmt.py | 73 -
python/tvm/script/{ => parser_v1}/tir/ty.py | 2 +-
python/tvm/script/{ => parser_v1}/utils.py | 0
python/tvm/te/operation.py | 29 +-
python/tvm/testing/__init__.py | 2 -
python/tvm/testing/plugin.py | 10 +
python/tvm/testing/tir.py | 45 +-
python/tvm/tir/__init__.py | 4 +-
python/tvm/tir/analysis/analysis.py | 27 +
python/tvm/tir/buffer.py | 28 +-
python/tvm/tir/expr.py | 2 +
python/tvm/tir/function.py | 7 -
python/tvm/tir/op.py | 107 +-
python/tvm/tir/schedule/_type_checker.py | 40 +-
python/tvm/tir/schedule/analysis.py | 18 +
python/tvm/tir/schedule/schedule.py | 200 +-
python/tvm/tir/tensor_intrin/arm_cpu.py | 6 +-
python/tvm/tir/tensor_intrin/cuda.py | 12 +-
python/tvm/tir/tensor_intrin/hexagon.py | 138 +-
python/tvm/tir/tensor_intrin/rocm.py | 1 +
python/tvm/tir/tensor_intrin/x86.py | 5 +-
python/tvm/tir/transform/transform.py | 11 +
python/tvm/topi/adreno/__init__.py | 1 +
python/tvm/topi/adreno/reduction.py | 69 +
python/tvm/topi/adreno/utils.py | 31 +-
python/tvm/topi/arm_cpu/arm_utils.py | 58 +-
python/tvm/topi/arm_cpu/conv2d.py | 6 +-
python/tvm/topi/arm_cpu/conv2d_gemm.py | 23 +-
python/tvm/topi/arm_cpu/conv2d_int8.py | 12 +-
python/tvm/topi/arm_cpu/depthwise_conv2d.py | 13 +-
.../topi/arm_cpu/mprofile/dsp/tensordot_conv2ds.py | 39 +-
python/tvm/topi/cuda/conv2d_alter_op.py | 56 +-
python/tvm/topi/cuda/conv2d_nhwc_winograd.py | 4 +-
python/tvm/topi/cuda/conv2d_transpose.py | 2 +-
python/tvm/topi/cuda/conv2d_winograd.py | 35 +-
python/tvm/topi/cuda/reduction.py | 20 +-
python/tvm/topi/hexagon/qnn/__init__.py | 2 +
python/tvm/topi/hexagon/qnn/adaptive_avg_pool1d.py | 120 +
python/tvm/topi/hexagon/qnn/nn.py | 179 +-
.../topi/hexagon/qnn/qdepthwise_conv2d_slice.py | 217 ++
python/tvm/topi/hexagon/slice_ops/dwconv2d.py | 5 +-
python/tvm/topi/hexagon/tensor_intrin.py | 84 +-
python/tvm/topi/hexagon/utils.py | 26 +
python/tvm/topi/image/grid_sample.py | 18 +-
python/tvm/topi/math.py | 58 +
python/tvm/topi/nn/conv2d.py | 352 +-
python/tvm/topi/nn/dense.py | 4 +-
python/tvm/topi/transform.py | 7 +
python/tvm/topi/utils.py | 41 +-
python/tvm/topi/x86/batch_matmul.py | 8 +-
python/tvm/topi/x86/dense.py | 14 +-
python/tvm/topi/x86/dense_alter_op.py | 35 +-
python/tvm/topi/x86/tensor_intrin.py | 14 +-
src/arith/conjunctive_normal_form.cc | 26 +-
src/arith/const_fold.h | 3 +-
src/arith/constraint_extract.cc | 39 +-
src/arith/constraint_extract.h | 31 +-
src/arith/ir_mutator_with_analyzer.cc | 12 +-
src/arith/ir_visitor_with_analyzer.cc | 4 +-
src/arith/ir_visitor_with_analyzer.h | 6 +-
src/arith/modular_set.cc | 13 +-
src/arith/narrow_predicate_expression.cc | 219 ++
src/arith/narrow_predicate_expression.h | 57 +
src/arith/rewrite_simplify.cc | 341 +-
src/arith/rewrite_simplify.h | 21 +
src/arith/transitive_comparison_analyzer.cc | 170 +-
src/arith/unwrap_vector_expr.cc | 90 +
.../{constraint_extract.h => unwrap_vector_expr.h} | 30 +-
src/auto_scheduler/utils.h | 4 +-
src/contrib/hybrid/codegen_hybrid.cc | 4 +-
src/driver/driver_api.cc | 17 +-
src/meta_schedule/database/json_database.cc | 3 +-
src/meta_schedule/database/memory_database.cc | 3 +-
.../measure_callback/update_cost_model.cc | 4 +-
src/meta_schedule/module_equality.cc | 24 +
src/meta_schedule/module_equality.h | 4 +
src/meta_schedule/mutator/mutator.cc | 2 +
src/meta_schedule/postproc/postproc.cc | 12 +-
.../postproc/rewrite_cooperative_fetch.cc | 30 +
src/meta_schedule/postproc/rewrite_layout.cc | 159 +-
.../postproc/rewrite_unbound_block.cc | 3 +-
src/meta_schedule/postproc/verify_gpu_code.cc | 2 +
src/meta_schedule/schedule/cpu/winograd.cc | 101 +
.../auto_bind.cc => schedule/cuda/thread_bind.cc} | 166 +-
src/meta_schedule/schedule/cuda/winograd.cc | 163 +
.../schedule/generic/winograd.cc} | 37 +-
src/meta_schedule/schedule/x86/.gitignore | 0
.../schedule_rule/apply_custom_rule.cc | 93 +
src/meta_schedule/schedule_rule/auto_bind.cc | 143 +-
src/meta_schedule/schedule_rule/auto_inline.cc | 37 +
.../schedule_rule/multi_level_tiling.cc | 4 +-
.../multi_level_tiling_tensor_core.cc | 10 +
src/meta_schedule/schedule_rule/schedule_rule.cc | 86 +-
src/meta_schedule/schedule_rule/winograd.cc | 249 --
.../search_strategy/evolutionary_search.cc | 18 +-
.../space_generator/post_order_apply.cc | 49 +-
.../space_generator/space_generator.cc | 18 +-
src/meta_schedule/task_scheduler/gradient_based.cc | 17 +-
src/meta_schedule/trace_apply.cc | 259 ++
src/meta_schedule/trace_apply.h | 48 +
src/meta_schedule/utils.h | 61 +-
src/printer/tir_text_printer.cc | 14 +-
src/printer/tvmscript_printer.cc | 158 +-
src/relay/analysis/util.cc | 3 +-
