commit abd884945d66755c239d54ea258a4f366c0e23ae
parent 69600129406d3e2a7d2e06d6567325c158f22ced
Author: Virgil Dupras <hsoft@hardcoded.net>
Date: Sun, 4 Dec 2022 08:46:20 -0500
asm/uxntal: preparing the battleground
Diffstat:
8 files changed, 496 insertions(+), 3 deletions(-)
diff --git a/fs/asm/uxntal.c b/fs/asm/uxntal.c
@@ -0,0 +1,474 @@
+/* Source: https://git.sr.ht/~rabbits/uxn
+ * License: /license/dll-uxn.txt
+ */
+// TODO: Doesn't compile yet. See TODOs below
+
+#define TRIM $100
+#define LENGTH $10000
+
+typedef unsigned char Uint8;
+typedef char Sint8;
+typedef unsigned short Uint16;
+
+typedef struct {
+ // TODO: allow multiple decls per line
+ char name[$40];
+ char items[$40*$40];
+ Uint8 len;
+} Macro;
+
+typedef struct {
+ char name[$40];
+ Uint16 addr;
+ Uint16 refs;
+} Label;
+
+typedef struct {
+ char name[$40];
+ char rune;
+ Uint16 addr;
+} Reference;
+
+typedef struct {
+ Uint8 data[LENGTH];
+ unsigned int ptr;
+ unsigned int length;
+ Uint16 llen;
+ Uint16 mlen;
+ Uint16 rlen;
+ Label labels[$400];
+ Macro macros[$100];
+ Reference refs[$800];
+ char scope[$40];
+} Program;
+
+Program p;
+
+// TODO: compile error
+static char* ops[$20] = {
+ "LIT"0, "INC"0, "POP"0, "NIP"0, "SWP"0, "ROT"0, "DUP"0, "OVR"0,
+ "EQU"0, "NEQ"0, "GTH"0, "LTH"0, "JMP"0, "JCN"0, "JSR"0, "STH"0,
+ "LDZ"0, "STZ"0, "LDR"0, "STR"0, "LDA"0, "STA"0, "DEI"0, "DEO"0,
+ "ADD"0, "SUB"0, "MUL"0, "DIV"0, "AND"0, "ORA"0, "EOR"0, "SFT"0
+};
+
+static int scmp(char *a, char *b, int len) { int i = 0; while(a[i] == b[i]) if(!a[i] || ++i >= len) return 1; return 0; } /* string compare */
+static int sihx(char *s) { int i = 0; char c; while((c = s[i++])) if(!(c >= '0' && c <= '9') && !(c >= 'a' && c <= 'f')) return 0; return i > 1; } /* string is hexadecimal */
+static int shex(char *s) { int n = 0, i = 0; char c; while((c = s[i++])) if(c >= '0' && c <= '9') n = n * 16 + (c - '0'); else if(c >= 'a' && c <= 'f') n = n * 16 + 10 + (c - 'a'); return n; } /* string to num */
+static int slen(char *s) { int i = 0; while(s[i]) i++; return i; } /* string length */
+static int spos(char *s, char c) { Uint8 i = 0, j; while((j = s[i++])) if(j == c) return i; return -1; } /* character position */
+static char *scpy(char *src, char *dst, int len) { int i = 0; while((dst[i] = src[i]) && i < len - 2) i++; dst[i + 1] = 0; return dst; } /* string copy */
+static char *scat(char *dst, const char *src) { char *ptr = dst + slen(dst); while(*src) *ptr++ = *src++; *ptr = 0; return dst; } /* string cat */
+
+static int parse(char *w, int hdl);
+
+static int
+error(const char *name, const char *msg)
+{
+ fprintf(name, msg, "%s: %s\n", ConsoleOut());
+ return 0;
+}
+
+static char *
+sublabel(char *src, char *scope, char *name)
+{
+ return scat(scat(scpy(scope, src, $40), "/"), name);
+}
+
+static Macro *
+findmacro(char *name)
+{
+ int i;
+ for(i = 0; i < p.mlen; i++)
