[flagged]
It's a play on the classic pop hit song "Girls Just Wanna Have Fun"
Girls just wanna have fast MPMC queues with bounded waiting
11–20 of 57 posts
Re: Girls just wanna have fast MPMC queues with bounded waiting
#12[flagged]
Why should anyone care what comments your agent has on code?
Re: Girls just wanna have fast MPMC queues with bounded waiting
#13[flagged]
> Or add section to explain most common agent comments.
Shouldn't your agent explain its own comments? why would the author of a fast queue care what your agent says?
Re: Girls just wanna have fast MPMC queues with bounded waiting
#14Earlier quoted context omitted.
Thanks. I jumped at the headline. I'd be happy with wait-free MPSC. I haven't checked in for a while. Have there been any breakthroughs in low-complexity wait-free queues in the past 10 years?
The closest thing I know of, is that there was a concurrent queue algo called LCRQ It originally required double-width CAS, but IIRC in recent years someone figured out how to remove this to make it more portable Best reference I could find from cursory google: https://ppopp23.sigplan.org/details/PPoPP-2023-papers/2/The-...
https://nikitakoval.org/publications/ppopp23-lprq.pdf seems to be the paper in question.
Re: Girls just wanna have fast MPMC queues with bounded waiting
#15[flagged]
> Or add section to explain most common agent comments. Shouldn't your agent explain its own comments? why would the author of a fast queue care what your agent says?
Because it would not even pass initial code review for most developers. Most people use short review prompt, with yes/no answers.
Imagine the code compilers (or some analysis tool) gives several concurrency and memory warnings. It has easy workaround (just annotate strange code, with links to explanations that this is workaround for low level bugs).
I am too tired of shitty "safe" Rust code, with 'unsafe' section around every library call (not case here, just an example)! Be clear with that, it takes 15 minutes and 10 cents!
This project could have correct concurrent code and design, but around much narrower definitions. But most people will not go too deep with review to find it!
Re: Girls just wanna have fast MPMC queues with bounded waiting
#16Earlier quoted context omitted.
> Or add section to explain most common agent comments. Shouldn't your agent explain its own comments? why would the author of a fast queue care what your agent says?
Because it would not even pass initial code review for most developers. Most people use short review prompt, with yes/no answers. Imagine the code compilers (or some analysis tool) gives several concurrency and memory warnings. It has easy workaround (just annotate strange code, with links to explanations that this is workaround for low level bugs). I am too tired of shitty "safe" Rust code, with 'unsafe' section aro…
How to get access to your agent?
Re: Girls just wanna have fast MPMC queues with bounded waiting
#17On the topic of lock free data structures I found this one on a SPSC very interesting too https://david.alvarezrosa.com/posts/optimizing-a-lock-free-r... taking it from 12M to 305M ops/s
That looks like a rewrite of my earlier work: https://rigtorp.se/ringbuffer/
Re: Girls just wanna have fast MPMC queues with bounded waiting
#18Here's my widely used implementation of this approach in C++: https://github.com/rigtorp/MPMCQueue
Re: Girls just wanna have fast MPMC queues with bounded waiting
#19[dead]