src/relay/backend/annotate_used_memory.cc | 4 +
src/relay/backend/aot/aot_lower_main.cc | 7 +-
src/relay/backend/aot/create_executor_metadata.cc | 86 +
src/relay/backend/aot/create_executor_metadata.h | 50 +
src/relay/backend/aot/create_function_metadata.cc | 124 +
src/relay/backend/aot/create_function_metadata.h | 49 +
src/relay/backend/aot_executor_codegen.cc | 2 +-
.../backend/contrib/arm_compute_lib/codegen.cc | 2 +-
src/relay/backend/contrib/cmsisnn/relay_to_tir.cc | 2 +-
src/relay/backend/contrib/constant_transforms.cc | 10 +-
src/relay/backend/contrib/constant_transforms.h | 9 -
src/relay/backend/contrib/ethosn/codegen.cc | 85 +-
src/relay/backend/contrib/ethosn/codegen_ethosn.h | 18 +
.../backend/contrib/ethosn/convert_equivalent.cc | 4 +-
.../backend/contrib/ethosn/inline_partitions.cc | 126 +
.../contrib/example_target_hooks/relay_to_tir.cc | 2 +-
.../backend/contrib/libtorch/libtorch_codegen.cc | 11 +-
src/relay/backend/graph_plan_memory.cc | 245 +-
src/relay/backend/task_extraction.cc | 80 +-
src/relay/backend/te_compiler.cc | 13 +-
src/relay/backend/te_compiler_cache.cc | 56 +-
src/relay/backend/te_compiler_cache.h | 15 +-
src/relay/backend/token_allocator.cc | 201 ++
src/relay/backend/token_allocator.h | 161 +
src/relay/backend/utils.cc | 2 +-
src/relay/backend/utils.h | 2 +
src/relay/collage/mock_cost_estimator.h | 10 +
src/relay/ir/dataflow_matcher.cc | 24 +-
src/relay/op/image/grid_sample.cc | 2 +-
src/relay/op/make_op.h | 4 +
src/relay/op/nn/nn.cc | 1 +
src/relay/op/nn/nn.h | 37 +-
src/relay/op/tensor/reduce.cc | 37 +-
src/relay/op/tensor/transform.cc | 130 +
src/relay/qnn/op/add.cc | 3 +-
src/relay/qnn/op/convolution_transpose.cc | 10 +-
src/relay/qnn/op/dense.cc | 10 +-
src/relay/qnn/op/mul.cc | 3 +-
src/relay/qnn/op/requantize.cc | 20 +-
src/relay/qnn/op/subtract.cc | 3 +-
src/relay/qnn/utils.cc | 31 +
src/relay/qnn/utils.h | 17 +
src/relay/quantize/realize.cc | 18 +-
src/relay/quantize/realize.h | 2 -
src/relay/transforms/annotate_texture_storage.cc | 10 +-
.../transforms/auto_scheduler_layout_rewrite.cc | 2 +-
src/relay/transforms/div_to_mul.cc | 86 +
.../transforms/fake_quantization_to_integer.cc | 2 +-
src/relay/transforms/fold_constant.cc | 20 +-
.../fold_constant.h} | 30 +-
.../transforms/meta_schedule_layout_rewrite.cc | 2 +-
src/relay/transforms/pattern_utils.h | 36 +
src/relay/transforms/remove_standalone_reshapes.cc | 4 +-
src/relay/transforms/simplify_expr.cc | 212 +-
src/relay/transforms/to_mixed_precision.cc | 12 +-
src/runtime/aot_executor/aot_executor.cc | 4 +-
src/runtime/builtin_fp16.cc | 8 +
src/runtime/contrib/arm_compute_lib/acl_utils.cc | 2 +
src/runtime/contrib/ethosn/ethosn_device.cc | 70 +-
src/runtime/contrib/ethosn/ethosn_runtime.h | 33 +
.../crt/host/{Makefile => Makefile.template} | 5 +-
src/runtime/crt/host/main.cc | 2 +-
src/runtime/crt/host/microtvm_api_server.py | 46 +-
src/runtime/dso_library.cc | 14 +
src/runtime/hexagon/hexagon_buffer.cc | 18 +-
src/runtime/hexagon/hexagon_buffer_manager.h | 16 +-
src/runtime/hexagon/hexagon_common.cc | 13 +-
src/runtime/hexagon/hexagon_common.h | 5 +-
src/runtime/hexagon/hexagon_device_api.cc | 39 +-
src/runtime/hexagon/hexagon_device_api.h | 17 +-
.../hexagon/{hexagon_hmx.cc => hexagon_htp.cc} | 52 +-
.../hexagon/{hexagon_hmx.h => hexagon_htp.h} | 32 +-
src/runtime/hexagon/hexagon_hvx.cc | 27 +-
src/runtime/hexagon/hexagon_hvx.h | 13 +
src/runtime/hexagon/hexagon_power_manager.cc | 108 +
.../{hexagon_hmx.h => hexagon_power_manager.h} | 37 +-
src/runtime/hexagon/hexagon_thread_manager.cc | 99 +-
src/runtime/hexagon/hexagon_thread_manager.h | 57 +-
src/runtime/hexagon/hexagon_user_dma.cc | 61 +-
src/runtime/hexagon/hexagon_user_dma.h | 2 +-
src/runtime/hexagon/hexagon_vtcm_pool.cc | 3 +-
src/runtime/hexagon/hexagon_vtcm_pool.h | 12 +
src/runtime/hexagon/ops/conv2d_fp16_hvx.cc | 2 +-
src/runtime/hexagon/ops/conv_utils.cc | 4 +-
src/runtime/hexagon/profiler/README.md | 99 +
src/runtime/hexagon/profiler/lwp_handler.S | 115 +
src/runtime/hexagon/profiler/prof_utils.cc | 78 +
.../hexagon/profiler/prof_utils.h} | 27 +-
src/runtime/hexagon/rpc/android_bash.sh.template | 3 +-
src/runtime/hexagon/rpc/hexagon/rpc_server.cc | 49 +-
src/runtime/hexagon/rpc/simulator/rpc_server.cc | 27 +
src/runtime/hexagon/rpc/simulator/session.cc | 2 -
src/runtime/logging.cc | 7 +
src/runtime/opencl/opencl_module.cc | 2 +-
src/runtime/opencl/opencl_wrapper/README.md | 25 +
.../opencl/opencl_wrapper/opencl_wrapper.cc | 574 ++++
src/runtime/rocm/rocm_device_api.cc | 3 +-
src/runtime/threading_backend.cc | 17 +-
src/script/ir_builder/tir/frame.cc | 15 +-
src/script/ir_builder/tir/ir.cc | 78 +-
