path: root/miniany/README.html

<h1>CC - a self-hosting, bootstrappable, minimal C compiler</h1>
<h2>Introduction</h2>
<p>On the never-ending quest of a minimal system I found Swieros and C4 (the C compiler in 4 functions). Inspired and intrigued I started to implement my own.</p>
<p>For abaos (a small operating system of mine, also in C) I cloned the minimal C library, so we can build a freestanding version of C4.</p>
<p>C4 serves as a test whether my own CC is minimal enough and doesn't use silly functions. Additionally C4 as well as CC are compiled both in a (on Linux) hosted version and a freestanding version. We use a series of compilers like gcc, clang, tcc and pcc to make sure that we are not using silly C constructs.</p>
<p>In order to be able to port easily we make almost no use of system calls, the ones we need are:</p>
<ul>
<li>brk: for malloc/free, change the start address of the heap segment of the process, if the OS only assigns a single static space, then brk results in a NOP.</li>
<li>exit: terminate the process, return does not always work in all combinations (for instance with pcc on Linux). Can be a NOP, we don't require any trickery as <i>atext </i>and we don't use buffering anywhere (for instance flushing stdout on exit).</li>
<li>read/write: read from stdin linearly, write to stdout linearly, this is essentially a model using an input and an output tape. Those two functions must really exist. This basically eliminates the need for a file system which we might not have during early bootstrapping.</li>
</ul>
<p>Similarly we simplify the C language to not use certain features which can cause trouble when bootstrapping:</p>
<ul>
<li>variable arguments: though simple in principle (just some pointers into the stack if you use a stack for function parameters), it is not typesafe. And the only example in practice it's really heavily used for is in printf-like functions.</li>
<li>preprocessor: it needs a filesystem, we take this outside of the compiler by feeding it an (eventually) concatenated list of *.c files.</li>
<li>two types: int and char, so we can interpret memory as words or as bytes.</li>
</ul>
<h2>Local version of C4</h2>
<p>The local version of C4 has the following adaoptions and extensions:</p>
<ul>
<li>switch statement from the <i>switch-and-struct</i>s branch, adapted c4 itself to use switch statements instead of if's (as in the <i>switch-and-structs </i>branch)</li>
<li>struct support from <i>switch-and-structs</i></li>
<li>constants like <i>EO</i>F, <i>EXIT_SUCCES</i>S, <i>NUL</i>L</li>
<li>standard C block comments along to c++ end of line ones</li>
<li>negative enum initializers</li>
<li>do/while loops</li>
<li>more C functions like <i>isspac</i>e, <i>get</i>c, <i>strcm</i>p</li>
<li>some simplified functions for printing like <i>putstring</i>, <i>putin</i>t, <i>putn</i>l</li>
<li>strict C89 conformance, mainly use standard comment blocks, also removed some warnings</li>
<li>some casts around malloc and memset to fit to non-void freestanding-libc</li>
<li>converted printf to putstring/putint/putnl and some helper functions for error reporting like error()</li>
<li>removed all memory leaks</li>
<li>de-POSIX-ified, no open/read/close, use getchar from stdin only (don't assume the existence of a file system), this also means we had to create sort of an old style tape-file with FS markers to separate the files piped to c4.</li>
</ul>
<p><i>Note: </i>only too late I discovered that there was a C5 version of the same compiler, which would maybe have served better as a basis.</p>
<h2>Examples</h2>
<h3>Running on the host system using the hosts C compiler</h3>
<p>Compiled in either hosted (host libc) or freestanding (our own libc, currently IA-32 Linux kernel only syscalls):</p>
<p><code>./build.sh cc hostcc hosted d ./build.sh cc hostcc freestanding d ./cc \&lt; test1.c \&gt; test1.asm</code>
