Hacker News new | ask | show | jobs
by rbmm 835 days ago
several years ago i do own implementation of SRW/PushLocks - of course based on original NeillC code.. https://github.com/rbmm/SRW_ALT/tree/main/PushLock-ALT

they have slightly worse performance compared to MS, when high contention, but in test with this OP case - work well. in my test i not found bugs in implementation, but of course can not be sure that they not exist, very complex logic really. nobody test this, however very simply replace SRW calls to my implementation, by macros in h file