src/support/libinfo.cc | 1 +
src/target/intrin_rule.cc | 87 +-
src/target/llvm/codegen_hexagon.cc | 23 +
src/target/llvm/codegen_llvm.cc | 8 +-
src/target/llvm/intrin_rule_nvptx.cc | 3 +
src/target/llvm/intrin_rule_rocm.cc | 3 +
src/target/llvm/llvm_instance.cc | 2 +
src/target/parsers/cpu.cc | 5 +
src/target/source/codegen_c.cc | 4 +-
src/target/source/codegen_cuda.cc | 4 +-
src/target/source/intrin_rule_aocl.cc | 6 +
src/target/source/intrin_rule_cuda.cc | 3 +
src/target/source/intrin_rule_metal.cc | 3 +
src/target/source/intrin_rule_opencl.cc | 3 +
src/target/source/intrin_rule_vhls.cc | 3 +
src/target/spirv/codegen_spirv.cc | 4 +-
src/target/spirv/intrin_rule_spirv.cc | 6 +
src/target/stackvm/codegen_stackvm.cc | 4 +-
src/target/tag.cc | 9 +-
src/target/target_kind.cc | 3 +
src/te/operation/compute_op.cc | 4 +-
src/te/operation/create_primfunc.cc | 43 +-
src/te/operation/create_primfunc.h | 8 +-
src/tir/analysis/block_access_region_detector.cc | 4 +-
src/tir/analysis/control_flow_graph.cc | 1647 ++++++++++
src/tir/analysis/control_flow_graph.h | 653 ++++
src/tir/analysis/device_constraint_utils.cc | 22 +-
src/tir/analysis/estimate_flops.cc | 4 +-
src/tir/analysis/stmt_finding.cc | 150 +
src/tir/analysis/verify_gpu_code.cc | 13 +
src/tir/contrib/ethosu/passes.cc | 33 +-
src/tir/ir/data_type_rewriter.cc | 391 ++-
src/tir/ir/function.cc | 10 +-
src/tir/ir/stmt.cc | 4 +-
src/tir/ir/stmt_functor.cc | 14 +-
src/tir/op/builtin.cc | 14 +-
src/tir/op/op.cc | 11 +-
src/tir/schedule/analysis.h | 9 -
src/tir/schedule/analysis/analysis.cc | 80 +-
src/tir/schedule/analysis/reducer.cc | 20 +-
src/tir/schedule/concrete_schedule.cc | 25 +
src/tir/schedule/concrete_schedule.h | 3 +
src/tir/schedule/ir_comparator.cc | 247 +-
src/tir/schedule/primitive.h | 25 +
src/tir/schedule/primitive/blockize_tensorize.cc | 15 +
src/tir/schedule/primitive/cache_index.cc | 484 +++
src/tir/schedule/primitive/cache_read_write.cc | 37 +-
src/tir/schedule/primitive/compute_inline.cc | 46 +-
src/tir/schedule/primitive/get_block_loop.cc | 27 +-
.../schedule/primitive/layout_transformation.cc | 462 +--
src/tir/schedule/primitive/rolling_buffer.cc | 474 +++
src/tir/schedule/schedule.cc | 5 +
src/tir/schedule/state.cc | 7 +-
src/tir/schedule/traced_schedule.cc | 26 +
src/tir/schedule/traced_schedule.h | 3 +
src/tir/schedule/utils.h | 46 +
src/tir/transforms/bf16_legalize.cc | 29 -
src/tir/transforms/common_subexpr_elim_tools.cc | 6 +-
src/tir/transforms/compact_buffer_region.cc | 4 +-
src/tir/transforms/coproc_sync.cc | 4 +-
src/tir/transforms/flatten_buffer.cc | 16 +-
src/tir/transforms/inject_virtual_thread.cc | 6 +-
src/tir/transforms/ir_utils.cc | 2 +-
src/tir/transforms/legalize_packed_calls.cc | 4 +-
src/tir/transforms/lift_attr_scope.cc | 4 +-
src/tir/transforms/loop_partition.cc | 6 +-
src/tir/transforms/lower_async_dma.cc | 14 +-
src/tir/transforms/lower_match_buffer.cc | 4 +-
src/tir/transforms/lower_tvm_builtin.cc | 15 -
src/tir/transforms/make_packed_api.cc | 4 +-
src/tir/transforms/narrow_datatype.cc | 148 +-
.../plan_update_buffer_allocation_location.cc | 6 +-
src/tir/transforms/profile_instrumentation.cc | 293 ++
src/tir/transforms/remove_no_op.cc | 4 +-
src/tir/transforms/simplify.cc | 120 +-
src/tir/transforms/storage_access.cc | 4 +-
src/tir/transforms/storage_flatten.cc | 201 +-
src/tir/transforms/thread_storage_sync.cc | 31 +-
src/tir/transforms/vectorize_loop.cc | 6 +-
src/tir/usmp/algo/hill_climb.cc | 78 +-
src/tir/usmp/transform/assign_pool_info.cc | 4 +-
.../convert_pool_allocations_to_offsets.cc | 10 +-
src/tir/usmp/transform/create_io_allocates.cc | 5 +-
src/tir/usmp/utils.cc | 6 +-
.../hexagon/hexagon_thread_manager_tests.cc | 45 +-
.../cpp-runtime/hexagon/hexagon_user_dma_tests.cc | 151 +-
.../cpp-runtime/hexagon/hexagon_vtcm_pool_tests.cc | 80 +
tests/cpp-runtime/opencl/opencl_timer_test.cc | 4 +-
tests/cpp/arith_simplify_test.cc | 22 +
tests/cpp/data_type_rewriter_test.cc | 2 +-
tests/cpp/relay/backend/graph_plan_token_alloc.cc | 351 ++
tests/cpp/runtime/contrib/ethosn/inference_test.cc | 74 +
tests/lint/check_file_type.py | 3 +
tests/lint/check_request_hook.py | 74 +-
tests/lint/cpplint.sh | 3 +-
tests/lint/pylint.sh | 19 +-
tests/python/ci/test_ci.py | 142 +-
tests/python/ci/test_tvmbot.py | 8 +-
tests/python/ci/test_utils.py | 2 +
tests/python/conftest.py | 37 +-
.../contrib/test_arm_compute_lib/test_add.py | 4 +-
.../contrib/test_arm_compute_lib/test_network.py | 25 +-
tests/python/contrib/test_dnnl.py | 5 +-
tests/python/contrib/test_ethosn/infrastructure.py | 43 +-
tests/python/contrib/test_ethosn/test_addition.py | 22 +-