+ if(scmp(p.macros[i].name, name, $40))
+ return &p.macros[i];
+ return NULL;
+}
+
+static Label *
+findlabel(char *name)
+{
+ int i;
+ for(i = 0; i < p.llen; i++)
+ if(scmp(p.labels[i].name, name, $40))
+ return &p.labels[i];
+ return NULL;
+}
+
+static Uint8
+findopcode(char *s)
+{
+ int i;
+ for(i = 0; i < $20; i++) {
+ int m = 0;
+ if(!scmp(ops[i], s, 3))
+ // TODO: implement continue
+ // continue;
+ {}
+ if(!i) i |= (1 << 7); /* force keep for LIT */
+ while(s[3 + m]) {
+ if(s[3 + m] == '2')
+ i |= (1 << 5); /* mode: short */
+ else if(s[3 + m] == 'r')
+ i |= (1 << 6); /* mode: return */
+ else if(s[3 + m] == 'k')
+ i |= (1 << 7); /* mode: keep */
+ else
+ return 0; /* failed to match */
+ m++;
+ }
+ return i;
+ }
+ return 0;
+}
+
+static int
+freadword(char *buf, int hdl)
+{
+ int i, c = 0;
+ while ((c > 0) && (c <= ' ')) c = fgetc(hdl);
+ if (c < 0) /* EOF */ return 0;
+ for (i=0; i < $3f; i++) {
+ // TODO: add CC test for this, I'm really not sure it's going to work.
+ *buf++ = c;
+ c = fgetc(hdl)
+ if ((c > 0) && (c <= ' ')) {
+ *buf = 0;
+ return 1;
+ }
+ }
+ return 0; // too long
+}
+
+static int
+makemacro(char *name, int hdl)
+{
+ Macro *m;
+ char word[$40];
+ if(findmacro(name))
+ return error("Macro duplicate", name);
+ if(sihx(name) && slen(name) % 2 == 0)
+ return error("Macro name is hex number", name);
+ if(findopcode(name) || scmp(name, "BRK", 4) || !slen(name))
+ return error("Macro name is invalid", name);
+ if(p.mlen == $100)
+ return error("Macros limit exceeded", name);
+ m = &p.macros[p.mlen++];
+ scpy(name, m->name, $40);
+ while(freadword(word, hdl) == 1) {
+ // TODO: implement continue
+ if(word[0] == '{') /* continue; */ {}
+ if(word[0] == '}') break;
+ if(word[0] == '%')
+ return error("Macro error", name);
+ if(m->len >= $40)
+ return error("Macro size exceeded", name);
+ scpy(word, m->items[m->len++], $40);
+ }
+ return 1;
+}
+
+static int
+makelabel(char *name)
+{
+ Label *l;
+ if(findlabel(name))
+ return error("Label duplicate", name);
+ if(sihx(name) && (slen(name) == 2 || slen(name) == 4))
+ return error("Label name is hex number", name);
+ if(findopcode(name) || scmp(name, "BRK", 4) || !slen(name))
+ return error("Label name is invalid", name);
+ if(p.llen == $400)
+ return error("Labels limit exceeded", name);
+ l = &p.labels[p.llen++];
+ l->addr = p.ptr;
+ l->refs = 0;
+ scpy(name, l->name, $40);
+ return 1;
+}
+
+static int
+makereference(char *scope, char *label, Uint16 addr)
+{
+ char subw[$40], parent[$40];
+ int pos;
+ Reference *r;
+ if(p.rlen == $800)
+ return error("References limit exceeded", label);
+ r = &p.refs[p.rlen++];
+ if(label[1] == '&')
+ scpy(sublabel(subw, scope, label + 2), r->name, $40);
+ else {
+ pos = spos(label + 1, '/');
+ if(pos > 0) {
+ Label *l;
+ if((l = findlabel(scpy(label + 1, parent, pos))))
+ l->refs++;
+ }
+ scpy(label + 1, r->name, $40);
