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 (develop branch only): multiple tests are now failing when forcing uvm usage. #570

Closed
crtrott opened this issue Dec 8, 2016 · 4 comments
Closed

Bug (develop branch only): multiple tests are now failing when forcing uvm usage. #570

crtrott opened this issue Dec 8, 2016 · 4 comments
Assignees
@hcedwar @crtrott
Labels
Bug Broken / incorrect code; it could be Kokkos' responsibility, or others’ (e.g., Trilinos)
Milestone
END 2016

Comments

@crtrott
Copy link
Member

crtrott commented Dec 8, 2016

We have now multiple tests failing with UVM being enabled:

generate_makefile.bash --with-cuda --with-openmp --with-cuda-options=force_uvm

[  FAILED  ] 21 tests, listed below:
[  FAILED  ] cuda.team_shared_request
[  FAILED  ] cuda.multi_level_scratch
[  FAILED  ] cuda.team_vector
[  FAILED  ] cuda.atomics
[  FAILED  ] cuda.atomic_operations
[  FAILED  ] cuda.long_reduce
[  FAILED  ] cuda.double_reduce
[  FAILED  ] cuda.long_reduce_dynamic
[  FAILED  ] cuda.double_reduce_dynamic
[  FAILED  ] cuda.long_reduce_dynamic_view
[  FAILED  ] cuda.scan
[  FAILED  ] cuda.team_scan
[  FAILED  ] cuda.team_long_reduce
[  FAILED  ] cuda.team_double_reduce
[  FAILED  ] cuda.reduction_deduction
[  FAILED  ] cuda.reducers
[  FAILED  ] cuda.md_range
[  FAILED  ] cuda.range_tag
[  FAILED  ] cuda.compiler_macros
[  FAILED  ] cuda.memory_pool
[  FAILED  ] cuda.tile_layout

This does not happen on master.
@crtrott crtrott added this to the END 2016 milestone Dec 8, 2016
@crtrott crtrott self-assigned this Dec 8, 2016
@crtrott crtrott added the Bug Broken / incorrect code; it could be Kokkos' responsibility, or others’ (e.g., Trilinos) label Dec 8, 2016
@crtrott
Copy link
Member Author

crtrott commented Dec 8, 2016

42087a8 is good.

@crtrott
Copy link
Member Author

crtrott commented Dec 8, 2016

bd389a7 bad
42087a8 good

@crtrott
Copy link
Member Author

crtrott commented Dec 8, 2016

So its the thing with the memory access matrix which screws it up.

hcedwar added a commit that referenced this issue Dec 8, 2016
@hcedwar
Fix (#570) memory accessibility matrix metafunction when using force-UVM
8fa71fe 
	--with-cuda-options=force_uvm
@crtrott
Copy link
Member Author

crtrott commented Dec 9, 2016

Args had a bugfix as well but had to go to a meeting.

@crtrott crtrott closed this as completed Dec 16, 2016
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@hcedwar hcedwar

@crtrott crtrott

Labels
Bug Broken / incorrect code; it could be Kokkos' responsibility, or others’ (e.g., Trilinos)
Projects
None yet
Milestone
END 2016
Development

No branches or pull requests
2 participants
@hcedwar @crtrott