tests/python/contrib/test_ethosn/test_codegen.py | 52 +
.../python/contrib/test_ethosn/test_concatenate.py | 11 +-
.../test_ethosn/test_convert_equivalents.py | 4 +-
.../contrib/test_ethosn/test_depth_to_space.py | 11 +-
.../contrib/test_ethosn/test_inline_partitions.py | 167 +
.../python/contrib/test_ethosn/test_leaky_relu.py | 11 +-
tests/python/contrib/test_ethosn/test_multiply.py | 11 +-
tests/python/contrib/test_ethosn/test_networks.py | 11 +-
.../contrib/test_ethosn/test_partition_params.py | 82 -
tests/python/contrib/test_ethosn/test_relu.py | 11 +-
.../python/contrib/test_ethosn/test_requantize.py | 18 +-
tests/python/contrib/test_ethosn/test_reshape.py | 19 +-
tests/python/contrib/test_ethosn/test_split.py | 11 +-
tests/python/contrib/test_ethosn/test_tanh.py | 11 +-
.../python/contrib/test_ethosn/test_topologies.py | 84 +-
.../contrib/test_ethosu/test_encode_constants.py | 61 +-
.../contrib/test_ethosu/test_hoist_allocates.py | 31 +-
.../contrib/test_ethosu/test_merge_constants.py | 209 +-
.../test_ethosu/test_remove_concatenates.py | 7 +-
.../contrib/test_ethosu/test_replace_conv2d.py | 48 +-
.../contrib/test_ethosu/test_replace_copy.py | 8 +-
tests/python/contrib/test_ethosu/test_scheduler.py | 4 +-
tests/python/contrib/test_hexagon/README_RPC.md | 2 +-
.../python/contrib/test_hexagon/benchmark_util.py | 6 +-
.../python/contrib/test_hexagon/infrastructure.py | 47 +-
.../{topi => metaschedule_e2e}/__init__.py | 2 +-
.../test_hexagon/metaschedule_e2e/export_models.py | 79 +
.../metaschedule_e2e/test_resnet50_fp16.py | 135 +
.../metaschedule_e2e/test_resnet50_int8.py | 539 ++++
tests/python/contrib/test_hexagon/pytest_util.py | 36 +-
.../test_hexagon/test_2d_physical_buffers.py | 5 +-
.../test_hexagon/test_async_dma_pipeline.py | 727 +++--
.../test_hexagon/test_benchmark_elemwise_add.py | 107 +-
.../test_hexagon/test_benchmark_maxpool2d.py | 194 +-
.../contrib/test_hexagon/test_cache_read_write.py | 222 --
.../test_hexagon/test_fixed_point_conversion.py | 13 +-
.../test_hexagon/test_fixed_point_multiply.py | 152 +-
tests/python/contrib/test_hexagon/test_launcher.py | 158 +
.../contrib/test_hexagon/test_memory_alloc.py | 25 +-
.../contrib/test_hexagon/test_meta_schedule.py | 165 +-
tests/python/contrib/test_hexagon/test_models.py | 73 +-
.../contrib/test_hexagon/test_parallel_hvx.py | 81 +-
.../test_hexagon/test_parallel_hvx_load_vtcm.py | 274 +-
.../contrib/test_hexagon/test_parallel_scalar.py | 63 +-
.../contrib/test_hexagon/test_run_unit_tests.py | 4 +
tests/python/contrib/test_hexagon/test_sigmoid.py | 43 +-
.../test_hexagon/test_software_pipeline_async.py | 274 +-
tests/python/contrib/test_hexagon/test_usmp.py | 10 +-
tests/python/contrib/test_hexagon/test_vtcm.py | 63 +
.../contrib/test_hexagon/test_vtcm_bandwidth.py | 56 +-
.../test_hexagon/test_wo_qnn_canonicalization.py | 191 +-
tests/python/contrib/test_hexagon/topi/__init__.py | 2 +-
.../test_hexagon/topi/slice_op}/__init__.py | 6 +-
.../topi/{ => slice_op}/test_argmax_slice.py | 4 +-
.../topi/{ => slice_op}/test_avg_pool2d_slice.py | 7 +-
.../topi/{ => slice_op}/test_cast_slice.py | 8 +-
.../{test_clip.py => slice_op/test_clip_slice.py} | 5 +-
.../topi/{ => slice_op}/test_conv2d_slice.py | 3 +-
.../topi/slice_op/test_depthwise_conv2d_slice.py | 340 ++
.../topi/{ => slice_op}/test_dequantize_slice.py | 4 +-
.../topi/{ => slice_op}/test_max_pool2d_slice.py | 7 +-
.../topi/{ => slice_op}/test_relu_slice.py | 4 +-
.../topi/{ => slice_op}/test_softmax_slice.py | 3 +-
.../topi/{ => slice_op}/test_tanh_slice.py | 5 +-
.../test_hexagon/topi/test_adaptive_avg_pool1d.py | 185 ++
.../topi/test_add_subtract_multiply.py | 274 +-
.../contrib/test_hexagon/topi/test_batch_matmul.py | 54 +-
.../test_hexagon/topi/test_conv2d_fp16_intrin.py | 199 +-
.../contrib/test_hexagon/topi/test_conv2d_nchw.py | 133 +-
.../contrib/test_hexagon/topi/test_conv2d_nhwc.py | 49 +-
.../test_hexagon/topi/test_conv2d_transpose.py | 83 +-
.../python/contrib/test_hexagon/topi/test_dense.py | 178 +-
.../test_hexagon/topi/test_depth_to_space.py | 52 +-
.../test_hexagon/topi/test_depthwise_conv2d.py | 273 +-
.../test_hexagon/topi/test_dwconv2d_slice.py | 314 --
tests/python/contrib/test_hexagon/topi/test_pad.py | 11 +-
.../contrib/test_hexagon/topi/test_pooling.py | 39 +-
.../contrib/test_hexagon/topi/test_quantize.py | 62 +-
.../contrib/test_hexagon/topi/test_reduce.py | 209 +-
.../contrib/test_hexagon/topi/test_reshape.py | 79 +-
.../contrib/test_hexagon/topi/test_resize2d.py | 146 +-
.../contrib/test_hexagon/topi/test_softmax.py | 121 +-
tests/python/driver/tvmc/conftest.py | 16 +