+ }
+ r->rune = label[0];
+ r->addr = addr;
+ return 1;
+}
+
+static int
+writebyte(Uint8 b)
+{
+ if(p.ptr < TRIM)
+ return error("Writing in zero-page", "");
+ else if(p.ptr > $ffff)
+ return error("Writing after the end of RAM", "");
+ else if(p.ptr < p.length)
+ return error("Memory overwrite", "");
+ p.data[p.ptr++] = b;
+ p.length = p.ptr;
+ return 1;
+}
+
+static int
+writeopcode(char *w)
+{
+ Uint8 res;
+ res = writebyte(findopcode(w));
+ return res;
+}
+
+static int
+writeshort(Uint16 s, int lit)
+{
+ if(lit)
+ if(!writebyte(findopcode("LIT2"))) return 0;
+ return writebyte(s >> 8) && writebyte(s & $ff);
+}
+
+static int
+writelitbyte(Uint8 b)
+{
+ if(!writebyte(findopcode("LIT"))) return 0;
+ if(!writebyte(b)) return 0;
+ return 1;
+}
+
+static int
+doinclude(const char *filename)
+{
+ int hdl;
+ char w[$40];
+ // TODO: add API to detect if file exists
+ hdl = fopen(filename);
+ while(freadword(w, hdl) == 1)
+ if(!parse(w, hdl))
+ return error("Unknown token", w);
+ fclose(hdl);
+ return 1;
+}
+
+static int
+parse(char *w, int hdl)
+{
+ int i;
+ char word[$40], subw[$40], c;
+ Macro *m;
+ if(slen(w) >= 63)
+ return error("Invalid token", w);
+ switch(w[0]) {
+ case '(': /* comment */
+ if(slen(w) != 1) fprintf(w, "-- Malformed comment: %s\n", ConsoleOut());
+ i = 1; /* track nested comment depth */
+ while(freadword(word, hdl) == 1) {
+ if(slen(word) != 1)
+ continue;
+ else if(word[0] == '(')
+ i++;
+ else if(word[0] == ')' && --i < 1)
+ break;
+ }
+ break;
+ case '~': /* include */
+ if(!doinclude(w + 1))
+ return error("Invalid include", w);
+ break;
+ case '%': /* macro */
+ if(!makemacro(w + 1, hdl))
+ return error("Invalid macro", w);
+ break;
+ case '|': /* pad-absolute */
+ if(!sihx(w + 1))
+ return error("Invalid padding", w);
+ p.ptr = shex(w + 1);
+ break;
+ case '$': /* pad-relative */
+ if(!sihx(w + 1))
+ return error("Invalid padding", w);
+ p.ptr += shex(w + 1);
+ break;
+ case '@': /* label */
+ if(!makelabel(w + 1))
+ return error("Invalid label", w);
+ scpy(w + 1, p.scope, $40);
+ break;
+ case '&': /* sublabel */
+ if(!makelabel(sublabel(subw, p.scope, w + 1)))
+ return error("Invalid sublabel", w);
+ break;
+ case '#': /* literals hex */
+ if(!sihx(w + 1) || (slen(w) != 3 && slen(w) != 5))
+ return error("Invalid hex literal", w);
+ if(slen(w) == 3) {
+ if(!writelitbyte(shex(w + 1))) return 0;
+ } else if(slen(w) == 5) {
+ if(!writeshort(shex(w + 1), 1)) return 0;
+ }
+ break;
+ case '_': /* raw byte relative */
+ makereference(p.scope, w, p.ptr);
+ if(!writebyte($ff)) return 0;
+ break;
+ case ',': /* literal byte relative */
+ makereference(p.scope, w, p.ptr);
+ if(!writelitbyte($ff)) return 0;
+ break;
+ case '-': /* raw byte absolute */
+ makereference(p.scope, w, p.ptr);
+ if(!writebyte($ff)) return 0;
+ break;
+ case '.': /* literal byte zero-page */
+ makereference(p.scope, w, p.ptr);
