Skip to content

Commit

Permalink
Extend POINTER transfer to any RANGE variable in a NRN_THREAD (BlueBr…
Browse files Browse the repository at this point in the history 
…ain/CoreNeuron#772)

* Extend POINTER from voltage to any RANGE variable.
* trajectory recording is after AFTER_SOLVE
  and for consistency with NEURON, after BEFORE_STEP as well.
* update coreneuron ringtest integration data to version 1.5
* Handle the checkpoint for the POINTER
* Initialize reporting interface only if there are reports

CoreNEURON Repo SHA: BlueBrain/CoreNeuron@f72026d
  • Loading branch information
@nrnhines
nrnhines authored Mar 10, 2022
1 parent 5512857 commit ea1b8ea
Show file tree
Hide file tree
Showing 10 changed files with 289 additions and 53 deletions.
7 changes: 5 additions & 2 deletions src/coreneuron/apps/main1.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -608,8 +608,11 @@ extern "C" int run_solve_core(int argc, char** argv) {
if (corenrn_param.report_buff_size != corenrn_param.report_buff_size_default) {
set_report_buffer_size(corenrn_param.report_buff_size);
}
setup_report_engine(min_report_dt, delay);
configs.clear();

if (!configs.empty()) {
setup_report_engine(min_report_dt, delay);
configs.clear();
}

// call prcellstate for prcellgid
call_prcellstate_for_prcellgid(corenrn_param.prcellgid, compute_gpu, 0);
Expand Down
4 changes: 3 additions & 1 deletion src/coreneuron/io/nrn2core_direct.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@
#pragma once

#include <iostream>
#include <vector>

extern "C" {
// The callbacks into nrn/src/nrniv/nrnbbcore_write.cpp to get
Expand Down Expand Up @@ -56,7 +57,8 @@ extern int (*nrn2core_get_dat2_mech_)(int tid,
int dsz_inst,
int*& nodeindices,
double*& data,
int*& pdata);
int*& pdata,
std::vector<int>& pointer2type);

extern int (*nrn2core_get_dat2_3_)(int tid,
int nweight,
Expand Down
103 changes: 74 additions & 29 deletions src/coreneuron/io/nrn_checkpoint.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -88,6 +88,33 @@ void CheckPoints::write_checkpoint(NrnThread* nt, int nb_threads) const {
#endif
}

// Factor out the body of ion handling below as the same code
// handles POINTER
static int nrn_original_aos_index(int etype, int ix, NrnThread& nt, int** ml_pinv) {
// Determine ei_instance and ei from etype and ix.
// Deal with existing permutation and SoA.
Memb_list* eml = nt._ml_list[etype];
int ecnt = eml->nodecount;
int esz = corenrn.get_prop_param_size()[etype];
int elayout = corenrn.get_mech_data_layout()[etype];
// current index into eml->data is a function
// of elayout, eml._permute, ei_instance, ei, and
// eml padding.
int p = ix - (eml->data - nt._data);
assert(p >= 0 && p < eml->_nodecount_padded * esz);
int ei_instance, ei;
nrn_inverse_i_layout(p, ei_instance, ecnt, ei, esz, elayout);
if (elayout == Layout::SoA) {
if (eml->_permute) {
if (!ml_pinv[etype]) {
ml_pinv[etype] = inverse_permute(eml->_permute, eml->nodecount);
}
ei_instance = ml_pinv[etype][ei_instance];
}
}
return ei_instance * esz + ei;
}

void CheckPoints::write_phase2(NrnThread& nt) const {
FileHandler fh;

Expand Down Expand Up @@ -186,7 +213,7 @@ void CheckPoints::write_phase2(NrnThread& nt) const {
}

auto& memb_func = corenrn.get_memb_funcs();
// will need the ml_pinv inverse permutation of ml._permute for ions
// will need the ml_pinv inverse permutation of ml._permute for ions and POINTER
int** ml_pinv = (int**) ecalloc(memb_func.size(), sizeof(int*));

for (NrnThreadMembList* current_tml = nt.tml; current_tml; current_tml = current_tml->next) {
Expand Down Expand Up @@ -224,9 +251,10 @@ void CheckPoints::write_phase2(NrnThread& nt) const {

sz = nrn_prop_dparam_size_[type];
if (sz) {
int* d = soa2aos(ml->pdata, cnt, sz, layout, ml->_permute);
// need to update some values according to Datum semantics.
if (!nrn_is_artificial_[type])
int* d = soa2aos(ml->pdata, cnt, sz, layout, ml->_permute);
std::vector<int> pointer2type; // voltage or mechanism type (starts empty)
if (!nrn_is_artificial_[type]) {
for (int i_instance = 0; i_instance < cnt; ++i_instance) {
for (int i = 0; i < sz; ++i) {
int ix = i_instance * sz + i;
Expand All @@ -238,32 +266,33 @@ void CheckPoints::write_phase2(NrnThread& nt) const {
int p = pinv_nt[d[ix] - (nt._actual_diam - nt._data)];

