reynir/blog.robur.coop
1
0
Fork 0
forked from robur/blog.robur.coop
Code Pull requests Activity

Compare commits

base: reynir:yocaml2
Branches Tags
reynir:main
reynir:teeest
reynir:gh-pages
reynir:yocaml2
robur:gh-pages
robur:main
..
compare: reynir:main
Branches Tags
reynir:teeest
reynir:gh-pages
reynir:main
reynir:yocaml2
robur:gh-pages
robur:main

1 commit
yocaml2 ... main

Author SHA1 Message Date
The Robur Team
9b38bbc751 Pushed by YOCaml 2 2024-10-04 10:48:33 +00:00
50 changed files with 3232 additions and 822 deletions
Show stats Expand all files Collapse all files

3
.gitignore vendored
View file

@ -1,3 +0,0 @@
_build/
_site/
_cache

13
README.md
View file

@ -10,26 +10,25 @@ $ git clone git@git.robur.coop:robur/blog.robur.coop
$ cd blog.robur.coop/ $ cd blog.robur.coop/
$ opam pin add -yn . $ opam pin add -yn .
$ opam install --deps-only blogger $ opam install --deps-only blogger
$ dune exec src/watch.exe -- $ dune exec src/blogger.exe -- watch
``` ```
A little server run on `http://localhost:8000`. A little server run on `http://localhost:8888`.
The user can add an article into the `articles/` directory. The format is easy. The user can add an article into the `articles/` directory. The format is easy.
A simple header which starts with `---` and finish with `---`. Inside, you have A simple header which starts with `---` and finish with `---`. Inside, you have
a YAML description of the article where some fields are required: a YAML description of the article where some fields are required:
- `date` - `date`
- `title` - `article.title`
- `description` - `article.description`
- `tags` - `tags`
You can specify an `author` (with its `name`, `email` and `link`) or not. By You can specify an `author` (with its `name`, `email` and `link`) or not. By
default, we use `team@robur.coop`. If everything looks good, you can generate default, we use `team@robur.coop`. If everything looks good, you can generate
via the `blogger.exe` tool the generated website via: via the `blogger.exe` tool the generated website via:
```shell-session ```shell-session
$ dune exec src/push.exe -- push \ $ dune exec src/blogger.exe -- push \
-r git@git.robur.coop:robur/blog.robur.coop.git#gh-pages \ -r git@git.robur.coop:robur/blog.robur.coop.git#gh-pages
--host https://blog.robur.coop
[--name "The Robur team"] \ [--name "The Robur team"] \
[--email team@robur.coop] [--email team@robur.coop]
``` ```

278
articles/2024-02-03-python-str-repr.html Normal file
View file

@ -0,0 +1,278 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogPython&apos;s `str.__repr__()`
</title>
<meta name="description" content="Reimplementing Python string escaping in OCaml">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>Python&apos;s `str.__repr__()`</h1>
<ul class="tags-list"><li>OCaml</li><li>Python</li><li>unicode</li></ul><p>Sometimes software is written using whatever built-ins you find in your programming language of choice.
This is usually great!
However, it can happen that you depend on the precise semantics of those built-ins.
This can be a problem if those semantics become important to your software and you need to port it to another programming language.
This story is about Python and its <code>str.__repr()__</code> function.</p>
<p>The piece of software I was helping port to <a href="https://ocaml.org/">OCaml</a> was constructing a hash from the string representation of a tuple.
The gist of it was basically this:</p>
<pre><code class="language-python">def get_id(x):
id = (x.get_unique_string(), x.path, x.name)
return myhash(str(id))
</code></pre>
<p>In other words it's a Python tuple consisting of mostly strings but also a <code>PosixPath</code> object.
The way <code>str()</code> works is it calls the <code>__str__()</code> method on the argument objects (or otherwise <code>repr(x)</code>).
For Python tuples the <code>__str__()</code> method seems to print the result of <code>repr()</code> on each elemenet separated by a comma and a space and surrounded by parenthesis.
So good so far.
If we can precisely emulate <code>repr()</code> on strings and <code>PosixPath</code> it's easy.
In the case of <code>PosixPath</code> it's really just <code>'PosixPath('+repr(str(path))+')'</code>;
so in that case it's down to <code>repr()</code> on strings - which is <code>str.__repr__()</code>,</p>
<p>There had been a previous attempt at this that would use OCaml's string escape functions and surround the string with single quotes (<code>'</code>).
This works for some cases, but not if the string has a double quote (<code>&quot;</code>).
In that case OCaml would escape the double quote with a backslash (<code>\&quot;</code>) while python would not escape it.
So a regular expression substitution was added to replace the escape sequence with just a double quote.
This pattern of finding small differences between Python and OCaml escaping had been repeated,
and eventually I decided to take a more rigorous approach to it.</p>
<h2 id="what-is-a-string"><a class="anchor" aria-hidden="true" href="#what-is-a-string"></a>What is a string?</h2>
<p>First of all, what is a string? In Python? And in OCaml?
In OCaml a string is just a sequence of bytes.
Any bytes, even <code>NUL</code> bytes.
There is no concept of unicode in OCaml strings.<br>
In Python there is the <code>str</code> type which is a sequence of Unicode code points[^python-bytes].
I can recommend reading Daniel Bünzli's <a href="https://ocaml.org/p/uucp/13.0.0/doc/unicode.html#minimal">minimal introduction to Unicode</a>.
Already here there is a significant gap in semantics between Python and OCaml.
For many practical purposes we can get away with using the OCaml <code>string</code> type and treating it as a UTF-8 encoded Unicode string.
This is what I will do as in both the Python code and the OCaml code the data being read is a UTF-8 (or often only the US ASCII subset) encoded string.</p>
<h2 id="what-does-a-string-literal-look-like"><a class="anchor" aria-hidden="true" href="#what-does-a-string-literal-look-like"></a>What does a string literal look like?</h2>
<h3 id="ocaml"><a class="anchor" aria-hidden="true" href="#ocaml"></a>OCaml</h3>
<p>I will not dive too deep into the details of OCaml string literals, and focus mostly on how they are escaped by the language built-ins (<code>String.escaped</code>, <code>Printf.printf &quot;%S&quot;</code>).
Normal printable ASCII is printed as-is.
That is, letters, numbers and other symbols except for backslash and double quote.
There are the usual escape sequences <code>\n</code>, <code>\t</code>, <code>\r</code>, <code>\&quot;</code> and <code>\\</code>.
Any byte value can be represented with decimal notation <code>\032</code> or octal notation '\o040' or hexadecimal notation <code>\x20</code>.
The escape functions in OCaml has a preference for the decimal notation over the hexadecimal notation.
Finally I also want to mention the Unicode code point escape sequence <code>\u{3bb}</code> which represents the UTF-8 encoding of U+3BB.
While the escape functions do not use it, it will become handy later on.
Illegal escape sequences (escape sequences that are not recognized) will emit a warning but otherwise result in the escape sequence as-is.
It is common to compile OCaml programs with warnings-as-errors, however.</p>
<h3 id="python"><a class="anchor" aria-hidden="true" href="#python"></a>Python</h3>
<p>Python has a number of different string literals and string-like literals.
They all use single quote or double quote to delimit the string (or string-like) literals.
There is a preference towards single quotes in <code>str.__repr__()</code>.
You can also triple the quotes if you like to write a string that uses a lot of both quote characters.
This format is not used by <code>str.__repr__()</code> so I will not cover it further, but you can read about it in the <a href="https://docs.python.org/3/reference/lexical_analysis.html#strings">Python reference manual</a>.
The string literal can optionally have a prefix character that modifies what type the string literal is and how its content is interpreted.</p>
<p>The <code>r</code>-prefixed strings are called <em>raw strings</em>.
That means backslash escape sequences are not interpreted.
In my experiments they seem to be quasi-interpreted, however!
The string <code>r&quot;\&quot;</code> is considered unterminated!
But <code>r&quot;\&quot;&quot;</code> is fine as is interpreted as <code>'\\&quot;'</code>[^raw-escape-example].
Why this is the case I have not found a good explanation for.</p>
<p>The <code>b</code>-prefixed strings are <code>bytes</code> literals.
This is close to OCaml strings.</p>
<p>Finally there are the unprefixed strings which are <code>str</code> literals.
These are the ones we are most interested in.
They use the usual escape <code>\[ntr&quot;]</code> we know from OCaml as well as <code>\'</code>.
<code>\032</code> is <strong>octal</strong> notation and <code>\x20</code> is hexadecimal notation.
There is as far as I know <strong>no</strong> decimal notation.
The output of <code>str.__repr__()</code> uses the hexadecimal notation over the octal notation.
As Python strings are Unicode code point sequences we need more than two hexadecimal digits to be able to represent all valid &quot;characters&quot;.
Thus there are the longer <code>\u0032</code> and the longest <code>\U00000032</code>.</p>
<h2 id="intermezzo"><a class="anchor" aria-hidden="true" href="#intermezzo"></a>Intermezzo</h2>
<p>While studying Python string literals I discovered several odd corners of the syntax and semantics besides the raw string quasi-escape sequence mentioned earlier.
One fact is that Python doesn't have a separate character or Unicode code point type.
Instead, a character is a one element string.
This leads to some interesting indexing shenanigans: <code>&quot;a&quot;[0][0][0] == &quot;a&quot;</code>.
Furthermore, strings separated by spaces only are treated as one single concatenated string: <code>&quot;a&quot; &quot;b&quot; &quot;c&quot; == &quot;abc&quot;</code>.
These two combined makes it possible to write this unusual snippet: <code>&quot;a&quot; &quot;b&quot; &quot;c&quot;[0] == &quot;a&quot;</code>!
For byte sequences, or <code>b</code>-prefixed strings, things are different.
Indexing a bytes object returns the integer value of that byte (or character):</p>
<pre><code class="language-python">&gt;&gt;&gt; b&quot;a&quot;[0]
97
&gt;&gt;&gt; b&quot;a&quot;[0][0]
&lt;stdin&gt;:1: SyntaxWarning: 'int' object is not subscriptable; perhaps you missed a comma?
Traceback (most recent call last):
File &quot;&lt;stdin&gt;&quot;, line 1, in &lt;module&gt;
TypeError: 'int' object is not subscriptable
</code></pre>
<p>For strings <code>\x32</code> can be said to be shorthand for <code>&quot;\u0032&quot;</code> (or <code>&quot;\u00000032&quot;</code>).
But for bytes <code>&quot;\x32&quot; != &quot;\u0032&quot;</code>!
Why is this?!
Well, bytes is a byte sequence and <code>b&quot;\u0032&quot;</code> is not interpreted as an escape sequence and is instead <strong>silently</strong> treated as <code>b&quot;\\u0032&quot;</code>!
Writing <code>&quot;\xff&quot;.encode()</code> which encodes the string <code>&quot;\xff&quot;</code> to UTF-8 is <strong>not</strong> the same as <code>b&quot;\xff&quot;</code>.
The bytes <code>&quot;\xff&quot;</code> consist of a single byte with decimal value 255,
and the Unicode wizards reading will know that the Unicode code point 255 (or U+FF) is encoded in two bytes in UTF-8.</p>
<h2 id="where-is-the-python-code"><a class="anchor" aria-hidden="true" href="#where-is-the-python-code"></a>Where is the Python code?</h2>
<p>Finding the implementation of <code>str.__repr__()</code> turned out to not be so easy.
In the end I asked on the Internet and got a link to <a href="https://github.com/python/cpython/blob/963904335e579bfe39101adf3fd6a0cf705975ff/Objects/unicodeobject.c#L12245-L12405">cpython's <code>Objects/unicodeobject.c</code></a>.
And holy cow!
That's some 160 lines of C code with two loops, a switch statement and I don't know how many chained and nested if statements!
Meanwhile the OCaml implementation is a much less daunting 52 lines of which about a fifth is a long comment.
It also has two loops which each contain one much more tame match expression (roughly a C switch statement).
In both cases they first loop over the string to compute the size of the output string.
The Python implementation also counts the number of double quotes and single quotes as well as the highest code point value.
The latter I'm not sure why they do, but my guess it's so they can choose an efficient internal representation.
Then the Python code decides what quote character to use with the following algorithm:<br>
Does the string contain single quotes but no double quotes? Then use double quotes. Otherwise use single quotes.
Then the output size estimate is adjusted with the number of backslashes to escape the quote character chosen and the two quotes surrounding the string.</p>
<p>Already here it's clear that a regular expression substitution is not enough by itself to fix OCaml escaping to be Python escaping.
My first step then was to implement the algorithm only for US ASCII.
This is simpler as we don't have to worry much about Unicode, and I could implement it relatively quickly.
The first 32 characters and the last US ASCII character (DEL or <code>\x7f</code>) are considered non-printable and must be escaped.
I then wrote some simple tests by hand.
Then I discovered the OCaml <a href="https://github.com/zshipko/ocaml-py">py</a> library which provides bindings to Python from OCaml.
Great! This I can use to test my implementation against Python!</p>
<h2 id="how-about-unicode"><a class="anchor" aria-hidden="true" href="#how-about-unicode"></a>How about Unicode?</h2>
<p>For the non-ascii characters (or code points rather) they are either considered <em>printable</em> or <em>non-printable</em>.
For now let's look at what that means for the output.
A printable character is copied as-is.
That is, there is no escaping done.
Non-printable characters must be escaped, and python wil use <code>\xHH</code>, <code>\uHHHH</code> or <code>\UHHHHHHHH</code> depending on how many hexadecimal digits are necessary to represent the code point.
That is, the latin-1 subset of ASCII (<code>0x80</code>-<code>0xff</code>) can be represented using <code>\xHH</code> and neither <code>\u00HH</code> nor <code>\U000000HH</code> will be used etc.</p>
<h3 id="what-is-a-printable-unicode-character"><a class="anchor" aria-hidden="true" href="#what-is-a-printable-unicode-character"></a>What is a printable Unicode character?</h3>
<p>In the cpython <a href="https://github.com/python/cpython/blob/963904335e579bfe39101adf3fd6a0cf705975ff/Objects/unicodeobject.c#L12245-L12405">function</a> mentioned earlier they use the function <code>Py_UNICODE_ISPRINTABLE</code>.
I had a local clone of the cpython git repository where I ran <code>git grep Py_UNICODE_ISPRINTABLE</code> to find information about it.
In <a href="https://github.com/python/cpython/blob/963904335e579bfe39101adf3fd6a0cf705975ff/Doc/c-api/unicode.rst?plain=1#L257-L265">unicode.rst</a> I found a documentation string for the function that describes it to return false if the character is nonprintable with the definition of nonprintable as the code point being in the categories &quot;Other&quot; or &quot;Separator&quot; in the Unicode character database <strong>with the exception of ASCII space</strong> (U+20 or <code> </code>).</p>
<p>What are those &quot;Other&quot; and &quot;Separator&quot; categories?
Further searching for the function definition we find in <a href="https://github.com/python/cpython/blob/963904335e579bfe39101adf3fd6a0cf705975ff/Include/cpython/unicodeobject.h#L683"><code>Include/cpython/unicodeobject.h</code></a> the definition.
Well, we find <code>#define Py_UNICODE_ISPRINTABLE(ch) _PyUnicode_IsPrintable(ch)</code>.
On to <code>git grep _PyUnicode_IsPrintable</code> then.
That function is defined in <a href="https://github.com/python/cpython/blob/963904335e579bfe39101adf3fd6a0cf705975ff/Objects/unicodectype.c#L158-L163"><code>Objects/unicodectype.c</code></a>.</p>
<pre><code class="language-C">/* Returns 1 for Unicode characters to be hex-escaped when repr()ed,
0 otherwise.
All characters except those characters defined in the Unicode character
database as following categories are considered printable.
* Cc (Other, Control)
* Cf (Other, Format)
* Cs (Other, Surrogate)
* Co (Other, Private Use)
* Cn (Other, Not Assigned)
* Zl Separator, Line ('\u2028', LINE SEPARATOR)
* Zp Separator, Paragraph ('\u2029', PARAGRAPH SEPARATOR)
* Zs (Separator, Space) other than ASCII space('\x20').
*/
int _PyUnicode_IsPrintable(Py_UCS4 ch)
{
const _PyUnicode_TypeRecord *ctype = gettyperecord(ch);
return (ctype-&gt;flags &amp; PRINTABLE_MASK) != 0;
}
</code></pre>
<p>Ok, now we're getting close to something.
Searching for <code>PRINTABLE_MASK</code> we find in <a href="https://github.com/python/cpython/blob/963904335e579bfe39101adf3fd6a0cf705975ff/Tools/unicode/makeunicodedata.py#L450-L451"><code>Tools/unicode/makeunicodedata.py</code></a> the following line of code:</p>
<pre><code class="language-Python">if char == ord(&quot; &quot;) or category[0] not in (&quot;C&quot;, &quot;Z&quot;):
flags |= PRINTABLE_MASK
</code></pre>
<p>So the algorithm is really if the character is a space character or if its Unicode general category doesn't start with a <code>C</code> or <code>Z</code>.
This can be implemented in OCaml using the uucp library as follows:</p>
<pre><code class="language-OCaml">let py_unicode_isprintable uchar =
(* {[if char == ord(&quot; &quot;) or category[0] not in (&quot;C&quot;, &quot;Z&quot;):
flags |= PRINTABLE_MASK]} *)
Uchar.equal uchar (Uchar.of_char ' ')
||
let gc = Uucp.Gc.general_category uchar in
(* Not those categories starting with 'C' or 'Z' *)
match gc with
| `Cc | `Cf | `Cn | `Co | `Cs | `Zl | `Zp | `Zs -&gt; false
| `Ll | `Lm | `Lo | `Lt | `Lu | `Mc | `Me | `Mn | `Nd | `Nl | `No | `Pc | `Pd
| `Pe | `Pf | `Pi | `Po | `Ps | `Sc | `Sk | `Sm | `So -&gt;
true
</code></pre>
<p>After implementing unicode I expanded the tests to generate arbitrary OCaml strings and compare the results of calling my function and Python's <code>str.__repr__()</code> on the string.
Well, that didn't go quite well.
OCaml strings are just any byte sequence, and ocaml-py expects it to be a UTF-8 encoded string and fails on invalid UTF-8.
Then in qcheck you can &quot;assume&quot; a predicate which means if a predicate doesn't hold on the generated value then the test is skipped for that input.
So I implement a simple verification of UTF-8.
This is far from optimal because qcheck will generate a lot of invalid utf-8 strings.</p>
<p>The next test failure is some unassigned code point.
So I add to <code>py_unicode_isprintable</code> a check that the code point is assigned using <code>Uucp.Age.age uchar &lt;&gt; `Unassigned</code>.</p>
<p>Still, qcheck found a case I hadn't considered: U+61D.
My python version (Python 3.9.2 (default, Feb 28 2021, 17:03:44)) renders this as <code>'\u061'</code> while my OCaml function prints it as-is.
In other words my implementation considers it printable while python does not.
I try to enter this Unicode character in my terminal, but nothing shows up.
Then I look it up and its name is <code>ARABIC END OF TEXT MARKER</code>.
The general category according to uucp is <code>`Po</code>.
So this <strong>should</strong> be a printable character‽</p>
<p>After being stumped by this for a while I get the suspicion it may be dependent on the Python version.
I am still on Debian 11 and my Python version is far from being the latest and greatest.
I ask someone with a newer Python version to write <code>'\u061d'</code> in a python session.
And 'lo! It prints something that looks like <code>''</code>!
Online I figure out how to get the unicode version compiled into Python:</p>
<pre><code class="language-Python">&gt;&gt;&gt; import unicodedata
&gt;&gt;&gt; unicodedata.unidata_version
'13.0.0'
</code></pre>
<p>Aha! And with uucp we find that the unicode version that introduced U+61D to be 14.0:</p>
<pre><code class="language-OCaml"># Uucp.Age.age (Uchar.of_int 0x61D);;
- : Uucp.Age.t = `Version (14, 0)
</code></pre>
<p>My reaction is this is seriously some ungodly mess we are in.
Not only is the code that instigated this journey highly dependent on Python-specifics - it's also dependent on the specific version of unicode and thus the version of Python!</p>
<p>I modify our <code>py_unicode_isprintable</code> function to take an optional <code>?unicode_version</code> argument and replace the &quot;is this unassigned?&quot; check with the following snippet:</p>
<pre><code class="language-OCaml">let age = Uucp.Age.age uchar in
(match (age, unicode_version) with
| `Unassigned, _ -&gt; false
| `Version _, None -&gt; true
| `Version (major, minor), Some (major', minor') -&gt;
major &lt; major' || (major = major' &amp;&amp; minor &lt;= minor'))
</code></pre>
<p>Great! I modify the test suite to first detect the unicode version python uses and then pass that version to the OCaml function.
Now I can't find anymore failing test cases!</p>
<h2 id="epilogue"><a class="anchor" aria-hidden="true" href="#epilogue"></a>Epilogue</h2>
<p>What can we learn from this?
It is easy to say in hindsight that a different representation should have been chosen.
However, arriving at this insight takes time.
The exact behavior of <code>str.__repr__()</code> is poorly documented.
Reaching my understanding of <code>str.__repr__()</code> took hours of research and reading the C implementation.
It often doesn't seem to be worth it to spend so much time on research for a small function.
Technical debt is a real thing and often hard to predict.
Below is the output of <code>help(str.__repr__)</code>:</p>
<pre><code class="language-Python">__repr__(self, /)
Return repr(self)
</code></pre>
<p>Language and (standard) library designers could consider whether the slightly nicer looking strings are worth the added complexity users eventually are going to rely on - inadvertently or not.
I do think strings and bytes in Python are a bit too complex.
It is not easy to get a language lawyer[^language-lawyer] level understanding.
In my opinion it is a mistake to not at least print a warning if there are illegal escape sequences - especially considering there are escape sequences that are valid in one string literal but not another.</p>
<p>Unfortunately it is often the case that to get a precise specification it is necessary to look at the implementation.
For testing your implementation hand-written tests are good.
Testing against the original implementation is great, and if combined with property-based testing or fuzzing you may find failing test cases you couldn't dream up!
I certainly didn't see it coming that the output depends on the Unicode version.
As is said, testing can only show the presence of bugs, but with a, in a sense, limited domain like this function you can get pretty close to showing absence of bugs.</p>
<p>I enjoyed working on this.
Sure, it was frustrating and at times I discovered some ungodly properties, but it's a great feeling to study and understand something at a deeper level.
It may be the last time I need to understand Python's <code>str.__repr__()</code> this well, but if I do I now have the OCaml code and this blog post to reread.</p>
<p>If you are curious to read the resulting code you may find it on github at <a href="https://github.com/reynir/python-str-repr">github.com/reynir/python-str-repr</a>.
I have documented the code to make it more approachable and maintainable by others.
Hopefully it is not something that you need, but in case it is useful to you it is licensed under a permissive license.</p>
<p>If you have a project in OCaml or want to port something to OCaml and would like help from me and my colleagues at <a href="https://robur.coop/">Robur</a> please <a href="https://robur.coop/Contact">get in touch</a> with us and we will figure something out.</p>
<p>[^python-bytes]: There is as well the <code>bytes</code> type which is a byte sequence like OCaml's <code>string</code>.
The Python code in question is using <code>str</code> however.</p>
<p>[^raw-escape-example]: Note I use single quotes for the output. This is what Python would do. It would be equivalent to <code>&quot;\\\&quot;&quot;</code>.</p>
<p>[^language-lawyer]: <a href="http://catb.org/jargon/html/L/language-lawyer.html">A person, usually an experienced or senior software engineer, who is intimately familiar with many or most of the numerous restrictions and features (both useful and esoteric) applicable to one or more computer programming languages. A language lawyer is distinguished by the ability to show you the five sentences scattered through a 200-plus-page manual that together imply the answer to your question “if only you had thought to look there”.</a></p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/2024-02-03-python-str-repr.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-02-03 date: 2024-02-03
title: Python's `str.__repr__()` article.title: Python's `str.__repr__()`
description: Reimplementing Python string escaping in OCaml article.description: Reimplementing Python string escaping in OCaml
tags: tags:
- OCaml - OCaml
- Python - Python

229
articles/2024-08-21-OpenVPN-and-MirageVPN.html Normal file
View file

@ -0,0 +1,229 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogMirageVPN and OpenVPN
</title>
<meta name="description" content="Discoveries made implementing MirageVPN, a OpenVPN-compatible VPN library">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>MirageVPN and OpenVPN</h1>
<ul class="tags-list"><li>MirageVPN</li><li>OpenVPN</li><li>security</li></ul><p>At <a href="https://robur.coop/">Robur</a> we have been busy at work implementing our OpenVPN™-compatible MirageVPN software.
Recently we have implemented the <a href="https://blog.robur.coop/articles/miragevpn-server.html">server side</a>.
In order to implement this side of the protocol I studied parts of the OpenVPN™ source code and performed experiments to understand what the implementation does at the protocol level.
Studying the OpenVPN™ implementation has lead me to discover two security issues: CVE-2024-28882 and CVE-2024-5594.
In this article I will talk about the relevant parts of the protocol, and describe the security issues in detail.</p>
<p>A VPN establishes a secure tunnel in which (usually) IP packets are sent.
The OpenVPN protocol establishes a TLS tunnel[^openvpn-tls] with which key material and configuration options are negotiated.
Once established the TLS tunnel is used to exchange so-called control channel messages.
They are NUL-terminated (well, more on that later) text messages sent in a single TLS record frame (mostly, more on that later).</p>
<p>I will describe two (groups) of control channel messages (and a bonus control channel message):</p>
<ul>
<li><code>EXIT</code>, <code>RESTART</code>, and <code>HALT</code></li>
<li><code>PUSH_REQUEST</code> / <code>PUSH_REPLY</code></li>
<li>(<code>AUTH_FAILED</code>)</li>
</ul>
<p>The <code>EXIT</code>, <code>RESTART</code>, and <code>HALT</code> messages share similarity.
