Programmer Humor

parent

[–] Opisek@lemmy.world 7 points 1 year ago (1 child)

I absolutely love how easy multi threading and communication between threads is made in Go. Easily one of the biggest selling points.

parent

[–] firelizzard@programming.dev 1 point 1 year ago (1 child)

Key point: they're not threads, at least not in the traditional sense. That makes a huge difference under the hood.

parent

[–] Opisek@lemmy.world 1 point 1 year ago* (1 child)

Well, they're userspace threads. That's still concurrency just like kernel threads.

Also, it still uses kernel threads, just not for every single goroutine.

parent

[–] firelizzard@programming.dev 1 point 1 year ago (1 child)

What I mean is, from the perspective of performance they are very different. In a language like C where (p)threads are kernel threads, creating a new thread is only marginally less expensive than creating a new process (in Linux, not sure about Windows). In comparison creating a new 'user thread' in Go is exceedingly cheap. Creating 10s of thousands of goroutines is feasible. Creating 10s of thousands of threads is a problem.

Also, it still uses kernel threads, just not for every single goroutine.

This touches on the other major difference. There is zero connection between the number of goroutines a program spawns and the number of kernel threads it spawns. A program using kernel threads is relying on the kernel's scheduler which adds a lot of complexity and non-determinism. But a Go program uses the same number of kernel threads (assuming the same hardware and you don't mess with GOMAXPROCS) regardless of the number of goroutines it uses, and the goroutines are cooperatively scheduled by the runtime instead of preemptively scheduled by the kernel.

parent

[–] Opisek@lemmy.world 1 point 1 year ago*

Great details! I know the difference personally, but this is a really nice explanation for other readers.

About the last point though: I'm not sure Go always uses the maximum amount of kernel threads it is allowed to use. I read it spawns one on blocking syscalls, but I can't confirm that. I could imagine it would make sense for it to spawn them lazily and then keep around to lessen the overhead of creating it in case it's needed later again, but that is speculation.

Edit: I dove a bit deeper. It seems that nowadays it spawns as many kernel threads as CPU cores available plus additional ones for blocking syscalls. https://go.dev/doc/go1.5 https://docs.google.com/document/u/0/d/1At2Ls5_fhJQ59kDK2DFVhFu3g5mATSXqqV5QrxinasI/mobilebasic

parent

[–] Successful_Try543@feddit.org 23 points 1 year ago (3 children)

Does Python have the ability to specify loops that should be executed in parallel, as e.g. Matlab uses parfor instead of for?

hideshow 6 child comments

parent

[–] lime@feddit.nu 52 points 1 year ago (2 children)

python has way too many ways to do that. asyncio, future, thread, multiprocessing...

parent

[–] WolfLink@sh.itjust.works 41 points 1 year ago (1 child)

Of the ways you listed the only one that will actually take advantage of a multi core CPU is multiprocessing

parent

[–] lime@feddit.nu 11 points 1 year ago (1 child)

yup, that's true. most meaningful tasks are io-bound so "parallel" basically qualifies as "whatever allows multiple threads of execution to keep going". if you're doing numbercrunching in pythen without a proper library like pandas, that can parallelize your calculations, you're doing it wrong.

parent

[–] WolfLink@sh.itjust.works 8 points 1 year ago* (1 child)

I’ve used multiprocessing to squeeze more performance out of numpy and scipy. But yeah, resorting to multiprocessing is a sign that you should be dropping into something like Rust or a C variant.

parent

[–] itslilith@lemmy.blahaj.zone 1 point 1 year ago

Most numpy array functions already utilize multiple cores, because they're optimized and written in C

parent

[–] danhab99@programming.dev 9 points 1 year ago (1 child)

I've always hated object oriented multi threading. Goroutines (green threads) are just the best way 90% of the time. If I need to control where threads go I'll write it in rust.

parent

[–] lime@feddit.nu 7 points 1 year ago (2 children)

nothing about any of those libraries dictates an OO approach.

parent

[–] Buddahriffic@lemmy.world 2 points 1 year ago (1 child)

Unless it's java.

parent

[–] entropicdrift@lemmy.sdf.org 1 point 1 year ago

Meh, even Java has decent FP paradigm support these days. Just because you can do everything in an OO way in Java doesn't mean you need to.

parent

[–] danhab99@programming.dev 0 points 1 year ago (1 child)

If I have to put a thread object in a variable and call a method on it to start it then it's OO multi threading. I don't want to know when the thread spawns, I don't want to know what code it's running, and I don't want to know when it's done. I just want shit to happen at the same time (90% of the time)

parent

[–] lime@feddit.nu 4 points 1 year ago

the thread library is aping the posix thread interface with python semantics.

parent

[–] Midnitte@beehaw.org 9 points 1 year ago (1 child)

Experimentally, yes

parent

[–] Successful_Try543@feddit.org 4 points 1 year ago

Cool.

parent

[–] enemenemu@lemm.ee 8 points 1 year ago (2 children)

Are you still using matlab? Why? Seriously

parent

[–] Successful_Try543@feddit.org 18 points 1 year ago (1 child)