d[ix] = p; // relative to _actual_diam
} else if (s == -5) { // Assume pointer to membrane voltage
int p = pinv_nt[d[ix] - (nt._actual_v - nt._data)];
d[ix] = p; // relative to _actual_v
} else if (s >= 0 && s < 1000) { // ion
// determine ei_instance and ei
int etype = s;
Memb_list* eml = nt._ml_list[etype];
int ecnt = eml->nodecount;
int esz = nrn_prop_param_size_[etype];
int elayout = corenrn.get_mech_data_layout()[etype];
// current index into eml->data is a function
// of elayout, eml._permute, ei_instance, ei, and
// eml padding.
int p = d[ix] - (eml->data - nt._data);
int ei_instance, ei;
nrn_inverse_i_layout(p, ei_instance, ecnt, ei, esz, elayout);
if (elayout == Layout::SoA) {
if (eml->_permute) {
if (!ml_pinv[etype]) {
ml_pinv[etype] = inverse_permute(eml->_permute,
eml->nodecount);
}
ei_instance = ml_pinv[etype][ei_instance];
}
} else if (s == -5) { // POINTER
// loop over instances, then sz, means that we
// visit consistent with natural order of
// pointer2type

// Relevant code that this has to invert
// is permute/node_permute.cpp :: update_pdata_values with
// respect to permutation, and
// io/phase2.cpp :: Phase2::pdata_relocation
// with respect to that AoS -> SoA

// Step 1: what mechanism is d[ix] pointing to
int ptype = type_of_ntdata(nt, d[ix], i_instance == 0);
pointer2type.push_back(ptype);

// Step 2: replace d[ix] with AoS index relative to type
if (ptype == voltage) {
int p = pinv_nt[d[ix] - (nt._actual_v - nt._data)];
d[ix] = p; // relative to _actual_v
} else {
// Since we know ptype, the situation is
// identical to ion below. (which was factored
// out into the following function.
d[ix] = nrn_original_aos_index(ptype, d[ix], nt, ml_pinv);
}
d[ix] = ei_instance * esz + ei;
} else if (s >= 0 && s < 1000) { // ion
d[ix] = nrn_original_aos_index(s, d[ix], nt, ml_pinv);
}
#if CHKPNTDEBUG
if (s != -8) { // WATCH values change
Expand All @@ -273,8 +302,14 @@ void CheckPoints::write_phase2(NrnThread& nt) const {
#endif
}
}
}
fh.write_array<int>(d, cnt * sz);
delete[] d;
size_t s = pointer2type.size();
fh << s << " npointer\n";
if (s) {
fh.write_array<int>(pointer2type.data(), s);
}
}
}

Expand All @@ -298,7 +333,17 @@ void CheckPoints::write_phase2(NrnThread& nt) const {
} else {
Point_process* pnt = ps->pntsrc_;
assert(pnt);
output_vindex[i] = -(pnt->_i_instance * 1000 + pnt->_type);
int type = pnt->_type;
int ix = pnt->_i_instance;
if (nt._ml_list[type]->_permute) {
// pnt->_i_instance is the permuted index into pnt->_type
if (!ml_pinv[type]) {
Memb_list* ml = nt._ml_list[type];
ml_pinv[type] = inverse_permute(ml->_permute, ml->nodecount);
}
ix = ml_pinv[type][ix];
}
output_vindex[i] = -(ix * 1000 + type);
}
}
fh.write_array<int>(output_vindex, nt.n_presyn);
Expand Down
82 changes: 78 additions & 4 deletions src/coreneuron/io/phase2.cpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,8 @@ int (*nrn2core_get_dat2_mech_)(int tid,
int dsz_inst,
int*& nodeindices,
double*& data,
int*& pdata);
int*& pdata,
std::vector<int>& pointer2type);

int (*nrn2core_get_dat2_3_)(int tid,
int nweight,
Expand Down Expand Up @@ -172,7 +173,14 @@ void Phase2::read_file(FileHandler& F, const NrnThread& nt) {
if (dsz > 0) {
pdata = F.read_vector<int>(dsz * n);
}
tmls.emplace_back(TML{nodeindices, pdata, 0, {}, {}});
tmls.emplace_back(TML{nodeindices, pdata, mech_types[i], {}, {}});
if (dsz > 0) {
int sz = F.read_int();
if (sz) {
auto& p2t = tmls.back().pointer2type;
p2t = F.read_vector<int>(sz);
}
}
}
output_vindex = F.read_vector<int>(nt.n_presyn);
output_threshold = F.read_vector<double>(n_real_output);
Expand Down Expand Up @@ -314,8 +322,13 @@ void Phase2::read_direct(int thread_id, const NrnThread& nt) {
int* nodeindices_ = nullptr;
double* data_ = _data + offset;
int* pdata_ = const_cast<int*>(tml.pdata.data());
(*nrn2core_get_dat2_mech_)(
thread_id, i, dparam_sizes[type] > 0 ? dsz_inst : 0, nodeindices_, data_, pdata_);
(*nrn2core_get_dat2_mech_)(thread_id,
i,
dparam_sizes[type] > 0 ? dsz_inst : 0,
nodeindices_,
data_,
pdata_,
tml.pointer2type);
if (dparam_sizes[type] > 0)
dsz_inst++;
offset += nrn_soa_padded_size(nodecounts[i], layout) * param_sizes[type];
Expand Down Expand Up @@ -586,6 +599,16 @@ void Phase2::pdata_relocation(const NrnThread& nt, const std::vector<Memb_func>&
// -9 (diam), // or 0-999 (ion variables).
// Note that pdata has a layout and the // type block in nt.data into which
// it indexes, has a layout.

