(this example code uses x86 and MSAM inline assembly, I have this is GAS x86
and SPARC 32/64 assembly...)
This started off as a testing benchmark against SMR, RCU and some of my
other collectors, but I now have this incorporated into my library. This
proxy garbage collector has basically the exact same functionality as RCU.
It has finer granularity than RCU... It uses another one of my specialized
forms of atomic reference counting:
Any comments/questions/tricks/help/critiques/whatever is encouraged and will
be highly appreciated...
FWIW, this is one of the collectors that I have hooked up to my
high-performance memory allocator:
Is there any interest in this kind of stuff?
> Thank you.
> Is there any interest in this kind of stuff?
This can scale way better than virtually all existing "traditional