void SpringEmbedderFRExact::mainStep_sse3(ArrayGraph &C)
{
//#if (defined(OGDF_ARCH_X86) || defined(OGDF_ARCH_X64)) && !(defined(__GNUC__) && !defined(__SSE3__))
#ifdef OGDF_SSE3_EXTENSIONS
const int n = C.numberOfNodes();
#ifdef _OPENMP
const int work = 256;
const int nThreadsRep = min(omp_get_max_threads(), 1 + n*n/work);
const int nThreadsPrev = min(omp_get_max_threads(), 1 + n /work);
#endif
const double k = m_idealEdgeLength;
const double kSquare = k*k;
const double c_rep = 0.052 * kSquare; // 0.2 = factor for repulsive forces as suggested by Warshal
const double minDist = 10e-6;//100*DBL_EPSILON;
const double minDistSquare = minDist*minDist;
double *disp_x = (double*) System::alignedMemoryAlloc16(n*sizeof(double));
double *disp_y = (double*) System::alignedMemoryAlloc16(n*sizeof(double));
__m128d mm_kSquare = _mm_set1_pd(kSquare);
__m128d mm_minDist = _mm_set1_pd(minDist);
__m128d mm_minDistSquare = _mm_set1_pd(minDistSquare);
__m128d mm_c_rep = _mm_set1_pd(c_rep);
#pragma omp parallel num_threads(nThreadsRep)
{
double tx = m_txNull;
double ty = m_tyNull;
int cF = 1;
for(int i = 1; i <= m_iterations; i++)
{
// repulsive forces
#pragma omp for
for(int v = 0; v < n; ++v)
{
__m128d mm_disp_xv = _mm_setzero_pd();
__m128d mm_disp_yv = _mm_setzero_pd();
__m128d mm_xv = _mm_set1_pd(C.m_x[v]);
__m128d mm_yv = _mm_set1_pd(C.m_y[v]);
int u;
for(u = 0; u+1 < v; u += 2)
{
__m128d mm_delta_x = _mm_sub_pd(mm_xv, _mm_load_pd(&C.m_x[u]));
__m128d mm_delta_y = _mm_sub_pd(mm_yv, _mm_load_pd(&C.m_y[u]));
__m128d mm_distSquare = _mm_max_pd(mm_minDistSquare,
_mm_add_pd(_mm_mul_pd(mm_delta_x,mm_delta_x),_mm_mul_pd(mm_delta_y,mm_delta_y))
);
__m128d mm_t = _mm_div_pd(_mm_load_pd(&C.m_nodeWeight[u]), mm_distSquare);
mm_disp_xv = _mm_add_pd(mm_disp_xv, _mm_mul_pd(mm_delta_x, mm_t));
mm_disp_yv = _mm_add_pd(mm_disp_yv, _mm_mul_pd(mm_delta_y, mm_t));
//mm_disp_xv = _mm_add_pd(mm_disp_xv, _mm_mul_pd(mm_delta_x, _mm_div_pd(mm_kSquare,mm_distSquare)));
//mm_disp_yv = _mm_add_pd(mm_disp_yv, _mm_mul_pd(mm_delta_y, _mm_div_pd(mm_kSquare,mm_distSquare)));
}
int uStart = u+2;
if(u == v) ++u;
if(u < n) {
__m128d mm_delta_x = _mm_sub_sd(mm_xv, _mm_load_sd(&C.m_x[u]));
__m128d mm_delta_y = _mm_sub_sd(mm_yv, _mm_load_sd(&C.m_y[u]));
__m128d mm_distSquare = _mm_max_sd(mm_minDistSquare,
_mm_add_sd(_mm_mul_sd(mm_delta_x,mm_delta_x),_mm_mul_sd(mm_delta_y,mm_delta_y))
);
__m128d mm_t = _mm_div_sd(_mm_load_sd(&C.m_nodeWeight[u]), mm_distSquare);
mm_disp_xv = _mm_add_sd(mm_disp_xv, _mm_mul_sd(mm_delta_x, mm_t));
mm_disp_yv = _mm_add_sd(mm_disp_yv, _mm_mul_sd(mm_delta_y, mm_t));
//mm_disp_xv = _mm_add_sd(mm_disp_xv, _mm_mul_sd(mm_delta_x, _mm_div_sd(mm_kSquare,mm_distSquare)));
//mm_disp_yv = _mm_add_sd(mm_disp_yv, _mm_mul_sd(mm_delta_y, _mm_div_sd(mm_kSquare,mm_distSquare)));
}
for(u = uStart; u < n; u += 2)
{
__m128d mm_delta_x = _mm_sub_pd(mm_xv, _mm_load_pd(&C.m_x[u]));
__m128d mm_delta_y = _mm_sub_pd(mm_yv, _mm_load_pd(&C.m_y[u]));
__m128d mm_distSquare = _mm_max_pd(mm_minDistSquare,
_mm_add_pd(_mm_mul_pd(mm_delta_x,mm_delta_x),_mm_mul_pd(mm_delta_y,mm_delta_y))
