You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@tvm.apache.org by gi...@apache.org on 2022/04/04 02:24:26 UTC
[tvm] branch last-successful updated (daa068979a -> 966d018da8)
This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a change to branch last-successful
in repository https://gitbox.apache.org/repos/asf/tvm.git
from daa068979a Fix a small timer bug. (#10875)
add afe6793fb0 Bump pyxir version tp v0.3.5 to avoid bad cleanup error with pyxir and tensorflow 2.6 (#10858)
add 966d018da8 [PTX] `ldmatrix` builtin to accelerate copying data from shared memory to warp memory (#10855)
No new revisions were added by this update.
Summary of changes:
.../install/ubuntu_install_vitis_ai_packages_ci.sh | 2 +-
include/tvm/tir/builtin.h | 9 ++
src/target/source/codegen_cuda.cc | 26 ++++-
src/target/source/{ptx_mma.cc => ptx.cc} | 126 +++++++++++++++++----
src/target/source/{ptx_mma.h => ptx.h} | 38 +++++--
src/tir/op/builtin.cc | 3 +
.../python/contrib/test_vitis_ai/infrastructure.py | 1 -
tests/python/unittest/test_tir_ptx_ldmatrix.py | 101 +++++++++++++++++
8 files changed, 264 insertions(+), 42 deletions(-)
rename src/target/source/{ptx_mma.cc => ptx.cc} (81%)
rename src/target/source/{ptx_mma.h => ptx.h} (63%)
create mode 100644 tests/python/unittest/test_tir_ptx_ldmatrix.py