| Sysop: | Amessyroom |
|---|---|
| Location: | Fayetteville, NC |
| Users: | 23 |
| Nodes: | 6 (0 / 6) |
| Uptime: | 44:28:39 |
| Calls: | 583 |
| Calls today: | 1 |
| Files: | 1,138 |
| Messages: | 110,798 |
On 6/29/2025 9:16 AM, jseigh wrote:
On 6/28/25 18:41, Chris M. Thomasson wrote:
How does the technique adapt to different methods of creating linked
data structures?
Working through that.-a It's low priority work.-a There's a bunch of
Java stuff I need to get to.
On 7/4/25 17:56, Chris M. Thomasson wrote:
On 6/29/2025 9:16 AM, jseigh wrote:
On 6/28/25 18:41, Chris M. Thomasson wrote:
How does the technique adapt to different methods of creating linked
data structures?
Working through that.-a It's low priority work.-a There's a bunch of
Java stuff I need to get to.
I did figure out how to do certain operations on linked lists.-a It could work on other linked data structures provided the operations met certain criteria.-a Something to use if I can think of any interesting
applications.
On 7/20/2025 8:25 AM, jseigh wrote:
On 7/4/25 17:56, Chris M. Thomasson wrote:
On 6/29/2025 9:16 AM, jseigh wrote:
On 6/28/25 18:41, Chris M. Thomasson wrote:
How does the technique adapt to different methods of creating
linked data structures?
Working through that.-a It's low priority work.-a There's a bunch of
Java stuff I need to get to.
I did figure out how to do certain operations on linked lists.-a It could
work on other linked data structures provided the operations met certain
criteria.-a Something to use if I can think of any interesting
applications.
Iirc, you made an old post on c.p.t many years ago about a somewhat
general case using RCU and/or your atomic-ptr-plus proxy collector, for linked node based tree algorithms in the past? Is this related to it?
and asymmetric memory barriers of course.
https://threadnought.wordpress.com/2025/05/26/wait-free-hazard-pointers- using-std-atomics/
Looks like it runs faster without the conditional branch.
Not sure it's a new idea.-a The only wait-free version of hazard pointers I've seen so far involves a storage to storage move instruction which
might not be available on all platforms.
On 5/26/25 17:58, jseigh wrote:
and asymmetric memory barriers of course.
https://threadnought.wordpress.com/2025/05/26/wait-free-hazard-
pointers- using-std-atomics/
Looks like it runs faster without the conditional branch.
Not sure it's a new idea.-a The only wait-free version of hazard pointers
I've seen so far involves a storage to storage move instruction which
might not be available on all platforms.
For some reason I though restartable sequences (RSEQ) couldn't be
inlined but apparently it can.-a I did a quick and dirty perf test
of hp protect and it looks like it runs 3x faster.-a But this will
only work on linux and only with asymmetric memory barriers.
On 7/27/2025 10:13 AM, jseigh wrote:
On 5/26/25 17:58, jseigh wrote:
and asymmetric memory barriers of course.
https://threadnought.wordpress.com/2025/05/26/wait-free-hazard-
pointers- using-std-atomics/
Looks like it runs faster without the conditional branch.
Not sure it's a new idea.-a The only wait-free version of hazard pointers >>> I've seen so far involves a storage to storage move instruction which
might not be available on all platforms.
For some reason I though restartable sequences (RSEQ) couldn't be
inlined but apparently it can.-a I did a quick and dirty perf test
of hp protect and it looks like it runs 3x faster.-a But this will
only work on linux and only with asymmetric memory barriers.
Please excuse my ignorance, by why only on Linux? windows has asymmetric memory barriers...