Re:The Windows malloc() implementation from MSVCRT is slow
by dang on Saturday July 02, 2022 @05:19PM
(#31962531)
Discussed here:
Benchmarking Malloc with Doom 3 - https://news.ycombinator.com/item?id=31631352 - June 2022 (30 comments)
It's interesting that LLVM is suffering so horrifically using default malloc. I really wish the author did a deeper investigation into why exactly.