external/ton
1
0
Fork 0
mirror of https://github.com/ton-blockchain/ton synced 2025-03-09 15:40:10 +00:00
Code Issues Releases Wiki Activity

[Tolk] Refactor: get rid of split_vars, construct valid LET ops

Browse source 
In FunC (and in Tolk before), tensor vars (actually occupying
several stack slots) were represented as a single var in terms
or IR vars (Ops):
> var a = (1, 2);
> LET (_i) = (_1, _2)

Now, every tensor of N stack slots is represented as N IR vars.
> LET (_i, _j) = (_1, _2)

This will give an ability to control access to parts of a tensor
when implementing `tensorVar.0` syntax.
This commit is contained in:
tolk-vm 2024-12-18 19:26:26 +03:00
parent 989629a832
commit 565bc59735
No known key found for this signature in database
GPG key ID: 7905DD7FE0324B12
17 changed files with 100 additions and 217 deletions
Download patch file Download diff file Expand all files Collapse all files

87
tolk/abscode.cpp
View file

@ -31,16 +31,6 @@ void TmpVar::dump(std::ostream& os) const {
os << " : " << v_type << " (width ";
os << v_type->calc_width_on_stack();
os << ")";
if (coord > 0) {
os << " = _" << (coord >> 8) << '.' << (coord & 255);
} else if (coord < 0) {
int n = (~coord >> 8), k = (~coord & 0xff);
if (k) {
os << " = (_" << n << ".._" << (n + k - 1) << ")";
} else {
os << " = ()";
}
}
os << std::endl;
}
@ -51,7 +41,7 @@ void TmpVar::show(std::ostream& os, int omit_idx) const {
return;
}
}
os << '_' << idx;
os << '_' << ir_idx;
}
std::ostream& operator<<(std::ostream& os, const TmpVar& var) {
@ -182,47 +172,6 @@ void VarDescrList::show(std::ostream& os) const {
os << " ]\n";
}
void Op::split_vars(const std::vector<TmpVar>& vars) {
split_var_list(left, vars);
split_var_list(right, vars);
for (auto& op : block0) {
op.split_vars(vars);
}
for (auto& op : block1) {
op.split_vars(vars);
}
}
void Op::split_var_list(std::vector<var_idx_t>& var_list, const std::vector<TmpVar>& vars) {
int new_size = 0, changes = 0;
for (var_idx_t v : var_list) {
int c = vars.at(v).coord;
if (c < 0) {
++changes;
new_size += (~c & 0xff);
} else {
++new_size;
}
}
if (!changes) {
return;
}
std::vector<var_idx_t> new_var_list;
new_var_list.reserve(new_size);
for (var_idx_t v : var_list) {
int c = vars.at(v).coord;
if (c < 0) {
int n = (~c >> 8), k = (~c & 0xff);
while (k-- > 0) {
new_var_list.push_back(n++);
}
} else {
new_var_list.push_back(v);
}
}
var_list = std::move(new_var_list);
}
void Op::show(std::ostream& os, const std::vector<TmpVar>& vars, std::string pfx, int mode) const {
if (mode & 2) {
os << pfx << " [";
@ -444,26 +393,22 @@ void CodeBlob::print(std::ostream& os, int flags) const {
os << "-------- END ---------\n\n";
}
var_idx_t CodeBlob::create_var(TypePtr var_type, const LocalVarData* v_sym, SrcLocation location) {
vars.emplace_back(var_cnt, var_type, v_sym, location);
return var_cnt++;
}
bool CodeBlob::import_params(FormalArgList&& arg_list) {
if (var_cnt || in_var_cnt) {
return false;
std::vector<var_idx_t> CodeBlob::create_var(TypePtr var_type, const LocalVarData* v_sym, SrcLocation loc) {
std::vector<var_idx_t> ir_idx;
ir_idx.reserve(var_type->calc_width_on_stack());
if (const TypeDataTensor* t_tensor = var_type->try_as<TypeDataTensor>()) {
for (TypePtr item : t_tensor->items) {
std::vector<var_idx_t> nested = create_var(item, v_sym, loc);
ir_idx.insert(ir_idx.end(), nested.begin(), nested.end());
}
} else if (var_type != TypeDataVoid::create()) {
tolk_assert(var_type->calc_width_on_stack() == 1);
vars.emplace_back(var_cnt, var_type, v_sym, loc);
ir_idx.emplace_back(var_cnt);
var_cnt++;
}
std::vector<var_idx_t> list;
for (const auto& par : arg_list) {
TypePtr arg_type;
const LocalVarData* arg_sym;
SrcLocation arg_loc;
std::tie(arg_type, arg_sym, arg_loc) = par;
list.push_back(create_var(arg_type, arg_sym, arg_loc));
}
emplace_back(loc, Op::_Import, list);
in_var_cnt = var_cnt;
return true;
tolk_assert(static_cast<int>(ir_idx.size()) == var_type->calc_width_on_stack());
return ir_idx;
}
} // namespace tolk