Google Wants Rust in the Linux Kernel | Hacker News Reader

Google Wants Rust in the Linux Kernel

phoronix.com

24 pointszathan11 comments

This might be a little radical, but I really wish google stopped acting like they own anything they want because "they could buy you and your family".

It seems to me like this decision is being made purely because Rust is hot right now, and not because it makes so much sense that its the best decision moving forward.

Or maybe it's because people want to build software that is more secure and reliable (something which C does not have a great track record for)? Google are not the main instigator of this initiative, they are adding their support for it.

You mean like firefox ? OTOH there was a thread about coreutils written in Rust that crash reliably.

If the crashes are intentional panics caused because of detection of a memory safety issue, more crashes there are actually good; in a C program they would be silent buffer overflows or memory corruption that don't manifest until later, or may not manifest at all until it becomes a security issue.

Some more details here: https://lwn.net/Articles/860160/

> The Prossimo project has announced that it has contracted with Miguel Ojeda to work on Rust in the Linux kernel for the next year. Prossimo is a new name for the memory-safety projects being run by the Internet Security Research Group (ISRG), which is the organization behind the Let's Encrypt certificate authority (CA) project. Google provided the funds to enable Ojeda to work full-time on the project starting back in April.

And here: https://www.memorysafety.org/initiative/linux-kernel/

> ...Alex Gaynor and Miguel Ojeda have done some incredible work, both technically and in terms of making the case for Rust in the Linux kernel. Miguel's first kernel mailing list RFC has a lot of great information about what's being proposed.

> In April of 2021, were able to fund Miguel to work on Rust for Linux full time for one year.

> Sponsors: Google

I know a good number of people who will move away from Linux after Rust get in the kernel.

As someone who worked on kernel/OS code in C for many years, I'm very happy to see any language that doesn't sacrifice it's fit for embedded systems try to enter this space.

I think it was a shame that go didn't prioritize the lower-level use cases like its ancestors and I think not falling into a NIH syndrome is wise and mature of Google.

People will drop out because of change, and others will join or comeback. Life goes on.

What are those people's objection to Rust in the kernel? Do they prefer C's way of controlling and managing memory allocation?

Rust is an absolute no brainer in my eyes after writing it almost exclusively for embedded platforms for the past 8 months.. But sure, it’s not fit for all legacy systems with parts of Linux probably included in that list

Why ? Are they Users, kernel developers ? Move to what

time to fork it, pull out rust, and get that one active. that's what foss is all about.

Reply on news.ycombinator.com