This is an automated email from the ASF dual-hosted git repository.
tqchen pushed a change to branch unity-staging
in repository https://gitbox.apache.org/repos/asf/tvm.git
discard 2fe58003f8 [Unity][Pass] Operator legalization (#14029)
discard bf5ab8a8bb [Unity][TVMScript] Move tir/relax import in script out of
__init__.py (#14033)
discard 5b30ecef63 [Unity][Pass] Wellformed Analysis (#14032)
discard 5c565cbad7 [Unity][BlockBuilder] CallTE convert PrimValue args
(#14028)
discard 121f50ecae [Unity][Pass] Normalize Pass (#14031)
discard e0321a6583 [Unity] Relay -> Relax translator (#14026)
discard 75f133c813 [Unity][Pass][TuningAPI] Introduce TuningAPI and
MetaSchedule pass (#14014)
discard 7755b17338 [Unity][Pass] BindParams pass, FoldConstant pass (#14016)
discard e182fbcb14 [Unity][VM] Supporting "compiled" exec mode. (#14015)
discard 269e9551e6 [Unity][Pass] LambdaLift pass (#14012)
discard 93da12fcc4 [Unity][Pass] Operator Fusion Passes (#14001)
discard ea65c33747 [Unity] NestedMsg Support utility (#13995)
discard 194e3831d5 [Unity] Relax op: manipulation (#13989)
discard e2eff73148 [Unity] Relax op: search (#13992)
discard a68b668f31 [Unity] Relax op: linear algebra (#13988)
discard 1804d60b3e [Unity] Relax op: creation (#13984)
discard d187342ced [Unity] Relax op: neural networks (#13993)
discard 37041fec40 [Unity] Relax op: statistical (#13991)
discard 409c9d7c0c [Unity] Relax op: arithmetic, comparison (#13983)
discard 4e7e320776 [Unity] Relax op: image (#13994)
discard f069a1689b [Unity] Relax op: set (#13990)
discard 7efa6e674f [Unity] Relax op: datatype (#13986)
discard ddf3c2acc6 [Unity] Relax op: index (#13987)
discard 50aeca7b52 [Unity][TVMScript] Use explicit `R.shape` in TVMScript
(#13979)
discard b97501c9cc [Unity] e2e Relax minimum build flow (#13961)
discard 9f91acbc7c [Unity] Relax VM shape lowering pass (#13956)
discard 7854001f04 [Unity] Relax VM codegen (#13954)
discard 97bcff2b43 [Unity] Relax TVMScript Printer (#13944)
discard 5027cf108f [Unity] Relax TVMScript Parser. (#13932)
discard 9472b0524a [Unity] Relax BlockBuilder and ExprMutator (#13926)
discard 07a9b332b3 [Unity] Basic StructInfo Analysis and Expr construction
(#13916)
discard fd53ef5c5f [Unity][CI] Unity specific jenkins setup (do not upstream
to main) (#13910)
discard b5d6978de5 [Unity][IR] First-class StructInfo (#13907)
discard 93bb293ed5 [Unity] Relax expressions and types (#13901)
discard 51ec06a8a6 [Unity] Relax VM (#13878)
add ac57b013b2 [TIR] Update block flags and simplify predicate in
Reverse-Compute-Inline (#14030)
add 8613c79aa0 [TIR] Enable Host Func Attribute for PrimFunc (#14020)
add 054c11e101 [TOPI] Fix index dtype in topi strided_slice (#14022)
add 87bb8b1be8 [TIR] Introduce Pass InjectPTXLDG32 (#13973)
new 8bc90d5ae1 [Unity] Relax VM (#13878)
new 09a15d78b2 [Unity] Relax expressions and types (#13901)
new d3ee97dd19 [Unity][IR] First-class StructInfo (#13907)
new df06448690 [Unity][CI] Unity specific jenkins setup (do not upstream
to main) (#13910)
new 22d3ca5aff [Unity] Basic StructInfo Analysis and Expr construction
