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

Jump to bottom

Lockfree Ripser #31

Open
wants to merge 42 commits into
base: master
Choose a base branch
from
Open

Lockfree Ripser #31

wants to merge 42 commits into from

Conversation

mrzv
Copy link

@mrzv mrzv commented Jun 24, 2020

@ubauer I'm not actually suggesting to merge this, but just FYI. It's a parallel version of Ripser, parallelized using the algorithm in this paper.

CC: @grey-narn

@mrzv
Switch compressed_sparse_matrix to vector<vector<ValueType>>
eaf5900
@mrzv
Add USE_DENSE_PIVOTS
7d2babc
@mrzv
Fix compilation with coefficients
35ee0dd
@mrzv
Make cofacet_entries static thread_local
10a06fa
@mrzv
Add trivial_concurrent_hash_map; for now doesn't work with coefficients
adeb456
@mrzv
Make entry_t = index_t, when using coefficients, to enable atomic_ref
1eca2b0
@mrzv
Undo over-reserving pivot_column_index
44f6581
@mrzv
Switch compressed_sparse_matrix to a vector of atomic pointers to vector
d923928
@mrzv
Read column and check correctness of the read
caec39e
@mrzv
Store coefficient in the value of the hash map, not the key
cbfcf29
@mrzv
Restructure the code for parallelism
0882d65
@mrzv
Add i/atomic_ref_serial.hpp, included via a flag
1047cb2
@mrzv
Add USE_PARALLEL_STL for-loop parallelization
7e21cf8
@mrzv
Add logic to move reduction to the right
7ddbf8c
@mrzv
Add shuffled serial and hand-rolled chunking in foreach
8c5764d
@mrzv
Don't add diagonal element to working_reduction_column, when continui…
f04a19b 
…ng reduction
@mrzv
Make simplex_coboundary_enumerator::neighbor_{it,end} cache thread_lo…
7711c40 
…cal + add thread sanitizer to debug build
@mrzv
Disable emergent pair check in parallel
eb20913
@mrzv
trivial_concurrent_hash_map::find() guarantees that a value has been …
665ff4f 
…written
@mrzv
Don't output progress and pairs in parallel
f62afd4
@mrzv
Add --threads argument + indicate progress in parallel
7ff922f
@mrzv
Enable emergent pairs optimization in parallel
de75624
@mrzv
Add memory reclamation (for hand-rolled parallel foreach)
780ff3b
@mrzv
Add pair output in the parallel case
5eea68e
@mrzv
Add MemoryManager to serial versions of foreach
d9e0120
@mrzv
Add MemoryManager to parallel STL version of foreach
d391d5c
@mrzv
Pass memory manager by reference
4a1ec6a
@mrzv
Add parallel assembly of columns
2de8ca1
@mrzv
Add parallel_sort from Boost
a1ece3d
@mrzv
Fix compilation with parallel STL
fdb9a41
mrzv added 12 commits March 17, 2020 12:16
@mrzv
Add missing INDICATE_PROGRESS guards
fbdc42e
@mrzv
Indicate empty columns by nullptr, rather than empty vector
a793274
@mrzv
Fix resize bug in trivial_concurrent_hash_map + bump multiplier to 10
0c7e27a
@mrzv
Add necessary function to hash_map based on unordered_map
0ae052b
@mrzv
Add USE_TBB_HASHMAP option
fa93b2b
@mrzv
Don't allocate second hash map, if not printing pairs
cb34f33
@mrzv
Use the next power of 2 for size in trivial_concurrent_hash_map + bum…
f2bcbbd 
…p multiplier down to 8
@mrzv
Add USE_TBB option
b283b1f
@mrzv
Tweak the Makefile: add more targets + automatic dependencies
4bf9361
@mrzv
USE_CONCURRENT_PIVOTS -> USE_TRIVIAL_CONCURRENT_HASHMAP + reorder #el…
ba9be8f 
…if to actually use TBB hashmap, when requested
@mrzv
Fix memory reclamation by extending the wait by an extra epoch
1264577
@mrzv
Add a link to the paper that describes the algorithm
625be2c
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
1 participant
@mrzv