205 lines
8.4 KiB
C++
205 lines
8.4 KiB
C++
/*************************************************************************
|
|
* *
|
|
* Open Dynamics Engine, Copyright (C) 2001,2002 Russell L. Smith. *
|
|
* All rights reserved. Email: russ@q12.org Web: www.q12.org *
|
|
* *
|
|
* This library is free software; you can redistribute it and/or *
|
|
* modify it under the terms of EITHER: *
|
|
* (1) The GNU Lesser General Public License as published by the Free *
|
|
* Software Foundation; either version 2.1 of the License, or (at *
|
|
* your option) any later version. The text of the GNU Lesser *
|
|
* General Public License is included with this library in the *
|
|
* file LICENSE.TXT. *
|
|
* (2) The BSD-style license that is included with this library in *
|
|
* the file LICENSE-BSD.TXT. *
|
|
* *
|
|
* This library is distributed in the hope that it will be useful, *
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of *
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the files *
|
|
* LICENSE.TXT and LICENSE-BSD.TXT for more details. *
|
|
* *
|
|
*************************************************************************/
|
|
|
|
/*
|
|
* Vector scaling related code of ThreadedEquationSolverLDLT
|
|
* Copyright (c) 2017-2024 Oleh Derevenko, odar@eleks.com (change all "a" to "e")
|
|
*/
|
|
|
|
|
|
#include <ode/common.h>
|
|
#include <ode/matrix.h>
|
|
#include <ode/matrix_coop.h>
|
|
#include "config.h"
|
|
#include "threaded_solver_ldlt.h"
|
|
#include "threading_base.h"
|
|
#include "resource_control.h"
|
|
#include "error.h"
|
|
|
|
#include "fastvecscale_impl.h"
|
|
|
|
|
|
/*static */
|
|
void ThreadedEquationSolverLDLT::estimateCooperativeScalingVectorResourceRequirements(
|
|
dxResourceRequirementDescriptor *summaryRequirementsDescriptor,
|
|
unsigned allowedThreadCount, unsigned elementCount)
|
|
{
|
|
dxThreadingBase *threading = summaryRequirementsDescriptor->getrelatedThreading();
|
|
unsigned limitedThreadCount = restrictScalingVectorAllowedThreadCount(threading, allowedThreadCount, elementCount);
|
|
|
|
if (limitedThreadCount > 1)
|
|
{
|
|
doEstimateCooperativeScalingVectorResourceRequirementsValidated(summaryRequirementsDescriptor, allowedThreadCount, elementCount);
|
|
}
|
|
}
|
|
|
|
/*static */
|
|
void ThreadedEquationSolverLDLT::cooperativelyScaleVector(dxRequiredResourceContainer *resourceContainer, unsigned allowedThreadCount,
|
|
dReal *vectorData, const dReal *scaleData, unsigned elementCount)
|
|
{
|
|
dAASSERT(elementCount != 0);
|
|
|
|
dxThreadingBase *threading = resourceContainer->getThreadingInstance();
|
|
unsigned limitedThreadCount = restrictScalingVectorAllowedThreadCount(threading, allowedThreadCount, elementCount);
|
|
|
|
if (limitedThreadCount <= 1)
|
|
{
|
|
scaleLargeVector<SV_A_STRIDE, SV_D_STRIDE>(vectorData, scaleData, elementCount);
|
|
}
|
|
else
|
|
{
|
|
doCooperativelyScaleVectorValidated(resourceContainer, limitedThreadCount, vectorData, scaleData, elementCount);
|
|
}
|
|
}
|
|
|
|
/*static */
|
|
unsigned ThreadedEquationSolverLDLT::restrictScalingVectorAllowedThreadCount(
|
|
dxThreadingBase *threading, unsigned allowedThreadCount, unsigned elementCount)
|
|
{
|
|
unsigned limitedThreadCount = 1;
|
|
|
|
#if dCOOPERATIVE_ENABLED
|
|
const unsigned int blockStep = SV_BLOCK_SIZE; // Required by the implementation
|
|
unsigned scalingBlockCount = deriveScalingVectorBlockCount(elementCount, blockStep);
|
|
dIASSERT(deriveScalingVectorThreadCount(SV_COOPERATIVE_BLOCK_COUNT_MINIMUM - 1, 2) > 1);
|
|
|
|
if (scalingBlockCount >= SV_COOPERATIVE_BLOCK_COUNT_MINIMUM)
|
|
{
|
|
limitedThreadCount = threading->calculateThreadingLimitedThreadCount(allowedThreadCount, true);
|
|
}
|
|
#endif // #if dCOOPERATIVE_ENABLED
|
|
|
|
return limitedThreadCount;
|
|
}
|
|
|
|
/*static */
|
|
void ThreadedEquationSolverLDLT::doEstimateCooperativeScalingVectorResourceRequirementsValidated(
|
|
dxResourceRequirementDescriptor *summaryRequirementsDescriptor,
|
|
unsigned allowedThreadCount, unsigned elementCount)
|
|
{
|
|
unsigned simultaneousCallCount = 1 + (allowedThreadCount - 1);
|
|
|
|
sizeint scalingMemoryRequired = 0;
|
|
const unsigned scalingAlignmentRequired = 0;
|
|
|
|
unsigned featureRequirement = dxResourceRequirementDescriptor::STOCK_CALLWAIT_REQUIRED;
|
