Rust front-end merged in GCC trunk
gcc.gnu.org
gcc.gnu.org
¹https://github.com/rust-lang/rust/tree/master/compiler/rustc...
I think that's what the parent was replying to, not somehow claiming C/C++ people are insignificant.
Micropython can be counted as 0.25 or something :)
I will admit to not being familiar enough with Ruby to say anything about that
This comment is so blatantly wrong and detached from reality that it makes me wonder if it's just a good old fashioned troll.
You'd be hard-pressed to find any production programming language which does not have multiple independent implementations.
‶Toys″ like Go, OCaml, arguably Ruby & PHP, Perl, Erlang, Kotlin, ...
PHP has alternative VMs and specialized servers for serving it fast and in encrypted manner (they are closed sourced), tho.
Kotlin runs on JVM, which has at least three fully compliant implementations.
Erlang is a specialized language and telecommunications platform. It’s something different.
Well, if they are not public, it does not really have any influence on language specification.
> Kotlin runs on JVM, which has at least three fully compliant implementations.
And Rust runs on x86 which has myriad of implementations; still, we're talking about the language compiler, not the underlying runtime.
> Erlang is a specialized language and telecommunications platform. It’s something different.
How so? It's a language built over a runtime, just like Java. I don't see how the specifics of the runtime have to do with anything.
TruffleRuby is a thing, as is JRuby.
Implementing Erlang without implementing its VM and the whole infrastructure, which is a much larger task than merely a VM like Python's.
Any single-implementation language inevitably degrades to the point where the implementation becomes the specification. Shortly afterwards, people can only use the language if their tool chain, operating system, etc. closely match the understanding of the language authors. That's a bad situation to be in.
> Shortly afterwards, people can only use the language if their tool chain, operating system, etc. closely match the understanding of the language authors.
Perl, well-known for being an iffy language to install and running on a restricted set of OS/architectures.
Having two independent implementations is good for finding code depending on compiler bugs. The development will also highlight where the Rust language is not documented/specified enough yet.
The GCC implementation may end up with a different design, and perhaps faster compile times, or at least we'll know whether Rust is inherently slow to compile, or is that just rustc. Rustc supporting multiple back-ends can't have as close integration with GCC's back-end as GCC's own front-end.
I'd go as far as to claim that having multiple independent implementations is a basic requirement for any programming language which has any aspirations of being production-ready.
It seems you're oblivious to the fact that projects such as swc[1] exist, and have been adopted by projects such as deno.
[1] https://swc.rs/
> Elixir etc
I'm not familiar with Elixir nor am I in a googling mood. Nevertheless that was an awfully short list. Why is that?
Here is an interview with them: https://www.totaltypescript.com/rewriting-typescript-in-rust
What leads you to believe that the lack of alternatives leads to enabling "a lot of cool tooling"?
> and great new features are introduced quickly in the official implementation.
That has zero to do with existing only a single implementation, and everything to do with having developers working on unstable releases and a poor job thinking things though by writing stuff down both as proposals and specifications.
If there is a specification that was discussed and thought-through by competent individuals, nothing stops anyone in the world from reading the specs and implementing the same features.
Rust projects have settled on Cargo, which gives them uniform way of building, testing, getting dependencies, docs generation (https://docs.rs), IDE support, etc. (https://lib.rs/development-tools/cargo-plugins)
Contrast this with C which has several compilers and plenty of build systems to choose from, often more flexible and advanced than Cargo, but the fragmentation means that C projects are snowflakes, and tooling needs to be configured for each project and compiler combo. For every build system X there's someone who says it sucks and you should be using Y instead.
However, I don't think GCC will cause similar fragmentation problems for Rust, because there are already 100K rustc+Cargo packages in existence, so it has no choice but to follow and be compatible with them.
I used to work on the Haskell compiler GHC. Haskell has had many alternative implementations over the years. But they were often nowhere near the level of engineering quality of GHC, and nobody used them for nearly anything, unless they were graduate students exploring surface and language semantics. But many of them just used GHC instead, too. It wasn't for lack of trying either; lots of people wanted the standard for the language to work, I used to be on the standards committee for the language. So they were useful to some extent design space. But there are many, many, many other important concerns for a "production" programming language; robust feature support, good platform support, high quality code output, robust testing regime and code review, good library and API design, high quality debugging support, package management, editor support, all kinds of stuff. These also dovetail together; certain efforts at the semantic level are nice but if they don't have a good story for debugging or something, they might not make the cut.
