frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-21 05:34:18 +08:00
Code Issues Packages Projects Releases Wiki Activity

Nccl update to 2.25.1 for cuda 12.4-12.8 (#146073)

Browse Source 
Should resolve: https://github.com/pytorch/pytorch/issues/144768
We use one common nccl version for cuda builds 12.4-12.8 : ``NCCL_VERSION=v2.25.1-1``
For CUDA 11.8 we use legacy ``NCCL_VERSION=v2.21.1-1``
We use pinned version of NCCL rather then submodule.
Move nccl location from ``third_party/nccl/nccl`` to ``third_party/nccl``

Pull Request resolved: https://github.com/pytorch/pytorch/pull/146073
Approved by: https://github.com/Skylion007, https://github.com/malfet, https://github.com/kwen2501, https://github.com/fduwjj
This commit is contained in:
atalman
2025-02-14 21:23:19 +00:00
committed by PyTorch MergeBot
parent d38db94689
commit eecee5863e
19 changed files with 137 additions and 111 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
tools/build_pytorch_libs.py
View File

@ -2,12 +2,18 @@ from __future__ import annotations
import os
import platform
import subprocess
from glob import glob
from pathlib import Path
from .setup_helpers.cmake import CMake, USE_NINJA
from .setup_helpers.env import check_negative_env_flag, IS_64BIT, IS_WINDOWS
repo_root = Path(__file__).absolute().parent.parent
third_party_path = os.path.join(repo_root, "third_party")
def _get_vc_env(vc_arch: str) -> dict[str, str]:
try:
from setuptools import distutils # type: ignore[import]
@ -79,6 +85,31 @@ def _create_build_env() -> dict[str, str]:
return my_env
def read_nccl_pin() -> str:
nccl_file = "nccl-cu12.txt"
if os.getenv("DESIRED_CUDA", "").startswith("11") or os.getenv(
"CUDA_VERSION", ""
).startswith("11"):
nccl_file = "nccl-cu11.txt"
nccl_pin_path = os.path.join(
repo_root, ".ci", "docker", "ci_commit_pins", nccl_file
)
with open(nccl_pin_path) as f:
return f.read().strip()
def checkout_nccl() -> None:
release_tag = read_nccl_pin()
print(f"-- Checkout nccl release tag: {release_tag}")
nccl_basedir = os.path.join(third_party_path, "nccl")
if not os.path.exists(nccl_basedir):
subprocess.check_call(
["git", "clone", "https://github.com/NVIDIA/nccl.git", "nccl"],
cwd=third_party_path,
)
subprocess.check_call(["git", "checkout", release_tag], cwd=nccl_basedir)
def build_pytorch(
version: str | None,
cmake_python_library: str | None,
@ -88,6 +119,7 @@ def build_pytorch(
cmake: CMake,
) -> None:
my_env = _create_build_env()
checkout_nccl()
build_test = not check_negative_env_flag("BUILD_TEST")
cmake.generate(
version, cmake_python_library, build_python, build_test, my_env, rerun_cmake