Create a plain binary from the assembly code:</p>
<p><code>fasm test1.asm test1.bin</code>
Disassemble it to verify it's correctness:</p>
<p><code>ndisasm -b32 -o1000000h -a test1.bin</code>
You can choose <i>gcc</i>, <i>clang</i>, <i>tcc </i>or <i>pcc </i>as host compiler (<i>hostcc</i>).</p>
<h3>Running on the host in the C4 interpreter</h3>
<p>Running in C4 interpreter, again, the C4 program can be compiled in hosted or freestanding mode:</p>
<p><code>./build.sh c4 hostcc hosted d ./build.sh c4 hostcc freestanding d</code>
Here again you can choose the host compiler for compiling C4.</p>
<p>Then we have to create the standard input for C4 using:</p>
<p><code>echo -n -e &quot;\034&quot; \&gt; EOF cat cc.c EOF hello.c | ./c4 cat c4.c EOF cc.c EOF hello.c | ./c4 cat c4.c4 EOF c4.c EOF cc.c EOF hello.c | ./c4</code>
EOF contains the traditional FS (file separator) character in the ASCII character set. Every time c4/c4.c is invoked it reads exacly one input file up to the first FS character (or stops at the end of stdin).</p>
<p>We can also use <i>-s</i>, or <i>-d </i>on every level as follows:</p>
<p><code>cat cc.c EOF hello.c | ./c4 -d</code></p>
<h2>References</h2>
<p>Compiler construction in general:</p>
<ul>
<li><i>&quot;Compiler </i><i>Construction&quot;</i>&quot;, Niklaus Wirth</li>
<li><a href="https://github.com/DoctorWkt/acwj">https://github.com/DoctorWkt/acwj</a>: a nice series on building a C compiler, step by step with lots of good explanations</li>
<li><a href="https://www.engr.mun.ca/%7Etheo/Misc/exp">https://www.engr.mun.ca/~theo/Misc/exp</a>_parsing.htm#climbing, <a href="https://en.wikipedia.org/wiki/Operator-precedence">https://en.wikipedia.org/wiki/Operator-precedence</a>_parser#Precedence_climbing_method</li>
<li><a href="https://github.com/lotabout/write-a-C-interpreter/blob/master/tutorial/en/">https://github.com/lotabout/write-a-C-interpreter/blob/master/tutorial/en/</a>, tutorial based on C4 how to build a C interpreter, explains nicely details in C4.</li>
</ul>
<p>C4:</p>
<ul>
<li><a href="https://github.com/rswier/c4.git">https://github.com/rswier/c4.git</a>, <i>C4 </i><i>- </i><i>C </i><i>in </i><i>four </i><i>functions</i>, Robert Swierczek, minimalistic C compiler running on an emulator on the IR, inspiration for this project</li>
<li><a href="https://github.com/rswier/c4/blob/switch-and-structs/c4.c">https://github.com/rswier/c4/blob/switch-and-structs/c4.c</a>, c4 adaptions to provide switch and structs</li>
<li><a href="https://github.com/EarlGray/c4">https://github.com/EarlGray/c4</a>: a X86 JIT version of c4</li>
<li><a href="https://github.com/jserv/amacc">https://github.com/jserv/amacc</a>: based on C4, JIT or native code, for ARM, quite well documented, also very nice list of compiler resources on Github page</li>
</ul>
<p>Other minimal compilers and systems:</p>
<ul>
<li><a href="http://selfie.cs.uni-salzburg.at/">http://selfie.cs.uni-salzburg.at/</a>: C* self-hosting C compiler (also emulator, hypervisor) for RISCV, inspiration for what makes up a minimal C language</li>
<li><a href="http://www.iro.umontreal.ca/%7Efelipe/IFT2030-Automne2002/Complements/tinyc.c">http://www.iro.umontreal.ca/~felipe/IFT2030-Automne2002/Complements/tinyc.c</a>, Marc Feeley, really easy and much more readable, meant as educational compiler</li>
<li><a href="https://github.com/rswier/swieros.git">https://github.com/rswier/swieros.git</a>: c.c in swieros, Robert Swierczek</li>
</ul>
<p>Assembly:</p>
<ul>
<li><a href="https://github.com/felipensp/assembly/blob/master/x86/itoa.s">https://github.com/felipensp/assembly/blob/master/x86/itoa.s</a>, for putint (early debugging keyword)</li>
<li><a href="https://baptiste-wicht.com/posts/2011/11/print-strings-integers-intel-assembly.htm">https://baptiste-wicht.com/posts/2011/11/print-strings-integers-intel-assembly.htm</a> (earldy debugging keyword)</li>
</ul>