Really this seems like one of those requirements that's always impossible to satisfy in some sense, a kind of catch-22. If you're one of the 800lb gorilla programming languages (Java, C++, Python), you might have one or two "production" implementations and if you're lucky the ratio of usage is better than 80/20, but they exist and therefore you are a "production-ready" language. Nobody else has the money to staff engineers for multiple alternative languages (OCaml, Elixir, TypeScript, Go) despite them funding billions of dollars worth of software, so they still aren't "professional" and therefore do not qualify. I get it, and it's not for zero merit, but I think it's mostly just a cat-and-mouse game, at the end of the day.
gcc-rs existing of course has many other benefits beyond this; people know GCC so they can now have a free Rust compiler for their CPU ports, etc. This is still good. But really it's still the exception. I do think this is a good effort but I consider this sort of thing one of those basic folklore issues that's actually much more nuanced in practice.
Any time you hit a fork button on an Open Source programming language's github page, you get your own independent implementation. Hire a contractor and introduce any "extensions" you need.
[1] The good reasons, anyway. Please no GPL screaming.
[2] Something that IMHO rust has historically been kinda bad at. Even now there remains no clear specification I can find that explains exactly what behavior the borrow checker will admit vs. reject.
How is “we want a toolchain we can use as an library from non-GPL code” not a good reason?
The ability to link against LLVM and Clang has driven a massive advance in languages and tooling.
When there are a fews compilers with significant market share then developers are those who lose, lose that handy dev. experience of solid and consistent ecosystem
I hope it will be used only where necessary
It would be a C implementation for people who don't like C.
As others have mentioned, vendors have the standard practice to fork gcc for their hardware, but not necessarily rustc nor libgcc used by rustc_codegen_gcc. This means that rust will be possible on those platforms for free thanks to gcc-rs.
Another point is the bootstrapping of the compiler. rustc has a quite complex bootstrap process that depends on python, while afaik gcc only relies on shell scripts. It is much easier to port gcc than rustc to a new host platform. This is useful when you want to compile code directly on your target and don't want to rely on cross compilation. cross compiling is a tricky thing to do, and while rust fares pretty well there, as soon as it uses native C libraries, it all falls apart (e.g. try cross compiling a project that has dependencies on libss, libpq and the like). In such case, native compilation may be the only viable option. And most of the times, GCC is the first thing to be ported, so you would not need to wait for a rustc port.
People think differently, have different ideas for how to do things, need it to work for their particular use case, etc. Fragmentation and bloat are a result.
I am willing to bet it will happen rust as it gets used more and more. Then the treadmill will start again with a new language.
(Sorry if that’s jaded. I’ve been fighting with python build systems lately. What a nightmare)
One compiler for 99%, one package manager, one base classes liberary, one cli, etc.
That's what language standards are for.
The Ferrocene project is working on standardizing Rust and creating a safety-critical version of the Rust toolchain:
https://ferrous-systems.com/ferrocene/
The Ferrocene Rust draft spec is out:
As Rust continues to improve and be more widely used and adopted, there will be multiple implementations. As you've stated, it is inevitable as Rust gets popular. (For example, if Rust becomes popular enough to displace niches that C and C++ have occupied, when will Microsoft release a Rust toolchain? When will Intel release a Rust toolchain? And so forth.)
Standards help to better define what a Rust toolchain should do and what it means to be a conforming Rust implementation.
For example, C++ has a standard, but compliance is all over the place. Even Fortran has the same problem. The web also had a standards body (W3C) but eventually that was completely ignored and became irrelevant.
You are correct.
Compliance depends on the implementation developer / maintainer. If the standard exists and is agreed upon, at least users can petition the implementation developer / maintainer and say "Your implementation does not conform to section 3, paragraph 2, line 1 of the standard. The expected behavior is X, but your implementation does Y. Please fix it."
If a standard does not exist, then who is to say what the correct behavior should be?
If there is a standard, but no one follows it, then perhaps the standard needs to be adjusted. (Why is the standard not being followed? Were bad ideas standardized? Did the standards committee follow after one vendor's implementation to the detriment of other vendors' implementations? etc.)
