utf8decoder: Use byte-sized byte argument

This prevents passing a negative value out of a signed char by accident.
Also renders some casts in the code superfluous.

include/extern/utf8decoder.h

@@ -9,6 +9,6 @@
 #ifndef EXTERN_UTF8DECODER_H
 #define EXTERN_UTF8DECODER_H
 
-uint32_t decode(uint32_t *state, uint32_t *codep, uint32_t byte);
+uint32_t decode(uint32_t *state, uint32_t *codep, uint8_t byte);
 
 #endif /* EXTERN_UTF8DECODER_H */

src/asm/parser.y

@@ -61,7 +61,7 @@ static size_t strlenUTF8(const char *s)
 	uint32_t codep = 0;
 
 	while (*s) {
-		switch (decode(&state, &codep, (uint8_t)*s)) {
+		switch (decode(&state, &codep, *s)) {
 		case 1:
 			fatalerror("STRLEN: Invalid UTF-8 character\n");
 			break;
@@ -95,7 +95,7 @@ static void strsubUTF8(char *dest, const char *src, uint32_t pos, uint32_t len)
 
 	/* Advance to starting position in source string. */
 	while (src[srcIndex] && curPos < pos) {
-		switch (decode(&state, &codep, (uint8_t)src[srcIndex])) {
+		switch (decode(&state, &codep, src[srcIndex])) {
 		case 1:
 			fatalerror("STRSUB: Invalid UTF-8 character\n");
 			break;
@@ -113,7 +113,7 @@ static void strsubUTF8(char *dest, const char *src, uint32_t pos, uint32_t len)
 
 	/* Copy from source to destination. */
 	while (src[srcIndex] && destIndex < MAXSTRLEN && curLen < len) {
-		switch (decode(&state, &codep, (uint8_t)src[srcIndex])) {
+		switch (decode(&state, &codep, src[srcIndex])) {
 		case 1:
 			fatalerror("STRSUB: Invalid UTF-8 character\n");
 			break;

src/asm/util.c

@@ -69,7 +69,7 @@ size_t readUTF8Char(uint8_t *dest, char const *src)
 	size_t i = 0;
 
 	for (;;) {
-		if (decode(&state, &codep, (uint8_t)src[i]) == 1)
+		if (decode(&state, &codep, src[i]) == 1)
 			fatalerror("invalid UTF-8 character\n");
 
 		dest[i] = src[i];

src/extern/utf8decoder.c

@@ -38,7 +38,7 @@ static const uint8_t utf8d[] = {
 	  1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* s8 */
 };
 
-uint32_t decode(uint32_t *state, uint32_t *codep, uint32_t byte)
+uint32_t decode(uint32_t *state, uint32_t *codep, uint8_t byte)
 {
 	uint32_t type = utf8d[byte];
 

test/asm/invalid-utf-8.asm

@@ -0,0 +1,5 @@
+; This test tries to pass invalid UTF-8 through a macro argument
+; to exercise the lexer's reportGarbageChar
+m:MACRO \1
+ENDM
+	m <EFBFBD><EFBFBD>

test/asm/invalid-utf-8.err

@@ -0,0 +1,5 @@
+ERROR: invalid-utf-8.asm(4) -> invalid-utf-8.asm::m(3):
+    Unknown character 0xCF
+ERROR: invalid-utf-8.asm(4) -> invalid-utf-8.asm::m(3):
+    Unknown character 0xD3
+error: Assembly aborted (2 errors)!