They are all three used to signal to the client that it should disconnect[^disconnect] from the server.
<code>HALT</code> tells the client to disconnect and suggests the client should terminate.
<code>RESTART</code> also tells the client to disconnect and suggests the client can reconnect either to the same server or the next server if multiple are configured depending on flags in the message.
<code>EXIT</code> tells the <em>peer</em> that it is exiting and the <em>peer</em> should disconnect.
The last one can be sent by either the server or the client and is useful when the underlying transport is UDP.
It informs the peer that the sender is exiting and will (soon) not be receiving and ACK'ing messages; for UDP the peer would otherwise (re)send messages until a timeout.</p>
<p>Because the underlying transport can either be TCP or UDP the sender may have no guarantees that the message arrives.
OpenVPN's control channel implements a reliable layer with ACKs and retransmissions to work around that.
To accomodate this OpenVPN™ will wait five seconds before disconnecting to allow for retransmission of the exit message.</p>
<h3 id="the-bug"><a class="anchor" aria-hidden="true" href="#the-bug"></a>The bug</h3>
<p>While I was working on implementing more control channel message types I modified a client application that connects to a server and sends pings over the tunnel - instead of ICMPv4 echo requests I modified it to send the <code>EXIT</code> control channel message once a second.
In the server logs I saw that the server successfully received the <code>EXIT</code> message!
But nothing else happened.
The server just kept receiving <code>EXIT</code> messages but for some reason it never disconnected the client.</p>
<p>Curious about this behavior I dived into the OpenVPN™ source code and found that on each <code>EXIT</code> message it (re)schedules an exit (disconnect) timer! That is, every time the server receives an <code>EXIT</code> message it'll go &quot;OK! I'll shut down this connection in five seconds&quot; forgetting it promised to do so earlier, too.</p>
<h3 id="implications"><a class="anchor" aria-hidden="true" href="#implications"></a>Implications</h3>
<p>At first this seemed like a relatively benign bug.
What's the worst that could happen if a client says &quot;let's stop in five second! No, five seconds from now! No, five seconds from now!&quot; etc?
Well, it turns out the same timer is used when the server sends an exit message.
Ok, so what?
The client can hold open a resource it <em>was</em> authorized to use <em>longer</em>.
So we have a somewhat boring potential denial of service attack.</p>
<p>Then I learned more about the management interface.
The management interface is a text protocol to communicate with the OpenVPN server (or client) and query for information or send commands.
One command is the <code>client-kill</code> command.
The documentation says to use this command to &quot;[i]mmediately kill a client instance[...]&quot;.
In practice it sends an exit message to the client (either a custom one or the default <code>RESTART</code>).
I learnt that it shares code paths with the exit control messages to schedule an exit (disconnect)[^kill-immediately].
That is, <code>client-kill</code> schedules the same five second timer.</p>
<p>Thus a malicious client can, instead of exiting on receiving an exit or <code>RESTART</code> message, send back repeatedly <code>EXIT</code> to the server to reset the five second timer.
This way the client can indefinitely delay the exit/disconnect assuming sufficiently stable and responsive network.
This is suddenly not so good.
The application using the management interface might be enforcing a security policy which we can now circumvent!
The client might be a former employee in a company, and the security team might want to revoke access to the internal network for the former employee, and in that process uses <code>client-kill</code> to kick off all of his connecting clients.
The former employee, if prepared, can circumvent this by sending back <code>EXIT</code> messages repeatedly and thus keep unauthorized access.
Or a commercial VPN service may try to enforce a data transfer limit with the same mechanism which is then rather easily circumvented by sending <code>EXIT</code> messages.</p>
<p>Does anyone use the management interface in this way?
I don't know.
If you do or are aware of software that enforces policies this way please do reach out to <a href="https://reyn.ir/contact.html">me</a>.
It would be interesting to hear and discuss.
The OpenVPN security@ mailing list took it seriously enough to assign it CVE-2024-28882.</p>
<h2 id="openvpn-configuration-language"><a class="anchor" aria-hidden="true" href="#openvpn-configuration-language"></a>OpenVPN configuration language</h2>
<p>Next up we have <code>PUSH_REQUEST</code> / <code>PUSH_REPLY</code>.
As the names suggest it's a request/response protocol.
It is used to communicate configuration options from the server to the client.
These options include routes, ip address configuration, negotiated cryptographic algorithms.
The client signals it would like to receive configuration options from the server by sending the <code>PUSH_REQUEST</code> control channel message[^proto-push-request].
The server then sends a <code>PUSH_REPLY</code> message.</p>
<p>The format of a <code>PUSH_REPLY</code> message is <code>PUSH_REPLY,</code> followed by a comma separated list of OpenVPN configuration directives terminated by a NUL byte as in other control channel messages.
Note that this means pushed configuration directives cannot contain commas.</p>
<p>When implementing the <code>push</code> server configuration directive, which tells the server to send the parameter of <code>push</code> as a configuration option to the client in the <code>PUSH_REPLY</code>, I studied how exactly OpenVPN™ parses configuration options.
I learned some quirks of the configuration language which I find surprising and somewhat hard to implement.
I will not cover all corners of the configuration language.</p>
<p>In some sense you could say the configuration language of OpenVPN™ is line based.
At least, the first step to parsing configuration directives as OpenVPN 2.X does is to read one line at a time and parse it as one configuration directive[^inline-files].
A line is whatever <code>fgets()</code> says it is - this includes the newline if not at the end of the file[^configuration-newlines].
This is how it is for configuration files.
However, if it is a <code>PUSH_REPLY</code> a <em>&quot;line&quot;</em> is the text string up to a comma or the end of file (or, importantly, a NUL byte).
This &quot;line&quot; tokenization is done by repeatedly calling OpenVPN™'s <code>buf_parse(buf, ',', line, sizeof(line))</code> function.</p>
<pre><code class="language-C">/* file: src/openvpn/buffer.c */
bool
buf_parse(struct buffer *buf, const int delim, char *line, const int size)
{
bool eol = false;
int n = 0;
int c;
ASSERT(size &gt; 0);
do
{
c = buf_read_u8(buf);
if (c &lt; 0)
{
eol = true;
}
if (c &lt;= 0 || c == delim)
{
c = 0;
}
if (n &gt;= size)
{
break;
}
line[n++] = c;
}
while (c);
line[size-1] = '\0';
return !(eol &amp;&amp; !strlen(line));
}
</code></pre>
<p><code>buf_parse()</code> takes a <code>struct buffer*</code> which is a pointer to a byte array with a offset and length field, a delimiter character (in our case <code>','</code>), a destination buffer <code>line</code> and its length <code>size</code>.
It calls <code>buf_read_u8()</code> which returns the first character in the buffer and advances the offset and decrements the length, or returns <code>-1</code> if the buffer is empty.
In essence, <code>buf_parse()</code> &quot;reads&quot; from the buffer and copies over to <code>line</code> until it encounters <code>delim</code>, a NUL byte or the end of the buffer.
In that case a NUL byte is written to <code>line</code>.</p>
<p>What is interesting is that a NUL byte is effectively considered a delimiter, too, and that it is consumed by <code>buf_parse()</code>.
Next, let's look at how incoming control channel messages are handled (modified for brevity):</p>
<pre><code class="language-C">/* file: src/openvpn/forward.c (before fix) */
/*
* Handle incoming configuration
* messages on the control channel.
*/
static void
check_incoming_control_channel(struct context *c, struct buffer buf)
{
/* force null termination of message */
buf_null_terminate(&amp;buf);
/* enforce character class restrictions */
string_mod(BSTR(&amp;buf), CC_PRINT, CC_CRLF, 0);
if (buf_string_match_head_str(&amp;buf, &quot;AUTH_FAILED&quot;))
{
receive_auth_failed(c, &amp;buf);
}
else if (buf_string_match_head_str(&amp;buf, &quot;PUSH_&quot;))
{
incoming_push_message(c, &amp;buf);
}
/* SNIP */
}
</code></pre>
<p>First, the buffer is ensured to be NUL terminated by replacing the last byte with a NUL byte.
This is already somewhat questionable as it could make an otherwise invalid message valid.
Next, character class restrictions are &quot;enforced&quot;.
What this roughly does is removing non-printable characters and carriage returns and line feeds from the C string.
The macro <code>BSTR()</code> returns the underlying buffer behind the <code>struct buffer</code> with the offset added.
Notably, <code>string_mod()</code> works on (NUL terminated) C strings and not <code>struct buffer</code>s.
As an example, the string (with the usual C escape sequences):</p>
<pre><code>&quot;PUSH_REPLY,line \nfeeds\n,are\n,removed\n\000&quot;
</code></pre>
<p>becomes</p>
<pre><code>&quot;PUSH_REPLY,line feeds,are,removed\000ed\n\000&quot;
</code></pre>
<p>As you can see, if interpreted as a C string we have removed the line feeds.
But what is this at the end?
It is the same last 4 bytes from the original string.
More generally, it is the last N bytes from the original string if the original string has N line feeds (or other disallowed characters).</p>
<p>The whole buffer is still passed to the push reply parsing.
Remember that the &quot;line&quot; parser will not only consume commas as the line delimiter, but also NUL bytes!
This means the configuration directives are parsed as lines:</p>
<pre><code class="language-C">&quot;line feeds&quot;
&quot;are&quot;
&quot;removed&quot;
&quot;ed\n&quot;
</code></pre>
<p>With this technique we can now inject (almost; the exception is NUL) arbitrary bytes as configuration directive lines.
This is bad because the configuration directive is printed to the console if it doesn't parse.
As a proof of concept I sent a <code>PUSH_REPLY</code> with an embedded BEL character, and the OpenVPN™ client prints to console (abbreviated):</p>
<pre><code>Unrecognized option or missing or extra parameter(s): ^G
</code></pre>
<p>The <code>^G</code> is how the BEL character is printed in my terminal.
I was also able to hear an audible bell.</p>
<p>A more thorough explanation on how terminal escape sequences can be exploited can be found on <a href="https://www.gresearch.com/news/g-research-the-terminal-escapes/">G-Reasearch's blog</a>.</p>
<h3 id="the-fix"><a class="anchor" aria-hidden="true" href="#the-fix"></a>The fix</h3>
<p>The fix also is also a first step towards decoupling the control channel messaging from the TLS record frames.
First, the data is split on NUL bytes in order to get the control channel message(s), and then messages are rejected if they contain illegal characters.
This solves the vulnerability described previously.</p>
<p>Unfortunately, it turns out that especially for the <code>AUTH_FAILED</code> control channel message it is easy to create invalid messages:
If 2FA is implemented using the script mechanism sending custom messages they easily end with a newline asking the client to enter the verification code.
I believe in 2.6.12 the client tolerates trailing newline characters.</p>
<h2 id="conclusion"><a class="anchor" aria-hidden="true" href="#conclusion"></a>Conclusion</h2>
<p>The first bug, the timer rescheduling bug, is at least 20 years old!
It hasn't always been exploitable, but the bug itself goes back as far as the git history does.
I haven't attempted further software archeology to find the exact time of introduction.
Either way, it's old and gone unnoticed for quite a while.</p>
<p>I think this shows that diversity in implementations is a great way to exercise corner cases, push forward (protocol) documentation efforts and get thorough code review by motivated peers.
This work was funded by <a href="https://nlnet.nl/project/MirageVPN/">the EU NGI Assure Fund through NLnet</a>.
In my opinion, this shows that funding one open source project can have a positive impact on other open source projects, too.</p>
<p>[^openvpn-tls]: This is not always the case. It is possible to use static shared secret keys, but it is mostly considered deprecated.
[^disconnect]: I say &quot;disconnect&quot; even when the underlying transport is the connection-less UDP.
[^kill-immediately]: As the alert reader might have realized this is inaccurate. It does not kill the client &quot;immediately&quot; as it will wait five seconds after the exit message is sent before exiting. At best this will kill a cooperating client once it's received the kill message.
[^proto-push-request]: There is another mechanism to request a <code>PUSH_REPLY</code> earlier with less roundtrips, but let's ignore that for now. The exact message is <code>PUSH_REQUEST&lt;NUL-BYTE&gt;</code> as messages need to be NUL-terminated.
[^inline-files]: An exception being inline files which can span multiple lines. They vaguely resemble XML tags with an open <code>&lt;tag&gt;</code> and close <code>&lt;/tag&gt;</code> each on their own line with the data in between. I doubt these are sent in <code>PUSH_REPLY</code>s, but I can't rule out without diving into the source code that it isn't possible to send inline files.
[^configuration-newlines]: This results in the quirk that it is possible to sort-of escape a newline in a configuration directive. But since the line splitting is done <em>first</em> it's not possible to continue the directive on the next line! I believe this is mostly useless, but it is a way to inject line feeds in configuration options without modifying the OpenVPN source code.</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/2024-08-21-OpenVPN-and-MirageVPN.md → articles/2024-08-21-OpenVPN-and-MirageVPN.markdown
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-08-21 date: 2024-08-21
title: MirageVPN and OpenVPN article.title: MirageVPN and OpenVPN
description: Discoveries made implementing MirageVPN, a OpenVPN-compatible VPN library article.description: Discoveries made implementing MirageVPN, a OpenVPN-compatible VPN library
tags: tags:
- MirageVPN - MirageVPN
- OpenVPN - OpenVPN

145
articles/gptar.html Normal file
View file

@ -0,0 +1,145 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogGPTar
</title>
<meta name="description" content="Hybrid GUID partition table and tar archive">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>GPTar</h1>
<ul class="tags-list"><li>OCaml</li><li>gpt</li><li>tar</li><li>mbr</li><li>persistent storage</li></ul><p>At <a href="https://robur.coop/">Robur</a> we developed a piece of software for mirroring or exposing an <a href="https://opam.ocaml.org/">opam</a> repository.
We have it deployed at <a href="https://opam.robur.coop/">opam.robur.coop</a>, and you can use it as an alternative to opam.ocaml.org.
It is usually more up-to-date with the git <a href="https://github.com/ocaml/opam-repository">opam-repository</a> than opam.ocaml.org although in the past it suffered from <a href="https://blog.osau.re/articles/lwt_pause.html">occasional availability issues</a>.
I can recommend reading Hannes' post about <a href="https://hannes.robur.coop/Posts/OpamMirror">opam-mirror</a>.
This article is about adding a partition table to the disk <a href="https://hannes.robur.coop/Posts/OpamMirror#code-development-and-improvements">as used by opam-mirror</a>.
For background I can recommend reading the previously linked subsection of the opam-mirror article.</p>
<h2 id="the-opam-mirror-persistent-storage-scheme"><a class="anchor" aria-hidden="true" href="#the-opam-mirror-persistent-storage-scheme"></a>The opam-mirror persistent storage scheme</h2>
<p>Opam-mirror uses a single block device for its persistent storage.
On the block device it stores cached source code archives from the opam repository.
These are stored in a <a href="https://en.wikipedia.org/wiki/Tar_(computing)">tar archive </a> consisting of files whose file name is the sha256 checksum of the file contents.
Furthermore, at the end of the block device some space is allocated for dumping the cloned git state of the upstream (git) opam repository as well as caches storing maps from md5 and sha512 checksums respectively to the sha256 checksums.
The partitioning scheme is entirely decided by command line arguments.
In other words, there is no partition table on the disk image.</p>
<p>This scheme has the nice property that the contents of the tar archive can be inspected by regular tar utilities in the host system.
Due to the append-only nature of tar and in the presence of concurrent downloads a file written to the archive may be partial or corrupt.
Opam-mirror handles this by prepending a <code>pending/</code> directory to partial downloads and <code>to-delete/</code> directory for corrupt downloads.
If there are no files after the failed download in the tar archive the file can be removed without any issues.
Otherwise a delete would involve moving all subsequent files further back in the archive - which is too error prone to do robustly.
So using the tar utilities in the host we can inspect how much garbage has accumulated in the tar file system.</p>
<p>The big downside to this scheme is that since the disk partitioning is not stored on the disk the contents can easily become corrupt if the wrong offsets are passed on the command line.
Therefore I have for a long time been wanting to use an on-disk partition table.
The problem is both <a href="https://en.wikipedia.org/wiki/MBR_partition_table">MBR</a> and <a href="https://en.m.wikipedia.org/wiki/GUID_Partition_Table">GPT</a> (GUID Partition Table) store the table at the beginning of the disk.
If we write a partition table at the beginning it is suddenly not a valid tar archive anymore.
Of course, in Mirage we can just write and read the table at the end if we please, but then we lose the ability to inspect the partition table in the host system.</p>
<h2 id="gpt-header-as-tar-file-name"><a class="anchor" aria-hidden="true" href="#gpt-header-as-tar-file-name"></a>GPT header as tar file name</h2>
<p>My first approach, which turned out to be a dead end, was when I realized that a GPT header consists of 92 bytes at the beginning followed by reserved space for the remainder of the LBA.
The reserved space should be all zeroes, but it seems no one bothers to enforce this.
What's more is that a tar header starts with the file name in the first 100 bytes.
This got me thinking we could embed a GPT header inside a tar header by using the GPT header as the tar header file name!</p>
<p>I started working on implementing this, but I quickly realized that 1) the tar header has a checksum, and 2) the gpt header has a checksum as well.
Having two checksums that cover each other is tricky.
Updating one checksum affects the other checksum.
So I started reading a paper written by Martin Stigge et al. about <a href="https://sar.informatik.hu-berlin.de/research/publications/SAR-PR-2006-05/SAR-PR-2006-05_.pdf">reversing CRC</a> as the GPT header use CRC32 checksum.
I ended up writing <a href="https://github.com/reynir/gptar/commit/e8269c4959e98aa0cf339cf250ff4e6671fd344c">something</a> that I knew was incorrect.</p>
<p>Next, I realized the GPT header's checksum only covers the first 92 bytes - that is, the reserved space is not checksummed!
I find this and the fact that the reserved space should be all zeroes but no one checks odd about GPT.
This simplified things a lot as we don't have to reverse any checksums!
Then I implemented a test binary that produces a half megabyte disk image with a hybrid GPT and tar header followed by a tar archive with a file <code>test.txt</code> whose content is <code>Hello, World!</code>.
I had used the byte <code>G</code> as the link indicator.
In POSIX.1-1988 the link indicators <code>A</code>-<code>Z</code> are reserved for vendor specific extensions, and it seemed <code>G</code> was unused.
A mistake I made was to not update the tar header checksum - the <a href="https://github.com/mirage/ocaml-tar">ocaml-tar</a> library doesn't support this link indicator value so I had manually updated the byte value in the serialized header but forgot to update the checksum.
This was easily remediated as the checksum is a simple sum of the bytes in the header.
The changes made are viewable on <a href="https://github.com/reynir/gptar/compare/e8269c4959e98aa0cf339cf250ff4e6671fd344c...03a57a1cef17eeb66b0d883dbe441c9c4b9093bd">GitHub</a>.
I also had to work around a <a href="https://github.com/mirage/ocaml-tar/issues/144">bug in ocaml-tar</a>.
GNU tar was successfully able to list the archive.
A quirk is that the archive will start with a dummy file <code>GPTAR</code> which consists of any remaining space in the first LBA if the sector size is greater than 512 bytes followed by the partition table.</p>
<h2 id="protective-mbr"><a class="anchor" aria-hidden="true" href="#protective-mbr"></a>Protective MBR</h2>
<p>Unfortunately, neither fdisk nor parted recognized the GPT partition table.
I was able to successfully read the partition table using <a href="https://github.com/mirage/ocaml-gpt">ocaml-gpt</a> however.
This puzzled me.
Then I got a hunch: I had read about <a href="https://en.m.wikipedia.org/wiki/GUID_Partition_Table#Protective_MBR_(LBA_0)">protective MBRs</a> on the Wikipedia page on GPT.
I had always thought it was optional and not needed in a new system such as Mirage that doesn't have to care too much about legacy code and operating systems.</p>
<p>So I started comparing the layout of MBR and tar.
The V7 tar format only uses the first 257 bytes of the 512 byte block.
The V7 format is differentiated by the UStar, POSIX/pax and old GNU tar formats by not having the string <code>ustar</code> at byte offset 257[^tar-ustar].
The master boot record format starts with the bootstrap code area.
In the classic format it is the first 446 bytes.
In the modern standard MBR format the first 446 bytes are mostly bootstrap code too with the exception of a handful bytes at offset 218 or so which are used for a timestamp or so.
This section overlaps with the tar V7 linked file name field.
In both formats these bytes can be zero without any issues, thankfully.</p>
<p>This is great!
This means we can put a tar header in the bootstrap code area of the MBR and have it be a valid tar header <em>and</em> MBR record at the same time.
The protective MBR has one partition of type <code>0xEE</code> whose LBA starts at sector 1 and the number of LBAs should cover the whole disk, or be <code>0xFFFFFFFF</code> (maximum representable number in unsigned 32 bit).
In practice this means we can get away with only touching byte offsets 446-453 and 510-511 for the protective MBR.
The MBR does not have a checksum which also makes things easier.
Using this I could create a disk image that parted and fdisk recognized as a GPT partitioned disk!
With the caveat that they both reported that the backup GPT header was corrupt.
I had just copied the primary GPT header to the end of the disk.
It turns out that the alternate, or backup, GPT header should have the current LBA and backup LBA fields swapped (and the header crc32 recomputed).
I updated the ocaml-gpt code so that it can <a href="https://github.com/mirage/ocaml-gpt/pull/14">marshal alternate GPT headers</a>.</p>
<p>Finally we can produce GPT partitioned disks that can be inspected with tar utilities!</p>
<pre><code>$ /usr/sbin/parted disk.img print
WARNING: You are not superuser. Watch out for permissions.
Model: (file)
Disk /home/reynir/workspace/gptar/disk.img: 524kB
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Disk Flags: pmbr_boot
Number Start End Size File system Name Flags
1 17.4kB 19.5kB 2048B Real tar archive hidden
</code></pre>
<pre><code>$ tar -tvf disk.img
?r-------- 0/0 16896 1970-01-01 01:00 GPTAR unknown file type ‘G’
-r-------- 0/0 14 1970-01-01 01:00 test.txt
</code></pre>
<p>The <a href="https://github.com/reynir/gptar">code</a> is freely available on GitHub.</p>
<h2 id="future-work"><a class="anchor" aria-hidden="true" href="#future-work"></a>Future work</h2>
<p>One thing that bothers me a bit is the dummy file <code>GPTAR</code>.
By using the <code>G</code> link indicator GNU tar will print a warning about the unknown file type <code>G</code>,
but it will still extract the dummy file when extracting the archive.
I have been thinking about what tar header I could put in the MBR so tar utilities will skip the partition table but not try to extract the dummy file.
Ideas I've had is to:</p>
<ul>
<li>Pretend it is a directory with a non-zero file size which is nonsense.
I'm unsure what tar utilities would do in that case.
I fear not all implementations will skip to the next header correctly as a non-zero directory is nonsense.
I may give it a try and check how GNU tar, FreeBSD tar and ocaml-tar react.</li>
<li>Say it is a PAX extended header and use a nonsense tag or attribute whose value covers the GPT header and partition table.
The problem is the PAX extended header content format is <code>&lt;length&gt; &lt;tag&gt;=&lt;value&gt;\n</code> where <code>&lt;length&gt;</code> is the decimal string encoding of the length of <code>&lt;tag&gt;=&lt;value&gt;\n</code>.
In other words it must start with the correct length.
For sector size 512 this is a problem because the PAX extended header content would start with the GPT header which starts with the string <code>EFI PART</code>.
If the sector size is greater than 512 we can use the remaining space in LBA 0 to write a length, dummy tag and some padding.