tests/python/driver/tvmc/test_autotuner.py | 16 +
tests/python/driver/tvmc/test_compiler.py | 31 +-
tests/python/driver/tvmc/test_frontends.py | 17 +-
tests/python/driver/tvmc/test_transform.py | 74 +
tests/python/frontend/caffe/test_forward.py | 3 +
tests/python/frontend/onnx/test_forward.py | 281 +-
tests/python/frontend/paddlepaddle/test_forward.py | 9 +-
tests/python/frontend/pytorch/test_forward.py | 13 +-
tests/python/frontend/tflite/test_forward.py | 23 +
tests/python/integration/test_auto_tensorize.py | 1 +
.../relay/aot/test_aot_create_executor_metadata.py | 176 +
.../relay/aot/test_aot_create_function_metadata.py | 302 ++
tests/python/relay/aot/test_pass_aot_lower_main.py | 30 +-
.../opencl_texture/test_conv2d_nchw_texture.py | 193 ++
.../relay/opencl_texture/test_reduction_texture.py | 51 +
tests/python/relay/strategy/arm_cpu/test_conv2d.py | 133 +
.../relay/strategy/arm_cpu/test_conv2d_nchw.py | 110 -
.../relay/strategy/arm_cpu/test_conv2d_nhwc.py | 180 --
.../strategy/arm_cpu/test_depthwise_conv2d.py | 235 +-
.../strategy/arm_cpu/test_generalized_conv2d.py | 152 +
.../relay/strategy/arm_cpu/test_group_conv2d.py | 142 +-
tests/python/relay/test_dataflow_pattern.py | 33 +
tests/python/relay/test_op_level1.py | 41 +-
tests/python/relay/test_op_level10.py | 16 +-
tests/python/relay/test_op_level2.py | 4 +-
tests/python/relay/test_op_level3.py | 20 +
tests/python/relay/test_op_level4.py | 23 +-
tests/python/relay/test_op_level5.py | 5 +-
tests/python/relay/test_op_qnn_requantize.py | 495 +--
.../test_pass_fake_quantization_to_integer.py | 21 +
tests/python/relay/test_pass_fold_scale_axis.py | 51 +
tests/python/relay/test_pass_simplify_expr.py | 155 +-
tests/python/relay/test_pipeline_executor.py | 25 +-
tests/python/relay/test_to_mixed_precision.py | 178 +-
tests/python/target/test_arm_target.py | 37 +-
tests/python/topi/python/test_topi_conv2d_int8.py | 5 +-
.../python/test_topi_group_conv2d_transpose.py | 1 +
tests/python/topi/python/test_topi_image.py | 5 +-
tests/python/topi/python/test_topi_math.py | 4 +
tests/python/topi/python/test_topi_transform.py | 1 +
.../unittest/test_aot_legalize_packed_call.py | 38 +-
.../unittest/test_arith_canonical_simplify.py | 4 +
tests/python/unittest/test_arith_domain_touched.py | 24 +-
.../test_arith_narrow_predicate_expression.py | 87 +
.../python/unittest/test_arith_rewrite_simplify.py | 63 +-
.../python/unittest/test_auto_scheduler_feature.py | 16 +-
.../conftest.py => unittest/test_div_to_mul.py} | 22 +-
tests/python/unittest/test_index_map.py | 8 +-
tests/python/unittest/test_lower_build.py | 36 +-
.../test_meta_schedule_custom_rule_winograd_cpu.py | 206 --
...test_meta_schedule_custom_rule_winograd_cuda.py | 328 --
.../test_meta_schedule_measure_callback.py | 54 +-
.../test_meta_schedule_post_order_apply.py | 43 -
.../test_meta_schedule_postproc_rewrite_layout.py | 276 ++
...est_meta_schedule_postproc_rewrite_tensorize.py | 5 +-
.../test_meta_schedule_relay_integration.py | 188 +-
...eta_schedule_schedule_rule_apply_custom_rule.py | 66 +
...test_meta_schedule_schedule_rule_auto_inline.py | 115 +
.../test_meta_schedule_schedule_rule_mlt_tc.py | 39 +-
.../unittest/test_meta_schedule_search_strategy.py | 56 +
.../test_meta_schedule_space_cpu_winograd.py | 168 +
.../unittest/test_meta_schedule_space_cuda.py | 171 +-
.../test_meta_schedule_space_cuda_winograd.py | 355 ++
.../unittest/test_meta_schedule_task_scheduler.py | 85 +
.../unittest/test_meta_schedule_trace_apply.py | 3376 ++++++++++++++++++++
.../test_meta_schedule_vnni_integration.py | 16 +-
tests/python/unittest/test_te_create_primfunc.py | 43 +-
.../unittest/test_tir_analysis_stmt_finding.py | 54 +
.../python/unittest/test_tir_lower_match_buffer.py | 42 +-
.../unittest/test_tir_schedule_cache_index.py | 78 +
.../unittest/test_tir_schedule_cache_read_write.py | 76 +
.../unittest/test_tir_schedule_compute_inline.py | 178 ++
tests/python/unittest/test_tir_schedule_reindex.py | 2 +-
.../unittest/test_tir_schedule_rolling_buffer.py | 573 ++++
.../test_tir_schedule_state_cached_flags.py | 30 +
.../python/unittest/test_tir_schedule_tensorize.py | 46 +-
.../unittest/test_tir_schedule_transform_layout.py | 107 +-
.../python/unittest/test_tir_schedule_utilities.py | 2 +-
.../python/unittest/test_tir_te_extern_primfunc.py | 50 +-
.../unittest/test_tir_transform_flatten_buffer.py | 44 +-
.../test_tir_transform_inject_software_pipeline.py | 20 +-
.../test_tir_transform_inject_virtual_thread.py | 18 +-
.../unittest/test_tir_transform_loop_partition.py | 189 +-
...t_tir_transform_lower_cross_thread_reduction.py | 179 +-
