I have another lock-free algorithm on my FOSS project site
http://atomic-ptr-plus.sourceforge.net/
It's the rcpc (reference counted proxy collector) package.
It uses double wide compare and swap, load reserved/store conditional,
or single wide compare and swap where the former aren't available,
e.g. 64 bit sparc. Which pretty much makes it portable to practically
anything in any environment, kernel or user space.
One caveat. It's probably not safe to use the single wide compare and
swap for 32 bit words if there's a possibility that a read lock will be
held long enough for a 30 bit counter to wrap. But it's probably safe if
the reader does not preempt.
It's ported to 32 bit x86 (linux) and 32 bit ppc (OSX) which is the
only development platforms I have access to. I may port it back to
win32 where I did the initial prototype (it's a vc++ vs. gcc port).
--
Joe Seigh
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]