I may try this for a sector size of 4096, but I'm not happy that it doesn't work with sector size 512 which solo5 will default to.</li>
</ul>
<p>If you have other ideas what I can do please reach out!</p>
<p>[^tar-ustar]: This is somewhat simplified. There are some more nuances between the different formats, but for this purpose they don't matter much.</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/gptar.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-02-21 date: 2024-02-21
title: GPTar article.title: GPTar
description: Hybrid GUID partition table and tar archive article.description: Hybrid GUID partition table and tar archive
tags: tags:
- OCaml - OCaml
- gpt - gpt

263
articles/lwt_pause.html Normal file
View file

@ -0,0 +1,263 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogCooperation and Lwt.pause
</title>
<meta name="description" content="A disgression about Lwt and Miou">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>Cooperation and Lwt.pause</h1>
<ul class="tags-list"><li>OCaml</li><li>Scheduler</li><li>Community</li><li>Unikernel</li><li>Git</li></ul><p>Here's a concrete example of the notion of availability and the scheduler used
(in this case Lwt). As you may know, at Robur we have developed a unikernel:
<a href="https://git.robur.coop/robur/opam-mirror">opam-mirror</a>. It launches an HTTP service that can be used as an
OPAM overlay available from a Git repository (with <code>opam repository add &lt;name&gt; &lt;url&gt;</code>).</p>
<p>The purpose of such an unikernel was to respond to a failure of the official
repository which fortunately did not last long and to offer decentralisation
of such a service. You can use https://opam.robur.coop!</p>
<p>It was also useful at the Mirage retreat, where we don't usually have a
great internet connection. Caching packages for our OCaml users on the local
network has benefited us in terms of our Internet bill by allowing the OCaml
users to fetch opam packages over the local network instead of over the shared,
metered 4G Internet conncetion.</p>
<p>Finally, it's a unikernel that I also use on my server for my software
<a href="https://blog.osau.re/articles/reproducible.html">reproducibility service</a> in order to have an overlay for my
software like <a href="https://bob.osau.re/">Bob</a>.</p>
<p>In short, I advise you to use it, you can see its installation
<a href="https://blog.osau.re/articles/reproducible.html">here</a> (I think that in the context of a company, internally, it
can be interesting to have such a unikernel available).</p>
<p>However, this unikernel had a long-standing problem. We were already talking
about it at the Mirleft retreat, when we tried to get the repository from Git,
we had a (fairly long) unavailability of our HTTP server. Basically, we had to
wait ~10 min before the service offered by the unikernel was available.</p>
<h2 id="availability"><a class="anchor" aria-hidden="true" href="#availability"></a>Availability</h2>
<p>If you follow my <a href="https://blog.osau.re/tags/scheduler.html">articles</a>, as far as Miou is concerned, from
the outset I talk of the notion of availability if we were to make yet another
new scheduler for OCaml 5. We emphasised this notion because we had quite a few
problems on this subject and Lwt.</p>
<p>In this case, the notion of availability requires the scheduler to be able to
observe system events as often as possible. The problem is that Lwt doesn't
really offer this approach.</p>
<p>Indeed, Lwt offers a way of observing system events (<code>Lwt.pause</code>) but does not
do so systematically. The only time you really give the scheduler the
opportunity to see whether you can read or write is when you want to...
read or write...</p>
<p>More generally, it is said that Lwt's <strong>bind</strong> does not <em>yield</em>. In other words,
you can chain any number of functions together (via the <code>&gt;&gt;=</code> operator), but
from Lwt's point of view, there is no opportunity to see if an event has
occurred. Lwt always tries to go as far down your chain as possible:</p>
<ul>
<li>and finish your promise</li>
<li>or come across an operation that requires a system event (read or write)</li>
<li>or come across an <code>Lwt.pause</code> (as a <em>yield</em> point)</li>
</ul>
<p>Lwt is rather sparse in adding cooperation points besides <code>Lwt.pause</code> and
read/write operations, in contrast with Async where the bind operator is a
cooperation point.</p>
<h3 id="if-there-is-no-io-do-not-wrap-in-lwt"><a class="anchor" aria-hidden="true" href="#if-there-is-no-io-do-not-wrap-in-lwt"></a>If there is no I/O, do not wrap in Lwt</h3>
<p>It was (bad<sup><a href="#fn1">1</a></sup>) advice I was given. If a function doesn't do
I/O, there's no point in putting it in Lwt. At first glance, however, the idea
may be a good one. If you have a function that doesn't do I/O, whether it's in
the Lwt monad or not won't make any difference to the way Lwt tries to execute
it. Once again, Lwt should go as far as possible. So Lwt tries to solve both
functions in the same way:</p>
<pre><code class="language-ocaml">val merge : int array -&gt; int array -&gt; int array
let rec sort0 arr =
if Array.length arr &lt;= 1 then arr
else
let m = Array.length arr / 2 in
let arr0 = sort0 (Array.sub arr 0 m) in
let arr1 = sort0 (Array.sub arr m (Array.length arr - m)) in
merge arr0 arr1
let rec sort1 arr =
let open Lwt.Infix in
if Array.length arr &lt;= 1 then Lwt.return arr
else
let m = Array.length arr / 2 in
Lwt.both
(sort1 (Array.sub arr m (Array.length arr - m)))
(sort1 (Array.sub arr 0 m))
&gt;|= fun (arr0, arr1) -&gt;
merge arr0 arr1
</code></pre>
<p>If we trace the execution of the two functions (for example, by displaying our
<code>arr</code> each time), we see the same behaviour whether Lwt is used or not. However,
what is interesting in the Lwt code is the use of <code>both</code>, which suggests that
the processes are running <em>at the same time</em>.</p>
<p>&quot;At the same time&quot; does not necessarily suggest the use of several cores or &quot;in
parallel&quot;, but the possibility that the right-hand side may also have the
opportunity to be executed even if the left-hand side has not finished. In other
words, that the two processes can run <strong>concurrently</strong>.</p>
<p>But factually, this is not the case, because even if we had the possibility of
a point of cooperation (with the <code>&gt;|=</code> operator), Lwt tries to go as far as
possible and decides to finish the left part before launching the right part:</p>
<pre><code class="language-shell">$ ./a.out
sort0: [|3; 4; 2; 1; 7; 5; 8; 9; 0; 6|]
sort0: [|3; 4; 2; 1; 7|]
sort0: [|3; 4|]
sort0: [|2; 1; 7|]
sort0: [|1; 7|]
sort0: [|5; 8; 9; 0; 6|]
sort0: [|5; 8|]
sort0: [|9; 0; 6|]
sort0: [|0; 6|]
sort1: [|3; 4; 2; 1; 7; 5; 8; 9; 0; 6|]
sort1: [|3; 4; 2; 1; 7|]
sort1: [|3; 4|]
sort1: [|2; 1; 7|]
sort1: [|1; 7|]
sort1: [|5; 8; 9; 0; 6|]
sort1: [|5; 8|]
sort1: [|9; 0; 6|]
sort1: [|0; 6|]
</code></pre>
<hr>
<p><strong><tag id="fn1">1</tag></strong>: However, if you are not interested in availability
and would like the scheduler to try to resolve your promises as quickly as
possible, this advice is clearly valid.</p>
<h4 id="performances"><a class="anchor" aria-hidden="true" href="#performances"></a>Performances</h4>
<p>It should be noted, however, that Lwt has an impact. Even if the behaviour is
the same, the Lwt layer is not free. A quick benchmark shows that there is an
overhead:</p>
<pre><code class="language-ocaml">let _ =
let t0 = Unix.gettimeofday () in
for i = 0 to 1000 do let _ = sort0 arr in () done;
let t1 = Unix.gettimeofday () in
Fmt.pr &quot;sort0 %fs\n%!&quot; (t1 -. t0)
let _ =
let t0 = Unix.gettimeofday () in
Lwt_main.run @@ begin
let open Lwt.Infix in
let rec go idx = if idx = 1000 then Lwt.return_unit
else sort1 arr &gt;&gt;= fun _ -&gt; go (succ idx) in
go 0 end;
let t1 = Unix.gettimeofday () in
Fmt.pr &quot;sort1 %fs\n%!&quot; (t1 -. t0)
</code></pre>
<pre><code class="language-sh">$ ./a.out
sort0 0.000264s
sort1 0.000676s
</code></pre>
<p>This is the fairly obvious argument for not using Lwt when there's no I/O. Then,
if the Lwt monad is really needed, a simple <code>Lwt.return</code> at the very last
instance is sufficient (or, better, the use of <code>Lwt.map</code> / <code>&gt;|=</code>).</p>
<h4 id="cooperation-and-concrete-example"><a class="anchor" aria-hidden="true" href="#cooperation-and-concrete-example"></a>Cooperation and concrete example</h4>
<p>So <code>Lwt.both</code> is the one to use when we want to run two processes
&quot;at the same time&quot;. For the example, <a href="https://github.com/mirage/ocaml-git">ocaml-git</a> attempts <em>both</em> to
retrieve a repository and also to analyse it. This can be seen in this snippet
of <a href="https://github.com/mirage/ocaml-git/blob/a36c90404b149ab85f429439af8785bb1dde1bee/src/not-so-smart/smart_git.ml#L476-L481">code</a>.</p>
<p>In our example with ocaml-git, the problem &quot;shouldn't&quot; appear because, in this
case, both the left and right side do I/O (the left side binds into a socket
while the right side saves Git objects in your file system). So, in our tests
with <code>Git_unix</code>, we were able to see that the analysis (right-hand side) was
well executed and 'interleaved' with the reception of objects from the network.</p>
<h3 id="composability"><a class="anchor" aria-hidden="true" href="#composability"></a>Composability</h3>
<p>However, if we go back to our initial problem, we were talking about our
opam-mirror unikernel. As you might expect, there is no standalone MirageOS file
system (and many of our unikernels don't need one). So, in the case of
opam-mirror, we use the ocaml-git memory implementation: <code>Git_mem</code>.</p>
<p><code>Git_mem</code> is different in that Git objects are simply stored in a <code>Hashtbl</code>.
There is no cooperation point when it comes to obtaining Git objects from this
<code>Hashtbl</code>. So let's return to our original advice:</p>
<blockquote>
<p>don't wrap code in Lwt if it doesn't do I/O.</p>
</blockquote>
<p>And, of course, <code>Git_mem</code> doesn't do I/O. It does, however, require the process
to be able to work with Lwt. In this case, <code>Git_mem</code> wraps the results in Lwt
<strong>as late as possible</strong> (as explained above, so as not to slow down our
processes unnecessarily). The choice inevitably means that the right-hand side
can no longer offer cooperation points. And this is where our problem begins:
composition.</p>
<p>In fact, we had something like:</p>
<pre><code class="language-ocaml">let clone socket git =
Lwt.both (receive_pack socket) (analyse_pack git) &gt;&gt;= fun ((), ()) -&gt;
Lwt.return_unit
</code></pre>
<p>However, our <code>analyse_pack</code> function is an injection of a functor representing
the Git backend. In other words, <code>Git_unix</code> or <code>Git_mem</code>:</p>
<pre><code class="language-ocaml">module Make (Git : Git.S) = struct
let clone socket git =
Lwt.both (receive_pack socket) (Git.analyse_pack git) &gt;&gt;= fun ((), ()) -&gt;
Lwt.return_unit
end
</code></pre>
<p>Composability poses a problem here because even if <code>Git_unix</code> and <code>Git_mem</code>
offer the same function (so both modules can be used), the fact remains that one
will always offer a certain availability to other services (such as an HTTP
service) while the other will offer a Lwt function which will try to go as far
as possible quite to make other services unavailable.</p>
<p>Composing with one or the other therefore does not produce the same behavior.</p>
<h4 id="where-to-put-lwtpause"><a class="anchor" aria-hidden="true" href="#where-to-put-lwtpause"></a>Where to put <code>Lwt.pause</code>?</h4>
<p>In this case, our <code>analyse_pack</code> does read/write on the Git store. As far as
<code>Git_mem</code> is concerned, we said that these read/write accesses were just
accesses to a <code>Hashtbl</code>.</p>
<p>Thanks to <a href="https://hannes.robur.coop/">Hannes</a>' help, it took us an afternoon to work out where we
needed to add cooperation points in <code>Git_mem</code> so that <code>analyse_pack</code> could give
another service such as HTTP the opportunity to work. Basically, this series of
<a href="https://github.com/mirage/ocaml-git/pull/631/files">commits</a> shows where we needed to add <code>Lwt.pause</code>.</p>
<p>However, this points to a number of problems:</p>
<ol>
<li>it is not necessarily true that on the basis of composability alone (by
<em>functor</em> or by value), Lwt reacts in the same way</li>
<li>Subtly, you have to dig into the code to find the right opportunities where
to put, by hand, <code>Lwt.pause</code>.</li>
<li>In the end, Lwt has no mechanisms for ensuring the availability of a service
(this is something that must be taken into account by the implementer).</li>
</ol>
<h3 id="in-depth-knowledge-of-lwt"><a class="anchor" aria-hidden="true" href="#in-depth-knowledge-of-lwt"></a>In-depth knowledge of Lwt</h3>
<p>I haven't mentioned another problem we encountered with <a href="https://cambium.inria.fr/~agueneau/">Armael</a> when
implementing <a href="https://discuss.ocaml.org/t/ann-release-of-multipart-form-0-2-0/7704#memory-bound-implementation">multipart_form</a> where the use of stream meant that
Lwt didn't interleave the two processes and the use of a <em>bounded stream</em> was
required. Again, even when it comes to I/O, Lwt always tries to go as far as
possible in one of two branches of a <code>Lwt.both</code>.</p>
<p>This allows us to conclude that beyond the monad, Lwt has subtleties in its
behaviour which may be different from another scheduler such as Async (hence the
incompatibility between the two, which is not just of the <code>'a t</code> type).</p>
<h3 id="digression-on-miou"><a class="anchor" aria-hidden="true" href="#digression-on-miou"></a>Digression on Miou</h3>
<p>That's why we put so much emphasis on the notion of availability when it comes
to Miou: to avoid repeating the mistakes of the past. The choices that can be
made with regard to this notion in particular have a major impact, and can be
unsatisfactory to the user in certain cases (for example, so-called pure
calculations could take longer with Miou than with another scheduler).</p>
<p>In this sense, we have tried to constrain ourselves in the development of Miou
through the use of <code>Effect.Shallow</code> which requires us to always re-attach our
handler (our scheduler) as soon as an effect is produced, unlike <code>Effect.Deep</code>
which can re-use the same handler for several effects. In other words, and as
we've described here, <strong>an effect yields</strong>!</p>
<h2 id="conclusion"><a class="anchor" aria-hidden="true" href="#conclusion"></a>Conclusion</h2>
<p>As far as opam-mirror is concerned, we now have an unikernel that is available
even if it attempts to clone a Git repository and save Git objects in memory. At
least, an HTTP service can co-exist with ocaml-git!</p>
<p>I hope we'll be able to use it at <a href="https://retreat.mirage.io/">the next retreat</a>, which I invite
you to attend to talk more about Lwt, scheduler, Git and unikernels!</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

8
articles/lwt_pause.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-02-11 date: 2024-02-11
title: Cooperation and Lwt.pause article.title: Cooperation and Lwt.pause
description: article.description:
A disgression about Lwt and Miou A disgression about Lwt and Miou
tags: tags:
- OCaml - OCaml
@ -9,10 +9,6 @@ tags:
- Community - Community
- Unikernel - Unikernel
- Git - Git
author:
name: Romain Calascibetta
email: romain.calascibetta@gmail.com
link: https://blog.osau.re/
breaks: false breaks: false
--- ---

58
articles/miragevpn-ncp.html Normal file
View file

@ -0,0 +1,58 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogMirageVPN updated (AEAD, NCP)
</title>
<meta name="description" content="How we resurrected MirageVPN from its bitrot state">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>MirageVPN updated (AEAD, NCP)</h1>
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>VPN</li><li>security</li></ul><h2 id="updating-miragevpn"><a class="anchor" aria-hidden="true" href="#updating-miragevpn"></a>Updating MirageVPN</h2>
<p>As announced <a href="https://blog.robur.coop/articles/miragevpn.html">earlier this month</a>, we've been working hard over the last months on MirageVPN (initially developed in 2019, targeting OpenVPN™ 2.4.7, now 2.6.6). We managed to receive funding from <a href="https://www.assure.ngi.eu/">NGI Assure</a> call (via <a href="https://nlnet.nl">NLnet</a>). We've made over 250 commits with more than 10k lines added, and 18k lines removed. We closed nearly all old issues, and opened 100 fresh ones, of which we already closed more than half of them. :D</p>
<h3 id="actual-bugs-fixed-that-were-leading-to-non-working-miragevpn-applications"><a class="anchor" aria-hidden="true" href="#actual-bugs-fixed-that-were-leading-to-non-working-miragevpn-applications"></a>Actual bugs fixed (that were leading to non-working MirageVPN applications)</h3>
<p>In more detail, we had a specific configuration running over all the years, namely UDP mode with static keys (no TLS handshake, etc.). There were several issues (bitrot) that we encountered and solved along the path, amongst others:</p>
<ul>
<li>related to the <a href="https://github.com/robur-coop/miragevpn/pull/111">static-key mode and TCP/IP</a>,</li>
<li>the <a href="https://github.com/robur-coop/miragevpn/pull/98">order of ACK between the client and the server</a>,</li>
<li><a href="https://github.com/robur-coop/miragevpn/pull/110">outgoing TLS packets</a>.</li>
</ul>
<p>To avoid any future breakage while revising the code (cleaning it up, extending it), we are now building several unikernels as part of our CI system. We also have setup OpenVPN™ servers with various configurations that we periodically test with our new code (we'll also work on further automation thereof).</p>
<h3 id="new-features-aead-ciphers-supporting-more-configuration-primitives"><a class="anchor" aria-hidden="true" href="#new-features-aead-ciphers-supporting-more-configuration-primitives"></a>New features: AEAD ciphers, supporting more configuration primitives</h3>
<p>We added various configuration primitives, amongst them configuratble tls ciphersuites, minimal and maximal tls version to use, <a href="https://blog.robur.coop/articles/miragevpn.html">tls-crypt-v2</a>, verify-x509-name, cipher, remote-random, ...</p>
<p>From a cryptographic point of view, we are now supporting more <a href="https://github.com/robur-coop/miragevpn/pull/108">authentication hashes</a> via the configuration directive <code>auth</code>, namely the SHA2 family - previously, only SHA1 was supported, <a href="https://github.com/robur-coop/miragevpn/pull/125">AEAD ciphers</a> (AES-128-GCM, AES-256-GCM, CHACHA20-POLY1305) - previously only AES-256-CBC was supported.</p>
<h3 id="ncp---negotiation-of-cryptographic-parameters"><a class="anchor" aria-hidden="true" href="#ncp---negotiation-of-cryptographic-parameters"></a>NCP - Negotiation of cryptographic parameters</h3>
<p>OpenVPN™ has a way to negotiate cryptographic parameters, instead of hardcoding them in the configuration. The client can propose its supported ciphers, and other features (MTU, directly request a push message for IP configuration, use TLS exporter secret instead of the hand-crafted (TLS 1.0 based PRF), ...) once the TLS handshake has been completed.</p>
<p>We are now supporting this negotiation protocol, and have been working on the different extensions that are useful to us. Namely, transmitting the <a href="https://github.com/robur-coop/miragevpn/pull/121">supported ciphers</a>, <a href="https://github.com/robur-coop/miragevpn/pull/129">request push</a> (which deletes an entire round-trip), <a href="https://github.com/robur-coop/miragevpn/pull/163">TLS-exporter</a>. This will also be part of the <a href="https://git.robur.coop/robur/openvpn-spec">protocol specification</a> that we're working on while finishing our implementation.</p>
<h3 id="cleanups-and-refactorings"><a class="anchor" aria-hidden="true" href="#cleanups-and-refactorings"></a>Cleanups and refactorings</h3>
<p>We also took some time to cleanup our code base, removing <code>Lwt.fail</code> (which doesn't produce proper backtraces), using lzo from the decompress package (since that code has been upstreamed a while ago), remove unneeded dependencies (rresult, astring), avoiding <code>assert false</code> in pattern matches by improving types, improve the log output (include a timestamp, show log source, use colors).</p>
<h3 id="future"><a class="anchor" aria-hidden="true" href="#future"></a>Future</h3>
<p>There is still some work that we want to do, namely a QubesOS client implementation, an operators manual, extending our specification, resurrecting and adapting the server implementation, supporting more NCP features (if appropriate), etc. So stay tuned, we'll also provide reproducible binaries once we're ready.</p>
<p>Don't hesitate to reach out to us on <a href="https://github.com/robur-coop/miragevpn/issues">GitHub</a>, <a href="https://robur.coop/Contact">by mail</a> or me personally <a href="https://mastodon.social/@hannesm">on Mastodon</a> if you're stuck.</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/miragevpn-ncp.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2023-11-20 date: 2023-11-20
title: MirageVPN updated (AEAD, NCP) article.title: MirageVPN updated (AEAD, NCP)
description: article.description:
How we resurrected MirageVPN from its bitrot state How we resurrected MirageVPN from its bitrot state
tags: tags:
- OCaml - OCaml

57
articles/miragevpn-performance.html Normal file
View file

@ -0,0 +1,57 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogSpeeding up MirageVPN and use it in the wild
</title>
<meta name="description" content="Performance engineering of MirageVPN, speeding it up by a factor of 25.">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>Speeding up MirageVPN and use it in the wild</h1>
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>cryptography</li><li>security</li><li>VPN</li><li>performance</li></ul><p>As we were busy continuing to work on <a href="https://github.com/robur-coop/miragevpn">MirageVPN</a>, we got in touch with <a href="https://eduvpn.org">eduVPN</a>, who are interested about deploying MirageVPN. We got example configuration from their side, and <a href="https://github.com/robur-coop/miragevpn/pull/201">fixed</a> <a href="https://github.com/robur-coop/miragevpn/pull/168">some</a> <a href="https://github.com/robur-coop/miragevpn/pull/202">issues</a>, and also implemented <a href="https://github.com/robur-coop/miragevpn/pull/169">tls-crypt</a> - which was straightforward since we earlier spend time to implement <a href="https://blog.robur.coop/articles/miragevpn.html">tls-crypt-v2</a>.</p>
<p>In January, they gave MirageVPN another try, and <a href="https://github.com/robur-coop/miragevpn/issues/206">measured the performance</a> -- which was very poor -- MirageVPN (run as a Unix binary) provided a bandwith of 9.3Mb/s, while OpenVPN provided a bandwidth of 360Mb/s (using a VPN tunnel over TCP).</p>
<p>We aim at spending less resources for computing, thus the result was not satisfying for us. We re-read a lot of code, refactored a lot, and are now at ~250Mb/s.</p>
<h2 id="tooling-for-performance-engineering-of-ocaml"><a class="anchor" aria-hidden="true" href="#tooling-for-performance-engineering-of-ocaml"></a>Tooling for performance engineering of OCaml</h2>
<p>As a first approach we connected with the MirageVPN unix client &amp; OpenVPN client to a eduVPN server and ran speed tests using <a href="https://fast.com">fast.com</a>. This was sufficient to show the initial huge gap in download speeds between MirageVPN and OpenVPN. There is <em>a lot</em> of noise in this approach as there are many computers and routers involved in this setup, and it is hard to reproduce.</p>
<p>To get more reproducible results we set up a local VM with openvpn and iperf3 installed. On the host machine we can then connect to the VM's OpenVPN server and run iperf3 against the <em>VPN</em> ip address. This worked more reliably. However, it was still noisy and not suitable to measure single digit percentage changes in performance.
To better guide the performance engineering, we also developed <a href="https://github.com/robur-coop/miragevpn/pull/230">a microbenchmark</a> using OCaml tooling. This will setup a client and server without any input and output, and transfer data in memory.</p>
<p>We also re-read our code and used the Linux utility <a href="https://perf.wiki.kernel.org/index.php/Main_Page"><code>perf</code></a> together with <a href="https://github.com/brendangregg/FlameGraph">Flamegraph</a> to graph its output. This works nicely with OCaml programs (we're using the 4.14.1 compiler and runtime system). We did the performance engineering on Unix binaries, i.e. not on MirageOS unikernels - but the MirageVPN protocol is used in both scenarios - thus the performance improvements described here are also in the MirageVPN unikernels.</p>
<h2 id="takeaway-of-performance-engineering"><a class="anchor" aria-hidden="true" href="#takeaway-of-performance-engineering"></a>Takeaway of performance engineering</h2>
<p>The learnings of our performance engineering are in three areas:</p>
<ul>
<li>Formatting strings is computational expensive -- thus if in an error case a hexdump is produced of a packet, its construction must be delayed for when the error case is executed (we have <a href="https://github.com/robur-coop/miragevpn/pull/220">this PR</a> and <a href="https://github.com/robur-coop/miragevpn/pull/209">that PR</a>). Alain Frisch wrote a nice <a href="https://www.lexifi.com/blog/ocaml/note-about-performance-printf-and-format/#">blog post</a> at LexiFi about performance of <code>Printf</code> and <code>Format</code>[^lexifi-date].</li>
<li>Rethink allocations: fundamentally, only a single big buffer (to be send out) for each incoming packet should be allocated, not a series of buffers that are concatenated (see <a href="https://github.com/robur-coop/miragevpn/pull/217">this PR</a> and <a href="https://github.com/robur-coop/miragevpn/pull/219">that PR</a>). Additionally, not zeroing out the just allocated buffer (if it is filled with data anyways) removes some further instructions (see <a href="https://github.com/robur-coop/miragevpn/pull/218">this PR</a>). And we figured that appending to an empty buffer nevertheless allocated and copied in OCaml, so we worked on <a href="https://github.com/robur-coop/miragevpn/pull/214">this PR</a>.</li>
<li>Still an open topic is: we are in the memory-safe language OCaml, and we sometimes extract data out of a buffer (or set data in a buffer). Now, each operation lead to bounds checks (that we do not touch memory that is not allocated or not ours). However, if we just checked for the buffer being long enough (either by checking the length, or by allocating a specific amount of data), these bounds checks are superfluous. So far, we don't have an automated solution for this issue, but we are <a href="https://discuss.ocaml.org/t/bounds-checks-for-string-and-bytes-when-retrieving-or-setting-subparts-thereof/">discussing it in the OCaml community</a>, and are eager to find a solution to avoid unneeded computations.</li>
</ul>
<h2 id="conclusion"><a class="anchor" aria-hidden="true" href="#conclusion"></a>Conclusion</h2>
<p>To conclude: we already achieved a factor of 25 in performance by adapting the code in various ways. We have ideas to improve the performance even more in the future - we also work on using OCaml string and bytes, instead of off-the-OCaml-heap-allocated bigarrays (see <a href="https://blog.robur.coop/articles/speeding-ec-string.html">our previous article</a>, which provided some speedups).</p>
<p>Don't hesitate to reach out to us on <a href="https://github.com/robur-coop/miragevpn/issues">GitHub</a>, or <a href="https://robur.coop/Contact">by mail</a> if you're stuck.</p>
<p>We want to thank <a href="https://nlnet.nl">NLnet</a> for their funding (via <a href="https://www.assure.ngi.eu/">NGI assure</a>), and <a href="https://eduvpn.org">eduVPN</a> for their interest.</p>
<p>[^lexifi-date]: It has come to our attention that the blog post is rather old (2012) and that the implementation has been completely rewritten since then.</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

5
articles/miragevpn-performance.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-04-16 date: 2024-04-16
title: Speeding up MirageVPN and use it in the wild article.title: Speeding up MirageVPN and use it in the wild
description: article.description:
Performance engineering of MirageVPN, speeding it up by a factor of 25. Performance engineering of MirageVPN, speeding it up by a factor of 25.
tags: tags:
- OCaml - OCaml
@ -19,6 +19,7 @@ coauthors:
name: Reynir Björnsson name: Reynir Björnsson
email: reynir@reynir.dk email: reynir@reynir.dk
link: https://reyn.ir/ link: https://reyn.ir/
contribution: What is this field used for?