No, I'm not at university anymore.

parent

[–] enemenemu@lemm.ee 5 points 1 year ago (1 child)

Good for you

Poor prof

parent

[–] Successful_Try543@feddit.org 5 points 1 year ago* (last edited 1 year ago) (1 child)

We weren't doing any ressource extensive computations with Matlab, mainly just for teaching FEM, as we've had an extensive collection of scripts for that purpose, and pre- and some post processing.

parent

[–] twice_hatch@midwest.social 1 point 1 year ago

I don't like that they don't write their own algorithms in any other language. I was trying to understand low-pass filters a while back and so many web pages were like, "Call this MATLAB function" or "here's a code generator that puts out bad C for specific filter parameters" Like no, I want the algorithm explained to me...

parent

[–] Panties@lemmy.ca 7 points 1 year ago (1 child)

I was telling a colleague about how my department started using Rust for some parts of our projects lately. (normally Python was good enough for almost everything but we wanted to try it out)

They asked me why we're not using MATLAB. They were not joking. So, I can at least tell you their reasoning. It was their first programming language in university, it's safer and faster than Python, and it's quite challenging to use.

parent

[–] twice_hatch@midwest.social 4 points 1 year ago

"Just use MATLAB" - Someone with a kind heart who has never deployed anything to anything

parent

[–] AndrasKrigare@beehaw.org 14 points 1 year ago (1 child)

I think OP is making a joke about python's GIL, which makes it so even if you are explicitly multi threading, only one thread is ever running at a time, which can defeat the point in some circumstances.

parent

[–] lime@feddit.nu 1 point 1 year ago* (last edited 1 year ago) (1 child)

no, they're just saying python is slow. even without the GIL python is not multithreaded. the thread library doesn't use OS threads so even a free-threaded runtime running "parallel" code is limited to one thread.

apparently not!

parent

[–] AndrasKrigare@beehaw.org 3 points 1 year ago (2 children)

If what you said were true, wouldn't it make a lot more sense for OP to be making a joke about how even if the source includes multi threading, all his extra cores are wasted? And make your original comment suggesting a coding issue instead of a language issue pretty misleading?

But what you said is not correct. I just did a dumb little test

import threading 
import time

def task(name):
  time.sleep(600)

t1 = threading.Thread(target=task, args=("1",))
t2 = threading.Thread(target=task, args=("2",))
t3 = threading.Thread(target=task, args=("3",))

t1.start()
t2.start()
t3.start()

And then ps -efT | grep python and sure enough that python process has 4 threads. If you want to be even more certain of it you can strace -e clone,clone3 python ./threadtest.py and see that it is making clone3 syscalls.

parent

[–] lime@feddit.nu 3 points 1 year ago*

is this stackless?

anyway, that's interesting! i was under the impression that they eschewed os threads because of the gil. i've learned something.

parent

[–] anton@lemmy.blahaj.zone 2 points 1 year ago* (last edited 1 year ago) (1 child)

~~Now do computation in those threads and realize that they all wait on the GIL giving you single core performance on computation and multi threaded performance on io.~~

parent

[–] AndrasKrigare@beehaw.org 4 points 1 year ago (2 children)

Correct, which is why before I had said

I think OP is making a joke about python's GIL, which makes it so even if you are explicitly multi threading, only one thread is ever running at a time, which can defeat the point in some circumstances.

parent

[–] anton@lemmy.blahaj.zone 3 points 1 year ago

Ups, my attention got trapped by the code and I didn't properly read the comment.

parent

[–] thisisnotgoingwell@programming.dev 1 point 1 year ago* (1 child)

Isn't that what threading is? Concurrency always happens on single core. Parallelism is when separate threads are running on different cores. Either way, while the post is meant to be humorous, understanding the difference is what prevents people from picking up the topic. It's really not difficult. Most reasons to bypass the GIL are IO bound, meaning using threading is perfectly fine. If things ran on multiple cores by default it would be a nightmare with race conditions.

parent

https://en.m.wikipedia.org/wiki/Thread_(computing)

[–] AndrasKrigare@beehaw.org 1 point 1 year ago

I haven't heard of that being what threading is, but that threading is about shared resourcing and memory space and not any special relationship with the scheduler.

Per the wiki:

On a multiprocessor or multi-core system, multiple threads can execute in parallel, with every processor or core executing a separate thread simultaneously; on a processor or core with hardware threads, separate software threads can also be executed concurrently by separate hardware threads.

I also think you might be misunderstanding the relationship between concurrency and parallelism; they are not mutually exclusive. Something can be concurrent through parallelism, as the wiki page has (emphasis mine):

Concurrency refers to the ability of a system to execute multiple tasks through simultaneous execution or time-sharing (context switching), sharing resources and managing interactions.

https://en.m.wikipedia.org/wiki/Concurrency_(computer_science)

parent

[–] groknull@programming.dev 5 points 1 year ago

I initially read this as “all programmers are single-threaded” and thought to myself, “yeah, that tracks”