changa 3.5
|
Data and parameters for requesting gravity calculations on the GPU. More...
#include <HostCUDA.h>
Public Attributes | |
cudaStream_t | stream |
CudaMultipoleMoments * | d_localMoments |
for accessing device memory | |
CudaMultipoleMoments * | d_remoteMoments |
CompactPartData * | d_localParts |
CompactPartData * | d_remoteParts |
VariablePartData * | d_localVars |
size_t | sMoments |
size_t | sCompactParts |
size_t | sVarParts |
void * | list |
can either be a ILCell* or an ILPart* | |
int * | bucketMarkers |
int * | bucketStarts |
int * | bucketSizes |
int | numInteractions |
int | numBucketsPlusOne |
void * | tp |
void * | missedNodes |
pointer to off-processor Node/Particle buffer. | |
void * | missedParts |
size_t | sMissed |
Size of the off-processor data buffer. | |
int * | affectedBuckets |
these buckets were finished in this work request | |
void * | cb |
void * | state |
cudatype | fperiod |
bool | node |
is this a node or particle computation request? | |
bool | remote |
is this a remote or local computation? | |
Data and parameters for requesting gravity calculations on the GPU.
int* _CudaRequest::bucketMarkers |
index in the cell or particle list for the cells/particles for a given bucket
int* _CudaRequest::bucketSizes |
number of particles in a bucket on the GPU
int* _CudaRequest::bucketStarts |
index in the particle array on the GPU of the bucket
void* _CudaRequest::cb |
Callback
cudatype _CudaRequest::fperiod |
Size of periodic volume
int _CudaRequest::numBucketsPlusOne |
Number of buckets affected by this request
int _CudaRequest::numInteractions |
Total number of interactions in this request
void* _CudaRequest::state |
Pointer to state of walk that created this request
cudaStream_t _CudaRequest::stream |
CUDA stream to handle memory operations and kernel launches for this request Allocation and deallocation of the stream is handled by the DataManager
void* _CudaRequest::tp |
Pointer to TreePiece that made this request