--- ---
As we were busy continuing to work on [MirageVPN](https://github.com/robur-coop/miragevpn), we got in touch with [eduVPN](https://eduvpn.org), who are interested about deploying MirageVPN. We got example configuration from their side, and [fixed](https://github.com/robur-coop/miragevpn/pull/201) [some](https://github.com/robur-coop/miragevpn/pull/168) [issues](https://github.com/robur-coop/miragevpn/pull/202), and also implemented [tls-crypt](https://github.com/robur-coop/miragevpn/pull/169) - which was straightforward since we earlier spend time to implement [tls-crypt-v2](https://blog.robur.coop/articles/miragevpn.html). As we were busy continuing to work on [MirageVPN](https://github.com/robur-coop/miragevpn), we got in touch with [eduVPN](https://eduvpn.org), who are interested about deploying MirageVPN. We got example configuration from their side, and [fixed](https://github.com/robur-coop/miragevpn/pull/201) [some](https://github.com/robur-coop/miragevpn/pull/168) [issues](https://github.com/robur-coop/miragevpn/pull/202), and also implemented [tls-crypt](https://github.com/robur-coop/miragevpn/pull/169) - which was straightforward since we earlier spend time to implement [tls-crypt-v2](https://blog.robur.coop/articles/miragevpn.html).

46
articles/miragevpn-server.html Normal file
View file

@ -0,0 +1,46 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogMirageVPN server
</title>
<meta name="description" content="Announcement of our MirageVPN server.">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>MirageVPN server</h1>
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>cryptography</li><li>security</li><li>VPN</li></ul><p>It is a great pleasure to finally announce that we have finished a server implementation for MirageVPN (OpenVPN™-compatible). This allows to setup a very robust VPN network on both the client and the server side.</p>
<p>As announced last year, <a href="https://blog.robur.coop/articles/miragevpn.html">MirageVPN</a> is a reimplemtation of OpenVPN™ in OCaml, with <a href="https://mirage.io">MirageOS</a> unikernels.</p>
<h2 id="why-a-miragevpn-server"><a class="anchor" aria-hidden="true" href="#why-a-miragevpn-server"></a>Why a MirageVPN server?</h2>
<p>Providing Internet services with programming languages that have not much safety requires a lot of discipline by the developers to avoid issues which may lead to exploitable services that are attacked (and thus will circumvent any security goals). Especially services that are critical for security and privacy, it is crucial to avoid common memory safety pitfalls.</p>
<p>Some years back, when we worked on the client implementation, we also drafted a server implementation. The reasoning was that a lot of the code was already there, and just a few things needed to be developed to allow clients to connect there.</p>
<p>Now, we spend several months to push our server implementation into a state where it is usable and we are happy for everyone who wants to test it. We also adapted the modern ciphers we recently implemented for the client, and also tls-crypt and tls-crypt-v2 for the server implementation.</p>
<p>The overall progress was tracked in <a href="https://github.com/robur-coop/miragevpn/issues/15">this issue</a>. We developed, next to the MirageOS unikernel, also a test server that doesn't use any tun interface.</p>
<p>Please move along to our handbook with the <a href="https://robur-coop.github.io/miragevpn-handbook/miragevpn_server.html">chapter on MirageVPN server</a>.</p>
<p>If you encounter any issues, please open an issue at <a href="https://github.com/robur-coop/miragevpn">the repository</a>.</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

5
articles/miragevpn-server.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-06-17 date: 2024-06-17
title: MirageVPN server article.title: MirageVPN server
description: article.description:
Announcement of our MirageVPN server. Announcement of our MirageVPN server.
tags: tags:
- OCaml - OCaml
@ -18,6 +18,7 @@ coauthors:
name: Reynir Björnsson name: Reynir Björnsson
email: reynir@reynir.dk email: reynir@reynir.dk
link: https://reyn.ir/ link: https://reyn.ir/
contribution: What is this field used for?
--- ---
It is a great pleasure to finally announce that we have finished a server implementation for MirageVPN (OpenVPN™-compatible). This allows to setup a very robust VPN network on both the client and the server side. It is a great pleasure to finally announce that we have finished a server implementation for MirageVPN (OpenVPN™-compatible). This allows to setup a very robust VPN network on both the client and the server side.

120
articles/miragevpn.html Normal file
View file

@ -0,0 +1,120 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogMirageVPN &amp; tls-crypt-v2
</title>
<meta name="description" content="How we implementated tls-crypt-v2 for miragevpn">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>MirageVPN &amp; tls-crypt-v2</h1>
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>VPN</li><li>security</li></ul><p>In 2019 <a href="https://robur.coop/">Robur</a> started working on a <a href="https://github.com/robur-coop/miragevpn/">OpenVPN™-compatible implementation in OCaml</a>.
The project was funded for 6 months in 2019 by <a href="https://prototypefund.de">prototypefund</a>.
In late 2022 we applied again for funding this time to the <a href="https://www.assure.ngi.eu/">NGI Assure</a> open call, and our application was eventually accepted.
In this blog post I will explain why reimplementing the OpenVPN™ protocol in OCaml is a worthwhile effort, and describe the Miragevpn implementation and in particular the <code>tls-crypt-v2</code> mechanism.</p>
<h2 id="what-even-is-openvpn™"><a class="anchor" aria-hidden="true" href="#what-even-is-openvpn™"></a>What even is OpenVPN™?</h2>
<p><a href="https://openvpn.net/">OpenVPN™</a> is a protocol and software implementation to provide <a href="https://en.wikipedia.org/wiki/Virtual_private_network">virtual private networks</a>: computer networks that do not exist in hardware and are encrypted and tunnelled through existing networks.
Common use cases for this is to provide access to internal networks for remote workers, and for routing internet traffic through another machine for various reasons e.g. when using untrusted wifi, privacy from a snooping ISP, circumventing geoblock etc.</p>
<p>It is a protocol that has been worked on and evolved over the decades.
OpenVPN™ has a number of modes of operations as well as a number of options in the order of hundreds.
The modes can be categorized into two main categories: static mode and TLS mode.
The former mode uses static symmetric keys, and will be removed in the upcoming OpenVPN™ 2.7 (community edition).
I will not focus on static mode in this post.
The latter uses separate data &amp; control channels where the control channel uses TLS - more on that later.</p>
<h3 id="why-reimplement-it-and-why-in-ocaml"><a class="anchor" aria-hidden="true" href="#why-reimplement-it-and-why-in-ocaml"></a>Why reimplement it? And why in OCaml?</h3>
<p>Before diving into TLS mode and eventually tls-crypt-v2 it's worth to briefly discuss why we spend time reimplementing the OpenVPN™ protocol.
You may ask yourself: why not just use the existing tried and tested implementation?</p>
<p>OpenVPN™ community edition is implemented in the C programming language.
It heavily uses the OpenSSL library[^mbedtls] which is as well written in C and has in the past had some notable security vulnerabilities.
Many vulnerabilities and bugs in C can be easily avoided in other languages due to bounds checking and stricter and more expressive type systems.
The state machine of the protocol can be more easily be expressed in OCaml, and some properties of the protocol can be encoded in the type system.</p>
<p>Another reason is <a href="https://mirage.io/">Mirage OS</a>, a library operating system implemented in OCaml.
We work on the Mirage project and write applications (unikernels) using Mirage.
In many cases it would be desirable to be able to connect to an existing VPN network[^vpn-network],
or be able to offer a VPN network to clients using OpenVPN™.</p>
<p>Consider a VPN provider:
The VPN provider runs many machines that run an operating system in order to run the user-space OpenVPN™ service.
There are no <em>real</em> users on the system, and a lot of unrelated processes and legacy layers are around that are not needed.
With a Mirage OS unikernel, which is basically a statically linked binary and operating system such a setup becomes simpler with fewer layers.
With <a href="https://robur.coop/Projects/Reproducible_builds">reproducible builds</a> deployment and updates will be straightforward.</p>
<p>Another very interesting example is a unikernel for <a href="https://www.qubes-os.org/">Qubes OS</a> that we have planned.
Qubes OS is an operating system with a high focus on security.
It offers an almost seamless experience of running applications in different virtual machines on the same machine.
The networking provided to a application (virtual machine) can be restricted to only go through the VPN.
It is possible to use OpenVPN™ for such a setup, but that requires running OpenVPN™ in a full Linux virtual machine.
With Mirage OS the resource footprint is typically much smaller than an equivalent application running in a Linux virtual machine; often the memory footprint is smaller by an order.</p>
<p>Finally, while it's not an explicit goal of ours, reimplementing a protocol without an explicit specification can help uncover bugs and things that need better documentation in the original implementation.</p>
<h3 id="tls-mode"><a class="anchor" aria-hidden="true" href="#tls-mode"></a>TLS mode</h3>
<p>There are different variants of TLS mode, but what they share is separate &quot;control&quot; channel and &quot;data&quot; channel.
The control channel is used to do a TLS handshake, and with the established TLS session data channel encryption keys, username/password authentication, etc. is negotiated.
Once this dance has been performed and data channel encryption keys have been negotiated the peers can exchange IP packets over the data channel.</p>
<p>Over the years a number of mechanisms has been implemented to protect the TLS stack from being exposed to third parties, protect against denial of service attacks and to hide information exchanged during a TLS handshake such as certificates (which was an isue before TLS 1.3).
These are known as <code>tls-auth</code>, <code>tls-crypt</code> and <code>tls-crypt-v2</code>.
The <code>tls-auth</code> mechanism adds a pre-shared key for hmac authentication on the control channel.
This makes it possible for an OpenVPN™ server to reject early clients that don't know the shared key before any TLS handshakes are performed.
In <code>tls-crypt</code> the control channel is encrypted as well as hmac authenticated using a pre-shared key.
Common to both is that the pre-shared key is shared between the server and all clients.
For large deployments this significantly reduces the usefulness - the key is more likely to be leaked the greate the number of clients who share this key.</p>
<h3 id="tls-crypt-v2"><a class="anchor" aria-hidden="true" href="#tls-crypt-v2"></a>tls-crypt-v2</h3>
<p>To improve on <code>tls-crypt</code>, <code>tls-crypt-v2</code> uses one pre-shared key per client.
This could be a lot of keys for the server to keep track of, so instead of storing all the client keys on the server the server has a special tls-crypt-v2 server key that is used to <em><a href="https://en.wikipedia.org/wiki/Key_wrap">wrap</a></em> the client keys.
That is, each client has their own client key as well as the client key wrapped using the server key.
The protocol is then extended so the client in the first message appends the wrapped key <em>unencrypted</em>.
The server can then decrypt and verify the client key and decrypt the rest of the packet.
Then the client and server use the client key just as in <code>tls-crypt</code>.</p>
<p>This is great!
Each client can have their own key, and the server doesn't need to keep a potentially large database of client keys.
What if the client's key is leaked?
A detail I didn't mention is that the wrapped key contains metadata.
By default this is the current timestamp, but it is possible on creation to put any (relative short) binary data in there as the metadata.
The server can then be configured to check the metadata by calling a script.</p>
<p>An issue exists that an initial packet takes up resources on the server because the server needs to</p>
<ol>
<li>decrypt the wrapped key, and</li>
<li>keep the unwrapped key and other data in memory while waiting for the handshake to complete.</li>
</ol>
<p>This can be abused in an attack very similar to a TCP <a href="https://en.wikipedia.org/wiki/SYN_flood">SYN flood</a>.
Without <code>tls-crypt-v2</code> OpenVPN uses a specially crafted session ID (a 64 bit identifier) to avoid this issue similar to <a href="https://en.wikipedia.org/wiki/SYN_cookies">SYN cookies</a>.
To address this in OpenVPN 2.6 the protocol for <code>tls-crypt-v2</code> was extended yet further with a 'HMAC cookie' mechanism.
The client sends the same packet as before, but uses a sequence number <code>0x0f000001</code> instead of <code>1</code> to signal support of this mechanism.
The server responds in a similar manner with a sequence number of <code>0x0f000001</code> and the packet is appended with a tag-length-value encoded list of flags.
At the moment only one tag and one value is defined which signifies the server supports HMAC cookies - this seems unnecessarily complex, but is done to allow future extensibility.
Finally, if the server supports HMAC cookies, the client sends a packet where the wrapped key is appended in cleartext.
The server is now able to decrypt the third packet without having to keep the key from the first packet around and can verify the session id.</p>
<h2 id="cool-lets-deploy-it"><a class="anchor" aria-hidden="true" href="#cool-lets-deploy-it"></a>Cool! Let's deploy it!</h2>
<p>Great!
We build on a daily basis unikernels in our <a href="https://builds.robur.coop/">reproducible builds setup</a>.
At the time of writing we have published a <a href="https://builds.robur.coop/job/miragevpn-router">Miragevpn router unikernel</a> acting as a client.
For general instructions on running Mirage unikernels see our <a href="https://robur.coop/Projects/Reproducible_builds">reproducible builds</a> blog post.
The unikernel will need a block device containing the OpenVPN™ configuration and a network device.
More detailed instructions Will Follow Soon™!
Don't hesitate to reach out to us on <a href="https://github.com/robur-coop/miragevpn/issues">GitHub</a>, <a href="https://robur.coop/Contact">by mail</a> or me personally <a href="https://bsd.network/@reynir">on Mastodon</a> if you're stuck.</p>
<p>[^mbedtls]: It is possible to compile OpenVPN™ community edition with Mbed TLS instead of OpenSSL which is written in C as well.</p>
<p>[^vpn-network]: I use the term &quot;VPN network&quot; to mean the virtual private network itself. It is a bit odd because the 'N' in 'VPN' is 'Network', but without disambiguation 'VPN' could refer to the network itself, the software or the service.</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/miragevpn.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2023-11-14 date: 2023-11-14
title: MirageVPN & tls-crypt-v2 article.title: MirageVPN & tls-crypt-v2
description: article.description:
How we implementated tls-crypt-v2 for miragevpn How we implementated tls-crypt-v2 for miragevpn
tags: tags:
- OCaml - OCaml

83
articles/qubes-miragevpn.html Normal file
View file

@ -0,0 +1,83 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogqubes-miragevpn, a MirageVPN client for QubesOS
</title>
<meta name="description" content="A new OpenVPN client for QubesOS">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>qubes-miragevpn, a MirageVPN client for QubesOS</h1>
<ul class="tags-list"><li>OCaml</li><li>vpn</li><li>unikernel</li><li>QubesOS</li></ul><p>We are pleased to announce the arrival of a new unikernel:
<a href="https://github.com/robur-coop/qubes-miragevpn">qubes-miragevpn</a>. The latter is the result of work begun
several months ago on <a href="https://github.com/robur-coop/miragevpn">miragevpn</a>.</p>
<p>Indeed, with the ambition of completing our unikernel suite and the success of
<a href="https://github.com/mirage/qubes-mirage-firewall">qubes-mirage-firewall</a> - as well as the general aims of
QubesOS - we thought it would be a good idea to offer this community a unikernel
capable of acting as an OpenVPN client, from which other virtual machines (app
qubes) can connect so that all their connections pass through the OpenVPN
tunnel.</p>
<h2 id="qubesos--mirageos"><a class="anchor" aria-hidden="true" href="#qubesos--mirageos"></a>QubesOS &amp; MirageOS</h2>
<p>Unikernels and QubesOS have always been a tempting idea for users in the sense
that a network application (such as a firewall or VPN client) could be smaller
than a Linux kernel: no keyboard, mouse, wifi management, etc. Just network
management via virtual interfaces should suffice.</p>
<p>In this case, the unikernel corresponds to this ideal where, starting from a
base (<a href="https://github.com/Solo5/solo5">Solo5</a>) that only allows the strictly necessary (reading and
writing on a virtual interface or block device) and building on top of it all
the application logic strictly necessary to the objective we wish to achieve
reduces, in effect, drastically:</p>
<ol>
<li>the unikernel's attack surface</li>
<li>its weight</li>
<li>its memory usage</li>
</ol>
<p>We won't go into all the work that's been done to maintain and improve
<a href="https://github.com/mirage/qubes-mirage-firewall">qubes-mirage-firewall</a> over the last 10
years<sup><a href="#fn1">1</a></sup>, but it's clear that this particular unikernel has
found its audience, who aren't necessarily OCaml and MirageOS aficionados.</p>
<p>In other words, <a href="https://github.com/mirage/qubes-mirage-firewall">qubes-mirage-firewall</a> may well be a
fine example of what can actually be done with MirageOS, and of real utility.</p>
<hr>
<p><tag id="fn1"><strong>1</strong></tag>: <a href="https://github.com/marmarek">marmarek</a>, <a href="https://github.com/yomimono">Mindy</a> or
<a href="https://github.com/mato">mato</a> were (and still are) heavily involved in the work between QubesOS
and MirageOS. We'd also like to thank them, because if we're able to continue
this adventure, it's also thanks to them.</p>
<h2 id="qubesos--miragevpn"><a class="anchor" aria-hidden="true" href="#qubesos--miragevpn"></a>QubesOS &amp; MirageVPN</h2>
<p>So, after a lengthy development phase for MirageVPN, we set about developing a
unikernel for QubesOS to offer an OpenVPN client as an operating system. We'd
like to give special thanks to <a href="https://github.com/palainp">Pierre Alain</a>, who helped us to better
understand QubesOS and its possibilities.</p>
<p>The unikernel is available here: https://github.com/robur-coop/qubes-miragevpn
A tutorial has just been created to help QubesOS users install and configure
such an unikernel: https://robur-coop.github.io/miragevpn-handbook/</p>
<p>In the same way as <a href="https://github.com/mirage/qubes-mirage-firewall">qubes-mirage-firewall</a>, we hope to
offer a solution that works and expand the circle of MirageOS and unikernel
users!</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/qubes-miragevpn.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-06-24 date: 2024-06-24
title: qubes-miragevpn, a MirageVPN client for QubesOS article.title: qubes-miragevpn, a MirageVPN client for QubesOS
description: A new OpenVPN client for QubesOS article.description: A new OpenVPN client for QubesOS
tags: tags:
- OCaml - OCaml
- vpn - vpn

82
articles/speeding-ec-string.html Normal file
View file

@ -0,0 +1,82 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogSpeeding elliptic curve cryptography
</title>
<meta name="description" content="How we improved the performance of elliptic curves by only modifying the underlying byte array">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>Speeding elliptic curve cryptography</h1>
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>cryptography</li><li>security</li></ul><p>TL;DR: replacing cstruct with string, we gain a factor of 2.5 in performance.</p>
<h2 id="mirage-crypto-ec"><a class="anchor" aria-hidden="true" href="#mirage-crypto-ec"></a>Mirage-crypto-ec</h2>
<p>In April 2021 We published our implementation of <a href="https://hannes.robur.coop/Posts/EC">elliptic curve cryptography</a> (as <code>mirage-crypto-ec</code> opam package) - this is DSA and DH for NIST curves P224, P256, P384, and P521, and also Ed25519 (EdDSA) and X25519 (ECDH). We use <a href="https://github.com/mit-plv/fiat-crypto/">fiat-crypto</a> for the cryptographic primitives, which emits C code that by construction is correct (note: earlier we stated &quot;free of timing side-channels&quot;, but this is a huge challenge, and as <a href="https://discuss.systems/@edwintorok/111925959867297453">reported by Edwin Török</a> likely impossible on current x86 hardware). More C code (such as <code>point_add</code>, <code>point_double</code>, and further 25519 computations including tables) have been taken from the BoringSSL code base. A lot of OCaml code originates from our TLS 1.3 work in 2018, where Etienne Millon, Nathan Rebours, and Clément Pascutto interfaced <a href="https://github.com/mirage/fiat/">elliptic curves for OCaml</a> (with the goal of being usable with MirageOS).</p>
<p>The goal of mirage-crypto-ec was: develop elliptic curve support for OCaml &amp; MirageOS quickly - which didn't leave much time to focus on performance. As time goes by, our mileage varies, and we're keen to use fewer resources - and thus fewer CPU time and a smaller memory footprint is preferable.</p>
<h2 id="memory-allocation-and-calls-to-c"><a class="anchor" aria-hidden="true" href="#memory-allocation-and-calls-to-c"></a>Memory allocation and calls to C</h2>
<p>OCaml uses managed memory with a generational copying collection. To safely call a C function at any point in time when the arguments are OCaml values (memory allocated on the OCaml heap), it is crucial that while the C function is executed, the arguments should stay at the same memory location, and not being moved by the GC. Otherwise the C code may be upset retrieving wrong data or accessing unmapped memory.</p>
<p>There are several strategies to achieve this, ranging from &quot;let's use another memory area where the GC doesn't mess around with&quot;, &quot;do not run any GC while executing the C code&quot; (read further in the OCaml <a href="https://v2.ocaml.org/releases/4.14/htmlman/intfc.html#ss:c-direct-call">cheaper C calls</a> manual), &quot;deeply copy the arguments to a non-moving memory area before executing C code&quot;, and likely others.</p>
<p>For our elliptic curve operations, the C code is pretty simple - there are no memory allocations happening in C, neither are exceptions raised. Also, the execution time of the code is constant and pretty small.</p>
<h2 id="ocaml-cstruct"><a class="anchor" aria-hidden="true" href="#ocaml-cstruct"></a>ocaml-cstruct</h2>
<p>In the <a href="https://mirage.io">MirageOS</a> ecosystem, a core library is <a href="https://github.com/mirage/ocaml-cstruct">cstruct</a> - which purpose is manifold: provide ppx rewriters to define C structure layouts in OCaml (getter/setter functions are generated), as well as enums; also a fundamental idea is to use OCaml bigarray which is non-moving memory not allocated on the OCaml heap but directly by calling <code>malloc</code>. The memory can even be page-aligned, as required by some C software, such as Xen. Convenient functionality, such as &quot;retrieve a big-endian unsigned 32 bit integer from offset X in this buffer&quot; are provided as well.</p>
<p>But there's a downside to it - as time moves along, Xen is no longer the only target for MirageOS, and other virtualization mechanisms (such as KVM / virtio) do not require page-aligned memory ranges that are retained at a given memory address. It also turns out that cstruct spends a lot of time in bounds checks. Another huge downside is that OCaml tooling (such as statmemprof) was for a long time (maybe still is not?) unaware of out-of-OCaml-GC allocated memory (cstruct uses bigarray as underlying buffer). Freeing up the memory requires finalizers to be executed - after all pretty tedious (expensive) and against the OCaml runtime philosophy.</p>
<p>As time moves forward, also the OCaml standard library got support for (a) strings are immutable byte vectors now (since 4.06 - released in 2017 -- there's as well an interface for mutable/immutable cstruct, but that is not used as far as I can tell), (b) retrieve a certain amount of octets in a string or byte as (unsigned) integer number (since 4.08 - released in 2019, while some additional functionality is only available in 4.13).</p>
<p>Still, bigarrays are necessary in certain situations - if you need to have a non-moving (shared) area of memory, as in the Xen interface, but also e.g. when you compute in parallel in different processes, or when you need mmap()ed files.</p>
<h2 id="putting-it-together"><a class="anchor" aria-hidden="true" href="#putting-it-together"></a>Putting it together</h2>
<p>Already in October 2021, Romain <a href="https://github.com/mirage/mirage-crypto/pull/146">proposed</a> to not use cstruct, but bytes for mirage-crypto-ec. The PR was sitting around since there were benchmarks missing, and developer time was small. But recently, Virgile Robles <a href="https://github.com/mirage/mirage-crypto/pull/191">proposed</a> another line of work to use pre-computed tables for NIST curves to speed up the elliptic curve cryptography. Conducting performance evaluation resulted that the &quot;use bytes instead of cstruct&quot; combined with pre-computed tables made a huge difference (factor of 6) compared to the latest release.</p>
<p>To ease reviewing changes, we decided to focus on landing the &quot;use bytes instead of cstruct&quot; first, and gladly Pierre Alain had already rebased the existing patch onto the latest release of mirage-crypto-ec. We also went further and use string where applicable instead of bytes. For safety reasons we also introduced an API layer which (a) allocates a byte vector for the result (b) calls the primitive, and (c) transforms the byte vector into an immutable string. This API is more in line with functional programming (immutable values), and since allocations and deallocations of values are cheap, there's no measurable performance decrease.</p>
<p>All the changes are internal, there's no external API that needs to be adjusted - still there's at the API boundary one conversion of cstruct to string (and back for the return value) done.</p>
<p>We used <code>perf</code> to construct some flame graphs (of the ECDSA P256 sign), shown below.</p>
<p><img src="../images/trace-cstruct-440.svg" alt="Flamegraph of ECDSA sign with cstruct" ></p>
<p>The flame graph of P256 ECDSA sign using the mirage-crypto release 0.11.2. The majority of time is spent in &quot;do_sign&quot;, which calls <code>inv</code> (inversion), <code>scalar_mult</code> (majority of time), and <code>x_of_finite_point_mod_n</code>. The scalar multiplication spends time in <code>add</code>, <code>double</code> and <code>select</code>. Several towers starting at <code>Cstruct.create_919</code> are visible.</p>
<p>With PR#146, the flame graph looks different:</p>
<p><img src="../images/trace-string-770.svg" alt="Flamegraph of ECDSA sign with string" ></p>
<p>Now, the allocation towers do not exist anymore. The time of a sign operation is spend in <code>inv</code>, <code>scalar_mult</code>, and <code>x_of_finite_point_mod_n</code>. There's still room for improvements in these operations.</p>
<h2 id="performance-numbers"><a class="anchor" aria-hidden="true" href="#performance-numbers"></a>Performance numbers</h2>
<p>All numbers were gathered on a Lenovo X250 laptop with a Intel i7-5600U CPU @ 2.60GHz. We used OCaml 4.14.1 as compiler. The baseline is OpenSSL 3.0.12. All numbers are in operations per second.</p>
<p>NIST P-256</p>
<p>| op | 0.11.2 | PR#146 | speedup | OpenSSL | speedup |
| - | - | - | - | - | - |
| sign | 748 | 1806 | 2.41x | 34392 | 19.04x |
| verify | 285 | 655 | 2.30x | 12999 | 19.85x |
| ecdh | 858 | 1785 | 2.08x | 16514 | 9.25x |</p>
<p>Curve 25519</p>
<p>| op | 0.11.2 | PR#146 | speedup | OpenSSL | speedup |
| - | - | - | - | - | - |
| sign | 10713 | 11560 | 1.08x | 21943 | 1.90x |
| verify | 7600 | 8314 | 1.09x | 7081 | 0.85x |
| ecdh | 12144 | 13457 | 1.11x | 26201 | 1.95x |</p>
<p>Note: to re-create the performance numbers, you can run <code>openssl speed ecdsap256 ecdhp256 ed25519 ecdhx25519</code> - for the OCaml site, use <code>dune bu bench/speed.exe --rel</code> and <code>_build/default/bench/speed.exe ecdsa-sign ecdsa-verify ecdh-share</code>.</p>
<p>The performance improvements are up to 2.5 times compared to the latest mirage-crypto-ec release (look at the 4th column). In comparison to OpenSSL, we still lack a factor of 20 for the NIST curves, and up to a factor of 2 for 25519 computations (look at the last column).</p>
<p>If you have ideas for improvements, let us know via an issue, eMail, or a pull request :) We started to <a href="https://github.com/mirage/mirage-crypto/issues/193">gather some</a> for 25519 by comparing our code with changes in BoringSSL over the last years.</p>
<p>As a spoiler, for P-256 sign there's another improvement of around 4.5 with <a href="https://github.com/mirage/mirage-crypto/pull/191">Virgile's PR</a> using pre-computed tables also for NIST curves.</p>
<h2 id="the-road-ahead-for-2024"><a class="anchor" aria-hidden="true" href="#the-road-ahead-for-2024"></a>The road ahead for 2024</h2>
<p>Remove all cstruct, everywhere, apart from in mirage-block-xen and mirage-net-xen ;). It was a fine decision in the early MirageOS days, but from a performance point of view, and for making our packages more broadly usable without many dependencies, it is time to remove cstruct. Earlier this year we already <a href="https://github.com/mirage/ocaml-tar/pull/137">removed cstruct from ocaml-tar</a> for similar reasons.</p>
<p>Our MirageOS work is only partially funded, we cross-fund our work by commercial contracts and public (EU) funding. We are part of a non-profit company, you can make a (tax-deducable - at least in the EU) <a href="https://aenderwerk.de/donate/">donation</a> (select &quot;DONATION robur&quot; in the dropdown menu).</p>
<p>We're keen to get MirageOS deployed in production - if you would like to do that, don't hesitate to reach out to us via eMail team at robur.coop</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/speeding-ec-string.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-02-13 date: 2024-02-13
title: Speeding elliptic curve cryptography article.title: Speeding elliptic curve cryptography
description: article.description:
How we improved the performance of elliptic curves by only modifying the underlying byte array How we improved the performance of elliptic curves by only modifying the underlying byte array
tags: tags:
- OCaml - OCaml

405
articles/tar-release.html Normal file
View file

@ -0,0 +1,405 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogThe new Tar release, a retrospective
</title>
<meta name="description" content="A little retrospective to the new Tar release and changes">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a href="https://blog.robur.coop/index.html">Back to index</a>
<article>
<h1>The new Tar release, a retrospective</h1>
<ul class="tags-list"><li>OCaml</li><li>Cstruct</li><li>functors</li></ul><p>We are delighted to announce the new release of <code>ocaml-tar</code>. A small library for
reading and writing tar archives in OCaml. Since this is a major release, we'll
take the time in this article to explain the work that's been done by the
cooperative on this project.</p>
<p>Tar is an <strong>old</strong> project. Originally written by David Scott as part of Mirage,
this project is particularly interesting for building bridges between the tools
we can offer and what already exists. Tar is, in fact, widely used. So we're
both dealing with a format that's older than I am (but I'm used to it by email)
and a project that's been around since... 2012 (over 10 years!).</p>
<p>But we intend to maintain and improve it, since we're using it for the
<a href="https://hannes.robur.coop/Posts/OpamMirror">opam-mirror</a> project among other things - this unikernel is to
provide an opam-repository &quot;tarball&quot; for opam when you do <code>opam update</code>.</p>
<h2 id="cstructt--bytes"><a class="anchor" aria-hidden="true" href="#cstructt--bytes"></a><code>Cstruct.t</code> &amp; bytes</h2>
<p>As some of you may have noticed, over the last few months we've begun a fairly
substantial change to the Mirage ecosystem, replacing the use of <code>Cstruct.t</code> in
key places with bytes/string.</p>
<p>This choice is based on 2 considerations:</p>
<ul>
<li>we came to realize that <code>Cstruct.t</code> could be very costly in terms of
performance</li>
<li><code>Cstruct.t</code> remains a &quot;Mirage&quot; structure; outside the Mirage ecosystem, the
use of <code>Cstruct.t</code> is not so &quot;obvious&quot;.</li>
</ul>
<p>The pull-request is available here: https://github.com/mirage/ocaml-tar/pull/137.
