annilt

8 ms·

1.
Linux Kernel Memory Barriers(kernel.org)
1 points by annilt 3y ago|0 comments
2.
by annilt 3y ago
Is there a way to detect AI generated content reliably? It feels like at some point most people will use AI to generate content and then alter it as they wish. (This is what I do when I'm writing code). Maybe this is already a common p
3.
Tutorial: Writing a bare metal operating system for Raspberry Pi 4(github.com)
3 points by annilt 3y ago|0 comments
4.
A monolithic operating system in 512 bytes of x86 machine code(github.com)
246 points by annilt 3y ago|30 comments
5.
A book about the Linux kernel and its insides(github.com)
3 points by annilt 3y ago|0 comments
6.
A daemon that turns your webcam into a light sensor(github.com)
2 points by annilt 3y ago|0 comments
7.
by annilt 3y ago
I think if there is no good solution, doing nothing might be the best thing. We generally forget this option, in many cases, there are always ways to side step and avoid the problem.
8.
An incremental parsing system for programming tools(github.com)
2 points by annilt 3y ago|0 comments
9.
by annilt 3y ago
I'm not sure if they are aware of all the variables. %20 reduction only on one gender does not make much sense to me.
10.
ReactOS – A free Windows-compatible Operating System(github.com)
69 points by annilt 3y ago|38 comments
11.
by annilt 6y ago
So after witnessing all these problems, more projects will start closed source or with restrictive licenses(GPLs) instead of OSS or more permissive licenses. This is how OSS community is being degraded. A few tech giants are disrupting enti
12.
by annilt 6y ago
Just want to add one thing, x86 has stronger memory “semantics”. So, it doesn’t have to work that way behind the scenes, just at the end of the block, it has to appear it worked that way. So, x86 does reordering, store combining etc a lot.
13.
by annilt 6y ago
Not a web developer but I feel like text only/small websites should be promoted. Unnecessary images on this page, weird backlink stuff between comments.. and this is a website about web development. Not looking good to me(I’m a dumb us
14.
by annilt 6y ago
Yes, you’re right. I meant local hidden variables.
15.
by annilt 6y ago
Then explain it / give us a link please.
16.
by annilt 6y ago
The article is about that. Bell inequalities proved there isn’t ‘predestined’ effect(If you mean ‘hidden variables’ by that)
17.
by annilt 6y ago
Accepted as instant.
18.
by annilt 6y ago
I’m not blaming developers. Indeed extra features are more valuable(for users/marketing etc.) than performance most of the time. So, it’s completely normal. More hardware will lead to more features and bloat. There is no one to blame o
19.
by annilt 6y ago
Problem is these softwares can be made as fast as they were before but instead of investing time on performance/resource usage, developers see more value on extra features. So you end up with a fully featured but slower than ideal soft
20.
by annilt 6y ago
He uses Chrome as an example. You can think of a website instead of it, its ‘start up’ time always matters and almost all websites are slower than 10 years ago
21.
by annilt 6y ago
Search engines should punish slow/big/bloated websites or promote light ones. Is google doing something for that? It is so annoying to spend time on web nowadays. I find myself skipping search results when website looks bloated to
22.
by annilt 6y ago
Afaik all spectre/meltdown family attacks are possible even with a single OoO execution. But increasing the quantity(adding more complexity) might cause new batch of security issues ofcourse.
23.
by annilt 6y ago
They do something similar for ‘loops’. CPU doesn’t decode same instructions over and over again, just using them from ‘decoded instruction cache’ which has capacity around 1500 bytes.
24.
by annilt 6y ago
Out of order execution makes your code run faster like 20x-50x. You can’t produce a CPU without it and compete in the market. There are many variations of spectre and meltdown bugs, some of them might be fixed in the CPU and some in the OS.
25.
by annilt 6y ago
x86 has total store order, all stores are queued(store buffer) in the CPU and then pushed to the cache subsystem. Last item has to wait previous ones. Arm doesn’t have TSO, so CPU can reorder stores in the queue or issue stores out of order
26.
by annilt 6y ago
In addition to FJCVTZS instruction, they added more floating point execution units and registers. Also did lower latency of pipelined fp instructions
27.
by annilt 6y ago
I think it says refcounting twice faster for swift on M1 compared to swift on x86. It is possible because if you don’t have TSO(total store order), you can employ atomic operations without waiting/flushing previous items in the CPU’s s
28.
by annilt 6y ago
I meant no use-after-free bugs in production, otherwise we find a lot in development with daily tests etc. but looks like we catch them pretty effectively. It works good for us but doesn’t mean it’ll work for all other projects, so yeah I c
29.
by annilt 6y ago
If it works for you, then it is okay. It is not ‘a little effort’ for me to worry about someone else might use this pointer mistakenly, so I need to think about that all the time. It shifts my focus from problem solving to preventing future
30.
by annilt 6y ago
It’s added to the standard because macro version has undefined behavior. (yet it still works on any compiler without a problem)

More ›