);
__m128d mm_t = _mm_div_pd(_mm_load_pd(&C.m_nodeWeight[u]), mm_distSquare);
mm_disp_xv = _mm_add_pd(mm_disp_xv, _mm_mul_pd(mm_delta_x, mm_t));
mm_disp_yv = _mm_add_pd(mm_disp_yv, _mm_mul_pd(mm_delta_y, mm_t));
//mm_disp_xv = _mm_add_pd(mm_disp_xv, _mm_mul_pd(mm_delta_x, _mm_div_pd(mm_kSquare,mm_distSquare)));
//mm_disp_yv = _mm_add_pd(mm_disp_yv, _mm_mul_pd(mm_delta_y, _mm_div_pd(mm_kSquare,mm_distSquare)));
}
if(u < n) {
__m128d mm_delta_x = _mm_sub_sd(mm_xv, _mm_load_sd(&C.m_x[u]));
__m128d mm_delta_y = _mm_sub_sd(mm_yv, _mm_load_sd(&C.m_y[u]));
__m128d mm_distSquare = _mm_max_sd(mm_minDistSquare,
_mm_add_sd(_mm_mul_sd(mm_delta_x,mm_delta_x),_mm_mul_sd(mm_delta_y,mm_delta_y))
);
__m128d mm_t = _mm_div_sd(_mm_load_sd(&C.m_nodeWeight[u]), mm_distSquare);
mm_disp_xv = _mm_add_sd(mm_disp_xv, _mm_mul_sd(mm_delta_x, mm_t));
mm_disp_yv = _mm_add_sd(mm_disp_yv, _mm_mul_sd(mm_delta_y, mm_t));
//mm_disp_xv = _mm_add_sd(mm_disp_xv, _mm_mul_sd(mm_delta_x, _mm_div_sd(mm_kSquare,mm_distSquare)));
//mm_disp_yv = _mm_add_sd(mm_disp_yv, _mm_mul_sd(mm_delta_y, _mm_div_sd(mm_kSquare,mm_distSquare)));
}
mm_disp_xv = _mm_hadd_pd(mm_disp_xv,mm_disp_xv);
mm_disp_yv = _mm_hadd_pd(mm_disp_yv,mm_disp_yv);
_mm_store_sd(&disp_x[v], _mm_mul_sd(mm_disp_xv, mm_c_rep));
_mm_store_sd(&disp_y[v], _mm_mul_sd(mm_disp_yv, mm_c_rep));
}
// attractive forces
#pragma omp single
for(int e = 0; e < C.numberOfEdges(); ++e)
{
int v = C.m_src[e];
int u = C.m_tgt[e];
double delta_x = C.m_x[v] - C.m_x[u];
double delta_y = C.m_y[v] - C.m_y[u];
double dist = max(minDist, sqrt(delta_x*delta_x + delta_y*delta_y));
disp_x[v] -= delta_x * dist / k;
disp_y[v] -= delta_y * dist / k;
disp_x[u] += delta_x * dist / k;
disp_y[u] += delta_y * dist / k;
}
// limit the maximum displacement to the temperature (m_tx,m_ty)
__m128d mm_tx = _mm_set1_pd(tx);
__m128d mm_ty = _mm_set1_pd(ty);
#pragma omp for nowait
for(int v = 0; v < n-1; v += 2)
{
__m128d mm_disp_xv = _mm_load_pd(&disp_x[v]);
__m128d mm_disp_yv = _mm_load_pd(&disp_y[v]);
__m128d mm_dist = _mm_max_pd(mm_minDist, _mm_sqrt_pd(
_mm_add_pd(_mm_mul_pd(mm_disp_xv,mm_disp_xv),_mm_mul_pd(mm_disp_yv,mm_disp_yv))
));
_mm_store_pd(&C.m_x[v], _mm_add_pd(_mm_load_pd(&C.m_x[v]),
_mm_mul_pd(_mm_div_pd(mm_disp_xv, mm_dist), _mm_min_pd(mm_dist,mm_tx))
));
_mm_store_pd(&C.m_y[v], _mm_add_pd(_mm_load_pd(&C.m_y[v]),
_mm_mul_pd(_mm_div_pd(mm_disp_yv, mm_dist), _mm_min_pd(mm_dist,mm_ty))
));
}
#pragma omp single nowait
{
if(n % 2) {
int v = n-1;
double dist = max(minDist, sqrt(disp_x[v]*disp_x[v] + disp_y[v]*disp_y[v]));
C.m_x[v] += disp_x[v] / dist * min(dist,tx);
C.m_y[v] += disp_y[v] / dist * min(dist,ty);
}
}
cool(tx,ty,cF);
#pragma omp barrier
}
}
System::alignedMemoryFree(disp_x);
System::alignedMemoryFree(disp_y);
#else
mainStep(C);
#endif
}
void SpringEmbedderFR::call(GraphAttributes &AG)
{
const Graph &G = AG.constGraph();
if(G.empty())
return;
// all edges straight-line
AG.clearAllBends();
GraphCopy GC;
GC.createEmpty(G);
// compute connected component of G
