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.

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

[PHI]Fix cub bugs when use PHI shared lib in CUDA12 #57820

Merged
merged 5 commits into from
Oct 11, 2023

Conversation

YuanRisheng
Copy link
Contributor

@YuanRisheng YuanRisheng commented Sep 27, 2023

PR types

Bug fixes

PR changes

Others

Description

Pcard-67164
修复在cuda12环境下使用phi动态库,三方库cub造成的reduce精度问题,修复方案是使用nvidia cccl代码库来替代cub,顺便优化了creation.py的部分代码

@paddle-bot
Copy link

paddle-bot bot commented Sep 27, 2023

你的PR提交成功,感谢你对开源项目的贡献!
请关注后续CI自动化测试结果,详情请参考Paddle-CI手册
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

@paddle-ci-bot
Copy link

paddle-ci-bot bot commented Oct 6, 2023

Sorry to inform you that 780e10f's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually.

risemeup1
risemeup1 previously approved these changes Oct 10, 2023
Copy link
Contributor

@risemeup1 risemeup1 left a comment

Choose a reason for hiding this comment

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

LGTM

@@ -372,9 +372,12 @@ if(WITH_ONNXRUNTIME)
endif()

if(WITH_GPU)
if(${CMAKE_CUDA_COMPILER_VERSION} LESS 11.0)
if(${CMAKE_CUDA_COMPILER_VERSION} LESS 11.1)
Copy link
Contributor

Choose a reason for hiding this comment

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

image

前面是11.0的版本,这里是11.1,前后版本不一致会不会有问题?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

理论上排除不了cuda11.0有问题的可能性,不过从之前cub.cmake中的代码中可以推断cuda11.6以下使用cub1.8.0问题应该不大,而且我们没有cuda11.0的流水线机器去测试,先把代码合入吧,后续有问题再进一步解决
image

Copy link
Contributor Author

Choose a reason for hiding this comment

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

另外这里之所以改的原因是,cccl支持cuda11.1及以上版本

chenwhql
chenwhql previously approved these changes Oct 10, 2023
@YuanRisheng YuanRisheng dismissed stale reviews from chenwhql and risemeup1 via e7bf795 October 10, 2023 03:59
@YuanRisheng YuanRisheng merged commit 8044ca4 into PaddlePaddle:develop Oct 11, 2023
Frida-a pushed a commit to Frida-a/Paddle that referenced this pull request Oct 14, 2023
* fix cuda12 bugs

* fix windows bugs

* move cccl to third_party

* perfect code

* fix cuda bugs
jiahy0825 pushed a commit to jiahy0825/Paddle that referenced this pull request Oct 16, 2023
* fix cuda12 bugs

* fix windows bugs

* move cccl to third_party

* perfect code

* fix cuda bugs
danleifeng pushed a commit to danleifeng/Paddle that referenced this pull request Nov 14, 2023
* fix cuda12 bugs

* fix windows bugs

* move cccl to third_party

* perfect code

* fix cuda bugs
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants