Aditya K Kamath

Incorrect semantics in grid_barrier.cuh

2

In "/cub/grid/grid_barrier.cuh" I noticed some incorrect semantics being used. A flag "d_sync" is used to communicate between threadblocks and enforce syncing. In lines 98 and 120, LOAD_CG operations are used...

Data race in louvain_enactor.cuh

5

I found a data race in gunrock/app/louvain/louvain_enactor.cuh. In line 469 community_sizes[v] and community_sizes[n_comm] are being read, while in lines 482 and 483, atomic operations are performed on community_sizes, creating a...

Data race in sm_enactor.cuh

10

In sm_enactor.cuh there seems to be a data race between a load and an atomic operation. This occurs on lines 108 and 112. -->if (subgraphs[src] >= constrain[0]) { return true;...

MPI calls hang with random node allocation

2

MPI calls hang with random node allocation -------------------------- Assuming a topology has been set up (I used the default dragonfly), the following config causes the system to hang: ``` ep...

Race during insertion of key/values

8

I noticed that there were races in the code. For example, in line 45 of src/concurrent_map/warp: uint32_t src_unit_data = (next == SlabHashT::A_INDEX_POINTER) ? *(getPointerFromBucket(src_bucket, laneId)) : *(getPointerFromSlab(next, laneId)); Here a...