phaneron/Kelsidavis-WoWee
1
0
Fork 0
mirror of https://github.com/Kelsidavis/WoWee.git synced 2026-05-07 09:33:51 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

fix(wob+wom): reject corrupted header counts before allocating

Browse source 
Adds upfront sanity bounds to both WoB and WOM load:
  WOM: vert<=1M, index<=4M, tex<=1K
  WOB: groups<=4K, portals<=8K, doodads<=64K
Real WoW models stay well under these limits (M2 vert is uint16 anyway).
Without these checks a corrupted header could trigger a multi-GB
allocation and OOM the process before we finish reading the body. Also
caps name length to 1KB on WoB load (already done on WOM).
This commit is contained in:
Kelsi 2026-05-06 05:42:50 -07:00
parent c05d421c29
commit 90289ba48b
2 changed files with 17 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

8
src/pipeline/wowee_model.cpp
View file

@ -39,6 +39,14 @@ WoweeModel WoweeModelLoader::load(const std::string& basePath) {
f.read(reinterpret_cast<char*>(&model.boundRadius), 4);
f.read(reinterpret_cast<char*>(&model.boundMin), 12);
f.read(reinterpret_cast<char*>(&model.boundMax), 12);
// Sanity bounds. Real M2 models cap at 65k vertices (uint16 indices) and
// typically 64 textures. Reject obviously corrupted counts before we
// try to allocate huge vertex/index buffers.
if (vertCount > 1'000'000 || indexCount > 4'000'000 || texCount > 1024) {
LOG_ERROR("WOM header rejected (verts=", vertCount,
" indices=", indexCount, " textures=", texCount, "): ", basePath);
return WoweeModel{};
}
// Bound sanity — radius drives M2 culling, min/max drive collision AABBs.
// NaN/inf would either cull-out the model or crash the cull math.