3 * Fully source-based bootstrapping
5 ** R6RS-like scheme interpreter
6 This first part is prototyped in C by the mes.c core and Scheme
7 bootstrap code in module/. Of course, while mes.c is pretty small it
8 cannot serve as a fully source-based solution.
10 The initial idea was to have the minimal core support LISP-1.5 (or
11 something very close to that as a tribute to John McCarthy) and extend
12 eval/apply from LISP-1.5 source with define, define-macro etc. and
13 metamorphose into R6RS. It seemed to work but performance of the
14 LISP-intepreted RRS was so bad (~1000x slower than initial LISP-1.5)
15 that this track was abandoned after the initial ANNOUNCE.
17 The route changed, trying to strike a balance between core size and
18 performance: still writing as much as possible in Scheme but having a
19 mescc compiler that takes not more than some seconds to run.
21 Now that portable R6RS syntax-case runs and mes.c has grown to
22 ~1200LOC with another ~300LOC of optional C code, some effort must
23 probably be directed into making that smaller.
25 ** Move mes.c into hex?
26 One idea is to use OrianJ's amazing self-hosting [[https://github.com/oriansj/stage0][stage0]] hex assembler
27 and minimal bootstrap binaries and rewrite the mes.c core to directly
28 bootstrap into Scheme.
30 ** Rewrite mes.c and generate hex?
31 Another idea (thanks Rutger!) is to rewrite the mes.c core in a
32 C/Assembly variant and thave mescc produce the simple, annotated
37 mes.c is ~1500 lines (~10,000LOC Assembly) which seems much too big to
38 start translating it to assembly/hex.
40 ** Actually do something useful, build: [[https://en.wikipedia.org/wiki/Tiny_C_Compiler][Tiny C Compiler]]
41 * OLD: Booting from LISP-1.5 into Mes
43 Mes started out experimenting with booting from a hex-coded minimal
44 LISP-1.5 (prototype in mes.c), into an almost-RRS Scheme.
46 When EOF is read, the LISP-1.5 machine calls loop2 from loop2.mes,
47 which reads the rest of stdin and takes over control. The functions
48 readenv, eval and apply-env in mes.mes introduced define, define-macro
49 quasiquote and macro expansion.
51 While this works, it's amazingly slow. We implemented a full reader
52 in mes.c, which makes running mes:apply-env mes:eval somewhat
53 bearable, still over 1000x slower than running mes.c.
55 Bootstrapping has been removed and mes.c implements enough of RRS to
56 run a macro-based define-syntax and syntax-rules.
58 loop.mes and mes.mes are unused and lagging behind. Probably it's not
59 worth considering this route without a VM. GNU Epsilon is taking the
60 more usual VM-route to provide multiple personas. While that sounds
61 neat, Lisp/Scheme, bootstrapping and trusted binaries are probably not
62 in scope as there is no mention of such things; only ML is mentioned
63 while Guile is used for bootstrapping.
65 * Assorted ideas and info
67 *** [[https://savannah.gnu.org/projects/nyacc][nyacc]]
68 *** PEG: [[http://piumarta.com/software/peg/][parse C using PEG]]
69 *** [[https://en.wikipedia.org/wiki/Tiny_C_Compiler][Tiny C Compiler]]
70 *** [[http://www.t3x.org/subc/index.html][Sub C]]
71 *** [[https://groups.google.com/forum/#!topic/comp.lang.lisp/VPuX0VsjTTE][C intepreter in LISP/Scheme/Python]]
74 *** [[http://www.tldp.org/HOWTO/Assembly-HOWTO/linux.html][Assembly HOWTO]]
75 *** System call clue bat
76 Basically, you issue an int 0x80, with the __NR_syscallname number
77 (from asm/unistd.h) in eax, and parameters (up to six) in ebx, ecx,
78 edx, esi, edi, ebp respectively.
81 *** [[http://www.muppetlabs.com/~breadbox/software/tiny/][Small ELF programs]]
82 *** [[http://www.cirosantilli.com/elf-hello-world/][Elf hello world]]
84 ** SC - c as s-expressions
85 sc: http://sph.mn/content/3d3
87 *** [[http://www.scheme-reports.org/][Scheme Reports]]
88 *** [[ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-349.pdf][Scheme - Report on Scheme]]
89 *** [[ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-452.pdf][RRS - Revised Report on Scheme]]