The discussion can be interesting in discovering common bugs (uninitialized
buffer, invalid access). There's also a small benchmark to support our initial
intuition<sup><a href="#fn1">1</a></sup>.</p>
<p>But this PR can also be an opportunity to understand the existence of
<code>Cstruct.t</code> in the Mirage ecosystem and the reasons for this historic choice.</p>
<h3 id="cstructt-as-a-non-moveable-data"><a class="anchor" aria-hidden="true" href="#cstructt-as-a-non-moveable-data"></a><code>Cstruct.t</code> as a non-moveable data</h3>
<p>I've already <a href="https://discuss.ocaml.org/t/buffered-io-bytes-vs-bigstring/8978/3">made</a> a list of pros/cons when it comes to
bigarrays. Indeed, <code>Cstruct.t</code> is based on a bigarray:</p>
<pre><code class="language-ocaml">type buffer = (char, Bigarray.int8_unsigned_elt, Bigarray.c_layout) Bigarray.Array1.t
type t =
{ buffer : buffer
; off : int
; len : int }
</code></pre>
<p>The experienced reader may rightly wonder why Cstruct.t is a bigarray with <code>off</code>
and <code>len</code>. First, we need to clarify what a bigarray is for OCaml.</p>
<p>A bigarray is a somewhat special value in OCaml. This value is allocated in the
C heap. In other words, its contents are not in OCaml's garbage collector, but
exist outside it. The first (and very important) implication of this feature is
that the contents of a bigarray do not move (even if the GC tries to defragment
the memory). This feature has several advantages:</p>
<ul>
<li>in parallel programming, it can be very interesting to use a bigarray knowing
that, from the point of view of the 2 processes, the position of the bigarray
will never change - this is essentially what <a href="https://github.com/rdicosmo/parmap">parmap</a> does (before
OCaml 5).</li>
<li>for calculations such as checksum or hash, it can be interesting to use a
bigarray. The calculation would not be interrupted by the GC since the
bigarray does not move. The calculation can therefore be continued at the same
point, which can help the CPU to better predict the next stage of the
calculation. This is what <a href="https://github.com/mirage/digestif">digestif</a> offers and what
<a href="https://github.com/mirage/decompress">decompress</a> requires.</li>
<li>for one reason or another, particularly when interacting with something other
than OCaml, you need to offer a memory zone that cannot move. This is
particularly true for unikernels as Xen guests (where the <em>net device</em>
corresponds to a fixed memory zone with which we need to interact) or
<a href="https://ocaml.org/manual/5.2/api/Unix.html#1_Mappingfilesintomemory">mmap</a>.</li>
<li>there are other subtleties more related to the way OCaml compiles. For
example, using bigarray layouts to manipulate &quot;bigger words&quot; can really have
an impact on performance, as <a href="https://github.com/robur-coop/utcp/pull/29">this PR</a> on <a href="https://github.com/robur-coop/utcp">utcp</a> shows.</li>
<li>finally, it may be useful to store sensitive information in a bigarray so as
to have the opportunity to clean up this information as quickly as possible
(ensuring that the GC has not made a copy) in certain situations.</li>
</ul>
<p>All these examples show that bigarrays can be of real interest as long as
<strong>their uses are properly contextualized</strong> - which ultimately remains very
specific. Our experience of using them in Mirage has shown us their advantages,
but also, and above all, their disadvantages:</p>
<ul>
<li>keep in mind that bigarray allocation uses either a system call like <code>mmap</code> or
<code>malloc()</code>. The latter, compared with what OCaml can offer, is slow. As soon
as you need to allocate bytes/strings smaller than
<a href="https://github.com/ocaml/ocaml/blob/744006bfbfa045cc1ca442ff7b52c2650d2abe32/runtime/alloc.c#L175"><code>(256 * words)</code></a>, these values are allocated in the minor heap,
which is incredibly fast to allocate (3 processor instructions which can be
predicted very well). So, preferring to allocate a 10-byte bigarray rather
than a 10-byte <code>bytes</code> penalizes you enormously.</li>
<li>since the bigarray exists in the C heap, the GC has a special mechanism for
knowing when to <code>free()</code> the zone as soon as the value is no longer in use.
Reference-counting is used to then allocate &quot;small&quot; values in the OCaml heap
and use them to manipulate <em>indirectly</em> the bigarray.</li>
</ul>
<h4 id="ownership-proxy-and-gc"><a class="anchor" aria-hidden="true" href="#ownership-proxy-and-gc"></a>Ownership, proxy and GC</h4>
<p>This last point deserves a little clarification, particularly with regard to the
<code>Bigarray.sub</code> function. This function will not create a new, smaller bigarray
and copy what was in the old one to the new one (as <code>Bytes.sub</code>/<code>String.sub</code>
does). In fact, OCaml will allocate a &quot;proxy&quot; of your bigarray that represents a
subfield. This is where <em>reference-counting</em> comes in. This proxy value needs
the initial bigarray to be manipulated. So, as long as proxies exist, the GC
cannot <code>free()</code> the initial bigarray.</p>
<p>This poses several problems:</p>
<ul>
<li>the first is the allocation of these proxies. They can help us to manipulate
the initial bigarray in several places without copying it, but as time goes
by, these proxies could be very expensive</li>
<li>the second is GC intervention. You still need to scan the bigarray, in a
particular way, to know whether or not to keep it. This particular scan, once
again in time immemorial, was not all that common.</li>
<li>the third concerns bigarray ownership. Since we're talking about proxies, we
can imagine 2 competing tasks having access to the same bigarray.</li>
</ul>
<p>As far as the first point is concerned, <code>Bigarray.sub</code> could still be &quot;slow&quot; for
small data since it was, <em>de facto</em> (since a bigarray always has a finalizer -
don't forget reference counting!), allocated in the major heap. And, in truth,
this is perhaps the main reason for the existence of Cstruct! To have a &quot;proxy&quot;
to a bigarray allocated in the minor heap (and, be fast). But since
<a href="https://github.com/ocaml/ocaml/pull/92">Pierre Chambart's PR#92</a>, the problem is no more.</p>
<p>The second point, on the other hand, is still topical, even if we can see that
<a href="https://github.com/ocaml/ocaml/pull/1738">considerable efforts</a> have been made. What we see every
day on our unikernels is <a href="https://github.com/ocaml/ocaml/issues/7750">the pressure</a> that can be put on
the GC when it comes to bigarrays. Indeed, bigarrays use memory and making the C
heap cohabit with the OCaml heap inevitably comes at a cost. As far as
unikernels are concerned, which have a more limited memory than an OCaml
application, we reach this limit rather quickly and we therefore ask the GC to
work more specifically on our 10 or 20 byte bigarrays...</p>
<p>Finally, the third point can be the toughest. On several occasions, we've
noticed competing accesses on our bigarrays that we didn't want (for example,
<code>http-lwt-client</code> had <a href="https://github.com/robur-coop/http-lwt-client/pull/16">this problem</a>). In our experience,
it's very difficult to observe and know that there is indeed an unauthorized
concurrent access changing the contents of our buffer. In this respect, the
question remains open as regards <code>Cstruct.t</code> and the possibility of encoding
ownership of a <code>Cstruct.t</code> in the type to prevent unauthorized access.
<a href="https://github.com/mirage/ocaml-cstruct/pull/237">This PR</a> is interesting to see all the discussions that have taken
place on this subject<sup><a href="#fn2">2</a></sup>.</p>
<p>It should be noted that, with regard to the third point, the problem also
applies to bytes and the use of <code>Bytes.unsafe_to_string</code>!</p>
<h3 id="conclusion-about-cstruct"><a class="anchor" aria-hidden="true" href="#conclusion-about-cstruct"></a>Conclusion about Cstruct</h3>
<p>We hope we've been thorough enough in our experience with Cstruct. If we go back
to the initial definition of our <code>Cstruct.t</code> shown above and take all the
history into account, it becomes increasingly difficult to argue for a
<strong>systematic</strong> use of Cstruct in our unikernels. In fact, the question of
<code>Cstruct.t</code> versus bytes/string remains completely open.</p>
<p>It's worth noting that the original reasons for <code>Cstruct.t</code> are no longer really
relevant if we consider how OCaml has evolved. It should also be noted that this
systematic approach to using <code>Cstruct.t</code> rather than bytes/string has cost us.</p>
<p>This is not to say that <code>Cstruct.t</code> is obsolete. The library is very good and
offers an API where manipulating bytes to extract information such as a TCP/IP
packet remains more pleasant than directly using bytes (even if, here too,
<a href="https://github.com/ocaml/ocaml/pull/1864">efforts</a> have been made).</p>
<p>As far as <code>ocaml-tar</code> is concerned, what really counts is the possibility for
other projects to use this library without requiring <code>Cstruct.t</code> - thus
facilitating its adoption. In other words, given the advantages/disadvantages of
<code>Cstruct.t</code>, we felt it would be a good idea to remove this dependency.</p>
<hr />
<p><tag id="fn1"><strong>1</strong></tag>: It should be noted that the benchmark also concerns
compression. In this case, we use <code>decompress</code>, which uses bigarrays. So there's
some copying involved (from bytes to bigarrays)! But despite this copying, it
seems that the change is worthwhile.</p>
<p><tag id="fn2"><strong>2</strong></tag>: It reminds me that we've been experimenting with
capabilities and using the type system to enforce certain characteristics. To
date, <code>Cstruct_cap</code> has not been used anywhere, which raises a real question
about the advantages/disadvantages in everyday use.</p>
<h2 id="functors"><a class="anchor" aria-hidden="true" href="#functors"></a>Functors</h2>
<p>This is perhaps the other point of the Mirage ecosystem that is also the subject
of debate. Functors! Before we talk about functors, we need to understand their
relevance in the context of Mirage.</p>
<p>Mirage transforms an application into an operating system. What's the difference
between a &quot;normal&quot; application and a unikernel: the &quot;subsystem&quot; with which you
interact. In this case, a normal application will interact with the host system,
whereas a unikernel will have to interact with the Solo5 <em>mini-system</em>.</p>
<p>What Mirage is trying to offer is the ability for an application to transform
itself into either without changing a thing! Mirage's aim is to <strong>inject</strong> the
subsystem into your application. In this case:</p>
<ul>
<li>inject <code>unix.cmxa</code> when you want a Mirage application to become a simple
executable</li>
<li>inject <a href="https://github.com/mirage/ocaml-solo5">ocaml-solo5</a> when you want to produce a unikernel</li>
</ul>
<p>So we're not going to talk about the pros and cons of this approach here, but
consider this feature as one that requires us to use functors.</p>
<p>Indeed, what's the best way in OCaml to inject one implementation into another:
functors? There are definite advantages here too, but we're going to concentrate
on one in particular: the expressiveness of types at module level (which can be
used as arguments to our functors).</p>
<p>For example, did you know that OCaml has a dependent type system?</p>
<pre><code class="language-ocaml">type 'a nat = Zero : zero nat | Succ : 'a nat -&gt; 'a succ nat
and zero = |
and 'a succ = S
module type T = sig type t val v : t nat end
module type Rec = functor (T:T) -&gt; T
module type Nat = functor (S:Rec) -&gt; functor (Z:T) -&gt; T
module Zero = functor (S:Rec) -&gt; functor (Z:T) -&gt; Z
module Succ = functor (N:Nat) -&gt; functor (S:Rec) -&gt; functor (Z:T) -&gt; S(N(S)(Z))
module Add = functor (X:Nat) -&gt; functor (Y:Nat) -&gt; functor (S:Rec) -&gt; functor (Z:T) -&gt; X(S)(Y(S)(Z))
module One = Succ(Zero)
module Two_a = Add(One)(One)
module Two_b = Succ(One)
module Z : T with type t = zero = struct
type t = zero
let v = Zero
end
module S (T:T) : T with type t = T.t succ = struct
type t = T.t succ
let v = Succ T.v
end
module A = Two_a(S)(Z)
module B = Two_b(S)(Z)
type ('a, 'b) refl = Refl : ('a, 'a) refl
let _ : (A.t, B.t) refl = Refl (* 1+1 == succ 1 *)
</code></pre>
<p>The code is ... magical, but it shows that two differently constructed modules
(<code>Two_a</code> &amp; <code>Two_b</code>) ultimately produce the same type, and OCaml is able to prove
this equality. Above all, the example shows just how powerful functors can be.
But it also shows just how difficult functors can be to understand and use.</p>
<p>In fact, this is one of Mirage's biggest drawbacks: the overuse of functors
makes the code difficult to read and understand. It can be difficult to deduce
in your head the type that results from an application of functors, and the
constraints associated with it... (yes, I don't use <code>merlin</code>).</p>
<p>But back to our initial problem: injection! In truth, the functor is a
fly-killing sledgehammer in most cases. There are many other ways of injecting
what the system would be (and how to do a <code>read</code> or <code>write</code>) into an
implementation. The best example, as <a href="https://discuss.ocaml.org/t/best-practices-and-design-patterns-for-supporting-concurrent-io-in-libraries/15001/4?u=dinosaure">@nojb pointed out</a>, is of
course <a href="https://github.com/mirleft/ocaml-tls">ocaml-tls</a> - this answer also shows a contrast between the
functor approach (with <a href="https://github.com/mirage/ocaml-cohttp">CoHTTP</a> for example) and the &quot;pure value-passing
interface&quot; of <code>ocaml-tls</code>.</p>
<p>What's more, we've been trying to find other approaches for injecting the system
we want for several years now. We can already list several:</p>
<ul>
<li><code>ocaml-tls</code>' &quot;value-passing&quot; approach, of course, but also <code>decompress</code></li>
<li>of course, there's the passing of <a href="https://github.com/mirage/colombe/blob/07cd4cf134168ecd841924ee7ddda1a9af8fbd5a/src/sigs.ml#L13-L16">a record</a> (a sort of
mini-module with fewer possibilities with types, but which does the job - a
poor man's functor, in short) which would have the functions to perform the
system's operations</li>
<li><a href="https://github.com/dinosaure/mimic">mimic</a> can be used to inject a module as an implementation of a
flow/stream according to a resolution mechanism (DNS, <code>/etc/services</code>, etc.) -
a little closer to the idea of <em>runtime-resolved implicit implementations</em></li>
<li>there are, of course, the variants (but if we go back to 2010, this solution
wasn't so obvious) popularized by <a href="https://github.com/dbuenzli/ptime">ptime</a>/<a href="https://github.com/dbuenzli/mtime">mtime</a>, <code>digestif</code> &amp;
<a href="https://github.com/ocaml/dune/pull/1207">dune</a></li>
<li>and finally, <a href="https://github.com/mirage/decompress/blob/c8301ba674e037b682338958d6d0bb5c42fd720e/lib/lzo.ml#L164-L175">GADTs</a>, which describe what the process should
do, then let the user implement the <code>run</code> function according to the system.</li>
</ul>
<p>In short, based on this list and the various experiments we've carried out on a
number of projects, we've decided to remove the functors from <code>ocaml-tar</code>! The
crucial question now is: which method to choose?</p>
<h3 id="the-best-answers"><a class="anchor" aria-hidden="true" href="#the-best-answers"></a>The best answers</h3>
<p>There's no real answer to that, and in truth it depends on what level of
abstraction you're at. In fact, you'd like to have a fairly simple method of
abstraction from the system at the start and at the lowest level, to end up
proposing a functor that does all the <em>ceremony</em> (the glue between your
implementation and the system) at the end - that's what <a href="https://github.com/mirage/ocaml-git">ocaml-git</a>
does, for example.</p>
<p>The abstraction you choose also depends on how the process is going to work. As
far as streams/protocols are concerned, the <code>ocaml-tls</code>/<code>decompress</code> approach
still seems the best. But when it comes to introspecting a file/block-device, it
may be preferable to use a GADT that will force the user to implement an
arbitrary memory access rather than consume a sequence of bytes. In short, at
this stage, experience speaks for itself and, just as we were wrong about
functors, we won't be advising you to use this or that solution.</p>
<p>But based on our experience of <code>ocaml-tls</code> &amp; <code>decompress</code> with LZO (which
requires arbitrary access to the content) and the way Tar works, we decided to
use a &quot;value-passing&quot; approach (to describe when we need to read/write) and a
GADT to describe calculations such as:</p>
<ul>
<li>iterating over the files/folders contained in a Tar document</li>
<li>producing a Tar file according to a &quot;dispenser&quot; of inputs</li>
</ul>
<pre><code class="language-ocaml">val decode : decode_state -&gt; string -&gt;
decode_state *
* [ `Read of int
| `Skip of int
| `Header of Header.t ] option
* Header.Extended.t option
(** [decode state] returns a new state and what the user should do next:
- [`Skip] skip bytes
- [`Read] read bytes
- [`Header hdr] do something according the last header extracted
(like stream-out the contents of a file). *)
type ('a, 'err) t =
| Really_read : int -&gt; (string, 'err) t
| Read : int -&gt; (string, 'err) t
| Seek : int -&gt; (unit, 'err) t
| Bind : ('a, 'err) t * ('a -&gt; ('b, 'err) t) -&gt; ('b, 'err) t
| Return : ('a, 'err) result -&gt; ('a, 'err) t
| Write : string -&gt; (unit, 'err) t
</code></pre>
<p>However, and this is where we come back to OCaml's limitations and where
functors could help us: higher kinded polymorphism!</p>
<h3 id="higher-kinded-polymorphism"><a class="anchor" aria-hidden="true" href="#higher-kinded-polymorphism"></a>Higher kinded Polymorphism</h3>
<p>If we return to our functor example above, there's one element that may be of
interest: <code>T with type t = T.t succ</code></p>
<p>In other words, add a constraint to a signature type. A constraint often seen
with Mirage (but deprecated now according to <a href="https://github.com/mirage/mirage/issues/1004#issue-507517315">this issue</a>) is the
type <code>io</code> and its constraint: <code>type 'a io</code>, <code>with type 'a io = 'a Lwt.t</code>.</p>
<p>So we had this type in Tar. The problem is that our GADT can't understand that
sometimes it will have to manipulate <em>Lwt</em> values, sometimes <em>Async</em> or
sometimes <em>Eio</em> (or <em>Miou</em>!). In other words: how do we compose our <code>Bind</code> with
the <code>Bind</code> of these three targets? The difficulty lies above all in history?
Supporting this library requires us to assume a certain compatibility with
applications over which we have no control. What's more, we need to maintain
support for all three libraries without imposing one.</p>
<hr />
<p>A small disgression at this stage seems important to us, as we've been working
in this way for over 10 years. Of course, despite all the solutions mentioned
above, not depending on a system (and/or a scheduler) also allows us to ensure
the existence of libraries like Tar over more than a decade! The OCaml ecosystem
is changing, and choosing this or that library to facilitate the development of
an application has implications we might regret 10 years down the line (for
example... <code>Cstruct.t</code>!). So, it can be challenging to ensure compatibility with
all systems, but the result is libraries steeped in the experience and know-how
of many developers!</p>
<hr />
<p>So, and this is why we talk about Higher Kinded Polymorphism, how do we abstract
the <code>t</code> from <code>'a t</code> (to replace it with <code>Lwt.t</code> or even with a type such as
<code>type 'a t = 'a</code>)? This is where we're going to use the trick explained in
<a href="https://www.cl.cam.ac.uk/~jdy22/papers/lightweight-higher-kinded-polymorphism.pdf">this paper</a>. The trick is to consider a &quot;new type&quot; that will represent our
monad (lwt or async) and inject/project a value from this monad to something
understandable by our GADT: <code>High : ('a, 't) io -&gt; ('a, 't) t</code>.</p>
<pre><code class="language-ocaml">type ('a, 't) io
type ('a, 'err, 't) t =
| Really_read : int -&gt; (string, 'err, 't) t
| Read : int -&gt; (string, 'err, 't) t
| Seek : int -&gt; (unit, 'err, 't) t
| Bind : ('a, 'err, 't) t * ('a -&gt; ('b, 'err, 't) t) -&gt; ('b, 'err, 't) t
| Return : ('a, 'err) result -&gt; ('a, 'err, 't) t
| Write : string -&gt; (unit, 'err, 't) t
| High : ('a, 't) io -&gt; ('a, 'err, 't) t
</code></pre>
<p>Next, we need to create this new type according to the chosen scheduler. Let's
take <em>Lwt</em> as an example:</p>
<pre><code class="language-ocaml">module Make (X : sig type 'a t end) = struct
type t (* our new type *)
type 'a s = 'a X.t
external inj : 'a s -&gt; ('a, t) io = &quot;%identity&quot;
external prj : ('a, t) io -&gt; 'a s = &quot;%identity&quot;
end
module L = Make(Lwt)
let rec run
: type a err. (a, err, L.t) t -&gt; (a, err) result Lwt.t
= function
| High v -&gt; Ok (L.prj v)
| Return v -&gt; Lwt.return v
| Bind (x, f) -&gt;
run x &gt;&gt;= fun value -&gt; run (f value)
| _ -&gt; ...