// For faster search of tmls[i].type == type, use a map.
// (perhaps would be better to replace tmls so that we can use tmls[type].
std::map<int, size_t> type2itml;
for (size_t i = 0; i < tmls.size(); ++i) {
if (tmls[i].pointer2type.size()) {
type2itml[tmls[i].type] = i;
}
}

for (auto tml = nt.tml; tml; tml = tml->next) {
int type = tml->index;
int layout = corenrn.get_mech_data_layout()[type];
Expand Down Expand Up @@ -615,8 +638,23 @@ void Phase2::pdata_relocation(const NrnThread& nt, const std::vector<Memb_func>&
nt._actual_diam - nt._data, cnt, pdata, i, szdp, layout, nt.end);
break;
case -5: // pointer assumes a pointer to membrane voltage
// or mechanism data in this thread. The value of the
// pointer on the NEURON side was analyzed by
// nrn_dblpntr2nrncore which returned the
// mechanism index and type. At this moment the index
// is in pdata and the type is in tmls[type].pointer2type.
// However the latter order is according to the nested
// iteration for nodecount { for szdp {}}
// Also the nodecount POINTER instances of mechanism
// might possibly point to differnt range variables.
// Therefore it is not possible to use transform_int_data
// and the transform must be done one at a time.
// So we do nothing here and separately iterate
// after this loop instead of the former voltage only
/**
transform_int_data(
nt._actual_v - nt._data, cnt, pdata, i, szdp, layout, nt.end);
**/
break;
default:
if (s >= 0 && s < 1000) { // ion
Expand All @@ -637,6 +675,36 @@ void Phase2::pdata_relocation(const NrnThread& nt, const std::vector<Memb_func>&
}
}
}
// Handle case -5 POINTER transformation (see comment above)
auto search = type2itml.find(type);
if (search != type2itml.end()) {
auto& ptypes = tmls[type2itml[type]].pointer2type;
assert(ptypes.size());
size_t iptype = 0;
for (int iml = 0; iml < cnt; ++iml) {
for (int i = 0; i < szdp; ++i) {
int s = semantics[i];
if (semantics[i] == -5) { // POINTER
int* pd = pdata + nrn_i_layout(iml, cnt, i, szdp, layout);
int ix = *pd; // relative to elem0
int ptype = ptypes[iptype++];
if (ptype == voltage) {
nrn_assert((ix >= 0) && (ix < nt.end));
int elem0 = nt._actual_v - nt._data;
*pd = elem0 + ix;
} else {
Memb_list* pml = nt._ml_list[ptype];
int pcnt = pml->nodecount;
int psz = corenrn.get_prop_param_size()[ptype];
nrn_assert((ix >= 0) && (ix < pcnt * psz));
int elem0 = pml->data - nt._data;
*pd = elem0 + nrn_param_layout(ix, ptype, pml);
}
}
}
}
ptypes.clear();
}
}
}
}
Expand Down Expand Up @@ -1079,9 +1147,15 @@ void Phase2::populate(NrnThread& nt, const UserParams& userParams) {
#endif

// specify the ml->_permute and sort the nodeindices
// Have to calculate all the permute before updating pdata in case
// POINTER to data of other mechanisms exist.
for (auto tml = nt.tml; tml; tml = tml->next) {
if (tml->ml->nodeindices) { // not artificial
permute_nodeindices(tml->ml, p);
}
}
for (auto tml = nt.tml; tml; tml = tml->next) {
if (tml->ml->nodeindices) { // not artificial
permute_ml(tml->ml, tml->index, nt);
}
}
Expand Down
1 change: 1 addition & 0 deletions src/coreneuron/io/phase2.hpp
View file
Original file line number Diff line number Diff line change
Expand Up @@ -113,6 +113,7 @@ class Phase2 {
int type;
std::vector<int> iArray;
std::vector<double> dArray;
std::vector<int> pointer2type;
};
std::vector<TML> tmls;
std::vector<int> output_vindex;
Expand Down
Loading

0 comments on commit ea1b8ea

Please sign in to comment.