.../unittest/test_tir_transform_narrow_datatype.py | 33 +-
.../unittest/test_tir_transform_profiling_instr.py | 340 ++
.../unittest/test_tir_transform_remove_assume.py | 6 +-
...test_tir_transform_renormalize_split_pattern.py | 42 +-
.../python/unittest/test_tir_transform_simplify.py | 878 ++++-
.../unittest/test_tir_transform_thread_sync.py | 36 +-
...ransform_convert_pool_allocations_to_offsets.py | 72 -
.../python/unittest/test_tvmscript_error_report.py | 707 ++--
.../unittest/test_tvmscript_ir_builder_tir.py | 30 +-
.../unittest/test_tvmscript_parser_evaluator.py | 4 +-
.../test_tvmscript_parser_ir.py} | 22 +-
.../unittest/test_tvmscript_parser_source.py | 4 +-
tests/python/unittest/test_tvmscript_parser_tir.py | 63 +
tests/python/unittest/test_tvmscript_roundtrip.py | 33 +-
tests/python/unittest/test_tvmscript_spans.py | 2 +-
.../python/unittest/test_tvmscript_syntax_sugar.py | 62 +-
tests/scripts/ci.py | 11 +-
tests/scripts/release/gather_prs.py | 3 +-
tests/scripts/release/make_notes.py | 2 +
tests/scripts/setup-pytest-env.sh | 2 +-
tests/scripts/task_build.py | 4 +-
tests/scripts/task_build_adreno_bins.sh | 1 +
tests/scripts/task_java_unittest.sh | 35 +-
tests/scripts/task_python_ethosn_tests.sh | 7 +-
tests/scripts/task_python_frontend.sh | 2 +-
tests/scripts/task_python_microtvm.sh | 1 +
web/emcc/wasm_runtime.cc | 10 +-
772 files changed, 34865 insertions(+), 10995 deletions(-)
create mode 160000 3rdparty/OpenCL-Headers
rename ci/scripts/{ => github}/__init__.py (94%)
rename ci/scripts/{ => github}/github_cc_reviewers.py (95%)
rename ci/scripts/{ => github}/github_commenter.py (95%)
rename ci/scripts/{ => github}/github_docs_comment.py (100%)
rename ci/scripts/{ => github}/github_pr_comment.py (94%)
rename ci/scripts/{ => github}/github_skipped_tests_comment.py (75%)
rename ci/scripts/{ => github}/github_tag_teams.py (97%)
rename ci/scripts/{ => github}/github_tvmbot.py (98%)
rename ci/scripts/{ => github}/ping_reviewers.py (96%)
rename ci/scripts/{ => github}/update_branch.py (96%)
rename ci/scripts/{ => jenkins}/check_pr.py (100%)
rename ci/scripts/{ => jenkins}/cmd_utils.py (95%)
rename ci/scripts/{ => jenkins}/determine_docker_images.py (98%)
rename ci/scripts/{ => jenkins}/git_change_docker.sh (73%)
copy ci/scripts/{ => jenkins}/git_change_docs.sh (100%)
rename ci/scripts/{ => jenkins}/git_skip_ci.py (100%)
rename ci/scripts/{ => jenkins}/git_skip_ci_globs.py (100%)
rename ci/scripts/{ => jenkins}/git_utils.py (100%)
rename ci/scripts/{ => jenkins}/http_utils.py (100%)
rename ci/scripts/{ => jenkins}/open_docker_update_pr.py (99%)
rename ci/scripts/{ => jenkins}/pytest_ids.py (100%)
rename ci/scripts/{ => jenkins}/pytest_wrapper.py (98%)
rename ci/scripts/{ => jenkins}/retry.sh (94%)
rename ci/scripts/{ => jenkins}/should_rebuild_docker.py (100%)
rename ci/scripts/{ => jenkins}/should_run_slow_tests.py (95%)
copy docker/install/{ubuntu_install_core.sh => ubuntu_install_nrfjprog.sh}
(53%)
copy docker/install/{ubuntu_install_vela.sh => ubuntu_setup_tz.sh} (90%)
create mode 100644 gallery/how_to/work_with_microtvm/micro_pytorch.py
create mode 100644 include/tvm/meta_schedule/schedule/cpu/.gitignore
rename src/meta_schedule/schedule_rule/auto_bind.h =>
include/tvm/meta_schedule/schedule/cuda/thread_bind.h (57%)
copy src/target/parsers/cpu.cc =>
include/tvm/meta_schedule/schedule/generic/winograd.h (62%)
create mode 100644 include/tvm/meta_schedule/schedule/x86/.gitignore
create mode 100644 include/tvm/tir/data_type_rewriter.h
create mode 100755 python/tvm/contrib/hexagon/hexagon_profiler.py
create mode 100644 python/tvm/contrib/hexagon/profiling/process_lwp_data.py
copy {tests/python/contrib/test_hexagon/topi =>
python/tvm/meta_schedule/schedule}/__init__.py (90%)
copy {tests/python/contrib/test_hexagon/topi =>
python/tvm/meta_schedule/schedule/cpu}/__init__.py (92%)
copy {tests/python/contrib/test_hexagon/topi =>
python/tvm/meta_schedule/schedule/cuda}/__init__.py (91%)
copy {tests/python/contrib/test_hexagon/topi =>
python/tvm/meta_schedule/schedule/generic}/__init__.py (92%)
copy {tests/python/contrib/test_hexagon/topi =>
python/tvm/meta_schedule/schedule/x86}/__init__.py (92%)
rename ci/scripts/git_change_docs.sh =>
python/tvm/meta_schedule/schedule_rule/apply_custom_rule.py (53%)
mode change 100755 => 100644
delete mode 100644 python/tvm/meta_schedule/testing/conv2d_winograd_cpu.py
delete mode 100644 python/tvm/meta_schedule/testing/conv2d_winograd_cuda.py
create mode 100644 python/tvm/meta_schedule/testing/validate_database.py
create mode 100644 python/tvm/meta_schedule/trace_apply.py