- They aren't going to fork or extend the language; they'll work with the normal Rust language evolution process.
- They're treating rustc as the reference for correct Rust. This doesn't mean there can't be bugs in either rustc or gccrs, and rustc will deeply appreciate bug reports found through this level of inspection of its behavior, but nonetheless this is a stated goal of the gccrs folks.
- They plan to explicitly provide messaging, potentially even in compiler error messages, to the effect of "if gccrs doesn't compile a crate that does compile on rustc, that's a bug in gccrs, not a bug in that crate; do not report it as a bug in that crate or ask crate developers to work around gccrs limitations".
So, I think the gccrs folks are doing everything they can to avoid fragmentation, short of the project not existing at all.
But rustc itself can be far enough behind rustc, as the MSRV debate shows.
On the other hand, I acknowledge that LLVM kickstarted GCC development and made it better in thousand ways better, agile and flexible-minded (in terms of developers maintaining it).
Similarly, having two Rust compilers, nudging each other to be better is a great way to make a language more general, widely accepted, resilient and more free & open. Also Rust well be much more stressed and tested as a result.
> I hope it will be used only where necessary
I hope both will be used widely, and make Rust a better language with much better accessibility.
There is - between langs.
Compiler engineers can use ideas from other lang's compilers.
C# e.g
Programming languages with single compilers tend to be not used in my area of work and research (HPC and Scientific Computing). Hence, I'd rather have multiple compatible, yet independently implemented compilers, per language.
Yeah, it's not like Google (and dozens of others) has/have an internal fork of Linux, which never gets merged back. I'm not certain the GPL butters any parsnips here.
> Hence, I'd rather have multiple compatible, yet independently implemented compilers, per language.
You identified it. The problem is compatibility, and again Linux is a good negative example where ostensibly we have a language spec and multiple compilers, but Linux for years would only compile on one compiler.
Multiple compilers and "competition" doesn't solve compatibility. It makes it more difficult. Although I'm sure there will be benefits, I'm not certain the GPL/GCC will make compatibility any easier. It hasn't so far.
Doesn't matter for completely internal stuff. My concern is public code requiring non-public toolchains to compile. I have experienced this enough during the decades. I don't want to fight with this again.
Google doesn't want to touch Linux for their devices anymore, so they're building Fuchsia, but let's not digress...
> Multiple compilers and "competition" doesn't solve compatibility. It makes it more difficult. Although I'm sure there will be benefits, I'm not certain GPL will make compatibility easier. It hasn't so far.
What GPL brings to the table is strict openness, not compatibility. I can share my code and say that "It builds right on this $GPLd_Compiler", and people can get it and build it. This is what I like to bring to the table for anything I release.
I also wonder whether people would be this reactive to this issue if $company announced a closed source compiler with strict rustc compatibility.
I kinda wonder where this has been an issue with LLVM re: non-public toolchains, and public code where it wouldn't also be an issue with GCC. FWIW I definitely could see it being an issue in HPC/graphics/ML.
> I also wonder whether people would be this reactive to this issue if $company announced a closed source compiler with strict rustc compatibility.
Call me dense -- I guess I don't understand precisely what you're getting at. I don't think anyone would use a closed source compiler with strict rustc compatibility without a pretty big carrot. I think the issue here is strictly compatibility/divergence. Although I'm personally less of a fan of GCC/GPL/FSF, I think a new compiler is great so long as it doesn't (completely) ruin some of the nice things about a single implementation system, and why rust_codegen_gcc seems much more appealing.
Many Rust people don't want to live in the C/C++ compiler world, because they don't have to. "Oops this code won't build with GCC" is a battle they'd just as soon avoid.
We wanted to use these thing in newer systems, but it was not always possible, and made our lives way harder than it should be. At the end of the day, you buy vendor's hardware and want to couple it with more modern software, but the vendor doesn't want this for some unknown reason. Planned obsolescence, maybe.
Now, consider that a company pulls the same shenanigans, but by moving the magic to the compiler. All source is open, but can't be compiled because, while the code valid, its correct compilation needs a specific compiler behavior, and the vendor who opened the source cannot share the compiler in source form due to "trade secrets". They may require you to send the code in to "make sure that it's fine", or "sign an agreement" to get the compiler, for a fee, maybe.