</code></pre>
<p>So, as you can see, it's a real trick to avoid doing at home without a
companion. Indeed, the use of <code>%identity</code> corresponds to an <code>Obj.magic</code>! So even
if the <code>io</code> type is exposed (to let the user derive Tar for their own system),
this trick is not exposed for other packages, and we instead suggest helpers
such as:</p>
<pre><code class="language-ocaml">val lwt : 'a Lwt.t -&gt; ('a, 'err, lwt) t
val miou : 'a -&gt; ('a, 'err, miou) t
</code></pre>
<p>But this way, Tar can always be derived from another system, and the process for
extracting entries from a Tar file is the same for <strong>all</strong> systems!</p>
<h2 id="conclusion"><a class="anchor" aria-hidden="true" href="#conclusion"></a>Conclusion</h2>
<p>This Tar release isn't as impressive as this article, but it does sum up all the
work we've been able to do over the last few months and years. We hope that our
work is appreciated and that this article, which sets out all the thoughts we've
had (and still have), helps you to better understand our work!</p>
</article>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

4
articles/tar-release.md
View file

@ -1,7 +1,7 @@
--- ---
date: 2024-08-15 date: 2024-08-15
title: The new Tar release, a retrospective article.title: The new Tar release, a retrospective
description: A little retrospective to the new Tar release and changes article.description: A little retrospective to the new Tar release and changes
tags: tags:
- OCaml - OCaml
- Cstruct - Cstruct

137
atom.xml Normal file
View file

@ -0,0 +1,137 @@
<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
<id>https://blog.robur.coop/atom.xml</id>
<title type="text">The Robur&apos;s blog</title>
<generator uri="https://github.com/xhtmlboi/yocaml" version="2">YOCaml</generator>
<updated>2024-08-21T00:00:00Z</updated>
<author>
<name>The Robur Team</name>
</author>
<link href="https://blog.robur.coop/"/>
<link href="https://blog.robur.coop/atom.xml" rel="self"/>
<entry>
<id>https://blog.robur.coop//articles/2024-08-21-OpenVPN-and-MirageVPN.html</id>
<title type="text">MirageVPN and OpenVPN</title>
<updated>2024-08-21T00:00:00Z</updated>
<summary type="text">Discoveries made implementing MirageVPN, a OpenVPN-compatible VPN library</summary>
<link href="https://blog.robur.coop//articles/2024-08-21-OpenVPN-and-MirageVPN.html" rel="alternate" title="MirageVPN and OpenVPN"/>
<category term="MirageVPN"/>
<category term="OpenVPN"/>
<category term="security"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/tar-release.html</id>
<title type="text">The new Tar release, a retrospective</title>
<updated>2024-08-15T00:00:00Z</updated>
<summary type="text">A little retrospective to the new Tar release and changes</summary>
<link href="https://blog.robur.coop//articles/tar-release.html" rel="alternate" title="The new Tar release, a retrospective"/>
<category term="OCaml"/>
<category term="Cstruct"/>
<category term="functors"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/qubes-miragevpn.html</id>
<title type="text">qubes-miragevpn, a MirageVPN client for QubesOS</title>
<updated>2024-06-24T00:00:00Z</updated>
<summary type="text">A new OpenVPN client for QubesOS</summary>
<link href="https://blog.robur.coop//articles/qubes-miragevpn.html" rel="alternate" title="qubes-miragevpn, a MirageVPN client for QubesOS"/>
<category term="OCaml"/>
<category term="vpn"/>
<category term="unikernel"/>
<category term="QubesOS"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/miragevpn-server.html</id>
<title type="text">MirageVPN server</title>
<updated>2024-06-17T00:00:00Z</updated>
<summary type="text">Announcement of our MirageVPN server.</summary>
<link href="https://blog.robur.coop//articles/miragevpn-server.html" rel="alternate" title="MirageVPN server"/>
<category term="OCaml"/>
<category term="MirageOS"/>
<category term="cryptography"/>
<category term="security"/>
<category term="VPN"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/miragevpn-performance.html</id>
<title type="text">Speeding up MirageVPN and use it in the wild</title>
<updated>2024-04-16T00:00:00Z</updated>
<summary type="text">Performance engineering of MirageVPN, speeding it up by a factor of 25.</summary>
<link href="https://blog.robur.coop//articles/miragevpn-performance.html" rel="alternate" title="Speeding up MirageVPN and use it in the wild"/>
<category term="OCaml"/>
<category term="MirageOS"/>
<category term="cryptography"/>
<category term="security"/>
<category term="VPN"/>
<category term="performance"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/gptar.html</id>
<title type="text">GPTar</title>
<updated>2024-02-21T00:00:00Z</updated>
<summary type="text">Hybrid GUID partition table and tar archive</summary>
<link href="https://blog.robur.coop//articles/gptar.html" rel="alternate" title="GPTar"/>
<category term="OCaml"/>
<category term="gpt"/>
<category term="tar"/>
<category term="mbr"/>
<category term="persistent storage"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/speeding-ec-string.html</id>
<title type="text">Speeding elliptic curve cryptography</title>
<updated>2024-02-13T00:00:00Z</updated>
<summary type="text">
How we improved the performance of elliptic curves by only modifying the underlying byte array
</summary>
<link href="https://blog.robur.coop//articles/speeding-ec-string.html" rel="alternate" title="Speeding elliptic curve cryptography"/>
<category term="OCaml"/>
<category term="MirageOS"/>
<category term="cryptography"/>
<category term="security"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/lwt_pause.html</id>
<title type="text">Cooperation and Lwt.pause</title>
<updated>2024-02-11T00:00:00Z</updated>
<summary type="text">A disgression about Lwt and Miou</summary>
<link href="https://blog.robur.coop//articles/lwt_pause.html" rel="alternate" title="Cooperation and Lwt.pause"/>
<category term="OCaml"/>
<category term="Scheduler"/>
<category term="Community"/>
<category term="Unikernel"/>
<category term="Git"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/2024-02-03-python-str-repr.html</id>
<title type="text">Python&apos;s `str.__repr__()`</title>
<updated>2024-02-03T00:00:00Z</updated>
<summary type="text">Reimplementing Python string escaping in OCaml</summary>
<link href="https://blog.robur.coop//articles/2024-02-03-python-str-repr.html" rel="alternate" title="Python&apos;s `str.__repr__()`"/>
<category term="OCaml"/>
<category term="Python"/>
<category term="unicode"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/miragevpn-ncp.html</id>
<title type="text">MirageVPN updated (AEAD, NCP)</title>
<updated>2023-11-20T00:00:00Z</updated>
<summary type="text">How we resurrected MirageVPN from its bitrot state</summary>
<link href="https://blog.robur.coop//articles/miragevpn-ncp.html" rel="alternate" title="MirageVPN updated (AEAD, NCP)"/>
<category term="OCaml"/>
<category term="MirageOS"/>
<category term="VPN"/>
<category term="security"/>
</entry>
<entry>
<id>https://blog.robur.coop//articles/miragevpn.html</id>
<title type="text">MirageVPN &amp; tls-crypt-v2</title>
<updated>2023-11-14T00:00:00Z</updated>
<summary type="text">How we implementated tls-crypt-v2 for miragevpn</summary>
<link href="https://blog.robur.coop//articles/miragevpn.html" rel="alternate" title="MirageVPN &amp; tls-crypt-v2"/>
<category term="OCaml"/>
<category term="MirageOS"/>
<category term="VPN"/>
<category term="security"/>
</entry>
</feed>

638
bin/blog.ml
View file

@ -1,638 +0,0 @@
open Yocaml
let is_empty_list = function [] -> true | _ -> false
module Date = struct
type month =
| Jan
| Feb
| Mar
| Apr
| May
| Jun
| Jul
| Aug
| Sep
| Oct
| Nov
| Dec
type day_of_week = Mon | Tue | Wed | Thu | Fri | Sat | Sun
type year = int
type day = int
type hour = int
type min = int
type sec = int
type t = {
year : year
; month : month
; day : day
; hour : hour
; min : min
; sec : sec
}
let invalid_int x message =
Data.Validation.fail_with ~given:(string_of_int x) message
let month_from_int x =
if x > 0 && x <= 12 then
Result.ok
[| Jan; Feb; Mar; Apr; May; Jun; Jul; Aug; Sep; Oct; Nov; Dec |].(x - 1)
else invalid_int x "Invalid month value"
let year_from_int x =
if x >= 0 then Result.ok x else invalid_int x "Invalid year value"
let is_leap year =
if year mod 100 = 0 then year mod 400 = 0 else year mod 4 = 0
let days_in_month year month =
match month with
| Jan | Mar | May | Jul | Aug | Oct | Dec -> 31
| Feb -> if is_leap year then 29 else 28
| _ -> 30
let day_from_int year month x =
let dim = days_in_month year month in
if x >= 1 && x <= dim then Result.ok x
else invalid_int x "Invalid day value"
let hour_from_int x =
if x >= 0 && x < 24 then Result.ok x else invalid_int x "Invalid hour value"
let min_from_int x =
if x >= 0 && x < 60 then Result.ok x else invalid_int x "Invalid min value"
let sec_from_int x =
if x >= 0 && x < 60 then Result.ok x else invalid_int x "Invalid sec value"
let ( let* ) = Result.bind
let make ?(time = (0, 0, 0)) ~year ~month ~day () =
let hour, min, sec = time in
let* year = year_from_int year in
let* month = month_from_int month in
let* day = day_from_int year month day in
let* hour = hour_from_int hour in
let* min = min_from_int min in
let* sec = sec_from_int sec in
Result.ok { year; month; day; hour; min; sec }
let validate_from_datetime_str str =
let str = String.trim str in
match
Scanf.sscanf_opt str "%04d%c%02d%c%02d%c%02d%c%02d%c%02d"
(fun year _ month _ day _ hour _ min _ sec ->
((hour, min, sec), year, month, day))
with
| None -> Data.Validation.fail_with ~given:str "Invalid date format"
| Some (time, year, month, day) -> make ~time ~year ~month ~day ()
let validate_from_date_str str =
let str = String.trim str in
match
Scanf.sscanf_opt str "%04d%c%02d%c%02d" (fun year _ month _ day ->
(year, month, day))
with
| None -> Data.Validation.fail_with ~given:str "Invalid date format"
| Some (year, month, day) -> make ~year ~month ~day ()
let validate =
let open Data.Validation in
string & (validate_from_datetime_str / validate_from_date_str)
let month_to_int = function
| Jan -> 1
| Feb -> 2
| Mar -> 3
| Apr -> 4
| May -> 5
| Jun -> 6
| Jul -> 7
| Aug -> 8
| Sep -> 9
| Oct -> 10
| Nov -> 11
| Dec -> 12
let dow_to_int = function
| Mon -> 0
| Tue -> 1
| Wed -> 2
| Thu -> 3
| Fri -> 4
| Sat -> 5
| Sun -> 6
let compare_date a b =
let cmp = Int.compare a.year b.year in
if Int.equal cmp 0 then
let cmp = Int.compare (month_to_int a.month) (month_to_int b.month) in
if Int.equal cmp 0 then Int.compare a.day b.day else cmp
else cmp
let compare_time a b =
let cmp = Int.compare a.hour b.hour in
if Int.equal cmp 0 then
let cmp = Int.compare a.min b.min in
if Int.equal cmp 0 then Int.compare a.sec b.sec else cmp
else cmp
let compare a b =
let cmp = compare_date a b in
if Int.equal cmp 0 then compare_time a b else cmp
let pp_date ppf { year; month; day; _ } =
Format.fprintf ppf "%04d-%02d-%02d" year (month_to_int month) day
let month_value = function
| Jan -> 0
| Feb -> 3
| Mar -> 3
| Apr -> 6
| May -> 1
| Jun -> 4
| Jul -> 6
| Aug -> 2
| Sep -> 5
| Oct -> 0
| Nov -> 3
| Dec -> 5
let day_of_week { year; month; day; _ } =
let yy = year mod 100 in
let cc = (year - yy) / 100 in
let c_code = [| 6; 4; 2; 0 |].(cc mod 4) in
let y_code = (yy + (yy / 4)) mod 7 in
let m_code =
let v = month_value month in
if is_leap year && (month = Jan || month = Feb) then v - 1 else v
in
let index = (c_code + y_code + m_code + day) mod 7 in
[| Sun; Mon; Tue; Wed; Thu; Fri; Sat |].(index)
let normalize ({ year; month; day; hour; min; sec } as dt) =
let day_of_week = day_of_week dt in
let open Data in
record
[
("year", int year); ("month", int (month_to_int month)); ("day", int day)
; ("hour", int hour); ("min", int min); ("sec", int sec)
; ("day_of_week", int (dow_to_int day_of_week))
; ("human", string (Format.asprintf "%a" pp_date dt))
]
let to_archetype_date_time { year; month; day; hour; min; sec } =
let time = (hour, min, sec) in
let month = month_to_int month in
Result.get_ok (Archetype.Datetime.make ~time ~year ~month ~day ())
end
module Page = struct
let entity_name = "Page"
class type t = object ('self)
method title : string option
method charset : string option
method description : string option
method tags : string list
method with_host : string -> 'self
method get_host : string option
end
class page ?title ?description ?charset ?(tags = []) () =
object (_ : #t)
method title = title
method charset = charset
method description = description
method tags = tags
val host = None
method with_host v = {< host = Some v >}
method get_host = host
end
let neutral = Result.ok @@ new page ()
let validate fields =
let open Data.Validation in
let+ title = optional fields "title" string
and+ description = optional fields "description" string
and+ charset = optional fields "charset" string
and+ tags = optional_or fields ~default:[] "tags" (list_of string) in
new page ?title ?description ?charset ~tags ()
let validate =
let open Data.Validation in
record validate
end
module Author = struct
class type t = object
method name : string
method link : string
method email : string
method avatar : string option
end
let gravatar email =
let tk = String.(lowercase_ascii (trim email)) in
let hs = Digest.(to_hex (string tk)) in
"https://www.gravatar.com/avatar/" ^ hs
class author ~name ~link ~email ?(avatar = gravatar email) () =
object (_ : #t)
method name = name
method link = link
method email = email
method avatar = Some avatar
end
let validate fields =
let open Data.Validation in
let+ name = required fields "name" string
and+ link = required fields "link" string
and+ email = required fields "email" string
and+ avatar = optional fields "avatar" string in
match avatar with
| None -> new author ~name ~link ~email ()
| Some avatar -> new author ~name ~link ~email ~avatar ()
let validate =
let open Data.Validation in
record validate
let normalize obj =
let open Data in
record
[
("name", string obj#name); ("link", string obj#link)
; ("email", string obj#email); ("avatar", option string obj#avatar)
]
end
let robur_coop =
new Author.author
~name:"The Robur Team" ~link:"https://robur.coop/"
~email:"team@robur.coop" ()
module Article = struct
let entity_name = "Article"
class type t = object ('self)
method title : string
method description : string
method charset : string option
method tags : string list
method date : Date.t
method author : Author.t
method co_authors : Author.t list
method with_host : string -> 'self
method get_host : string option
end
class article ~title ~description ?charset ?(tags = []) ~date ~author
?(co_authors = []) () =
object (_ : #t)
method title = title
method description = description
method charset = charset
method tags = tags
method date = date
method author = author
method co_authors = co_authors
val host = None
method with_host v = {< host = Some v >}
method get_host = host
end
let title p = p#title
let description p = p#description
let tags p = p#tags
let date p = p#date
let neutral =
Data.Validation.fail_with ~given:"null" "Cannot be null"
|> Result.map_error (fun error ->
Required.Validation_error { entity = entity_name; error })
let validate fields =
let open Data.Validation in
let+ title = required fields "title" string
and+ description = required fields "description" string
and+ charset = optional fields "charset" string
and+ tags = optional_or fields ~default:[] "tags" (list_of string)
and+ date = required fields "date" Date.validate
and+ author =
optional_or fields ~default:robur_coop "author" Author.validate
and+ co_authors =
optional_or fields ~default:[] "co-authors" (list_of Author.validate)
in
new article ~title ~description ?charset ~tags ~date ~author ~co_authors ()
let validate =
let open Data.Validation in
record validate
let normalize obj =
Data.
[
("title", string obj#title); ("description", string obj#description)
; ("date", Date.normalize obj#date); ("charset", option string obj#charset)
; ("tags", list_of string obj#tags)
; ("author", Author.normalize obj#author)
; ("co-authors", list_of Author.normalize obj#co_authors)
; ("host", option string obj#get_host)
]
end
module Articles = struct
class type t = object ('self)
method title : string option
method description : string option
method articles : (Path.t * Article.t) list
method with_host : string -> 'self
method get_host : string option
end
class articles ?title ?description articles =
object (_ : #t)
method title = title
method description = description
method articles = articles
val host = None
method with_host v = {< host = Some v >}
method get_host = host
end
let sort_by_date ?(increasing = false) articles =
List.sort
(fun (_, articleA) (_, articleB) ->
let r = Date.compare articleA#date articleB#date in
if increasing then r else ~-r)
articles
let fetch (module P : Required.DATA_PROVIDER) ?increasing
?(filter = fun x -> x) ?(on = `Source) ~where ~compute_link path =
Task.from_effect begin fun () ->
let open Eff in
let* files = read_directory ~on ~only:`Files ~where path in
let+ articles =
List.traverse
(fun file ->
let url = compute_link file in
let+ metadata, _content =
Eff.read_file_with_metadata (module P) (module Article) ~on file
in
(url, metadata))
files
in
articles |> sort_by_date ?increasing |> filter end
let compute_index (module P : Required.DATA_PROVIDER) ?increasing
?(filter = fun x -> x) ?(on = `Source) ~where ~compute_link path =
let open Task in
(fun x -> (x, ()))
|>> second
(fetch (module P) ?increasing ~filter ~on ~where ~compute_link path)
>>> lift (fun (v, articles) ->
new articles ?title:v#title ?description:v#description articles)
let normalize (ident, article) =
let open Data in
record (("url", string @@ Path.to_string ident) :: Article.normalize article)
let normalize obj =
let open Data in
[
("articles", list_of normalize obj#articles)
; ("has_articles", bool @@ is_empty_list obj#articles)
; ("title", option string obj#title)
; ("description", option string obj#description)
; ("host", option string obj#get_host)
]
end
module Make_with_target (S : sig
val source : Path.t
val target : Path.t
end) =
struct
let source_root = S.source
module Source = struct
let css = Path.(source_root / "css")
let js = Path.(source_root / "js")
let images = Path.(source_root / "images")
let articles = Path.(source_root / "articles")
let index = Path.(source_root / "pages" / "index.md")
let templates = Path.(source_root / "templates")
let template file = Path.(templates / file)
let binary = Path.rel [ Sys.argv.(0) ]
let cache = Path.(source_root / "_cache")
end
module Target = struct
let target_root = S.target
let pages = target_root
let articles = Path.(target_root / "articles")
let rss1 = Path.(target_root / "rss1.xml")
let rss2 = Path.(target_root / "feed.xml")
let atom = Path.(target_root / "atom.xml")
let as_html into file =
file |> Path.move ~into |> Path.change_extension "html"
end
let target = Target.target_root
let process_css_files =
Action.copy_directory ~into:Target.target_root Source.css
let process_js_files =
Action.copy_directory ~into:Target.target_root Source.js
let process_images_files =
Action.copy_directory ~into:Target.target_root Source.images
let process_article ~host file =
let file_target = Target.(as_html articles file) in
let open Task in
Action.write_static_file file_target
begin
Pipeline.track_file Source.binary
>>> Yocaml_yaml.Pipeline.read_file_with_metadata (module Article) file
>>* (fun (obj, str) -> Eff.return (obj#with_host host, str))
>>> Yocaml_cmarkit.content_to_html ()
>>> Yocaml_jingoo.Pipeline.as_template
(module Article)
(Source.template "article.html")
>>> Yocaml_jingoo.Pipeline.as_template
(module Article)
(Source.template "layout.html")
>>> drop_first ()
end
let process_articles ~host =
Action.batch ~only:`Files ~where:(Path.has_extension "md") Source.articles
(process_article ~host)
let process_index ~host =
let file = Source.index in
let file_target = Target.(as_html pages file) in
let open Task in
let compute_index =
Articles.compute_index
(module Yocaml_yaml)
~where:(Path.has_extension "md")
~compute_link:(Target.as_html @@ Path.abs [ "articles" ])
Source.articles
in
Action.write_static_file file_target
begin
Pipeline.track_files [ Source.binary; Source.articles ]
>>> Yocaml_yaml.Pipeline.read_file_with_metadata (module Page) file
>>> Yocaml_cmarkit.content_to_html ()
>>> first compute_index
>>* (fun (obj, str) -> Eff.return (obj#with_host host, str))
>>> Yocaml_jingoo.Pipeline.as_template ~strict:true
(module Articles)
(Source.template "index.html")
>>> Yocaml_jingoo.Pipeline.as_template ~strict:true
(module Articles)
(Source.template "layout.html")
>>> drop_first ()
end
let feed_title = "The Robur's blog"
let site_url = "https://blog.robur.coop/"
let feed_description = "The Robur cooperative blog"
let fetch_articles =
let open Task in
Pipeline.track_files [ Source.binary; Source.articles ]
>>> Articles.fetch
(module Yocaml_yaml)
~where:(Path.has_extension "md")
~compute_link:(Target.as_html @@ Path.abs [ "articles" ])
Source.articles
let rss1 =
let from_articles ~title ~site_url ~description ~feed_url () =
let open Yocaml_syndication in
Rss1.from ~title ~url:feed_url ~link:site_url ~description
@@ fun (path, article) ->
let title = Article.title article in
let link = site_url ^ Yocaml.Path.to_string path in
let description = Article.description article in
Rss1.item ~title ~link ~description
in
let open Task in
Action.write_static_file Target.rss1
begin
fetch_articles
>>> from_articles ~title:feed_title ~site_url
~description:feed_description
~feed_url:"https://blog.robur.coop/rss1.xml" ()
end
let rss2 =
let open Task in
let from_articles ~title ~site_url ~description ~feed_url () =
let open Yocaml_syndication in
lift
begin
fun articles ->
let last_build_date =
List.fold_left
begin
fun acc (_, elt) ->
let v = Date.to_archetype_date_time (Article.date elt) in
match acc with
| None -> Some v
| Some a ->
if Archetype.Datetime.compare a v > 0 then Some a
else Some v
end
None articles
|> Option.map Datetime.make
in
let feed =
Rss2.feed ?last_build_date ~title ~link:site_url ~url:feed_url
~description
begin
fun (path, article) ->
let title = Article.title article in
let link = site_url ^ Path.to_string path in
let guid = Rss2.guid_from_link in
let description = Article.description article in
let pub_date =
Datetime.make
(Date.to_archetype_date_time (Article.date article))
in
Rss2.item ~title ~link ~guid ~description ~pub_date ()
end
articles
in
Xml.to_string feed
end
in
Action.write_static_file Target.rss2
begin
fetch_articles
>>> from_articles ~title:feed_title ~site_url
~description:feed_description
~feed_url:"https://blog.robur.coop/feed.xml" ()
end
let atom =
let open Task in
let open Yocaml_syndication in
let authors = Yocaml.Nel.singleton @@ Person.make "The Robur Team" in
let from_articles ?(updated = Atom.updated_from_entries ()) ?(links = [])
?id ~site_url ~authors ~title ~feed_url () =
let id = Option.value ~default:feed_url id in
let feed_url = Atom.self feed_url in
let base_url = Atom.link site_url in
let links = base_url :: feed_url :: links in
Atom.from ~links ~updated ~title ~authors ~id
begin
fun (path, article) ->
let title = Article.title article in
let content_url = site_url ^ Yocaml.Path.to_string path in
let updated =
Datetime.make (Date.to_archetype_date_time (Article.date article))
in
let categories = List.map Category.make (Article.tags article) in
let summary = Atom.text (Article.description article) in
let links = [ Atom.alternate content_url ~title ] in
Atom.entry ~links ~categories ~summary ~updated ~id:content_url
~title:(Atom.text title) ()
end
in
Action.write_static_file Target.atom
begin
fetch_articles
>>> from_articles ~site_url ~authors ~title:(Atom.text feed_title)
~feed_url:"https://blog.robur.coop/atom.xml" ()
end
let process_all ~host =
let open Eff in
Action.restore_cache ~on:`Source Source.cache
>>= process_css_files >>= process_js_files >>= process_images_files
>>= process_articles ~host >>= process_index ~host >>= rss1 >>= rss2 >>= atom
>>= Action.store_cache ~on:`Source Source.cache
end
module Make (S : sig
val source : Path.t
end) =
Make_with_target (struct
include S
let target = Path.(source / "_site")
end)

14
bin/blog.mli
View file

@ -1,14 +0,0 @@
module Make_with_target (_ : sig
val source : Yocaml.Path.t
val target : Yocaml.Path.t
end) : sig
val target : Yocaml.Path.t
val process_all : host:string -> unit Yocaml.Eff.t
end
module Make (_ : sig
val source : Yocaml.Path.t
end) : sig
val target : Yocaml.Path.t
val process_all : host:string -> unit Yocaml.Eff.t
end

23
bin/dune
View file

@ -1,23 +0,0 @@
(executable
(name watch)
(libraries
yocaml
yocaml_syndication
yocaml_yaml
yocaml_jingoo
yocaml_cmarkit
yocaml_unix))
(executable
(name push)
(libraries
fmt.tty
logs.fmt
git-unix
yocaml
yocaml_git
yocaml_syndication
yocaml_yaml
yocaml_jingoo
yocaml_cmarkit
yocaml_unix))

60
bin/push.ml
View file

@ -1,60 +0,0 @@
let reporter ppf =
let report src level ~over k msgf =
let k _ =
over ();
k ()
in
let with_metadata header _tags k ppf fmt =
Format.kfprintf k ppf
("%a[%a]: " ^^ fmt ^^ "\n%!")
