Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Translating All C to Rust (TRACTOR)

darpa.mil

Post reply on HN
Print view View original

231–240 of 403 posts

Re: Translating All C to Rust (TRACTOR)

#231

Post by andrewflnr »

sam0x17 wrote: post #50
mike_hearn wrote: post #10

That sounds ... hard. Especially as idiomatic Rust as written by skilled programmers looks nothing like C, and most interesting code is written in C++ anyway. Isn't it equivalent to statically determining the lifetimes of all allocations in the C program, including those that are implemented using custom allocators or which cross into proprietary libraries? There's been a lot of research into this sort of thing over…

speaking of hard, the DOE actually funds a project that has been around for 20+ years now (ROSE) that involves (among other things) doing static analysis on and automatically translating between C/C++/Cuda and even high level languages like Python as well as HPC variants of C/C++. They have a combined AST that supports all of those languages with the same set of node types essentially. Quite cool. I got to work on it…

> a combined AST that supports all of those languages with the same set of node types essentially.

I can't believe that works at all. I'll take a look for sure.

Re: Translating All C to Rust (TRACTOR)

#232

Post by woodruffw »

hgs3 wrote: post #226
woodruffw wrote: post #184

Earlier quoted context omitted.

"You aren't trying hard enough" isn't a serious approach to security: if it was, we wouldn't require seatbelts in cars or health inspections in restaurants. (It's also not clear that they aren't trying hard enough: Google, Apple, etc. have billions of dollars riding on the safety of their products, but still largely fail to produce memory-safe C and C++ codebases.)

In the case of OpenSSL, Big Tech clearly neglected proper support until after the Heartbleed vulnerability. Prior to Heartbleed, the OpenSSL Software Foundation only received about $2K annually in donations and employed just one full-time employee [1]. Given the projects critical role in internet security, Big Techs neglect raises concerns about their quality assurance practices for less critical projects. The OpenSS…

OpenSSL is not a great example, either before or after funding — it’s a notoriously poorly architected codebase with multiple layers of flawed abstractions. I meant things more like Chromium, WebKit, etc.: these have dozens to hundreds of professional top-bracket C and C++ developers working on them, and they still can’t avoid memory corruption bugs.

Re: Translating All C to Rust (TRACTOR)

#233

Post by b20000 »

“ the software engineering community has reached a consensus” … hahaha no sorry, I don’t think so

Your priority should be to learn how to build better software and not force a new language onto people.

do you remember the age old saying about nature and fools?

Re: Translating All C to Rust (TRACTOR)

#234

Post by Sytten »

Would be nice if they could first hire all the smart engineers (that Mozilla laid off) to continue working on the language itself.

Async is still a half finished mess even for people that use it every day. And that is my main annoyance but there are many (trait specialization, orphan rule limits, HKT, etc.)

Re: Translating All C to Rust (TRACTOR)

#235

Post by Thro4l31 »

The problems come with maintaining the translated code bases:

1. A code base written in C and a team of C engineers that have a good mental model of the code base to be able to maintain it.

2. An automatically translated Rust code base. Potentially (I'd say probably, but that is just my gut feeling) harder to read and understand than the original one.

3. Now you need a team of Rust engineers that have a good mental model of the code base that was generated.

If you already have that team of Rust engineers, I'd rather let them rewrite the code manually as they can improve it and have the correct mental model from the start.

Re: Translating All C to Rust (TRACTOR)

#236

Post by fragmede »

m0llusk wrote: post #133
programd wrote: post #64

Earlier quoted context omitted.

> > today cars drive themselves > You can attach about a hundred asterisks to that. Not in San Francisco. There are about 300 Waymo cars safely driving in one of the most difficult urban environments around (think steep hills, fog, construction, crazy traffic, crazy drivers, crazier pedestrians). Five years ago this was "someday" science-fiction. Frankly I trust them much more then human drivers and envision a future…

Opinions about automated driving systems vary. Just from my own experience doing business all around San Francisco I have seen at least a half dozen instances of Waymo vehicles making unsafe maneuvers. Responders have told me and local government officials that Waymo vehicles frequently fail to acknowledge emergency situations or respond to driving instructions. Driving is a social exercise which requires understandi…

they're not perfect, sure, but they're out there, just driving around all autonomously and all, contrary to GGP's assertion that they don't exist.

Re: Translating All C to Rust (TRACTOR)

#237

Post by TestingWithEdd »

the8472 wrote: post #107
mike_hearn wrote: post #10

That sounds ... hard. Especially as idiomatic Rust as written by skilled programmers looks nothing like C, and most interesting code is written in C++ anyway. Isn't it equivalent to statically determining the lifetimes of all allocations in the C program, including those that are implemented using custom allocators or which cross into proprietary libraries? There's been a lot of research into this sort of thing over…

Write tests for your C code. Run c2rust (mechanical translation), including the tests. Let a LLM/MCTS/verifier loop go to town. Verifier here means it passes compiler checks, tests, santiziers and miri. Additional training data can be generated by running mrustc or by inlining unsafe code (from std/core/leaf crates) into safe code and running semantics-preserving mechanical refactorings on the code. This can be close…

Essentially neural program synthesis

Re: Translating All C to Rust (TRACTOR)

#238

Post by indolering »

woodruffw wrote: post #129
fsckboy wrote: post #125

Earlier quoted context omitted.

in this case it seems to me the hard task that DARPA has chosen is to get me to forget how much they spent on pushing Ada.

I can't find any clear references to DARPA (or ARPA) being involved in Ada's development. It was a DoD program but, well, the DoD is notoriously large and multi-headed. (But even if DARPA was involved in Ada: I think it's clear, at this point, that Ada has been a resounding success in a small number of domains without successfully breaking into general-purpose adoption. I don't have a particular value judgment associ…

Too lazy to look it up, but I'm pretty sure DARPA was involved and certain that DoD contracta prioritized ADA for a long time.

Re: Translating All C to Rust (TRACTOR)

#239

Post by TestingWithEdd »

programd wrote: post #64
Calavar wrote: post #47

Earlier quoted context omitted.

> today cars drive themselves You can attach about a hundred asterisks to that. If anything, I think self the failure to hit L5 driving after billions of dollars and millions of man hours invested is probably reflective of how automatic C to Rust translation will go. We'll cruise 90% of the way, but the last 10% will prove insurmountable with current technology. Think about the number of C programs in the wild that r…

> > today cars drive themselves > You can attach about a hundred asterisks to that. Not in San Francisco. There are about 300 Waymo cars safely driving in one of the most difficult urban environments around (think steep hills, fog, construction, crazy traffic, crazy drivers, crazier pedestrians). Five years ago this was "someday" science-fiction. Frankly I trust them much more then human drivers and envision a future…

Limited to specific areas during specific hours, and have caused crashes (at least when I lived there till last summer).

Re: Translating All C to Rust (TRACTOR)

#240

Post by sam0x17 »

andrewflnr wrote: post #231
sam0x17 wrote: post #50

Earlier quoted context omitted.

speaking of hard, the DOE actually funds a project that has been around for 20+ years now (ROSE) that involves (among other things) doing static analysis on and automatically translating between C/C++/Cuda and even high level languages like Python as well as HPC variants of C/C++. They have a combined AST that supports all of those languages with the same set of node types essentially. Quite cool. I got to work on it…

> a combined AST that supports all of those languages with the same set of node types essentially. I can't believe that works at all. I'll take a look for sure.

Most of what they use it for is static analysis, but the funding comes from its ability to translate old simulation code to HPC-ready code. I think they even support fortran IIRC
Post reply on HN