I don't want to live in this world, or leave anyone accidentally in that state.
> Many Rust people don't want to live in the C/C++ compiler world, because they don't have to. "Oops this code won't build with GCC" is a battle they'd just as soon avoid.
As someone else noted, gccrs people say that "We strive for rustc compatibility. We will not extend or mangle the language, and consider rustc as our test suite." This is a pretty strong commitment to "many implementations, single behavior" promise.
I started my computing journey in a very open and flexible ecosystem, and with every stopgap put as in the name of security, this openness and flexibility eroded step by step. With this pace, programming and development will be confined to corporations which design the hardware and OS running on them, and PCs will be consoles with keyboards.
I don't want to live in such future. This is why FSF/GPL is important, for me.
I dont think it is valid concern in $currentYear
Nowadays languages like e.g c# are partially created by community and foundations (.net foundation) on github
So not only isnt it fully tied to company, but also it is oss.
>Hence, I'd rather have multiple compatible, yet independently implemented compilers, per language.
I dont because it permanently decreases developer experience and the value added is not something that could not be achieved with single compiler
I mean if you see room for improvement then go ahead and make PR instead of creating yet another compiler with its own bugs, quirks and wtfs
I don't care about $currentYear. My concern is about $nextDecade earliest, and future doesn't look bright from my perspective.
> Nowadays languages like e.g c# are partially created by community and foundations (.net foundation) on github
A foundation created by Microsoft for a Microsoft's cash cow language on Microsoft's OSS entanglement platform. Yes.
> but also it is oss.
I'm not talking about OSS, I'm talking about Free Software.
> I dont because it permanently decreases developer experience and the value added is not something that could not be achieved with single compiler
This is not what happened with clang vs g++/gcc. Also we have Intel's, Microsoft's and Portland groups compilers side by side for decades.
> I mean if you see room for improvement then go ahead and make PR instead of creating yet another compiler with its own bugs, quirks and wtfs
I prefer implementing language correctly in a compiler to force others to the same, spec-compliant behavior.
Otherwise compilers and languages divert, and lack of alternatives make everything much more complicated in the long run.
What? Cpps ecosystem is the most developer hostile ecosystem that Ive ever used
I wouldnt want to write in it even if my salary was multiplied by 1.5
A lot of compilers where all of them have different lang. features implemented, various bugs, quirks, wtfs, various perf. characteristics (e.g intel's generating faster code)
Long as hell compilation times. This is fundamental problem yet still that diversity didnt manage to solve it
Many package managers
And mediocre lang by modern standards on top of that. Minefield++ would fit it better.
>Otherwise compilers and languages divert, and lack of alternatives make everything much more complicated in the long run.
Theres still competition between langs
I bet Rust helped cpp improve more than gcc clang diversity
I also dont fully understand why you treat lang as a standard instead of product
The ecosystem's state is different depending on your perspective and what you are trying to achieve with it. While C++ is touted as a general purpose language, it's overkill for most of today's tasks. I use C++ to write high performance and low level code. It's my favorite language to work with, but I'm not using it blindly for everything.
As a person who worked with C++ a lot, and still developing something (which I'll eventually open source) for almost a decade, I can say that most of the things you say are not completely correct.
> different lang. features implemented, various bugs, quirks, wtfs, various perf. characteristics...
You can opt to not use any of the compiler extensions and have a completely portable code. Said code compiles in GCC, LLVM on macOS and possibly in MSVC, and uses CPU at its highest potential (verified on Linux with perf, in turn verified by timing on macOS). Intel's compiler creates fast code for Intel's processors, but it needs intel's patched libraries, etc. It's not worth it most of the time, and GCC already creates pretty fast code. In my discipline compiler cross-compatibility is king. So Intel's is not that important for our case.
> Long as hell compilation times. This is fundamental problem yet still that diversity didnt manage to solve it.
You can re-compile only the changed parts in any compiler for ages. This shaves a lot of time from compilation and linking times. Again this is neither novel, nor new.
> Many package managers
The idea of packages born with Java for enterprise software and became ubiquitous after that. Languages have their own package managers to prevent this, but it doesn't always work. Having an "official" package manager doesn't prevent from bringing in a "New and Improved (TM)" package manager to the ecosystem. Also, Rust is a young language. C++ is well, 35 years old? It's inevitable that we have better solutions for the issues we have at hand.