Logs_fmt.pp_header (level, header)
Fmt.(styled `Magenta string)
(Logs.Src.name src)
in
msgf @@ fun ?header ?tags fmt -> with_metadata header tags k ppf fmt
in
{ Logs.report }
let () = Fmt_tty.setup_std_outputs ~style_renderer:`Ansi_tty ~utf_8:true ()
let () = Logs.set_reporter (reporter Fmt.stdout)
(* let () = Logs.set_level ~all:true (Some Logs.Debug) *)
let author = ref "The Robur Team"
let email = ref "team@robur.coop"
let message = ref "Pushed by YOCaml 2"
let remote = ref "git@git.robur.coop:robur/blog.robur.coop.git#gh-pages"
let host = ref "https://blog.robur.coop"
module Source = Yocaml_git.From_identity (Yocaml_unix.Runtime)
let usage =
Fmt.str
"%s [--message <message>] [--author <author>] [--email <email>] -r \
<repository>"
Sys.argv.(0)
let specification =
[
("--message", Arg.Set_string message, "The commit message")
; ("--email", Arg.Set_string email, "The email used to craft the commit")
; ("-r", Arg.Set_string remote, "The Git repository")
; ("--author", Arg.Set_string author, "The Git commit author")
; ("--host", Arg.Set_string host, "The host where the blog is available")
]
let () =
Arg.parse specification ignore usage;
let author = !author
and email = !email
and message = !message
and remote = !remote in
let module Blog = Blog.Make_with_target (struct
let source = Yocaml.Path.rel []
let target = Yocaml.Path.rel []
end) in
Yocaml_git.run
(module Source)
(module Pclock)
~context:`SSH ~author ~email ~message ~remote
(fun () -> Blog.process_all ~host:!host)
|> Lwt_main.run
|> Result.iter_error (fun (`Msg err) -> invalid_arg err)

15
bin/watch.ml
View file

@ -1,15 +0,0 @@
let port = ref 8000
let usage = Fmt.str "%s [--port <port>]" Sys.argv.(0)
let specification =
[ ("--port", Arg.Set_int port, "The port where we serve the website") ]
module Dest = Blog.Make (struct
let source = Yocaml.Path.rel []
end)
let () =
Arg.parse specification ignore usage;
let host = Fmt.str "http://localhost:%d" !port in
Yocaml_unix.serve ~level:`Info ~target:Dest.target ~port:!port
@@ fun () -> Dest.process_all ~host

18
blogger.opam
View file

@ -5,7 +5,7 @@ maintainer: "romain.calascibetta@gmail.com"
authors: [ "The XHTMLBoy <xhtmlboi@gmail.com>" ] authors: [ "The XHTMLBoy <xhtmlboi@gmail.com>" ]
build: [ build: [
[ "dune" "subst" ] {dev} [ "dune" "subst" ]
[ "dune" "build" "-p" name "-j" jobs ] [ "dune" "build" "-p" name "-j" jobs ]
[ "dune" "runtest" "-p" name ] {with-test} [ "dune" "runtest" "-p" name ] {with-test}
[ "dune" "build" "@doc" "-p" name ] {with-doc} [ "dune" "build" "@doc" "-p" name ] {with-doc}
@ -18,7 +18,7 @@ dev-repo: "git://github.com/dinosaure/blogger.git"
bug-reports: "https://github.com/dinosaure/blogger/issues" bug-reports: "https://github.com/dinosaure/blogger/issues"
depends: [ depends: [
"ocaml" { >= "5.1.0" } "ocaml" { >= "4.11.1" }
"dune" { >= "2.8" } "dune" { >= "2.8" }
"preface" { >= "0.1.0" } "preface" { >= "0.1.0" }
"logs" {>= "0.7.0" } "logs" {>= "0.7.0" }
@ -27,19 +27,7 @@ depends: [
"yocaml" "yocaml"
"yocaml_unix" "yocaml_unix"
"yocaml_yaml" "yocaml_yaml"
"yocaml_cmarkit" "yocaml_cmark"
"yocaml_git" "yocaml_git"
"yocaml_jingoo" "yocaml_jingoo"
"yocaml_syndication"
]
pin-depends: [
["yocaml.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
["yocaml_runtime.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
["yocaml_unix.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
["yocaml_yaml.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
["yocaml_cmarkit.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
["yocaml_git.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
["yocaml_jingoo.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
["yocaml_syndication.dev" "git+https://gitlab.com/funkywork/yocaml.git#c2809182a59571a863d6ad14a77f720f6fa577dc" ]
] ]

95
feed.xml Normal file
View file

@ -0,0 +1,95 @@
<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
<title>The Robur&apos;s blog</title>
<link>https://blog.robur.coop/</link>
<description><![CDATA[The Robur cooperative blog]]></description>
<atom:link href="https://blog.robur.coop/feed.xml" rel="self" type="application/rss+xml"/>
<lastBuildDate>Wed, 21 Aug 2024 00:00:00 GMT</lastBuildDate>
<docs>https://www.rssboard.org/rss-specification</docs>
<generator>YOCaml</generator>
<item>
<title>MirageVPN and OpenVPN</title>
<link>https://blog.robur.coop//articles/2024-08-21-OpenVPN-and-MirageVPN.html</link>
<description>
<![CDATA[Discoveries made implementing MirageVPN, a OpenVPN-compatible VPN library]]>
</description>
<guid isPermaLink="true">https://blog.robur.coop//articles/2024-08-21-OpenVPN-and-MirageVPN.html</guid>
<pubDate>Wed, 21 Aug 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>The new Tar release, a retrospective</title>
<link>https://blog.robur.coop//articles/tar-release.html</link>
<description><![CDATA[A little retrospective to the new Tar release and changes]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/tar-release.html</guid>
<pubDate>Thu, 15 Aug 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>qubes-miragevpn, a MirageVPN client for QubesOS</title>
<link>https://blog.robur.coop//articles/qubes-miragevpn.html</link>
<description><![CDATA[A new OpenVPN client for QubesOS]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/qubes-miragevpn.html</guid>
<pubDate>Mon, 24 Jun 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>MirageVPN server</title>
<link>https://blog.robur.coop//articles/miragevpn-server.html</link>
<description><![CDATA[Announcement of our MirageVPN server.]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/miragevpn-server.html</guid>
<pubDate>Mon, 17 Jun 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>Speeding up MirageVPN and use it in the wild</title>
<link>https://blog.robur.coop//articles/miragevpn-performance.html</link>
<description>
<![CDATA[Performance engineering of MirageVPN, speeding it up by a factor of 25.]]>
</description>
<guid isPermaLink="true">https://blog.robur.coop//articles/miragevpn-performance.html</guid>
<pubDate>Tue, 16 Apr 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>GPTar</title>
<link>https://blog.robur.coop//articles/gptar.html</link>
<description><![CDATA[Hybrid GUID partition table and tar archive]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/gptar.html</guid>
<pubDate>Wed, 21 Feb 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>Speeding elliptic curve cryptography</title>
<link>https://blog.robur.coop//articles/speeding-ec-string.html</link>
<description>
<![CDATA[How we improved the performance of elliptic curves by only modifying the underlying byte array]]>
</description>
<guid isPermaLink="true">https://blog.robur.coop//articles/speeding-ec-string.html</guid>
<pubDate>Tue, 13 Feb 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>Cooperation and Lwt.pause</title>
<link>https://blog.robur.coop//articles/lwt_pause.html</link>
<description><![CDATA[A disgression about Lwt and Miou]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/lwt_pause.html</guid>
<pubDate>Sun, 11 Feb 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>Python&apos;s `str.__repr__()`</title>
<link>https://blog.robur.coop//articles/2024-02-03-python-str-repr.html</link>
<description><![CDATA[Reimplementing Python string escaping in OCaml]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/2024-02-03-python-str-repr.html</guid>
<pubDate>Sat, 03 Feb 2024 00:00:00 GMT</pubDate>
</item>
<item>
<title>MirageVPN updated (AEAD, NCP)</title>
<link>https://blog.robur.coop//articles/miragevpn-ncp.html</link>
<description><![CDATA[How we resurrected MirageVPN from its bitrot state]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/miragevpn-ncp.html</guid>
<pubDate>Mon, 20 Nov 2023 00:00:00 GMT</pubDate>
</item>
<item>
<title>MirageVPN &amp; tls-crypt-v2</title>
<link>https://blog.robur.coop//articles/miragevpn.html</link>
<description><![CDATA[How we implementated tls-crypt-v2 for miragevpn]]></description>
<guid isPermaLink="true">https://blog.robur.coop//articles/miragevpn.html</guid>
<pubDate>Tue, 14 Nov 2023 00:00:00 GMT</pubDate>
</item>
</channel>
</rss>

179
index.html Normal file
View file

@ -0,0 +1,179 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
Robur's blogIndex
</title>
<meta name="description" content="The famous root of the website">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/hl.css">
<link type="text/css" rel="stylesheet" href="https://blog.robur.coop/css/style.css">
<script src="https://blog.robur.coop/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="https://blog.robur.coop/feed.xml" title="blog.robur.coop">
</head>
<body>
<header>
<h1>blog.robur.coop</h1>
<blockquote>
The <strong>Robur</strong> cooperative blog.
</blockquote>
</header>
<main><a class="small-button rss" href="https://blog.robur.coop/feed.xml">RSS</a><p>The Robur blog.</p>
<h3>Essays and ramblings</h3>
<ol reversed class="list-articles"><li>
<div class="side">
<a href="https://reyn.ir/">
<img src="https://www.gravatar.com/avatar/54a15736b37879bc9708c1618a7cc130">
</a></div>
<div class="content">
<span class="date">2024-08-21</span>
<a href="https://blog.robur.coop/articles/2024-08-21-OpenVPN-and-MirageVPN.html">MirageVPN and OpenVPN</a><br />
<p>Discoveries made implementing MirageVPN, a OpenVPN-compatible VPN library</p>
<div class="bottom">
<ul class="tags-list"><li>MirageVPN</li><li>OpenVPN</li><li>security</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://blog.osau.re">
<img src="https://www.gravatar.com/avatar/e243d18f97471424ca390e85820797ac">
</a></div>
<div class="content">
<span class="date">2024-08-15</span>
<a href="https://blog.robur.coop/articles/tar-release.html">The new Tar release, a retrospective</a><br />
<p>A little retrospective to the new Tar release and changes</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>Cstruct</li><li>functors</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://blog.osau.re/">
<img src="https://www.gravatar.com/avatar/e243d18f97471424ca390e85820797ac">
</a></div>
<div class="content">
<span class="date">2024-06-24</span>
<a href="https://blog.robur.coop/articles/qubes-miragevpn.html">qubes-miragevpn, a MirageVPN client for QubesOS</a><br />
<p>A new OpenVPN client for QubesOS</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>vpn</li><li>unikernel</li><li>QubesOS</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://hannes.robur.coop">
<img src="https://www.gravatar.com/avatar/25558b4457cf73159f5427fdf2b4a717">
</a></div>
<div class="content">
<span class="date">2024-06-17</span>
<a href="https://blog.robur.coop/articles/miragevpn-server.html">MirageVPN server</a><br />
<p>Announcement of our MirageVPN server.</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>cryptography</li><li>security</li><li>VPN</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://hannes.robur.coop">
<img src="https://www.gravatar.com/avatar/25558b4457cf73159f5427fdf2b4a717">
</a></div>
<div class="content">
<span class="date">2024-04-16</span>
<a href="https://blog.robur.coop/articles/miragevpn-performance.html">Speeding up MirageVPN and use it in the wild</a><br />
<p>Performance engineering of MirageVPN, speeding it up by a factor of 25.</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>cryptography</li><li>security</li><li>VPN</li><li>performance</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://reyn.ir/">
<img src="https://www.gravatar.com/avatar/54a15736b37879bc9708c1618a7cc130">
</a></div>
<div class="content">
<span class="date">2024-02-21</span>
<a href="https://blog.robur.coop/articles/gptar.html">GPTar</a><br />
<p>Hybrid GUID partition table and tar archive</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>gpt</li><li>tar</li><li>mbr</li><li>persistent storage</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://hannes.robur.coop">
<img src="https://www.gravatar.com/avatar/25558b4457cf73159f5427fdf2b4a717">
</a></div>
<div class="content">
<span class="date">2024-02-13</span>
<a href="https://blog.robur.coop/articles/speeding-ec-string.html">Speeding elliptic curve cryptography</a><br />
<p>How we improved the performance of elliptic curves by only modifying the underlying byte array</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>cryptography</li><li>security</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://blog.osau.re/">
<img src="https://www.gravatar.com/avatar/e243d18f97471424ca390e85820797ac">
</a></div>
<div class="content">
<span class="date">2024-02-11</span>
<a href="https://blog.robur.coop/articles/lwt_pause.html">Cooperation and Lwt.pause</a><br />
<p>A disgression about Lwt and Miou</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>Scheduler</li><li>Community</li><li>Unikernel</li><li>Git</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://reyn.ir/">
<img src="https://www.gravatar.com/avatar/54a15736b37879bc9708c1618a7cc130">
</a></div>
<div class="content">
<span class="date">2024-02-03</span>
<a href="https://blog.robur.coop/articles/2024-02-03-python-str-repr.html">Python&apos;s `str.__repr__()`</a><br />
<p>Reimplementing Python string escaping in OCaml</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>Python</li><li>unicode</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://hannes.robur.coop">
<img src="https://www.gravatar.com/avatar/25558b4457cf73159f5427fdf2b4a717">
</a></div>
<div class="content">
<span class="date">2023-11-20</span>
<a href="https://blog.robur.coop/articles/miragevpn-ncp.html">MirageVPN updated (AEAD, NCP)</a><br />
<p>How we resurrected MirageVPN from its bitrot state</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>VPN</li><li>security</li></ul>
</div>
</div>
</li><li>
<div class="side">
<a href="https://reyn.ir/">
<img src="https://www.gravatar.com/avatar/54a15736b37879bc9708c1618a7cc130">
</a></div>
<div class="content">
<span class="date">2023-11-14</span>
<a href="https://blog.robur.coop/articles/miragevpn.html">MirageVPN &amp; tls-crypt-v2</a><br />
<p>How we implementated tls-crypt-v2 for miragevpn</p>
<div class="bottom">
<ul class="tags-list"><li>OCaml</li><li>MirageOS</li><li>VPN</li><li>security</li></ul>
</div>
</div>
</li></ol>
</main>
<footer>
<a href="https://github.com/xhtmlboi/yocaml">Powered by <strong>YOCaml</strong></a>
<br />
</footer>
<script>hljs.highlightAll();</script>
</body>
</html>

84
rss1.xml Normal file
View file

@ -0,0 +1,84 @@
<?xml version="1.0" encoding="utf-8"?>
<rdf:RDF xmlns="http://purl.org/rss/1.0/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#">
<channel rdf:about="https://blog.robur.coop/rss1.xml">
<title>The Robur&apos;s blog</title>
<link>https://blog.robur.coop/</link>
<description><![CDATA[The Robur cooperative blog]]></description>
<items>
<rdf:Seq>
<rdf:li resource="https://blog.robur.coop//articles/2024-08-21-OpenVPN-and-MirageVPN.html"/>
<rdf:li resource="https://blog.robur.coop//articles/tar-release.html"/>
<rdf:li resource="https://blog.robur.coop//articles/qubes-miragevpn.html"/>
<rdf:li resource="https://blog.robur.coop//articles/miragevpn-server.html"/>
<rdf:li resource="https://blog.robur.coop//articles/miragevpn-performance.html"/>
<rdf:li resource="https://blog.robur.coop//articles/gptar.html"/>
<rdf:li resource="https://blog.robur.coop//articles/speeding-ec-string.html"/>
<rdf:li resource="https://blog.robur.coop//articles/lwt_pause.html"/>
<rdf:li resource="https://blog.robur.coop//articles/2024-02-03-python-str-repr.html"/>
<rdf:li resource="https://blog.robur.coop//articles/miragevpn-ncp.html"/>
<rdf:li resource="https://blog.robur.coop//articles/miragevpn.html"/>
</rdf:Seq>
</items>
</channel>
<item rdf:about="https://blog.robur.coop//articles/2024-08-21-OpenVPN-and-MirageVPN.html">
<title>MirageVPN and OpenVPN</title>
<link>https://blog.robur.coop//articles/2024-08-21-OpenVPN-and-MirageVPN.html</link>
<description>
<![CDATA[Discoveries made implementing MirageVPN, a OpenVPN-compatible VPN library]]>
</description>
</item>
<item rdf:about="https://blog.robur.coop//articles/tar-release.html">
<title>The new Tar release, a retrospective</title>
<link>https://blog.robur.coop//articles/tar-release.html</link>
<description><![CDATA[A little retrospective to the new Tar release and changes]]></description>
</item>
<item rdf:about="https://blog.robur.coop//articles/qubes-miragevpn.html">
<title>qubes-miragevpn, a MirageVPN client for QubesOS</title>
<link>https://blog.robur.coop//articles/qubes-miragevpn.html</link>
<description><![CDATA[A new OpenVPN client for QubesOS]]></description>
</item>
<item rdf:about="https://blog.robur.coop//articles/miragevpn-server.html">
<title>MirageVPN server</title>
<link>https://blog.robur.coop//articles/miragevpn-server.html</link>
<description><![CDATA[Announcement of our MirageVPN server.]]></description>
</item>
<item rdf:about="https://blog.robur.coop//articles/miragevpn-performance.html">
<title>Speeding up MirageVPN and use it in the wild</title>
<link>https://blog.robur.coop//articles/miragevpn-performance.html</link>
<description>
<![CDATA[Performance engineering of MirageVPN, speeding it up by a factor of 25.]]>
</description>
</item>
<item rdf:about="https://blog.robur.coop//articles/gptar.html">
<title>GPTar</title>
<link>https://blog.robur.coop//articles/gptar.html</link>
<description><![CDATA[Hybrid GUID partition table and tar archive]]></description>
</item>
<item rdf:about="https://blog.robur.coop//articles/speeding-ec-string.html">
<title>Speeding elliptic curve cryptography</title>
<link>https://blog.robur.coop//articles/speeding-ec-string.html</link>
<description>
<![CDATA[How we improved the performance of elliptic curves by only modifying the underlying byte array]]>
</description>
</item>
<item rdf:about="https://blog.robur.coop//articles/lwt_pause.html">
<title>Cooperation and Lwt.pause</title>
<link>https://blog.robur.coop//articles/lwt_pause.html</link>
<description><![CDATA[A disgression about Lwt and Miou]]></description>
</item>
<item rdf:about="https://blog.robur.coop//articles/2024-02-03-python-str-repr.html">
<title>Python&apos;s `str.__repr__()`</title>
<link>https://blog.robur.coop//articles/2024-02-03-python-str-repr.html</link>
<description><![CDATA[Reimplementing Python string escaping in OCaml]]></description>
</item>
<item rdf:about="https://blog.robur.coop//articles/miragevpn-ncp.html">
<title>MirageVPN updated (AEAD, NCP)</title>
<link>https://blog.robur.coop//articles/miragevpn-ncp.html</link>
<description><![CDATA[How we resurrected MirageVPN from its bitrot state]]></description>
</item>
<item rdf:about="https://blog.robur.coop//articles/miragevpn.html">
<title>MirageVPN &amp; tls-crypt-v2</title>
<link>https://blog.robur.coop//articles/miragevpn.html</link>
<description><![CDATA[How we implementated tls-crypt-v2 for miragevpn]]></description>
</item>
</rdf:RDF>

332
src/blogger.ml Normal file
View file

@ -0,0 +1,332 @@
let caller = Sys.argv.(0)
let version = "%%VERSION%%"
let default_port = 8888
let default_target = Fpath.v "_site"
let program ~target =
let open Yocaml in
let* () = Task.move_javascript target in
let* () = Task.move_css target in
let* () = Task.move_images target in
let* () = Task.process_articles target in
let* () = Task.generate_feed target in
let* () = Task.generate_tags target in
Task.generate_index target
let local_build _quiet target =
Yocaml_unix.execute (program ~target:(Fpath.to_string target))
module SSH = struct
open Lwt.Infix
type error = Unix.error * string * string
type write_error = [ `Closed | `Error of Unix.error * string * string ]
let pp_error ppf (err, f, v) =
Fmt.pf ppf "%s(%s): %s" f v (Unix.error_message err)
let pp_write_error ppf = function
| `Closed -> Fmt.pf ppf "Connection closed by peer"
| `Error (err, f, v) -> Fmt.pf ppf "%s(%s): %s" f v (Unix.error_message err)
type flow = { ic : in_channel; oc : out_channel }
type endpoint = {
user : string;
path : string;
host : Unix.inet_addr;
port : int;
capabilities : [ `Rd | `Wr ];
}
let pp_inet_addr ppf inet_addr =
Fmt.string ppf (Unix.string_of_inet_addr inet_addr)
let connect { user; path; host; port; capabilities } =
let edn = Fmt.str "%s@%a" user pp_inet_addr host in
let cmd =
match capabilities with
| `Wr -> Fmt.str {sh|git-receive-pack '%s'|sh} path
| `Rd -> Fmt.str {sh|git-upload-pack '%s'|sh} path
in
let cmd = Fmt.str "ssh -p %d %s %a" port edn Fmt.(quote string) cmd in
try
let ic, oc = Unix.open_process cmd in
Lwt.return_ok { ic; oc }
with Unix.Unix_error (err, f, v) -> Lwt.return_error (`Error (err, f, v))
let read t =
let tmp = Bytes.create 0x1000 in
try
let len = input t.ic tmp 0 0x1000 in
if len = 0 then Lwt.return_ok `Eof
else Lwt.return_ok (`Data (Cstruct.of_bytes tmp ~off:0 ~len))
with Unix.Unix_error (err, f, v) -> Lwt.return_error (err, f, v)
let write t cs =
let str = Cstruct.to_string cs in
try
output_string t.oc str;
flush t.oc;
Lwt.return_ok ()
with Unix.Unix_error (err, f, v) -> Lwt.return_error (`Error (err, f, v))
let writev t css =
let rec go t = function
| [] -> Lwt.return_ok ()
| x :: r -> (
write t x >>= function
| Ok () -> go t r
| Error _ as err -> Lwt.return err)
in
go t css
let close t =
close_in t.ic;
close_out t.oc;
Lwt.return_unit
let shutdown t mode =
match mode with
| `read -> close_in t.ic ; Lwt.return_unit
| `write -> close_out t.oc ; Lwt.return_unit
| `read_write -> close t
end
let ssh_edn, ssh_protocol = Mimic.register ~name:"ssh" (module SSH)
let unix_ctx_with_ssh () =
let open Lwt.Infix in
Git_unix.ctx (Happy_eyeballs_lwt.create ()) >|= fun ctx ->
let open Mimic in
let k0 scheme user path host port capabilities =
match (scheme, Unix.gethostbyname host) with
| `SSH, { Unix.h_addr_list; _ } when Array.length h_addr_list > 0 ->
Lwt.return_some
{ SSH.user; path; host = h_addr_list.(0); port; capabilities }
| _ -> Lwt.return_none
in
ctx
|> Mimic.fold Smart_git.git_transmission
Fun.[ req Smart_git.git_scheme ]
~k:(function `SSH -> Lwt.return_some `Exec | _ -> Lwt.return_none)
|> Mimic.fold ssh_edn
Fun.
[
req Smart_git.git_scheme;
req Smart_git.git_ssh_user;
req Smart_git.git_path;
req Smart_git.git_hostname;
dft Smart_git.git_port 22;
req Smart_git.git_capabilities;
]
~k:k0
let run_git_config key = function
| Some value -> Some value
| None -> (
let open Bos in
let value = OS.Cmd.run_out Cmd.(v "git" % "config" % "--global" % key) in
match OS.Cmd.out_string value with
| Ok (value, _) -> Some value
| Error _ -> None)
let run_git_rev_parse default =
let open Bos in
let value = OS.Cmd.run_out
Cmd.(v "git" % "describe" % "--always" % "--dirty"
% "--exclude=*" % "--abbrev=0")
in
match OS.Cmd.out_string value with
| Ok (value, (_, `Exited 0)) -> value
| Ok (value, (run_info, _)) ->
Logs.warn (fun m -> m "Failed to get commit id: %a: %s"
Cmd.pp (OS.Cmd.run_info_cmd run_info)
value);
default
| Error `Msg e ->
Logs.warn (fun m -> m "Failed to get commit id: %s" e);
default
let get_name_and_email name email =
let name = run_git_config "user.name" name in
let email = run_git_config "user.email" email in
(name, email)
let name_and_email =
let name_arg =
let doc = "Name of the committer." in
Cmdliner.Arg.(value & opt (some string) None & info [ "name" ] ~doc)
in
let email_arg =
let doc = "Email of the committer." in
Cmdliner.Arg.(value & opt (some string) None & info [ "email" ] ~doc)
in
Cmdliner.Term.(const get_name_and_email $ name_arg $ email_arg)
let build_and_push _quiet remote (author, email) hook =
let fiber () =
let open Lwt.Syntax in
let commit_id = run_git_rev_parse "an unknown state" in
let comment = Printf.sprintf "Built from %s" commit_id in
let* ctx = unix_ctx_with_ssh () in
let* res =
Yocaml_git.execute
(module Yocaml_unix)
(module Pclock)
~ctx ?author ?email ~comment remote (program ~target:"")
in
match res with
| Error (`Msg err) -> Fmt.failwith "build-and-push: %s." err
| Ok () -> (
match hook with
| None -> Lwt.return_unit
| Some hook -> (
let open Lwt.Infix in
Http_lwt_client.request ~config:(`HTTP_1_1 Httpaf.Config.default)
~meth:`GET (Uri.to_string hook)
(fun _ () _ -> Lwt.return_unit)
()
>>= function
| Ok (_response, ()) -> Lwt.return_unit
| Error (`Msg err) -> failwith err))
in
Lwt_main.run (fiber ())
let watch quiet target potential_port =
let port = Option.value ~default:default_port potential_port in
let () = local_build quiet target in
let target = Fpath.to_string target in
let server = Yocaml_unix.serve ~filepath:target ~port (program ~target) in
Lwt_main.run server
let common_options = "COMMON OPTIONS"
let verbosity =
let open Cmdliner in
let env = Cmd.Env.info "BLOGGER_LOGS" in
Logs_cli.level ~docs:common_options ~env ()
let renderer =
let open Cmdliner in
let env = Cmd.Env.info "BLOGGER_FMT" in
Fmt_cli.style_renderer ~docs:common_options ~env ()
let utf_8 =
let open Cmdliner in
let doc = "Allow binaries to emit UTF-8 characters." in
let env = Cmd.Env.info "BLOGGER_UTF_8" in
Arg.(value & opt bool true & info [ "with-utf-8" ] ~doc ~env)
let reporter ppf =
let report src level ~over k msgf =
let k _ =
over ();
k ()
in
let with_metadata header _tags k ppf fmt =
Fmt.kpf k ppf
("%a[%a]: " ^^ fmt ^^ "\n%!")
