Skip to content

Fix build vllm kernel for the first time may encounter failure #54

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
yeahdongcn merged 2 commits into from
Apr 12, 2026
Merged

Fix build vllm kernel for the first time may encounter failure #54

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3fbaae4
fix build vllm kernel for the first time may encounter failure (file …
zichen-kuuga Apr 10, 2026
2c88b67
Bump version
zichen-kuuga Apr 10, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -255,7 +255,7 @@ See `src/torchada/_mapping.py` for the complete mapping table (380+ mappings).

```
# pyproject.toml or requirements.txt
torchada>=0.1.48
torchada>=0.1.49
```

### Step 2: Conditional Import
Expand Down
2 changes: 1 addition & 1 deletion README_CN.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -255,7 +255,7 @@ if torchada.is_gpu_device(device): # 在 CUDA 和 MUSA 上都能工作

```
# pyproject.toml 或 requirements.txt
torchada>=0.1.48
torchada>=0.1.49
```

### 步骤 2：条件导入
Expand Down
2 changes: 1 addition & 1 deletion benchmarks/benchmark_history.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@
"description": "Historical benchmark results for torchada performance tracking",
"results": [
{
"version": "0.1.48",
"version": "0.1.49",
"date": "2026-01-29",
"platform": "MUSA",
"pytorch_version": "2.7.1",
Expand Down
2 changes: 1 addition & 1 deletion pyproject.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"

[project]
name = "torchada"
version = "0.1.48"
version = "0.1.49"
description = "Adapter package for torch_musa to act exactly like PyTorch CUDA"
readme = "README.md"
license = {text = "MIT"}
Expand Down
2 changes: 1 addition & 1 deletion src/torchada/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,7 +24,7 @@
from torch.utils.cpp_extension import CUDAExtension, BuildExtension, CUDA_HOME
"""

__version__ = "0.1.48"
__version__ = "0.1.49"

from . import cuda, utils
from ._patch import apply_patches, get_original_init_process_group, is_patched
Expand Down
5 changes: 4 additions & 1 deletion src/torchada/utils/cpp_extension.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -624,7 +624,10 @@ def run(self):
if needs_porting:
source_dir = os.path.dirname(os.path.abspath(source))
dirs_to_port.add(source_dir)

# Sort directories by depth (deepest first) to ensure proper porting order
dirs_to_port = sorted(
dirs_to_port, key=lambda p: p.count("/"), reverse=True
)
# Port each unique directory
for cuda_dir in dirs_to_port:
self._port_directory(cuda_dir, mapping_rule)
Expand Down