> And mediocre lang by modern standards on top of that. Minefield++ would fit it better.
It's a language with unfettered low level access. It's a sharp knife (or fighter jet). Learn how to handle it, and it won't bite. It's never guaranteed to be fluffy like Go, Python or memory-safe (to a point) like Rust by adding hard barriers.
On the other hand, being semi-conscious about what you're doing prevents 99% of the problems in C++. Said code had memory leak once, during initial development, and fixed on the spot (I test my code with Valgrind), and is memory-sane since that day. That code is not something simple. It's a scientific application which makes your cores and memory controller saturate. It takes every bit of performance offered by your system and converts into science.
> Theres still competition between langs
Yes, but that's another matter, which is not subject of this discussion.
> I also dont fully understand why you treat lang as a standard instead of product.
Because a language is a language. Not different from mathematical notation or human language. Compiler is the product, which transpiles that language to another language (incl. machine code) to run on your PC, mouse, microwave or plant based biologic computer. Compiler is the product. Language is just a language. Sets of rules, a standard even (like C++ is an ISO standard).
I'm interested in this. Could you elaborate on why you are not using Rust? Because there is no competition in the language, or because there is not an alternate compiler, or because you don't like LLVM?
I don't want my software dependent on a non-GPL toolchain to ensure its long term sustainability. I also try to choose GPL licensed libraries, and include everything required to build my code (incl. external libraries) inside my repository.
In the end I want someone to be able to just clone the repo, and run make, or GCC, or what it needs to be run. So, the code is just dependent on the compiler and its standard libraries when obtained by someone.
Lastly, I find LLVM's supporters and general ecosystem slightly hostile towards GCC, and want to replace it anywhere and everywhere possible. I'm a free software advocate, and I prefer LLVM and GCC just to be interchangeable.
Hence I only use languages which GCC supports exclusively or has GPL compilers/interpreters. Working on HPC makes this easier for me.
I responded in a similar way to a conversation about Julia less than a week ago.
For people who have trouble understanding this reasoning, a message from the GCC mailing list (https://gcc.gnu.org/legacy-ml/gcc/2014-01/msg00247.html) provides some of the explanation, which I reproduce here:
In the free software movement, we campaign for the freedom of the users of computing. The values of free software are fundamentally different from the values of open source, which make "better code" the ultimate goal. If GCC were to change from a free compiler into a platform for nonfree compilers, it would no longer serve the goal of freedom very well. Therefore, we had to take care to prevent that.
(See http://www.gnu.org/philosophy/open-source-misses-the-point.h... for more explanation of the difference between free software and open source. See also https://thebaffler.com/salvos/the-meme-hustler for Evgeny Morozov's article on the same point.)
The Clang and LLVM developers reach different conclusions from ours because they do not share our values and goals. They object to the measures we have taken to defend freedom because they see the inconvenience of them and do not recognize (or don't care about) the need for them. I would guess they describe their work as "open source" and do not talk about freedom. They have been supported by Apple, the company which hates our freedom so much that its app store for the ithings _requires_ all apps to be nonfree. (*)
The nonfree compilers that are now based on LLVM prove that I was right -- that the danger was real. If I had "opened" up GCC code for use in nonfree combinations, that would not have prevented a defeat; rather, it would have caused that defeat to occur very soon.
For GCC to be replaced by another technically superior compiler that defended freedom equally well would cause me some personal regret, but I would rejoice for the community's advance. The existence of LLVM is a terrible setback for our community precisely because it is not copylefted and can be used as the basis for nonfree compilers -- so that all contribution to LLVM directly helps proprietary software as much as it helps us.
The cause of the setback is the existence of a non-copylefted compiler that therefore becomes the base for nonfree compilers. The identity of that compiler -- whether it be LLVM, GCC, or something else -- is a secondary detail. To make GCC available for such use would be throwing in the towel. If that enables GCC to "win", the victory would be hollow, because it would not be a victory for what really matters: users' freedom.
If you think we ought to "compromise" on this point, please see http://www.gnu.org/philosophy/compromise.html.
The only code that helps us and not our adversaries is copylefted code. Free software released under a pushover license is available for us to use, but available to our adversaries just as well. If you want your work to give freedom an advantage, use the leverage available to you -- copyleft your code. I invite those working on major add-ons to LLVM to release them under GNU GPL version-3-or-later.