+ if(!writelitbyte($ff)) return 0;
+ break;
+ case ':': /* raw short absolute */
+ case '=':
+ makereference(p.scope, w, p.ptr);
+ if(!writeshort($ffff, 0)) return 0;
+ break;
+ case ';': /* literal short absolute */
+ makereference(p.scope, w, p.ptr);
+ if(!writeshort($ffff, 1)) return 0;
+ break;
+ case '"': /* raw string */
+ i = 0;
+ while((c = w[++i]))
+ if(!writebyte(c)) return 0;
+ break;
+ case '[':
+ case ']':
+ if(slen(w) == 1) break; /* else fallthrough */
+ default:
+ /* opcode */
+ if(findopcode(w) || scmp(w, "BRK", 4)) {
+ if(!writeopcode(w)) return 0;
+ }
+ /* raw byte */
+ else if(sihx(w) && slen(w) == 2) {
+ if(!writebyte(shex(w))) return 0;
+ }
+ /* raw short */
+ else if(sihx(w) && slen(w) == 4) {
+ if(!writeshort(shex(w), 0)) return 0;
+ }
+ /* macro */
+ else if((m = findmacro(w))) {
+ for(i = 0; i < m->len; i++)
+ if(!parse(m->items[i], hdl))
+ return 0;
+ return 1;
+ } else
+ return error("Unknown token", w);
+ }
+ return 1;
+}
+
+static int
+resolve(void)
+{
+ Label *l;
+ int i;
+ for(i = 0; i < p.rlen; i++) {
+ Reference *r = &p.refs[i];
+ switch(r->rune) {
+ case '_':
+ if(!(l = findlabel(r->name)))
+ return error("Unknown relative reference", r->name);
+ p.data[r->addr] = (Sint8)(l->addr - r->addr - 2);
+ if((Sint8)p.data[r->addr] != (l->addr - r->addr - 2))
+ return error("Relative reference is too far", r->name);
+ l->refs++;
+ break;
+ case ',':
+ if(!(l = findlabel(r->name)))
+ return error("Unknown relative reference", r->name);
+ p.data[r->addr + 1] = (Sint8)(l->addr - r->addr - 3);
+ if((Sint8)p.data[r->addr + 1] != (l->addr - r->addr - 3))
+ return error("Relative reference is too far", r->name);
+ l->refs++;
+ break;
+ case '-':
+ if(!(l = findlabel(r->name)))
+ return error("Unknown absolute reference", r->name);
+ p.data[r->addr] = l->addr & $ff;
+ l->refs++;
+ break;
+ case '.':
+ if(!(l = findlabel(r->name)))
+ return error("Unknown zero-page reference", r->name);
+ p.data[r->addr + 1] = l->addr & $ff;
+ l->refs++;
+ break;
+ case ':':
+ case '=':
+ if(!(l = findlabel(r->name)))
+ return error("Unknown absolute reference", r->name);
+ p.data[r->addr] = l->addr >> $8;
+ p.data[r->addr + 1] = l->addr & $ff;
+ l->refs++;
+ break;
+ case ';':
+ if(!(l = findlabel(r->name)))
+ return error("Unknown absolute reference", r->name);
+ p.data[r->addr + 1] = l->addr >> $8;
+ p.data[r->addr + 2] = l->addr & $ff;
+ l->refs++;
+ break;
+ default:
+ return error("Unknown reference", r->name);
+ }
+ }
+ return 1;
+}
+
+static int
+assemble(int hdl)
+{
+ char w[$40];
+ scpy("on-reset", p.scope, $40);
+ while(freadword(w, hdl) == 1)
+ if(!parse(w, hdl))
+ return error("Unknown token", w);
+ return resolve();
+}
+
+static void
+review()
+{
+ int i;
+ for(i = 0; i < p.llen; i++)
+ if(p.labels[i].name[0] >= 'A' && p.labels[i].name[0] <= 'Z')
+ continue; /* Ignore capitalized labels(devices) */
