home.social

#multithreaded — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #multithreaded, aggregated by home.social.

  1. @ivantodorov hey keep posting, i have had malcolm running and eventually i will get to it and start looking at some dashboards and learning it - ids/ips is pretty complicated but i agree that it is worth it #multithreaded

  2. So, this might seem small, but it's a really cool step forward for #Verso browser (using the next gen, #Rust based, #multithreaded, memory safe @servo engine rather than #Chrome's monopoly #Blink or #Mozilla's #Gecko engines).

    They got tabs!

    Doesn't *every* browser have tabs? Sure, but others are basically just GUI's slapped on a couple of pre-existing engines. Verso is rebuilding the browser from the ground up using #Servo and #FOSS, even for the #UI.

    Early shots fired, #Blink/#Webkit!

  3. We just published a new 🎥 of #JCON EUROPE 2024!
    Watch Sebastian Konieczek talking about 'Handling concurrent access to shared resources in #Java'

    #Multithreaded applications are to be found everywhere. Think of #RESTful #micro services for …

    Watch it now: youtu.be/rMBV0iWgIfk

  4. When you spent waaaaay too much time debugging and doubting your #multithreaded code because the output is garbled and you suspect a race condition or cross thread writes but it turns out the input data was broken to begin with, you just needed the multithreaded code to be able to parse it at all in a finite time-frame... 🤦

    #Multithreading #programming #GarbageInGarbageOut

  5. Continuing our Schedule Highlights, @mariadb Chief Innovation Officer Daniel Black @danblack takes us through - a tool for C and C++ applications.

    In this tutorial, Daniel shows how rr records the entire state of the program, allowing it to be interrogated.

    You will need to have bare metal with rr compiled, or a with rr proven to work, for this

    🗓️ Schedule: 2024.everythingopen.au/schedul

    🎟️ Register: 2024.everythingopen.au/attend/

  6. PipelineC's new 'shared resource buses' are pretty neat: Multiple 'threads' of C function derived state machines can easily share resources (memory, compute, etc) by generating AXI-like buses and arbitration. This is still highly experimental but very powerful - reach out if you want to make something together! github.com/JulianKemmerer/Pipe

  7. Frame buffers are the "shared resource" in the recent Game of Life demo. PipelineC generates arbitration between multiple state machine 'threads' simultaneously reading+writing the buffers. What can you do with C code and a frame buffer? github.com/JulianKemmerer/Pipe

  8. 'C code' Game of Life in at 48FPS 🤓 PipelineC derived state machines and new shared resource buses make it super easy to adapt multi-threaded software C implementations for quick FPGA prototypes. Highly experimental but very powerful - reach out if you want to make something together! github.com/JulianKemmerer/Pipe

  9. Is there a library like Sidekiq, but runs entirely in memory and doesn't need redis? I want to write a generic multi-threaded (also want to support the async gem) worker pool with input and output queues, and am wondering if this already exists?
    #sidekiq #ruby #multithreaded #concurrency

  10. #AVR #prologue #generator update...

    I've just almost doubled my base interrupt rate - the hot path is a mere 20 cycles (plus 8 for taking the exception and returning).

    That's 571.4kirqs/sec - much better than the ~350k I was looking at yesterday.

    The penalty is that the *first* interrupt is a little slower, so if you're unlucky, and they come in timed "just wrong", performance is "capped" at ~350kirqs/s.

    That's still fast enough for high-speed serial comms, which is nice. - by 3x for a worst case (@115.2kbaud), or by very close to 5x otherwise.

    so, 115200 serial on a #multithreaded #kernel on an 8 bit MCU, with worst case CPU usage about 20%

    Even if all you did was read and throw away the byte, you'd need 1 cycle + 5 setup + 2 push + 2 pop + 3 return, total 13 cycles, so the "theoretical best" is only 1.5x as fast, and you'd get nothing else done in either case...

    #winning