|
summaryRequirementsDescriptor->mergeAnotherDescriptorIn(scalingMemoryRequired, scalingAlignmentRequired, simultaneousCallCount, featureRequirement);
|
|
}
|
|
|
|
/*static */
|
|
void ThreadedEquationSolverLDLT::doCooperativelyScaleVectorValidated(
|
|
dxRequiredResourceContainer *resourceContainer, unsigned allowedThreadCount,
|
|
dReal *vectorData, const dReal *scaleData, unsigned elementCount)
|
|
{
|
|
dIASSERT(allowedThreadCount > 1);
|
|
|
|
const unsigned int blockStep = SV_BLOCK_SIZE; // Required by the implementation
|
|
unsigned scalingBlockCount = deriveScalingVectorBlockCount(elementCount, blockStep);
|
|
dIASSERT(scalingBlockCount > 0U);
|
|
|
|
unsigned threadCountToUse = deriveScalingVectorThreadCount(scalingBlockCount - 1, allowedThreadCount);
|
|
dIASSERT(threadCountToUse > 1);
|
|
|
|
dCallWaitID completionWait = resourceContainer->getStockCallWait();
|
|
dAASSERT(completionWait != NULL);
|
|
|
|
atomicord32 blockCompletionProgress;
|
|
|
|
initializeCooperativelyScaleVectorMemoryStructures(blockCompletionProgress);
|
|
|
|
dCallReleaseeID calculationFinishReleasee;
|
|
ScaleVectorWorkerContext workerContext; // The variable must exist in the outer scope
|
|
|
|
workerContext.init(vectorData, scaleData, elementCount, blockCompletionProgress);
|
|
|
|
dxThreadingBase *threading = resourceContainer->getThreadingInstance();
|
|
threading->PostThreadedCall(NULL, &calculationFinishReleasee, threadCountToUse - 1, NULL, completionWait, &scaleVector_completion_callback, NULL, 0, "ScaleVector Completion");
|
|
threading->PostThreadedCallsGroup(NULL, threadCountToUse - 1, calculationFinishReleasee, &scaleVector_worker_callback, &workerContext, "ScaleVector Work");
|
|
|
|
participateScalingVector<blockStep, SV_A_STRIDE, SV_D_STRIDE>(vectorData, scaleData, elementCount, blockCompletionProgress);
|
|
|
|
threading->WaitThreadedCallExclusively(NULL, completionWait, NULL, "ScaleVector End Wait");
|
|
}
|
|
|
|
|
|
/*static */
|
|
int ThreadedEquationSolverLDLT::scaleVector_worker_callback(void *callContext, dcallindex_t dUNUSED(callInstanceIndex), dCallReleaseeID dUNUSED(callThisReleasee))
|
|
{
|
|
ScaleVectorWorkerContext *ptrContext = (ScaleVectorWorkerContext *)callContext;
|
|
|
|
scaleVector_worker(*ptrContext);
|
|
|
|
return 1;
|
|
}
|
|
|
|
/*static */
|
|
void ThreadedEquationSolverLDLT::scaleVector_worker(ScaleVectorWorkerContext &ref_context)
|
|
{
|
|
const unsigned blockStep = SV_BLOCK_SIZE;
|
|
|
|
participateScalingVector<blockStep, SV_A_STRIDE, SV_D_STRIDE>(ref_context.m_vectorData, ref_context.m_scaleData, ref_context.m_elementCount, *ref_context.m_ptrBlockCompletionProgress);
|
|
}
|
|
|
|
/*static */
|
|
int ThreadedEquationSolverLDLT::scaleVector_completion_callback(void *dUNUSED(callContext), dcallindex_t dUNUSED(callInstanceIndex), dCallReleaseeID dUNUSED(callThisReleasee))
|
|
{
|
|
return 1;
|
|
}
|
|
|
|
|
|
//////////////////////////////////////////////////////////////////////////
|
|
// Public interface functions
|
|
|
|
/*extern ODE_API */
|
|
void dScaleVector(dReal *a, const dReal *d, int n)
|
|
{
|
|
scaleLargeVector<1, 1>(a, d, n);
|
|
}
|
|
|
|
/*extern ODE_API_DEPRECATED ODE_API */
|
|
void dVectorScale(dReal *a, const dReal *d, int n)
|
|
{
|
|
scaleLargeVector<1, 1>(a, d, n);
|
|
}
|
|
|
|
|
|
/*extern ODE_API */
|
|
void dEstimateCooperativelyScaleVectorResourceRequirements(dResourceRequirementsID requirements,
|
|
unsigned maximalAllowedThreadCount, unsigned maximalElementCount)
|
|
{
|
|
dAASSERT(requirements != NULL);
|
|
|
|
dxResourceRequirementDescriptor *requirementsDescriptor = (dxResourceRequirementDescriptor *)requirements;
|
|
ThreadedEquationSolverLDLT::estimateCooperativeScalingVectorResourceRequirements(requirementsDescriptor, maximalAllowedThreadCount, maximalElementCount);
|
|
}
|
|
|
|
/*extern ODE_API */
|
|
void dCooperativelyScaleVector(dResourceContainerID resources, unsigned allowedThreadCount,
|
|
dReal *dataVector, const dReal *scaleVector, unsigned elementCount)
|
|
{
|
|
dAASSERT(resources != NULL);
|
|
|
|
dxRequiredResourceContainer *resourceContainer = (dxRequiredResourceContainer *)resources;
|
|
ThreadedEquationSolverLDLT::cooperativelyScaleVector(resourceContainer, allowedThreadCount, dataVector, scaleVector, elementCount);
|
|
}
|
|
|