copy python/tvm/script/{_parser/_core.py => parser/__init__.py} (86%)
rename python/tvm/script/{_parser => parser}/_core.py (84%)
rename python/tvm/script/{_parser => parser}/core/__init__.py (100%)
rename python/tvm/script/{_parser => parser}/core/diagnostics.py (88%)
rename python/tvm/script/{_parser => parser}/core/dispatch.py (100%)
rename python/tvm/script/{_parser => parser}/core/doc.py (100%)
rename python/tvm/script/{_parser => parser}/core/doc_core.py (100%)
rename python/tvm/script/{_parser => parser}/core/entry.py (83%)
rename python/tvm/script/{_parser => parser}/core/evaluator.py (100%)
rename python/tvm/script/{_parser => parser}/core/parser.py (97%)
rename python/tvm/script/{_parser => parser}/core/utils.py (63%)
copy docker/install/ubuntu_install_vela.sh =>
python/tvm/script/parser/ir/__init__.py (87%)
mode change 100755 => 100644
copy python/tvm/{relay/backend/aot.py => script/parser/ir/entry.py} (59%)
create mode 100644 python/tvm/script/parser/ir/parser.py
copy tests/python/contrib/test_opencl/conftest.py =>
python/tvm/script/parser/tir/__init__.py (72%)
create mode 100644 python/tvm/script/parser/tir/entry.py
create mode 100644 python/tvm/script/parser/tir/operation.py
create mode 100644 python/tvm/script/parser/tir/parser.py
copy python/tvm/script/{ => parser_v1}/__init__.py (100%)
rename python/tvm/script/{ => parser_v1}/_ffi_api.py (100%)
rename python/tvm/script/{ => parser_v1}/context_maintainer.py (98%)
rename python/tvm/script/{ => parser_v1}/diagnostics.py (100%)
rename python/tvm/script/{ => parser_v1}/meta_unparser.py (100%)
rename python/tvm/script/{ => parser_v1}/parser.py (99%)
rename python/tvm/script/{ => parser_v1}/registry.py (100%)
rename python/tvm/script/{ => parser_v1}/tir/__init__.py (88%)
rename python/tvm/script/{ => parser_v1}/tir/__init__.pyi (97%)
rename python/tvm/script/{ => parser_v1}/tir/intrin.py (79%)
rename python/tvm/script/{ => parser_v1}/tir/node.py (100%)
rename python/tvm/script/{ => parser_v1}/tir/prim_func.py (100%)
rename python/tvm/script/{ => parser_v1}/tir/scope_handler.py (100%)
rename python/tvm/script/{ => parser_v1}/tir/special_stmt.py (93%)
rename python/tvm/script/{ => parser_v1}/tir/ty.py (99%)
rename python/tvm/script/{ => parser_v1}/utils.py (100%)
create mode 100644 python/tvm/topi/adreno/reduction.py
create mode 100755 python/tvm/topi/hexagon/qnn/adaptive_avg_pool1d.py
create mode 100644 python/tvm/topi/hexagon/qnn/qdepthwise_conv2d_slice.py
create mode 100644 src/arith/narrow_predicate_expression.cc
create mode 100644 src/arith/narrow_predicate_expression.h
create mode 100644 src/arith/unwrap_vector_expr.cc
copy src/arith/{constraint_extract.h => unwrap_vector_expr.h} (57%)
create mode 100644 src/meta_schedule/schedule/cpu/winograd.cc
copy src/meta_schedule/{schedule_rule/auto_bind.cc =>
schedule/cuda/thread_bind.cc} (62%)
create mode 100644 src/meta_schedule/schedule/cuda/winograd.cc
copy src/{target/parsers/cpu.cc => meta_schedule/schedule/generic/winograd.cc}
(55%)
create mode 100644 src/meta_schedule/schedule/x86/.gitignore
create mode 100644 src/meta_schedule/schedule_rule/apply_custom_rule.cc
delete mode 100644 src/meta_schedule/schedule_rule/winograd.cc
create mode 100644 src/meta_schedule/trace_apply.cc
create mode 100644 src/meta_schedule/trace_apply.h
create mode 100644 src/relay/backend/aot/create_executor_metadata.cc
create mode 100644 src/relay/backend/aot/create_executor_metadata.h
create mode 100644 src/relay/backend/aot/create_function_metadata.cc
create mode 100644 src/relay/backend/aot/create_function_metadata.h
create mode 100644 src/relay/backend/contrib/ethosn/inline_partitions.cc
create mode 100644 src/relay/backend/token_allocator.cc
create mode 100644 src/relay/backend/token_allocator.h
create mode 100644 src/relay/transforms/div_to_mul.cc
copy src/relay/{backend/contrib/constant_transforms.h =>
transforms/fold_constant.h} (60%)
rename src/runtime/crt/host/{Makefile => Makefile.template} (94%)
rename src/runtime/hexagon/{hexagon_hmx.cc => hexagon_htp.cc} (66%)
copy src/runtime/hexagon/{hexagon_hmx.h => hexagon_htp.h} (68%)
create mode 100644 src/runtime/hexagon/hexagon_power_manager.cc
rename src/runtime/hexagon/{hexagon_hmx.h => hexagon_power_manager.h} (60%)
create mode 100644 src/runtime/hexagon/profiler/README.md
create mode 100644 src/runtime/hexagon/profiler/lwp_handler.S
create mode 100644 src/runtime/hexagon/profiler/prof_utils.cc
copy src/{target/parsers/cpu.cc => runtime/hexagon/profiler/prof_utils.h} (70%)
create mode 100644 src/runtime/opencl/opencl_wrapper/README.md
create mode 100644 src/runtime/opencl/opencl_wrapper/opencl_wrapper.cc
