Skip to content
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

Jump to bottom

TE with threading build #1092

Merged
merged 9 commits into from
Aug 12, 2024
Merged

TE with threading build #1092

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
4c760c8
added threading build back
phu0ngng Aug 9, 2024
a90721d
add options for pytorch and paddle extensions
phu0ngng Aug 9, 2024
81fea66
added messages
phu0ngng Aug 10, 2024
979b39b
Merge branch 'main' into build_threads
phu0ngng Aug 10, 2024
5f71ee7
Update transformer_engine/common/CMakeLists.txt
phu0ngng Aug 10, 2024
b7bc0c5
Merge branch 'main' into build_threads
phu0ngng Aug 10, 2024
8bc6bd2
renaming
phu0ngng Aug 12, 2024
e5e44e5
renaming
phu0ngng Aug 12, 2024
e8cbc5f
Merge branch 'main' into build_threads
phu0ngng Aug 12, 2024
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
3 changes: 2 additions & 1 deletion build_tools/paddle.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@
from pathlib import Path

import setuptools
import os

from .utils import cuda_version

Expand Down Expand Up @@ -62,7 +63,7 @@ def setup_paddle_extension(
print("Could not determine CUDA Toolkit version")
else:
if version >= (11, 2):
nvcc_flags.extend(["--threads", "4"])
nvcc_flags.extend(["--threads", os.getenv("NVTE_BUILD_THREADS_PER_JOB", "1")])
if version >= (11, 0):
nvcc_flags.extend(["-gencode", "arch=compute_80,code=sm_80"])
if version >= (11, 8):
Expand Down
2 changes: 1 addition & 1 deletion build_tools/pytorch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,7 +68,7 @@ def setup_pytorch_extension(
print("Could not determine CUDA Toolkit version")
else:
if version >= (11, 2):
nvcc_flags.extend(["--threads", "4"])
nvcc_flags.extend(["--threads", os.getenv("NVTE_BUILD_THREADS_PER_JOB", "1")])
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@phu0ngng could you document here the reason for switching from 4 to 1 as a default?

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

As @timmoon10 mentioned, users getting stuck in the build process (#976, #978, #1077).
Besides, the PyTorch Ext build with --thread 4 broke on my desktop. I discussed with Tim and we think it is better to have 1 as default so that all users have fewer build issues.
For "advanced" users, they are encouraged to play with NVTE_BUILD_THREADS_PER_JOB to best match their system.

if version >= (11, 0):
nvcc_flags.extend(["-gencode", "arch=compute_80,code=sm_80"])
if version >= (11, 8):
Expand Down
4 changes: 2 additions & 2 deletions build_tools/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,8 +37,8 @@ def get_max_jobs_for_parallel_build() -> int:
num_jobs = 0

# Check environment variable
if os.getenv("NVTE_MAX_BUILD_JOBS"):
num_jobs = int(os.getenv("NVTE_MAX_BUILD_JOBS"))
if os.getenv("NVTE_BUILD_MAX_JOBS"):
num_jobs = int(os.getenv("NVTE_BUILD_MAX_JOBS"))
elif os.getenv("MAX_JOBS"):
num_jobs = int(os.getenv("MAX_JOBS"))

Expand Down
16 changes: 16 additions & 0 deletions transformer_engine/common/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,22 @@ set(CMAKE_CUDA_STANDARD_REQUIRED ON)

project(transformer_engine LANGUAGES CUDA CXX)

set(BUILD_THREADS_PER_JOB $ENV{NVTE_BUILD_THREADS_PER_JOB})
if (NOT BUILD_THREADS_PER_JOB)
set(BUILD_THREADS_PER_JOB 1)
endif()
set(CMAKE_CUDA_FLAGS "${CMAKE_CUDA_FLAGS} --threads ${BUILD_THREADS_PER_JOB}")

if(DEFINED ENV{MAX_JOBS})
set(JOBS $ENV{MAX_JOBS})
elseif(DEFINED ENV{NVTE_BUILD_MAX_JOBS})
set(JOBS $ENV{NVTE_BUILD_MAX_JOBS})
else()
set(JOBS "max number of")
endif()

message(STATUS "Parallel build with ${JOBS} jobs and ${BUILD_THREADS_PER_JOB} threads per job")

if (CMAKE_BUILD_TYPE STREQUAL "Debug")
set(CMAKE_CUDA_FLAGS_DEBUG "${CMAKE_CUDA_FLAGS_DEBUG} -G")
endif()
Expand Down
Loading