Re: smrproxy v2

On 11/23/2024 8:10 AM, jseigh wrote:

On 11/21/24 15:17, jseigh wrote:

On 10/17/24 08:10, jseigh wrote:

I replaced the hazard pointer logic in smrproxy.  It's now wait-free
instead of mostly wait-free.  The reader lock logic after loading
the address of the reader lock object into a register is now 2
instructions a load followed by a store.  The unlock is same
as before, just a store.
>
It's way faster now.
>
It's on the feature/003 branch as a POC.   I'm working on porting
it to c++ and don't want to waste any more time on c version.
>
No idea of it's a new algorithm.  I suspect that since I use
the term epoch that it will be claimed that it's ebr, epoch
based reclamation, and that all ebr algorithms are equivalent.
Though I suppose you could argue it's qsbr if I point out what
the quiescent states are.
>

>
I got a port to c++ working now. There are 5 proxy implementations
1) smrproxy v2
2) arcproxy - reference counted proxy
3) rwlock based proxy
4) mutex based proxy
5) an unsafe proxy with no locking
>
The testcase is templated so you can use any of the
5 proxy implementations without rewriting for each proxy
type.  You can do apple to apple comparisons.  I
realize that's the complete antithesis of current
programming practices but there you have it.  :)
>
A bit of clean up and performance tuning now.
>

>
Ok, smrproxy lock/unlock is down to 0.6 nanoseconds now,
about what the C version was.

 Nice! Are you using pthread_getspecific or tss_get in you C version?