bool Programmer::loadJSON(JSONNode data) {
auto devices = data.find("devices");
if (devices == data.end()) {
Logger::log(ERR, "No devices found in Programmer");
return false;
}
auto it = devices->begin();
while (it != devices->end()) {
Device* device = new Device(it->name(), *it);
// We're overwriting data here, so make sure to actually delete the old data.
// if there is any
if (m_devices.count(device->getId()) > 0)
delete m_devices[device->getId()];
m_devices[device->getId()] = device;
it++;
}
auto c = data.find("captured");
if (c == data.end()) {
Logger::log(WARN, "No captured set found in Programmer");
captured = DeviceSet(m_rig);
}
else {
captured = DeviceSet(m_rig, *c);
}
return true;
}
Programmer::Programmer(Rig* rig) : m_rig(rig) {
const set<Device*>& devices = m_rig->getDeviceRaw();
for (Device* d : devices) {
m_devices[d->getId()] = new Device(d);
}
captured = DeviceSet(m_rig);
}
DeviceSet Rig::getAllDevices() {
return DeviceSet(this, m_devices);
}
void Programmer::clearCaptured() {
m_progMutex.lock();
captured = DeviceSet(m_rig);
m_progMutex.unlock();
}
int
main(int argc, char **argv)
{
double dt = 0;
long nvtk = 0;
char outnum[240];
long time_output = 0;
long flops = 0;
// double output_time = 0.0;
double next_output_time = 0;
double start_time = 0, end_time = 0;
double start_iter = 0, end_iter = 0;
double elaps = 0;
double avgMcps = 0;
long nAvgMcps = 0;
#ifdef MPI
MPI_Init(&argc, &argv);
DeviceSet();
#endif
if (H.mype == 1) fprintf(stdout, "Hydro starts.\n");
process_args(argc, argv, &H);
hydro_init(&H, &Hv);
// PRINTUOLD(H, &Hv);
cuAllocOnDevice(H);
// Allocate work space for 1D sweeps
allocate_work_space(H, &Hw, &Hvw);
// vtkfile(nvtk, H, &Hv);
if (H.dtoutput > 0) {
// outputs are in physical time not in time steps
time_output = 1;
next_output_time = next_output_time + H.dtoutput;
}
if (H.dtoutput > 0 || H.noutput > 0)
vtkfile(++nvtk, H, &Hv);
if (H.mype == 0)
fprintf(stdout, "Hydro starts main loop.\n");
cuPutUoldOnDevice(H, &Hv);
start_time = cclock();
// fprintf(stdout, "%lg %lg %d %d \n", H.t, H.tend, H.nstep, H.nstepmax);
while ((H.t < H.tend) && (H.nstep < H.nstepmax)) {
double iter_time = 0;
flopsAri = flopsSqr = flopsMin = flopsTra = 0;
start_iter = cclock();
outnum[0] = 0;
flops = 0;
if ((H.nstep % 2) == 0) {
cuComputeDeltat(&dt, H, &Hw, &Hv, &Hvw);
// fprintf(stdout, "dt=%lg\n", dt);
if (H.nstep == 0) {
dt = dt / 2.0;
}
if (H.nproc > 1) {
#ifdef MPI
double dtmin;
int uno = 1;
MPI_Allreduce(&dt, &dtmin, uno, MPI_DOUBLE, MPI_MIN, MPI_COMM_WORLD);
dt = dtmin;
#endif
}
}
if ((H.nstep % 2) == 0) {
cuHydroGodunov(1, dt, H, &Hv, &Hw, &Hvw);
} else {
cuHydroGodunov(2, dt, H, &Hv, &Hw, &Hvw);
}
end_iter = cclock();
iter_time = (double) (end_iter - start_iter);
H.nstep++;
H.t += dt;
{
double iter_time = (double) (end_iter - start_iter);
#ifdef MPI
long flopsAri_t, flopsSqr_t, flopsMin_t, flopsTra_t;
MPI_Allreduce(&flopsAri, &flopsAri_t, 1, MPI_LONG, MPI_SUM, MPI_COMM_WORLD);
MPI_Allreduce(&flopsSqr, &flopsSqr_t, 1, MPI_LONG, MPI_SUM, MPI_COMM_WORLD);
MPI_Allreduce(&flopsMin, &flopsMin_t, 1, MPI_LONG, MPI_SUM, MPI_COMM_WORLD);
MPI_Allreduce(&flopsTra, &flopsTra_t, 1, MPI_LONG, MPI_SUM, MPI_COMM_WORLD);
// if (H.mype == 1)
// printf("%ld %ld %ld %ld %ld %ld %ld %ld \n", flopsAri, flopsSqr, flopsMin, flopsTra, flopsAri_t, flopsSqr_t, flopsMin_t, flopsTra_t);
flops = flopsAri_t * FLOPSARI + flopsSqr_t * FLOPSSQR + flopsMin_t * FLOPSMIN + flopsTra_t * FLOPSTRA;
#else
flops = flopsAri * FLOPSARI + flopsSqr * FLOPSSQR + flopsMin * FLOPSMIN + flopsTra * FLOPSTRA;
#endif
nbFLOPS++;
if (flops > 0) {
if (iter_time > 1.e-9) {
double mflops = (double) flops / (double) 1.e+6 / iter_time;
MflopsSUM += mflops;
sprintf(outnum, "%s {%.2f Mflops %ld Ops} (%.3fs)", outnum, mflops, flops, iter_time);
}
} else {
sprintf(outnum, "%s (%.3fs)", outnum, iter_time);
}
}
if (iter_time > 1.e-9) {
double mcps = ((double) H.globnx * (double) H.globny) / iter_time / 1e6l;
if (H.nstep > 5) {
sprintf(outnum, "%s (%.1lf MC/s)", outnum, mcps);
nAvgMcps++;
avgMcps += mcps;
}
}
if (time_output == 0 && H.noutput > 0) {
if ((H.nstep % H.noutput) == 0) {
cuGetUoldFromDevice(H, &Hv);
vtkfile(++nvtk, H, &Hv);
sprintf(outnum, "%s [%04ld]", outnum, nvtk);
}
} else {
if (time_output == 1 && H.t >= next_output_time) {
cuGetUoldFromDevice(H, &Hv);
vtkfile(++nvtk, H, &Hv);
next_output_time = next_output_time + H.dtoutput;
sprintf(outnum, "%s [%04ld]", outnum, nvtk);
}
}
if (H.mype == 0) {
fprintf(stdout, "--> step=%-4ld %12.5e, %10.5e %s\n", H.nstep, H.t, dt, outnum);
fflush(stdout);
}
}
end_time = cclock();
hydro_finish(H, &Hv);
cuFreeOnDevice();
// Deallocate work space
deallocate_work_space(H, &Hw, &Hvw);
elaps = (double) (end_time - start_time);
timeToString(outnum, elaps);
if (H.mype == 0)
fprintf(stdout, "Hydro ends in %ss (%.3lf) <%.2lf MFlops>.\n", outnum, elaps, (float) (MflopsSUM / nbFLOPS));
if (H.mype == 0) {
avgMcps /= nAvgMcps;
fprintf(stdout, "Average MC/s: %.1lf\n", avgMcps);
}
#ifdef MPI
MPI_Finalize();
#endif
return 0;
}
