juli/OrcaSlicer
1
1
Fork 0
mirror of https://github.com/OrcaSlicer/OrcaSlicer.git synced 2026-10-15 20:01:31 +00:00
Code Issues Packages Projects Releases Wiki Activity

G-code: key the island tour cache on the island layout, never index past the islands

Browse Source 
The per-filament island tour was cached by island centroids only. A later layer with the
same centroids but fewer islands (thin walls, negative volumes) reused the stale visit
list, whose catch-all index pointed past the layer's islands, and extrude_perimeters read
freed memory (three fuzz crashes, planar and belt). The per-instance island layout is part
of the cache key and the use site never indexes past the islands.
This commit is contained in:
harrierpigeon
2026-10-02 01:00:36 -05:00
parent 6be7911e40
commit 6a6ceb2612
2 changed files with 30 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/libslic3r/GCode.cpp
+15 -6
View File
@@ -6491,6 +6491,7 @@ LayerResult GCode::process_layer(
std::vector<GCode::ObjectByExtruder> &objects_by_extruder = objects_by_extruder_it->second;
std::vector<InstanceToPrint> &instances = filament_plan.first;
std::vector<IslandOrderNode> nodes;
std::vector<std::pair<size_t, bool>> layout; // Per instance, see IslandOrderCacheEntry
std::vector<size_t> node_instances;
auto quantize_to_mm = [](const Point &pt) -> Point {
const coord_t grid = coord_t(scale_(1.));
@@ -6515,6 +6516,7 @@ LayerResult GCode::process_layer(
const size_t instance_idx = instances.size();
instances.emplace_back(object_by_extruder, layer_id, *print_object, instance_id,
print_object->instances()[instance_id].model_instance->get_labeled_id());
layout.emplace_back(islands.size(), ! islands.empty() && ! islands.back().by_region.empty());
const Point &shift = print_object->instances()[instance_id].shift;
const size_t first_node = nodes.size();
if (islands_chainable)
@@ -6534,8 +6536,9 @@ LayerResult GCode::process_layer(
// Reuse the cached tour while this filament's island layout is unchanged.
auto &cache_entry = m_ordering_cache[filament_id];
if (!(cache_entry.first == nodes)) {
cache_entry.first = nodes;
if (! (cache_entry.nodes == nodes && cache_entry.layout == layout)) {
cache_entry.nodes = nodes;
cache_entry.layout = layout;
Points node_points;
node_points.reserve(nodes.size());
for (const IslandOrderNode &node : nodes)
@@ -6568,12 +6571,12 @@ LayerResult GCode::process_layer(
// A visit without explicit islands already prints everything.
continue;
std::vector<ObjectByExtruder::Island> &islands = instances[i].object_by_extruder.islands;
if (!islands.back().by_region.empty())
if (! islands.empty() && ! islands.back().by_region.empty())
last_visit.islands.emplace_back(islands.size() - 1);
}
cache_entry.second = std::move(visits);
cache_entry.visits = std::move(visits);
}
filament_plan.second = cache_entry.second;
filament_plan.second = cache_entry.visits;
}
}
@@ -6896,7 +6899,13 @@ LayerResult GCode::process_layer(
// in this instance's frame after set_origin() above). Empty islands are skipped;
// the trailing catch-all island has no centroid to chain by and always goes last.
std::vector<ObjectByExtruder::Island> &islands = instance_to_print.object_by_extruder.islands;
std::vector<size_t> island_order = visit.islands;
std::vector<size_t> island_order;
island_order.reserve(visit.islands.size());
for (size_t idx : visit.islands) // Never index past the islands (see IslandOrderCacheEntry)
if (idx < islands.size())
island_order.emplace_back(idx);
else
BOOST_LOG_TRIVIAL(error) << "island tour refers to island " << idx << " of " << islands.size() << ", skipped";
if (island_order.empty()) {
island_order.reserve(islands.size());
if (layer_to_print.object_layer != nullptr && islands.size() == layer_to_print.object_layer->lslices.size() + 1) {