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

Bug for strassen code handling the fringes when multi-threaded #7

Open
jianyuh opened this issue Nov 4, 2016 · 1 comment
Open

Bug for strassen code handling the fringes when multi-threaded #7

jianyuh opened this issue Nov 4, 2016 · 1 comment
Assignees
@jianyuh
Labels
bug enhancement

Comments

@jianyuh
Copy link
Collaborator

jianyuh commented Nov 4, 2016

The output is incorrect when m,n,k is odd and multi-thread is on.
@jianyuh jianyuh added the bug label Nov 4, 2016
@jianyuh jianyuh self-assigned this Nov 4, 2016
@jianyuh jianyuh changed the title Bug for handling the fringes when multi-threaded Bug for strassen code handling the fringes when multi-threaded Nov 4, 2016
ChenhanYu pushed a commit that referenced this issue Nov 4, 2016
Merge pull request #8 from SudoNohup/master
32c51ac 
Solve Issue #7. We need chief thread
@jianyuh
Copy link
Collaborator Author

jianyuh commented Nov 8, 2016

By Chenhan:
Your dynamic_peeling function should take the thread info in the future. We can implement master thread or chief thread.

My comment:
Currently dynamic_peeling function is placed inside the OMP parallel region. We can only allocate one thread working on this if we invoke xgemm.

Two possible solutions:

  1. place dynamic_peeling outside the parallel region.
  2. invoke gxmm instead of xgemm.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@jianyuh jianyuh

Labels
bug enhancement
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@jianyuh