Logs_fmt.pp_header (level, header)
Fmt.(styled `Magenta string)
(Logs.Src.name src)
in
msgf @@ fun ?header ?tags fmt -> with_metadata header tags k ppf fmt
in
{ Logs.report }
let setup_logs utf_8 style_renderer level =
Fmt_tty.setup_std_outputs ~utf_8 ?style_renderer ();
Logs.set_level level;
Logs.set_reporter (reporter Fmt.stderr);
Option.is_none level
let setup_logs = Cmdliner.Term.(const setup_logs $ utf_8 $ renderer $ verbosity)
let man =
let open Cmdliner in
[ `S Manpage.s_authors; `P "blog.robur.coop" ]
let build_cmd =
let open Cmdliner in
let doc = Format.asprintf "Build the blog into the specified directory" in
let exits = Cmd.Exit.defaults in
let info = Cmd.info "build" ~version ~doc ~exits ~man in
let path_arg =
let doc =
Format.asprintf "Specify where we build the website (default: %a)"
Fpath.pp default_target
in
let arg = Arg.info ~doc [ "destination" ] in
Arg.(value & opt (conv (Fpath.of_string, Fpath.pp)) default_target & arg)
in
Cmd.v info Term.(const local_build $ setup_logs $ path_arg)
let watch_cmd =
let open Cmdliner in
let doc =
Format.asprintf
"Serve from the specified directory as an HTTP server and rebuild \
website on demand"
in
let exits = Cmd.Exit.defaults in
let path_arg =
let doc =
Format.asprintf "Specify where we build the website (default: %a)"
Fpath.pp default_target
in
let arg = Arg.info ~doc [ "destination" ] in
Arg.(value & opt (conv (Fpath.of_string, Fpath.pp)) default_target & arg)
in
let port_arg =
let doc = Format.asprintf "The port (default: %d)" default_port in
let arg = Arg.info ~doc [ "port"; "P"; "p" ] in
Arg.(value & opt (some int) None & arg)
in
let info = Cmd.info "watch" ~version ~doc ~exits ~man in
Cmd.v info Term.(const watch $ setup_logs $ path_arg $ port_arg)
let push_cmd =
let open Cmdliner in
let doc =
Format.asprintf
"Push the blog (from the specified directory) into a Git repository"
in
let exits = Cmd.Exit.defaults in
let remote_arg =
let remote =
let parser str =
match Smart_git.Endpoint.of_string str with
| Ok _ -> Ok str
| Error _ as err -> err
in
Arg.conv (parser, Fmt.string)
in
let doc = "The remote Git repository" in
let arg = Arg.info ~doc [ "r"; "remote" ] in
Arg.(required & opt (some remote) None & arg)
in
let hook_arg =
let doc = "The URL of the hook to update the unikernel" in
let arg = Arg.info ~doc [ "h"; "hook" ] in
let of_string str =
match Uri.of_string str with
| v -> Ok v
| exception _ -> Rresult.R.error_msgf "Invalid URI: %s" str
in
Arg.(value & opt (some (conv (of_string, Uri.pp))) None & arg)
in
let info = Cmd.info "push" ~version ~doc ~exits ~man in
Cmd.v info
Term.(
const build_and_push $ setup_logs $ remote_arg $ name_and_email $ hook_arg)
let cmd =
let open Cmdliner in
let sdocs = Manpage.s_common_options in
let doc = "Build, push or serve my personal website" in
let default_info = Cmd.info caller ~version ~doc ~sdocs ~man in
let default = Term.(ret (const (`Help (`Pager, None)))) in
Cmd.group ~default default_info [ build_cmd; watch_cmd; push_cmd ]
let () = exit @@ Cmdliner.Cmd.eval cmd

64
src/collection.ml Normal file
View file

@ -0,0 +1,64 @@
open Yocaml
let get_article (module V : Metadata.VALIDABLE) article_file =
let arr =
Build.read_file_with_metadata
(module V)
(module Model.Article)
article_file
in
let deps = Build.get_dependencies arr in
let task = Build.get_task arr in
let+ meta, _ = task () in
deps, (meta, Model.article_path article_file)
;;
let get_articles (module V : Metadata.VALIDABLE) path =
let* files = read_child_files path File.is_markdown in
let+ articles = Traverse.traverse (get_article (module V)) files in
let deps, effects = List.split articles in
Deps.Monoid.reduce deps, effects
;;
module Articles = struct
type t = (Model.Article.t * Filepath.t) list
let get_all (module V : Metadata.VALIDABLE) ?(decreasing = true) path =
let+ deps, articles = get_articles (module V) path in
let sorted_article = Model.Articles.sort ~decreasing articles in
Build.make deps (fun x -> return (x, sorted_article))
;;
end
module Tags = struct
module M = Map.Make (String)
let by_quantity ?(decreasing = true) (_, a) (_, b) =
let r = Int.compare $ List.length a $ List.length b in
if decreasing then ~-r else r
;;
let group metas =
List.fold_left
(fun accumulator (article, path) ->
List.fold_left
(fun map tag ->
match M.find_opt tag map with
| Some articles -> M.add tag ((article, path) :: articles) map
| None -> M.add tag [ article, path ] map)
accumulator
(Model.Article.tags article))
M.empty
metas
|> M.map
(List.sort (fun (a, _) (b, _) -> Model.Article.compare_by_date a b))
|> M.to_seq
|> List.of_seq
|> List.sort by_quantity
;;
let compute (module V : Metadata.VALIDABLE) path =
let+ deps, articles = get_articles (module V) path in
deps, group articles
;;
end

18
src/collection.mli Normal file
View file

@ -0,0 +1,18 @@
open Yocaml
module Articles : sig
type t = (Model.Article.t * Filepath.t) list
val get_all
: (module Metadata.VALIDABLE)
-> ?decreasing:bool
-> Filepath.t
-> ('a, 'a * t) Build.t Effect.t
end
module Tags : sig
val compute
: (module Metadata.VALIDABLE)
-> Filepath.t
-> (Deps.t * (string * (Model.Article.t * string) list) list) Effect.t
end

21
src/dune Normal file
View file

@ -0,0 +1,21 @@
(executable
(name blogger)
(libraries
logs
logs.fmt
logs.cli
fmt
fmt.tty
fmt.cli
cmdliner
preface
mirage-clock-unix
http-lwt-client
git-unix
cmarkit
yocaml
yocaml_yaml
yocaml_cmark
yocaml_unix
yocaml_git
yocaml_jingoo))

15
src/feed.ml Normal file
View file

@ -0,0 +1,15 @@
open Yocaml
let domain = "https://blog.robur.coop"
let feed_url = into domain "feed.xml"
let articles_to_items articles =
List.map
(fun (article, url) -> Model.Article.to_rss_item (into domain url) article)
articles
let make ((), articles) =
Yocaml.Rss.Channel.make ~title:"Robur's blog" ~link:domain ~feed_link:feed_url
~description:"The Robur cooperative blog" ~generator:"yocaml"
~webmaster:"team@robur.coop"
(articles_to_items articles)

1
src/feed.mli Normal file
View file

@ -0,0 +1 @@
val make : unit * Collection.Articles.t -> Yocaml.Rss.Channel.t

18
src/file.ml Normal file
View file

@ -0,0 +1,18 @@
open Yocaml
let is_css = with_extension "css"
let is_javascript = with_extension "js"
let is_image =
let open Preface.Predicate in
with_extension "png"
|| with_extension "svg"
|| with_extension "jpg"
|| with_extension "jpeg"
|| with_extension "gif"
;;
let is_markdown =
let open Preface.Predicate in
with_extension "md" || with_extension "markdown"
;;

6
src/file.mli Normal file
View file

@ -0,0 +1,6 @@
open Yocaml
val is_css : Filepath.t -> bool
val is_javascript : Filepath.t -> bool
val is_image : Filepath.t -> bool
val is_markdown : Filepath.t -> bool

256
src/model.ml Normal file
View file

@ -0,0 +1,256 @@
open Yocaml
let article_path file =
let filename = basename $ replace_extension file "html" in
filename |> into "articles"
let tag_path tag = add_extension tag "html" |> into "tags"
module Author = struct
type t = {
name : string;
link : string;
email : string;
avatar : string option;
}
let equal a b =
String.equal a.name b.name && String.equal a.link b.link
&& String.equal a.email b.email
&& Option.equal String.equal a.avatar b.avatar
let make name link email avatar = { name; link; email; avatar }
let from (type a) (module V : Metadata.VALIDABLE with type t = a) obj =
V.object_and
(fun assoc ->
let open Validate.Applicative in
make
<$> V.(required_assoc string) "name" assoc
<*> V.(required_assoc string) "link" assoc
<*> V.(required_assoc string) "email" assoc
<*> V.(optional_assoc string) "avatar" assoc)
obj
let default_user =
make "robur" "https://blog.robur.coop/" "team@robur.coop" None
let gravatar email =
let tk = String.(lowercase_ascii $ trim email) in
let hs = Digest.(to_hex $ string tk) in
"https://www.gravatar.com/avatar/" ^ hs
let inject (type a) (module D : Key_value.DESCRIBABLE with type t = a)
{ name; link; email; avatar } =
let avatar = match avatar with Some uri -> uri | None -> gravatar email in
D.
[
("name", string name);
("link", string link);
("email", string email);
("avatar", string avatar);
]
end
module Co_author = struct
type t = { author : Author.t; contribution : string }
let make author contribution = { author; contribution }
let from (type a) (module V : Metadata.VALIDABLE with type t = a) obj =
V.object_and
(fun assoc ->
let open Validate.Applicative in
make
<$> V.(required_assoc (Author.from (module V))) "author" assoc
<*> V.(required_assoc string) "contribution" assoc)
obj
let inject (type a) (module D : Key_value.DESCRIBABLE with type t = a)
{ author; contribution } =
D.
[
("author", object_ $ Author.inject (module D) author);
("contribution", string contribution);
]
end
module Article = struct
type t = {
article_title : string;
article_description : string;
tags : string list;
date : Date.t;
title : string option;
description : string option;
author : Author.t;
co_authors : Co_author.t list;
invited_article : bool;
}
let date { date; _ } = date
let tags { tags; _ } = tags
let escape_string str =
let renderer = Cmarkit_renderer.make () in
let buffer = Buffer.create (String.length str) in
let ctx = Cmarkit_renderer.Context.make renderer buffer in
Cmarkit_html.html_escaped_string ctx str;
Buffer.contents buffer
let to_rss_item url article =
let title = escape_string article.article_title in
let description = escape_string article.article_description in
Rss.(
Item.make ~title ~link:url ~pub_date:article.date ~description
~guid:(Guid.link url) ())
let make article_title article_description tags date title description author
co_authors =
let author = Option.value ~default:Author.default_user author in
let invited_article = not (Author.equal author Author.default_user) in
{
article_title;
article_description;
tags = List.map String.lowercase_ascii tags;
date;
title;
description;
author;
co_authors;
invited_article;
}
let from_string (module V : Metadata.VALIDABLE) = function
| None -> Validate.error $ Error.Required_metadata [ "Article" ]
| Some str ->
let open Validate.Monad in
V.from_string str
>>= V.object_and (fun assoc ->
let open Validate.Applicative in
make
<$> V.(required_assoc string) "article.title" assoc
<*> V.(required_assoc string) "article.description" assoc
<*> V.(optional_assoc_or ~default:[] (list_of string))
"tags" assoc
<*> V.required_assoc
(Metadata.Date.from (module V))
"date" assoc
<*> V.(optional_assoc string) "title" assoc
<*> V.(optional_assoc string) "description" assoc
<*> V.(optional_assoc (Author.from (module V))) "author" assoc
<*> V.(
optional_assoc_or ~default:[]
(list_of (Co_author.from (module V)))
"coauthors" assoc))
let inject (type a) (module D : Key_value.DESCRIBABLE with type t = a)
{
article_title;
article_description;
tags;
date;
title;
description;
author;
co_authors;
invited_article;
} =
let co_authors =
List.map (fun c -> D.object_ $ Co_author.inject (module D) c) co_authors
in
let has_co_authors = match co_authors with [] -> false | _ -> true in
D.
[
("root", string "..");
( "metadata",
object_
[
("title", string article_title);
("description", string article_description);
] );
("tags", list (List.map string tags));
("date", object_ $ Metadata.Date.inject (module D) date);
("author", object_ $ Author.inject (module D) author);
("coauthors", list co_authors);
("invited", boolean invited_article);
("has_coauthors", boolean has_co_authors);
]
@ Metadata.Page.inject (module D) (Metadata.Page.make title description)
let compare_by_date a b = Date.compare a.date b.date
end
module Articles = struct
type t = {
articles : (Article.t * string) list;
title : string option;
description : string option;
}
let make ?title ?description articles = { articles; title; description }
let title p = p.title
let description p = p.description
let articles p = p.articles
let set_articles new_articles p = { p with articles = new_articles }
let set_title new_title p = { p with title = new_title }
let set_description new_desc p = { p with description = new_desc }
let sort ?(decreasing = true) articles =
List.sort
(fun (a, _) (b, _) ->
let a_date = Article.date a and b_date = Article.date b in
let r = Date.compare a_date b_date in
if decreasing then ~-r else r)
articles
let sort_articles_by_date ?(decreasing = true) p =
{ p with articles = sort ~decreasing p.articles }
let inject (type a) (module D : Key_value.DESCRIBABLE with type t = a)
{ articles; title; description } =
( "articles",
D.list
(List.map
(fun (article, url) ->
D.object_
(("url", D.string url) :: Article.inject (module D) article))
articles) )
:: ("root", D.string ".")
:: (Metadata.Page.inject (module D) $ Metadata.Page.make title description)
end
let article_object (type a) (module D : Key_value.DESCRIBABLE with type t = a)
(article, url) =
D.object_ (("url", D.string url) :: Article.inject (module D) article)
module Tag = struct
type t = {
tag : string;
tags : (string * int) list;
articles : (Article.t * string) list;
title : string option;
description : string option;
}
let make ?title ?description tag articles tags =
{ tag; tags; articles = Articles.sort articles; title; description }
let inject (type a) (module D : Key_value.DESCRIBABLE with type t = a)
{ tag; tags; articles; title; description } =
("tag", D.string tag)
:: ("root", D.string "..")
:: ("articles", D.list (List.map (article_object (module D)) articles))
:: ( "tags",
D.list
(List.map
(fun (tag, n) ->
D.object_
[
("name", D.string tag);
("link", D.string (tag_path tag));
("number", D.integer n);
])
tags) )
:: (Metadata.Page.inject (module D) $ Metadata.Page.make title description)
end

55
src/model.mli Normal file
View file

@ -0,0 +1,55 @@
open Yocaml
val article_path : Filepath.t -> Filepath.t
val tag_path : string -> Filepath.t
module Article : sig
type t
val date : t -> Date.t
val tags : t -> string list
val to_rss_item : string -> t -> Rss.Item.t
val compare_by_date : t -> t -> int
include Metadata.INJECTABLE with type t := t
include Metadata.READABLE with type t := t
end
module Tag : sig
type t
val make
: ?title:string
-> ?description:string
-> string
-> (Article.t * string) list
-> (string * int) list
-> t
include Metadata.INJECTABLE with type t := t
end
module Articles : sig
type t
val make
: ?title:string
-> ?description:string
-> (Article.t * string) list
-> t
val sort
: ?decreasing:bool
-> (Article.t * string) list
-> (Article.t * string) list
val sort_articles_by_date : ?decreasing:bool -> t -> t
val articles : t -> (Article.t * string) list
val title : t -> string option
val description : t -> string option
val set_title : string option -> t -> t
val set_description : string option -> t -> t
val set_articles : (Article.t * string) list -> t -> t
include Metadata.INJECTABLE with type t := t
end

108
src/task.ml Normal file
View file

@ -0,0 +1,108 @@
open Yocaml
module Metaformat = Yocaml_yaml
module Markup = Yocaml_cmark
module Template = Yocaml_jingoo
let css_target target = "css" |> into target
let javascript_target target = "js" |> into target
let images_target target = "images" |> into target
let template file = add_extension file "html" |> into "templates"
let article_template = template "article"
let layout_template = template "layout"
let list_template = template "list_articles"
let article_target file target = Model.article_path file |> into target
let binary_update = Build.watch Sys.argv.(0)
let index_html target = "index.html" |> into target
let index_md = "index.md" |> into "pages"
let rss_feed target = "feed.xml" |> into target
let tag_file tag target = Model.tag_path tag |> into target
let tag_template = template "tag"
let move_css target =
process_files
[ "css" ]
File.is_css
(Build.copy_file ~into:(css_target target))
;;
let move_javascript target =
process_files
[ "js" ]
File.is_javascript
(Build.copy_file ~into:(javascript_target target))
;;
let move_images target =
process_files
[ "images" ]
File.is_image
(Build.copy_file ~into:(images_target target))
;;
let process_articles target =
process_files [ "articles" ] File.is_markdown (fun article_file ->
let open Build in
create_file
(article_target article_file target)
(binary_update
>>> Metaformat.read_file_with_metadata
(module Model.Article)
article_file
>>> Markup.content_to_html ~strict:false ()
>>> Template.apply_as_template (module Model.Article) article_template
>>> Template.apply_as_template (module Model.Article) layout_template
>>^ Stdlib.snd))
;;
let merge_with_page ((meta, content), articles) =
let title = Metadata.Page.title meta in
let description = Metadata.Page.description meta in
Model.Articles.make ?title ?description articles, content
;;
let generate_feed target =
let open Build in
let* articles_arrow =
Collection.Articles.get_all (module Metaformat) "articles"
in
create_file
(rss_feed target)
(binary_update >>> articles_arrow >>^ Feed.make >>^ Rss.Channel.to_rss)
;;
let generate_tags target =
let* deps, tags = Collection.Tags.compute (module Metaformat) "articles" in
let tags_string = List.map (fun (i, s) -> i, List.length s) tags in
let mk_meta tag articles () = Model.Tag.make tag articles tags_string, "" in
List.fold_left
(fun program (tag, articles) ->
let open Build in
program
>> create_file
(tag_file tag target)
(init deps
>>> binary_update
>>^ mk_meta tag articles
>>> Template.apply_as_template (module Model.Tag) tag_template
>>> Template.apply_as_template (module Model.Tag) layout_template
>>^ Stdlib.snd))
(return ())
tags
;;
let generate_index target =
let open Build in
let* articles_arrow =
Collection.Articles.get_all (module Metaformat) "articles"
in
create_file
(index_html target)
(binary_update
>>> Metaformat.read_file_with_metadata (module Metadata.Page) index_md
>>> Markup.content_to_html ~strict:false ()
>>> articles_arrow
>>^ merge_with_page
>>> Template.apply_as_template (module Model.Articles) list_template
>>> Template.apply_as_template (module Model.Articles) layout_template
>>^ Stdlib.snd)
;;

7
src/task.mli Normal file
View file

@ -0,0 +1,7 @@
val move_css : string -> unit Yocaml.Effect.t
val move_images : string -> unit Yocaml.Effect.t
val move_javascript : string -> unit Yocaml.Effect.t
val process_articles : string -> unit Yocaml.Effect.t
val generate_feed : string -> unit Yocaml.Effect.t
val generate_index : string -> unit Yocaml.Effect.t
val generate_tags : string -> unit Yocaml.Effect.t

8
templates/article.html
View file

@ -1,13 +1,13 @@
<a href="{{ host }}/index.html">Back to index</a> <a href="/index.html">Back to index</a>
<article> <article>
<h1>{{ title }}</h1> <h1>{{ metadata.title }}</h1>
<ul class="tags-list"> <ul class="tags-list">
{%- for tag in tags -%} {%- for tag in tags -%}
<li>{{ tag }}</li> <li><a href="/tags/{{ tag }}.html">{{ tag }}</a></li>
{%- endfor -%} {%- endfor -%}
</ul> </ul>
{%- autoescape false -%} {%- autoescape false -%}
{{ yocaml_body }} {{ body }}
{% endautoescape -%} {% endautoescape -%}
</article> </article>

31
templates/layout.html
View file

@ -1,3 +1,20 @@
{%- if metadata.title -%}
{%- set dash = " - " -%}
{%- set page_title = metadata.title -%}
{%- elseif title -%}
{%- set dash = " - " -%}
{%- set page_title = title -%}
{%- else -%}
{%- set dash = "" -%}
{%- set page_title = "" -%}
{%- endif %}
{% if metadata.description -%}
{%- set page_description = metadata.description -%}
{%- elseif description -%}
{%- set page_description = description -%}
{%- else -%}
{%- set page_description = "blog.robur.coop" -%}
{%- endif -%}
<!doctype html> <!doctype html>
<html lang="en"> <html lang="en">
<head> <head>
@ -5,13 +22,13 @@
<meta http-equiv="x-ua-compatible" content="ie=edge"> <meta http-equiv="x-ua-compatible" content="ie=edge">
<meta name="viewport" content="width=device-width, initial-scale=1"> <meta name="viewport" content="width=device-width, initial-scale=1">
<title> <title>
Robur's blog{{ dash }}{{ title }} Robur's blog{{ dash }}{{ page_title }}
</title> </title>
<meta name="description" content="{{ description }}"> <meta name="description" content="{{ page_description }}">
<link type="text/css" rel="stylesheet" href="{{ host }}/css/hl.css"> <link type="text/css" rel="stylesheet" href="{{ root }}/css/hl.css">
<link type="text/css" rel="stylesheet" href="{{ host }}/css/style.css"> <link type="text/css" rel="stylesheet" href="{{ root }}/css/style.css">
<script src="{{ host }}/js/hl.js"></script> <script src="{{ root }}/js/hl.js"></script>
<link rel="alternate" type="application/rss+xml" href="{{ host }}/feed.xml" title="blog.robur.coop"> <link rel="alternate" type="application/rss+xml" href="{{ root }}/feed.xml" title="blog.robur.coop">
</head> </head>
<body> <body>
<header> <header>
@ -22,7 +39,7 @@
</header> </header>
<main> <main>
{%- autoescape false -%} {%- autoescape false -%}
{{ yocaml_body }} {{ body }}
{% endautoescape -%} {% endautoescape -%}
</main> </main>
<footer> <footer>

12
templates/index.html → templates/list_articles.html
View file

@ -1,7 +1,7 @@
<a class="small-button rss" href="{{ host }}/feed.xml">RSS</a> <a class="small-button rss" href="./feed.xml">RSS</a>
{%- autoescape false -%} {%- autoescape false -%}
{{ yocaml_body }} {{ body }}
{% endautoescape -%} {% endautoescape -%}
<h3>Essays and ramblings</h3> <h3>Essays and ramblings</h3>
@ -20,13 +20,13 @@
{%- endfor -%} {%- endfor -%}
</div> </div>
<div class="content"> <div class="content">
<span class="date">{{ article.date.human }}</span> <span class="date">{{ article.date.canonical }}</span>
<a href="{{ host }}{{ article.url }}">{{ article.title }}</a><br /> <a href="{{ article.url }}">{{ article.metadata.title }}</a><br />
<p>{{ article.description }}</p> <p>{{ article.metadata.description }}</p>
<div class="bottom"> <div class="bottom">
<ul class="tags-list"> <ul class="tags-list">
{%- for tag in article.tags -%} {%- for tag in article.tags -%}
<li>{{ tag }}</li> <li><a href="/tags/{{ tag }}.html">{{ tag }}</a></li>
{%- endfor -%} {%- endfor -%}
</ul> </ul>
</div> </div>

3
update.sh
View file

@ -1,7 +1,6 @@
#!/bin/sh #!/bin/sh
opam exec -- dune exec src/push.exe -- opam exec -- dune exec src/blogger.exe -- push \
-r git@git.robur.coop:robur/blog.robur.coop.git#gh-pages \ -r git@git.robur.coop:robur/blog.robur.coop.git#gh-pages \
--host https://blog.robur.coop \
--name "The Robur team" \ --name "The Robur team" \
--email team@robur.coop --email team@robur.coop