Bart/bullet3
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

remove clFinish and copyToHost from GPU joint solver, performance is looking better now.

Browse Source
This commit is contained in:
erwin coumans
2013-07-09 11:19:16 -07:00
parent c4375a09e4
commit bd2cd7b2a7
3 changed files with 22 additions and 22 deletions
Download Patch File Download Diff File Expand all files Collapse all files

26
src/Bullet3OpenCL/RigidBody/b3GpuPgsJacobiSolver.cpp
View File

@@ -207,7 +207,7 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlySetup(b3OpenCLArray<b3Rigi
launcher.setBuffer(gpuBodies->getBufferCL());
launcher.setConst(numBodies);
launcher.launch1D(numBodies);
clFinish(m_gpuData->m_queue);
//clFinish(m_gpuData->m_queue);
// m_gpuData->m_gpuSolverBodies->copyToHost(m_tmpSolverBodyPool);
} else
@@ -264,20 +264,20 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlySetup(b3OpenCLArray<b3Rigi
launcher.setConst(numConstraints);
launcher.launch1D(numConstraints);
}
clFinish(m_gpuData->m_queue);
//clFinish(m_gpuData->m_queue);
if (batches.size()==0)
m_gpuData->m_gpuBatchConstraints->copyToHost(batchConstraints);
if (1)
{
m_gpuData->m_gpuConstraintInfo1->copyToHost(m_tmpConstraintSizesPool);
//m_gpuData->m_gpuConstraintInfo1->copyToHost(m_tmpConstraintSizesPool);
b3OpenCLArray<unsigned int> dst(m_gpuData->m_context,m_gpuData->m_queue);
dst.resize(numConstraints);
unsigned int total=0;
m_gpuData->m_prefixScan->execute(*m_gpuData->m_gpuConstraintInfo1,dst,numConstraints,&total);
unsigned int lastElem = m_gpuData->m_gpuConstraintInfo1->at(numConstraints-1);
b3AlignedObjectArray<unsigned int> dstHost;
dst.copyToHost(dstHost);
//b3AlignedObjectArray<unsigned int> dstHost;
//dst.copyToHost(dstHost);
totalNumRows = total+lastElem;
{
@@ -287,7 +287,7 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlySetup(b3OpenCLArray<b3Rigi
launcher.setBuffer(m_gpuData->m_gpuBatchConstraints->getBufferCL());
launcher.setConst(numConstraints);
launcher.launch1D(numConstraints);
clFinish(m_gpuData->m_queue);
//clFinish(m_gpuData->m_queue);
}
if (batches.size()==0)
m_gpuData->m_gpuBatchConstraints->copyToHost(batchConstraints);
@@ -346,12 +346,12 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlySetup(b3OpenCLArray<b3Rigi
launcher.setConst(infoGlobal.m_numIterations);
launcher.setConst(numConstraints);
launcher.launch1D(numConstraints);
clFinish(m_gpuData->m_queue);
//clFinish(m_gpuData->m_queue);
if (batches.size()==0)
m_gpuData->m_gpuBatchConstraints->copyToHost(batchConstraints);
//m_gpuData->m_gpuConstraintRows->copyToHost(verify);
m_gpuData->m_gpuConstraintRows->copyToHost(m_tmpSolverNonContactConstraintPool);
//m_gpuData->m_gpuConstraintRows->copyToHost(m_tmpSolverNonContactConstraintPool);
@@ -645,7 +645,7 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlyIterations(b3OpenCLArray<b
{
//only create the batches once.
//@todo: incrementally update batches when constraints are added/activated and/or removed/deactivated
bool createBatches = true;//batches.size()==0;
bool createBatches = batches.size()==0;
{
B3_PROFILE("GpuSolveGroupCacheFriendlyIterations");
if (createBatches)
@@ -701,7 +701,7 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlyIterations(b3OpenCLArray<b
launcher.setConst(numConstraintsInBatch);
launcher.launch1D(numConstraintsInBatch);
clFinish(m_gpuData->m_queue);
//clFinish(m_gpuData->m_queue);
} else
{
@@ -740,8 +740,8 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlyIterations(b3OpenCLArray<b
if (useGpu)
{
B3_PROFILE("copy to host");
m_gpuData->m_gpuSolverBodies->copyToHost(m_tmpSolverBodyPool);
//B3_PROFILE("copy to host");
//m_gpuData->m_gpuSolverBodies->copyToHost(m_tmpSolverBodyPool);
}
//int sz = sizeof(b3GpuSolverBody);
//printf("cpu sizeof(b3GpuSolverBody)=%d\n",sz);
@@ -971,7 +971,7 @@ b3Scalar b3GpuPgsJacobiSolver::solveGroupCacheFriendlyFinish(b3OpenCLArray<b3Rig
launcher.setBuffer(m_gpuData->m_gpuSolverBodies->getBufferCL());
launcher.setConst(numBodies);
launcher.launch1D(numBodies);
clFinish(m_gpuData->m_queue);
//clFinish(m_gpuData->m_queue);
// m_gpuData->m_gpuSolverBodies->copyToHost(m_tmpSolverBodyPool);
// m_gpuData->m_gpuBodies->copyToHostPointer(bodies,numBodies);
//m_gpuData->m_gpuBodies->copyToHost(testBodies);