This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a change to branch nightly-docker-update
in repository https://gitbox.apache.org/repos/asf/tvm.git
discard 02a67184a3 [ci][docker] Nightly Docker image update
add 600a2017a4 [MetaSchedule] Postproc: Rewrite-Layout (#11884)
add 98bf40fc0a [Relay] [PyTorch] Add aten::broadcast_tensors (#11863)
add 59fb4219ff [LLVM] Include LLVM headers in files that use them, not in
llvm_common.h (#11888)
add 6d6e070587 [AOT] Calculate used memory at the callsite of primitive
functions (#11208)
add 0dea2435cc [ci][docker] Nightly Docker image update
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 (02a67184a3)
\
N -- N -- N refs/heads/nightly-docker-update (0dea2435cc)
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.
No new revisions were added by this update.
Summary of changes:
include/tvm/meta_schedule/postproc.h | 7 +-
include/tvm/relay/transform.h | 13 +
python/tvm/auto_scheduler/__init__.py | 65 +--
python/tvm/meta_schedule/default_config.py | 1 +
python/tvm/meta_schedule/postproc/__init__.py | 5 +-
.../{verify_gpu_code.py => rewrite_layout.py} | 11 +-
python/tvm/relay/frontend/pytorch.py | 8 +
src/meta_schedule/postproc/rewrite_layout.cc | 183 +++++++++
src/relay/backend/annotate_used_memory.cc | 233 +++++++++++
src/relay/backend/aot_executor_codegen.cc | 2 +
src/relay/backend/liveness_analysis.cc | 232 +++++++++++
src/relay/backend/liveness_analysis.h | 270 +++++++++++++
src/relay/backend/vm/manifest_lifetimes.cc | 388 +-----------------
src/target/llvm/codegen_amdgpu.cc | 41 +-
src/target/llvm/codegen_arm.cc | 14 +-
src/target/llvm/codegen_blob.cc | 26 ++
src/target/llvm/codegen_blob.h | 8 +-
src/target/llvm/codegen_cpu.cc | 31 ++
src/target/llvm/codegen_cpu.h | 22 ++
src/target/llvm/codegen_hexagon.cc | 32 +-
src/target/llvm/codegen_llvm.cc | 63 ++-
src/target/llvm/codegen_llvm.h | 48 ++-
src/target/llvm/codegen_nvptx.cc | 45 ++-
src/target/llvm/codegen_params.cc | 8 +-
src/target/llvm/codegen_params.h | 7 +-
src/target/llvm/codegen_x86_64.cc | 17 +-
src/target/llvm/intrin_rule_hexagon.cc | 1 +
src/target/llvm/intrin_rule_llvm.cc | 1 +
src/target/llvm/intrin_rule_llvm.h | 5 +-
src/target/llvm/llvm_common.cc | 17 +
src/target/llvm/llvm_common.h | 58 +--
src/target/llvm/llvm_module.cc | 21 +
tests/python/frontend/pytorch/test_forward.py | 22 ++
tests/python/relay/test_used_memory_annotator.py | 434 +++++++++++++++++++++
.../test_meta_schedule_postproc_rewrite_layout.py | 91 +++++
35 files changed, 1921 insertions(+), 509 deletions(-)
copy python/tvm/meta_schedule/postproc/{verify_gpu_code.py =>
rewrite_layout.py} (78%)
create mode 100644 src/meta_schedule/postproc/rewrite_layout.cc
create mode 100644 src/relay/backend/annotate_used_memory.cc
create mode 100644 src/relay/backend/liveness_analysis.cc
create mode 100644 src/relay/backend/liveness_analysis.h
create mode 100644 tests/python/relay/test_used_memory_annotator.py
create mode 100644
tests/python/unittest/test_meta_schedule_postproc_rewrite_layout.py
