0ct0pu5/wesnoth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Color Range: fix possible uint8_t overflow

Browse source
This commit is contained in:
Charles Dang 2024-08-18 13:56:17 -04:00
parent 71e0d142d6
commit 9c1c442f31
1 changed files with 6 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
src/color_range.cpp
View file

@ -57,17 +57,17 @@ std::vector<color_t> recolor_range_impl(const color_range& new_range, const Cont
float old_ratio = static_cast<float>(old_avg) / reference_avg;
clist.emplace_back(
std::min<uint8_t>(255u, old_ratio * mid_c.r + (1 - old_ratio) * min_c.r),
std::min<uint8_t>(255u, old_ratio * mid_c.g + (1 - old_ratio) * min_c.g),
std::min<uint8_t>(255u, old_ratio * mid_c.b + (1 - old_ratio) * min_c.b)
std::min<uint32_t>(255u, old_ratio * mid_c.r + (1 - old_ratio) * min_c.r),
std::min<uint32_t>(255u, old_ratio * mid_c.g + (1 - old_ratio) * min_c.g),
std::min<uint32_t>(255u, old_ratio * mid_c.b + (1 - old_ratio) * min_c.b)
);
} else if(reference_avg != 255) {
float old_ratio = (255.0f - static_cast<float>(old_avg)) / (255.0f - reference_avg);
clist.emplace_back(
std::min<uint8_t>(255u, old_ratio * mid_c.r + (1 - old_ratio) * max_c.r),
std::min<uint8_t>(255u, old_ratio * mid_c.g + (1 - old_ratio) * max_c.r),
std::min<uint8_t>(255u, old_ratio * mid_c.b + (1 - old_ratio) * max_c.r)
std::min<uint32_t>(255u, old_ratio * mid_c.r + (1 - old_ratio) * max_c.r),
std::min<uint32_t>(255u, old_ratio * mid_c.g + (1 - old_ratio) * max_c.r),
std::min<uint32_t>(255u, old_ratio * mid_c.b + (1 - old_ratio) * max_c.r)
);
} else {
// Should never get here. Would imply old_avg > reference_avg = 255