+ else if(!p.labels[i].refs)
+ fprintf(p.labels[i].name, "-- Unused label: %s\n", ConsoleOut());
+ fprintf(
+ p.mlen, p.llen, p.length-TRIM
+ "Assembled %d bytes, %d labels, %d macros.\n", ConsoleOut());
+}
+
+int
+uxntal(int srchdl)
+{
+ if(!assemble(srchdl))
+ return !error("Assembly", "Failed to assemble rom.");
+ if(p.length <= TRIM)
+ return !error("Assembly", "Output rom is empty.");
+ fwrite(p.data + TRIM, p.length - TRIM, 1, StdOut());
+ return 0;
+}
diff --git a/fs/asm/uxntal.fs b/fs/asm/uxntal.fs
@@ -0,0 +1,2 @@
+?f<< /comp/c/lib.fs
+cc<< /asm/uxntal.c
diff --git a/fs/comp/c/lib.fs b/fs/comp/c/lib.fs
@@ -23,7 +23,7 @@
:c void memset(char *a, char c, unsigned int n);
: fgetc IO :getc ;
-:c char fgetc(int hdl);
+:c int fgetc(int hdl);
: fputc IO :putc ;
:c void fputc(char c, int hdl);
: fputback IO :putback ;
diff --git a/fs/doc/cc/lib.txt b/fs/doc/cc/lib.txt
@@ -44,7 +44,7 @@ int isdigit(char c)
The I/O API is a simple layer over the system I/O described at doc/io. `hdl` is
a regular I/O handle.
-char fgetc(int hdl) --> IO :getc
+int fgetc(int hdl) --> IO :getc
void fputc(char c, int hdl) --> IO :putc
void fputback(char c, int hdl) --> IO :putback
char* freadline(int hdl) --> IO :putback
diff --git a/fs/doc/cc/usage.txt b/fs/doc/cc/usage.txt
@@ -35,7 +35,7 @@ are a few differences:
yields 0. This is also true of the "x ? y : z" operator. Both y and z are
evaluated regardless of x.
* Number literals are the same as Dusk OS, so 12345, $1234 and 'X'. No 0x1234
- or 0o777.
+ or 0o777. No '\n', '\r', '\t', '\0' char literals.
* string literals are not null-terminated, but "counted strings". The exact same
format as system strings. You can create a null-terminated string literal by
appending a "0" next to the closing quote, for example "hello"0. This 0 shifts
@@ -57,6 +57,7 @@ are a few differences:
"a = (b = (c = 42))" and "a ? b ? c : d : e ? f : g" (seriously, you wanted
to write that?!?) must be written as "a ? (b ? c : d) : (e ? f : g)"
* The keyword "static" has a slightly different meaning. See below.
+* No multi-dimensional array.
## Pre-processor
diff --git a/fs/tests/asm/all.fs b/fs/tests/asm/all.fs
@@ -1,3 +1,4 @@
\ Run all asm test suites
f<< /tests/asm/i386.fs
f<< /tests/asm/arm.fs
+f<< /tests/asm/uxntal.fs
diff --git a/fs/tests/asm/uxntal.fs b/fs/tests/asm/uxntal.fs
@@ -0,0 +1,7 @@
+?f<< /tests/harness.fs
+\ Doesn't compile yet
+\ ?f<< /asm/uxntal.fs
+testbegin
+\ Tests for asm/uxntal.fs
+testend
+
diff --git a/license/dll-uxn.txt b/license/dll-uxn.txt
@@ -0,0 +1,8 @@
+Copyright (c) 2021 Devine Lu Linvega
+
+Permission to use, copy, modify, and distribute this software for any
+purpose with or without fee is hereby granted, provided that the above
+copyright notice and this permission notice appear in all copies.
+
+THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+WITH REGARD TO THIS SOFTWARE.
