Any distros providing a packages llgo?

<p>Waited an hour for compiling llvm-go to find out that llgo is what I needed to make llvm bc with llgo. Went to compile logo and found the instructions are suboptimal. </p> <p>Any one know of and distro I can just get it on in a package? I&#39;m just going to have to setup a build environment in docker. </p> <p>Ultimately I&#39;m trying to just expose an API in JavaScript that includes some CGO imports. I really don&#39;t want to maintain multiple code bases if I can avoid it. </p> <p>Side note: I did see an existing docker image, but it&#39;s already dated. Looking to be in the lastest and greatest.</p> <hr/>**评论：**<br/><br/>dlsniper: <pre><blockquote> <p>Ultimately I&#39;m trying to just expose an API in JavaScript that includes some CGO imports.</p> </blockquote> <p>Why do you need llgo and not use the Go toolchain for that?</p> <p>Also afaik llgo is not yet a blessed implementation and by the work it goes into it it looks like it still has a few legs to go before that (I might be wrong).</p></pre>igknighted: <pre><p>I want to evaluate the feasibility of using a webassembly port that I compile with llgo -&gt; emscripten. </p> <p>While it may not be blessed, I don&#39;t really think many people are even bothering with it. Personally reading over it, I believe the llgo impementation may be enough to bridge cryptography API implementations and am just aiming to test it out before making a decision to write multiple code bases. </p> <p>I&#39;m working on an open source project that I am aiming to provide wide support for, but I&#39;m only one guy. Being able to manage 1 code base and maintain high portability is pretty important to me. </p></pre>sbinet: <pre><p>gollvm might be a better horse to bet on than llgo (no disrespect to the llgo devs) gollvm has a few go googlers working on it...</p></pre>dlsniper: <pre><p>Oh, I see, that wasn&#39;t clear from the initial text.</p> <p>Maybe you can post the instructions / a Dockerfile which is useful to compile llgo and al that&#39;s needed for it? </p></pre>igknighted: <pre><p>This is what I was working from: <a href="http://llvm.org/svn/llvm-project/llgo/trunk/README.TXT" rel="nofollow">http://llvm.org/svn/llvm-project/llgo/trunk/README.TXT</a> </p> <p>When I went to build it I ended up being ultimately stuck on the last cmake. I&#39;m not really familiar with cmake. I&#39;m more of a <code>./configure &amp;&amp; make build</code> kinda guy. </p></pre>epiris: <pre><p>Stuck in what way? Build output may give me a hint at a quick glance. What Linux distro you run? I wouldn&#39;t compile llvm in a docker container myself, a lot of the defaults are gonna leave you resource starved. I&#39;ve never looked at or used llgo, but the link shows it as a top lvl prj for llvm so maybe the most recent Jenkins builds will be in the llvm lab. Will give you some hints maybe on any flags / dep version combinations that are known to work harmoniously. Best of luck, though crypto might not be the safest area to play with lol.</p></pre>daenney: <pre><p>Considering the start of the README that states llgo is not feature complete yet I doubt anyone has picked it up and packaged it.</p> <blockquote> <p>llgo is under active development. It compiles and passes most of the standard library test suite and a substantial portion of the gc test suite, but there are some corner cases that are known not to be handled correctly yet. Nevertheless it can compile modestly substantial programs (including itself; it is self hosting on x86-64 Linux).</p> </blockquote> <p>If you&#39;re looking to get Go code to run in WebAssembly though, <a href="https://github.com/go-interpreter/wagon" rel="nofollow">wagon</a> or <a href="https://github.com/tsavola/wag" rel="nofollow">wag</a> might interest you.</p></pre>igknighted: <pre><p>A lot of people have suggested that to other people, but that looks like an interpreter written in go for running llvm byte code. Did I miss something when checking that out? </p></pre>tiborvass: <pre><p>Almost: it is to run interpreted wasm code, not llvm byte code.</p></pre>igknighted: <pre><p>I see. So it&#39;s definitely not a solution for the end goal. I really like gopherjs. I just wish things didn&#39;t get slower when using encryption algorithms. </p></pre>