create mode 100644 src/tir/analysis/control_flow_graph.cc
create mode 100644 src/tir/analysis/control_flow_graph.h
create mode 100644 src/tir/analysis/stmt_finding.cc
create mode 100644 src/tir/schedule/primitive/cache_index.cc
create mode 100644 src/tir/schedule/primitive/rolling_buffer.cc
create mode 100644 src/tir/transforms/profile_instrumentation.cc
create mode 100644 tests/cpp/relay/backend/graph_plan_token_alloc.cc
create mode 100644 tests/cpp/runtime/contrib/ethosn/inference_test.cc
create mode 100644 tests/python/contrib/test_ethosn/test_codegen.py
create mode 100644 tests/python/contrib/test_ethosn/test_inline_partitions.py
delete mode 100644 tests/python/contrib/test_ethosn/test_partition_params.py
copy tests/python/contrib/test_hexagon/{topi => metaschedule_e2e}/__init__.py
(93%)
create mode 100644
tests/python/contrib/test_hexagon/metaschedule_e2e/export_models.py
create mode 100644
tests/python/contrib/test_hexagon/metaschedule_e2e/test_resnet50_fp16.py
create mode 100644
tests/python/contrib/test_hexagon/metaschedule_e2e/test_resnet50_int8.py
mode change 100644 => 100755
tests/python/contrib/test_hexagon/test_2d_physical_buffers.py
delete mode 100644 tests/python/contrib/test_hexagon/test_cache_read_write.py
create mode 100644 tests/python/contrib/test_hexagon/test_vtcm.py
rename {python/tvm/script/_parser =>
tests/python/contrib/test_hexagon/topi/slice_op}/__init__.py (92%)
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_argmax_slice.py (97%)
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_avg_pool2d_slice.py (99%)
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_cast_slice.py (97%)
rename tests/python/contrib/test_hexagon/topi/{test_clip.py =>
slice_op/test_clip_slice.py} (96%)
mode change 100755 => 100644
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_conv2d_slice.py (98%)
mode change 100755 => 100644
create mode 100644
tests/python/contrib/test_hexagon/topi/slice_op/test_depthwise_conv2d_slice.py
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_dequantize_slice.py (98%)
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_max_pool2d_slice.py (98%)
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_relu_slice.py (96%)
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_softmax_slice.py (98%)
rename tests/python/contrib/test_hexagon/topi/{ =>
slice_op}/test_tanh_slice.py (96%)
create mode 100755
tests/python/contrib/test_hexagon/topi/test_adaptive_avg_pool1d.py
mode change 100755 => 100644
tests/python/contrib/test_hexagon/topi/test_add_subtract_multiply.py
delete mode 100644
tests/python/contrib/test_hexagon/topi/test_dwconv2d_slice.py
mode change 100755 => 100644
tests/python/contrib/test_hexagon/topi/test_quantize.py
mode change 100755 => 100644
tests/python/contrib/test_hexagon/topi/test_resize2d.py
create mode 100644 tests/python/driver/tvmc/test_transform.py
create mode 100644 tests/python/relay/aot/test_aot_create_executor_metadata.py
create mode 100644 tests/python/relay/aot/test_aot_create_function_metadata.py
create mode 100644 tests/python/relay/opencl_texture/test_reduction_texture.py
create mode 100644 tests/python/relay/strategy/arm_cpu/test_conv2d.py
delete mode 100644 tests/python/relay/strategy/arm_cpu/test_conv2d_nchw.py
delete mode 100644 tests/python/relay/strategy/arm_cpu/test_conv2d_nhwc.py
create mode 100644
tests/python/relay/strategy/arm_cpu/test_generalized_conv2d.py
create mode 100644
tests/python/unittest/test_arith_narrow_predicate_expression.py
copy tests/python/{contrib/test_opencl/conftest.py =>
unittest/test_div_to_mul.py} (59%)
delete mode 100644
tests/python/unittest/test_meta_schedule_custom_rule_winograd_cpu.py
delete mode 100644
tests/python/unittest/test_meta_schedule_custom_rule_winograd_cuda.py
create mode 100644
tests/python/unittest/test_meta_schedule_schedule_rule_apply_custom_rule.py
create mode 100644
tests/python/unittest/test_meta_schedule_space_cpu_winograd.py
create mode 100644
tests/python/unittest/test_meta_schedule_space_cuda_winograd.py
create mode 100644 tests/python/unittest/test_meta_schedule_trace_apply.py
create mode 100644 tests/python/unittest/test_tir_analysis_stmt_finding.py
create mode 100644 tests/python/unittest/test_tir_schedule_cache_index.py
create mode 100644 tests/python/unittest/test_tir_schedule_rolling_buffer.py
create mode 100644 tests/python/unittest/test_tir_transform_profiling_instr.py
rename tests/python/{contrib/test_opencl/conftest.py =>
unittest/test_tvmscript_parser_ir.py} (70%)
create mode 100644 tests/python/unittest/test_tvmscript_parser_tir.py