* If a binary is made from published source code, but you can't install your binary of a modified version of that source code, the binary is proprietary even if the source code is free. (See http://www.gnu.org/philosophy/free-sw.html.) A binary in Apple's app store may be made from published free source code, but under Apple's rules and Apple's DRM, the binary can't be free.
I have to agree re: the parent. It's a perfectly fine opinion to hold, though it's one I don't share. I asked and he/she told me.
However, republishing an email by Richard Stallman, just like Richard Stallman, might seem a little excessive.
Why? I think it's because although just as it's fine to be an atheist, it always seems weird when someone brings it up at dinner parties. We were discussing the whys and wherefores of a new Rust compiler. I'm not sure any lengthy expressions of our licensing religion or lack thereof is a positive contribution.
First, GCC as a whole has never been technically superior to the commercial competition. The reason why GCC was popular outside the GNU crowd and received as much funded development is that it was the best compiler across a range of popular platforms and had the best price, too.
Second, one of the main drivers behind the creation of clang has been the hostility to making the C/C++ frontend of GCC usable as a standalone tool. Richard explicitly wanted to force everyone to either deal with the GPL and his interpretation of what is derived work, or shell off the bucks for the Edision Design Group. This was a calculated choice as no one would want to invest the time to write another C/C++ frontend, right? He intentionally ignored all other use cases for C/C++ parser as a library (Language server, anyone? Refactoring tools?) and thereby actually hurt the user base on the GNU platform just as well. It can not be understated how disruptive clang was as a set of libraries outside the LLVM/clang compiler.
Third, GCC had as middle and backend compiler always was moderately accessible, but by structure somewhat harder to access for researchers or people wanting on special purpose tools like optimized code generation for computational kernels on matrixes.
GCC was forced to acknowledge the competition by finally getting a plugin interface in the GCC 4.5 days, but at that point, the cat was out of the bag.
It means that crates.io will have crates which compile only with gcc and its extensions.
Unless you mean some specific chips?
The latest xtensa-esp8266-elf-gcc is version 5.2
You can see something of a summary over on https://builds.Debian.org/status/package.php?p=rustc&suite=s... the BD-Uninstallable entries are unsupported archs, though some of them might be due to rustc or bringup bugs (e.g. m68k is supposedly supported by llvm and rustc).
Then there’s embedded, on the more open side LLVM is slowly gaining ground but there’s also less open ecosystems where customising / forking gcc is pretty standard historically, and there’s not much LLVM can do.
If there are people who care about these historical architectures, they can do the work to maintain them in LLVM. That's how m68k support was added.
Or they can decide to use an existing compiler that already works.
There are also embedded applications, of course; plenty of crappy manufacturers still give you a GCC fork as your only compiler for their dedicated hardware. The ability to get Rust hooked into those compilers can help keep the ecosystem up to date.
Even companies maintaining their own software sometimes take a long time to add support to LLVM. Take for example Espressif's Xtensa fork that is only now starting to get merged (if https://discourse.llvm.org/t/rfc-request-for-upstream-tensil... is to be believed). Many patches are still waiting for review, so it'll be a while until LLVM finally supports Xtensa (and with it microcontrollers such as the ESP32). GCC has had Xtensa support for ages (Github lists xtensa.h going back all the way to 2002) and I doubt Espressif is the only company in this position.
"Just do the work" sounds easy but that work can take years. Getting modern Linux versions to compile may not be enough to drive companies to put in the effort; many may prefer to simply never update their kernels past the current LTS version again.
Adding a frontend to GCC works around this problem quite efficiently. It also makes it easier to port existing code to these platforms; sure, the lack of a borrow checker drops all guarantees Rust has been designed to provide, but if you're not writing any code, you don't need those anyway.
Non-standard C-like languages are used in computer graphics and GPU computing (OpenGL, Direct3D, OpenCL), but strictly speaking they don't count as C.
It was kept around for a long time after everyone went to OpenJDK, because it was the only project with certain unit tests for GCC code paths, when that was eventually sorted out, it was when they dropped it.
I'd also note that any Rust you compile with GCC probably shouldn't be dynamically linked with Rust code compiled with rustc.
That is one of the things that rustc is severely lacking.