(#13916)
new 98bb5282e5 [Unity] Relax BlockBuilder and ExprMutator (#13926)
new 66ffba8c95 [Unity] Relax TVMScript Parser. (#13932)
new eb026b74ea [Unity] Relax TVMScript Printer (#13944)
new 7dd45cc15e [Unity] Relax VM codegen (#13954)
new 998da7cf9c [Unity] Relax VM shape lowering pass (#13956)
new a86cc11c23 [Unity] e2e Relax minimum build flow (#13961)
new f197d45723 [Unity][TVMScript] Use explicit `R.shape` in TVMScript
(#13979)
new e6292a197c [Unity] Relax op: index (#13987)
new 6ed1b4bf72 [Unity] Relax op: datatype (#13986)
new 77a319e926 [Unity] Relax op: set (#13990)
new fcda9b6824 [Unity] Relax op: image (#13994)
new dbf121c293 [Unity] Relax op: arithmetic, comparison (#13983)
new c30576168b [Unity] Relax op: statistical (#13991)
new 8a04832e0e [Unity] Relax op: neural networks (#13993)
new b2b9e1278d [Unity] Relax op: creation (#13984)
new 323e99ce93 [Unity] Relax op: linear algebra (#13988)
new d805689640 [Unity] Relax op: search (#13992)
new ad1f8ac70b [Unity] Relax op: manipulation (#13989)
new d62e5b4419 [Unity] NestedMsg Support utility (#13995)
new 3f385aba94 [Unity][Pass] Operator Fusion Passes (#14001)
new c1902faa0a [Unity][Pass] LambdaLift pass (#14012)
new 99eef364ef [Unity][VM] Supporting "compiled" exec mode. (#14015)
new a48aa6bd7a [Unity][Pass] BindParams pass, FoldConstant pass (#14016)
new 54e15a94c7 [Unity][Pass][TuningAPI] Introduce TuningAPI and
MetaSchedule pass (#14014)
new 0bab3d400a [Unity] Relay -> Relax translator (#14026)
new d56e04c8cc [Unity][Pass] Normalize Pass (#14031)
new 170827f634 [Unity][BlockBuilder] CallTE convert PrimValue args
(#14028)
new e97060fd23 [Unity][Pass] Wellformed Analysis (#14032)
new d1d6cd784b [Unity][TVMScript] Move tir/relax import in script out of
__init__.py (#14033)
new 590d446863 [Unity][Pass] Operator legalization (#14029)
new d4c8f9bb0b [Unity][Op] Add ShapeExpr Tests for Reshape Op (#14035)
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 (2fe58003f8)
\
N -- N -- N refs/heads/unity-staging (d4c8f9bb0b)
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 36 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:
include/tvm/tir/builtin.h | 12 ++
include/tvm/tir/function.h | 7 ++
include/tvm/tir/transform.h | 6 +
include/tvm/topi/transform.h | 16 +--
src/driver/driver_api.cc | 12 ++
src/target/source/codegen_cuda.cc | 31 +++++
src/tir/op/builtin.cc | 2 +
src/tir/schedule/primitive/compute_inline.cc | 10 +-
src/tir/transforms/inject_ptx_ldg32.cc | 130 +++++++++++++++++++++
src/tir/transforms/primfunc_utils.cc | 4 +
tests/python/relax/test_op_manipulate.py | 9 ++
...ir_ptx_cp_async.py => test_inject_ptx_ldg32.py} | 56 ++++-----
tests/python/unittest/test_tir_host_func.py | 79 +++++++++++++
.../unittest/test_tir_schedule_compute_inline.py | 74 ++++++++++++
14 files changed, 413 insertions(+), 35 deletions(-)
create mode 100644 src/tir/transforms/inject_ptx_ldg32.cc
copy tests/python/unittest/{test_tir_ptx_cp_async.py =>
test_inject_ptx_ldg32.py} (55%)
create mode 100644 tests/python/unittest/test_tir_host_func.py