NodeArray<int> component(G);
int numCC = connectedComponents(G,component);
// intialize the array of lists of nodes contained in a CC
Array<List<node> > nodesInCC(numCC);
node v;
forall_nodes(v,G)
nodesInCC[component[v]].pushBack(v);
EdgeArray<edge> auxCopy(G);
Array<DPoint> boundingBox(numCC);
int i;
for(i = 0; i < numCC; ++i)
{
GC.initByNodes(nodesInCC[i],auxCopy);
GraphCopyAttributes AGC(GC,AG);
node vCopy;
forall_nodes(vCopy, GC) {
node vOrig = GC.original(vCopy);
AGC.x(vCopy) = AG.x(vOrig);
AGC.y(vCopy) = AG.y(vOrig);
}
// original
if (initialize(GC, AGC) == true)
{
for(int i = 1; i <= m_iterations; i++)
mainStep(GC, AGC);
}
cleanup();
// end original
node vFirst = GC.firstNode();
double minX = AGC.x(vFirst), maxX = AGC.x(vFirst),
minY = AGC.y(vFirst), maxY = AGC.y(vFirst);
forall_nodes(vCopy,GC) {
node v = GC.original(vCopy);
AG.x(v) = AGC.x(vCopy);
AG.y(v) = AGC.y(vCopy);
if(AG.x(v)-AG.width (v)/2 < minX) minX = AG.x(v)-AG.width(v) /2;
if(AG.x(v)+AG.width (v)/2 > maxX) maxX = AG.x(v)+AG.width(v) /2;
if(AG.y(v)-AG.height(v)/2 < minY) minY = AG.y(v)-AG.height(v)/2;
if(AG.y(v)+AG.height(v)/2 > maxY) maxY = AG.y(v)+AG.height(v)/2;
}
void SpringEmbedderFRExact::call(GraphAttributes &AG)
{
const Graph &G = AG.constGraph();
if(G.empty())
return;
// all edges straight-line
AG.clearAllBends();
ArrayGraph component(AG);
component.m_useNodeWeight = m_useNodeWeight;
EdgeArray<edge> auxCopy(G);
Array<DPoint> boundingBox(component.numberOfCCs());
int i;
for(i = 0; i < component.numberOfCCs(); ++i)
{
component.initCC(i);
if (component.numberOfNodes() >= 2)
{
initialize(component);
#ifdef OGDF_SSE3_EXTENSIONS
if(System::cpuSupports(cpufSSE3))
mainStep_sse3(component);
else
#endif
mainStep(component);
}
double minX, maxX, minY, maxY;
minX = maxX = component.m_x[0];
minY = maxY = component.m_y[0];
for(int vCopy = 0; vCopy < component.numberOfNodes(); ++vCopy) {
node v = component.original(vCopy);
AG.x(v) = component.m_x[vCopy];
AG.y(v) = component.m_y[vCopy];
if(AG.x(v)-AG.width (v)/2 < minX) minX = AG.x(v)-AG.width(v) /2;
if(AG.x(v)+AG.width (v)/2 > maxX) maxX = AG.x(v)+AG.width(v) /2;
if(AG.y(v)-AG.height(v)/2 < minY) minY = AG.y(v)-AG.height(v)/2;
if(AG.y(v)+AG.height(v)/2 > maxY) maxY = AG.y(v)+AG.height(v)/2;
}
minX -= m_minDistCC;
minY -= m_minDistCC;
for(int vCopy = 0; vCopy < component.numberOfNodes(); ++vCopy) {
node v = component.original(vCopy);
AG.x(v) -= minX;
AG.y(v) -= minY;
}
boundingBox[i] = DPoint(maxX - minX, maxY - minY);
}
Array<DPoint> offset(component.numberOfCCs());
TileToRowsCCPacker packer;
packer.call(boundingBox,offset,m_pageRatio);
// The arrangement is given by offset to the origin of the coordinate
// system. We still have to shift each node and edge by the offset
// of its connected component.
for(i = 0; i < component.numberOfCCs(); ++i)
{
const SList<node> &nodes = component.nodesInCC(i);
const double dx = offset[i].m_x;
const double dy = offset[i].m_y;
// iterate over all nodes in ith CC
for(node v : nodes)
{
AG.x(v) += dx;
AG.y(v) += dy;
}
}
}
